All abstracts for ICOS Science Conference 2026

6 Quantifying the annual atmospheric CO₂ burden from biomass burning and its transport time to monitoring sites

Poster

PP Musaid¹*, Vinu Valsala¹, Santanu Halder², Yogesh K. Tiwari¹

¹Indian Institute of Tropical Meteorology, Pune, India. ²Cooperative Institute for Research in Environmental Sciences, Boulder, USA

Session

Session 19: Understanding feedbacks between greenhouse gas exchange processes and climate variability using in situ observations, remote sensing, and machine learning

Abstract text

In the backdrop of increasing global temperatures, carbon dioxide (CO₂) emissions from biomass burning contribute to positive climate feedback under the failure of forest regrowth, leading to a long-term loss of carbon sinks. This study assesses biomass burning CO₂ contributions and transport times using the NIES-TM model (2003–2019) with GFEDv5 and GFAS emissions across 16 biomes. The TSG and TSM biomes contributed the highest, 10.74 ppm and 6.13 ppm respectively over 17 years with the GFED-based estimates. Trend analysis revealed a significant decreasing trend in TSG and TSM, and an increasing trend in BOF.

The mean transport time of CO₂ to monitoring stations ranged between 6 to 10 days, with regional variation influenced by biome proximity and atmospheric dynamics. Periodogram-based spectral analysis showed multiple peaks at several stations, indicating contributions from various biomes spread and seasonality. A Random Forest model trained on biome age and 850 hPa winds and proximity revealed zonal wind controls transport time, while proximity dominates in TCF and BOF. The study finds that annual fire driven CO₂ enhancement (> 1.0 ppm) exceeds ground-based detection uncertainty, matches satellite XCO₂ retrieval uncertainty and is detectable within 6–8 days. These quantifications imply the need for accurate fire emissions in top-down flux estimates and support the use of a 7 day back-trajectory window in regional inversions. Our findings underscore shifting patterns in burning biomes and emphasize the importance of considering the spatial extent of each biome when evaluating the impact of biomass burning on atmospheric CO₂ levels.

7 COCCON-Spain: Toward an Integrated Greenhouse Gas Observation System in Spain

Poster

Ayoze Álvarez-Hernández^1,2*, Omaira García², Eliezer Sepúlveda^1,2, Noémie Taquet^1,2, Iballa Cabello^1,2, Ramón Ramos², Frank Hase³, Darko Dubravika³, Carlos Alberti³, Jia Chen⁴, Moritz Makowski⁴, Antonio Alcántara², Virgilio Carreño², Pedro Pablo Rivas², Margarita Yela⁵, Olga Puentedura⁵, José Antonio Adame⁶, Gara Villalba-Méndez⁷, Roger Curcoll⁸, Joan Casals⁹, Abel Calle¹⁰, Ramiro González¹⁰, Pedro Martín-Mateos¹¹, Óscar Bonilla-Manrique¹¹, Aldo Moreno-Overyides¹¹, Joaquín Alonso-Montesinos¹², Jesús María Ballestrín-Bolea¹³, Francisco Javier Pérez¹⁴, María Teresa Garea¹⁴, Joan Girona¹⁵, Ramón Pascual¹⁵, Javier Martín-Vide¹⁶, Jorge Núñez¹⁶, Marc Guevara¹⁷, Paula Castesana¹⁷, Carlos Pérez García-Pando¹⁷, Carlos Torres²

¹Tragsatec, Madrid, Spain. ²Centro de Investigación Atmosférica de Izaña (CIAI), Agencia Estatal de Meteorología (AEMET), Santa Cruz de Tenerife, Spain. ³Institute of Meteorology and Climate Research (IMK-ASF), Karlsruhe Institute of Technology (KIT), Karlsruhe, Germany. ⁴Technical University of Munich (TUM), Munich, Germany. ⁵Área de Instrumentación e Investigación Atmosférica (AIIA), Instituto Nacional de Técnica Aeroespacial (INTA), Torrejón de Ardoz-Madrid, Spain. ⁶Estación de Sondeos Atmosféricos – El Arenosillo, Área de Investigación e Instrumentación Atmosférica (AIIA), Instituto Nacional de Técnica Aeroespacial (INTA), Huelva, Spain. ⁷Instituto de Ciencia y Tecnología Ambiental, Universidad Autónoma de Barcelona (ICTA-UAB), Barcelona, Spain. ⁸Institut de Tècniques Energètiques - Universitat Politècnica de Catalunya (INTE-UPC), Barcelona, Spain. ⁹HorPTA, Department of Agri-Food Engineering and Biotechnology, Universitat Politècnica de Catalunya - BarcelonaTech (UPC), Castelldefels, Spain. ¹⁰Grupo de Óptica Atmosférica (GOA), Universidad de Valladolid (UVa), Valladolid, Spain. ¹¹Departamento de Tecnología Electrónica, Universidad Carlos III de Madrid, Leganés, Spain. ¹²Grupo de Recursos Energéticos Solares, Climatología y Física de la Atmósfera, Universidad de Almería, Almería, Spain. ¹³Plataforma Solar de Almería (PSA), Centro de Investigaciones Energéticas, Medioambientales y Tecnológicas (CIEMAT), Almería, Spain. ¹⁴Delegación de AEMET en Galicia, Agencia Estatal de Meteorología (AEMET), A Coruña, Spain. ¹⁵Delegación de AEMET en Cataluña, Agencia Estatal de Meteorología (AEMET), Barcelona, Spain. ¹⁶Observatorio de Fabra, Reial Acadèmia de Ciències i Arts de Barcelona (RACAB), Barcelona, Spain. ¹⁷Centro de Supercomputación de Barcelona (BSC), Barcelona, Spain

Session

Session 23: Remote sensing and vertical profiling of atmospheric greenhouse gases for climate action

Abstract text

This work provides an overview of COCCON-Spain, an integrated GHG observation system that is being developed and implemented by the Spanish Meteorological Agency (AEMET). The system aims to partially address the current lack of continuous atmospheric GHG measurements in Spain, including the estimation of GHG emissions in the main urban–industrial hotspots, the metropolitan areas of Madrid and Barcelona.

COCCON-Spain will be the first permanent national-scale infrastructure dedicated to monitoring atmospheric GHG concentrations and emissions using ground-based remote sensing observations of total columns. It will work in close collaboration with other GHG initiatives, such as ICOS-Spain, in order to build a fully integrated GHG observation system in Spain. In this sense, COCCON-Spain will extend ICOS-Spain atmospheric monitoring to columnar observations, as well as to metropolitan areas and environments with specific conditions (e.g. low and high surface albedo). Of particular interest will be the co-location of ICOS-Spain and COCCON-Spain stations at Izaña Observatory, El Arenosillo, Valladolid, and the Barcelona urban network (ICOS Cities). The resulting integrated system will support the verification of national emission inventories and mitigation actions, as well as the validation of current and future GHG satellite missions.

In addition, COCCON-Spain will continue to participate in short-term campaigns aimed at quantifying GHG column concentrations and deriving emission fluxes in areas where such measurements have not yet been performed. To illustrate the potential of COCCON-Spain for such campaigns, we present two recent examples: measurements during the 2021 Tajogaite volcanic eruption (Canary Islands) and urban emissions in the Madrid metropolitan area.

8 Advancing Locally Developed Ocean Technologies for Blue Carbon Ecosystems Atlas in Africa

Poster

Peter Teye Busumprah*

Ocean Rock Base and Africa Ocean Alliance, Accra, Ghana

Session

Session 3: Blue carbon and seaweed: reforestation and cultivation

Abstract text

This initiative addresses UN Ocean Decade Challenges 5 and 8 by promoting innovative, locally developed technologies to conserve and restore blue carbon ecosystems across Africa. By integrating traditional knowledge with modern scientific approaches, the project enhances the capacity of African communities to monitor and manage mangroves, seagrasses, and saltmarshes key carbon sinks vital for climate mitigation. These technologies facilitate real-time data collection, support climate resilience, and promote sustainable blue economy practices. Additionally, the project advances the digital representation of Africa’s ocean biodiversity through the development of an African Ocean Biodiversity Atlas, providing an accessible platform for data sharing and decision-making. This effort fosters inclusive participation, strengthens regional collaboration, and contributes to global climate goals by harnessing indigenous innovation and digital tools to safeguard Africa’s vital blue carbon assets for future generations.

9 Climate vulnerability and adaptation among transhumant pastoralists in thenorth-western Himalayas

Poster

SHUBHAM*

DR YS PARMAR UNIVERSITY OF HORTICULTURE AND FORESTRY NAUNI SOLAN, KANGRA, India

Session

Session 11: Climate feedbacks from ecosystem management and natural disturbances

Abstract text

Pastoralist, dependent on nature, are increasingly facing livelihood challenges as climate patterns make weather events more unpredictable. A comprehensive assessment of climate vulnerability is therefore essential for designing targeted policy interventions. This study examines the climate vulnerability of transhumant pastoralist Gaddis, Kinnauras, and Lahaulas living in high-altitude regions of the north-western Himalayas. Primary data were collected from 275 pastoralist households across 11 development blocks. The analysis employed a Composite Climate Vulnerability Index based on indicators of exposure, sensitivity, and adaptive capacity, with weights derived using Principal Component Analysis. Inter-district comparisons indicate that overall vulnerability is not determined by exposure alone but is strongly influenced by household sensitivity and adaptive capacity. The results show that Kinnaur (0.54) is the most climate-vulnerable district, followed by Kangra (0.44), Chamba (0.41), and Lahaul and Spiti (0.40). Exposure was highest in Kangra (0.189), while sensitivity peaked in Kinnaur (0.370). Adaptive capacity was lowest in Kangra (–0.197) and relatively better in Kinnaur (–0.011). At the household level, poorer households with limited adaptive capacity were found to be highly vulnerable regardless of location. The average household size was 7.62, with the highest in Kinnaur (9.34), and the worker population constituted 37.8%. Literacy rates ranged from 66.1% in Chamba to 85.88% in Lahaul and Spiti. Pastoralism and agriculture accounted for 38.56% and 35.57% of employment, respectively. Key adaptation strategies included breed diversification, changes in migration timing, and regular veterinary care. Policy efforts should prioritize strengthening household adaptive capacity through investments in human, physical, and financial capital.

11 Data-driven prediction of urban vegetation cooling effects using machine learning and field observations

Poster

Terenzio Zenone¹*, Gabriele Guidolotti², Theodore Endreny³, Teresa Bertolini¹, Michele Mattioni², Emanuele Pallozzi⁴, Carlo Calfapietra²

¹CNR- IRET, Naples, Italy. ²CNR- IRET, Porano, Italy. ³Department of Environmental Resources Engineering, SUNY ESF, Syracuse, USA. ⁴CNR- IRET, Rome, Italy

Session

Session 6: Carbon cycle in the Mediterranean region: from the local to the regional scale

Abstract text

Rapid urbanization and growing evidence linking extreme heat events to adverse health outcomes highlight the importance of Urban Green Areas (UGAs) in providing ecosystem services that enhance human well-being. Among these services, air temperature cooling (ΔT°C) through evapotranspiration (ET) plays a critical role in mitigating heat-related risks. This study evaluates the capability of Machine Learning (ML) algorithms to predict ΔT°C in UGAs, supporting thermal regulation and informing sustainable urban planning.

Multi-year Eddy Covariance (EC) observations of ET were used to train and validate several ML models. Predictor importance analysis based on mean absolute Shapley values identified incoming shortwave radiation (Rg) as the most influential variable (Sh = 0.45), followed by vapor pressure deficit (VPD, Sh = 0.20), relative humidity (RH, Sh = 0.075), air temperature (AirT, Sh = 0.065), friction velocity (u*, Sh = 0.02), and wind speed (WS, Sh = 0.01). Bagging and Least-Squares Boosting (LSBoost) achieved the highest training performance (R² = 0.89 and 0.83), while Neural Networks, Gaussian Process Regression, and Support Vector Machines performed slightly lower (R² = 0.79–0.80).

Ten-fold cross-validation confirmed model robustness, and Taylor diagrams indicated strong agreement with observed data (σₙ ≈ 0.89; r ≈ 0.90). Testing over three years showed reduced accuracy, with Bagging and LSBoost maintaining superior performance (average R² ≈ 0.66–0.67). Midday predictions exhibited a slight overestimation (0.31°C ± 0.2). Overall, UGAs demonstrated a cooling potential of 2–4°C during summer, emphasizing their importance in climate adaptation strategies.

12 Measuring impact of school-based climate education in Eastern Uganda using adoption, learning gains, and student-led action

Poster

Hood Kagoda*

Girls for Climate Action, Jinja, Uganda. Uganda Developmement and Health Associates, Jinja, Uganda

Session

Session 30: Assessing impact in RIs

Abstract text

Climate education is widely promoted, but many programs struggle to show measurable impact beyond awareness. This presentation shares a practical impact measurement approach drawn from the Green Master Program (GMP), a school-based climate education initiative implemented in Eastern Uganda (2023–2025). The approach was designed for low-resource contexts where schools need simple, usable indicators that can be tracked each term.

We used a three-part framework to assess impact: (1) institutional adoption (school partnerships, integration into co-curricular routines, and continuity of implementation), (2) learning gains (termly pre- and post-tests to track changes in climate knowledge and attitudes), and (3) student-led action (observable climate club activities and school-community initiatives). Implementation included facilitator onboarding, Climate Captains leadership training, structured curriculum sessions, and regular school outreaches. Student engagement was reinforced through climate clubs, Orchard Week activities, eco-innovation challenges, and a student magazine (EcoBuzz) that showcased locally led solutions.

Monitoring results showed consistent improvements in climate knowledge across terms, stronger student ownership of climate action through active clubs, and increased visibility of practical eco-choices in schools (waste management, tree planting, and school gardening). The framework also highlighted what drives sustainability: strong school leadership support, simple tools for tracking outcomes, and activities that translate learning into routine practice.

We conclude with lessons for research infrastructures and climate education programs on how to communicate impact using a balanced mix of quantitative indicators and grounded narrative evidence that decision-makers can trust.

13 Latest Tools for Using Direct Flux Measurement Outside Academia:Harvesting Carbon 1.1 - A New Station-Level Protocol with FluxMapping

Poster

George Burba^1,2,3*, Stefan Metzger^4,5, Taylor Thomas², Sam Bower⁴

¹Daugherty Water for Food Global Institute, Lincoln, USA. ²LI-COR, Lincoln, USA. ³CarbonDew, Lincoln-Boulder, USA. ⁴AtmoFacts, Boulder, USA. ⁵CarbonDew, Boulder-Lincoln, USA

Session

Session 31: Flux measurements for immediate societal benefits

Abstract text

Continental-scale research infrastructures, along with smaller flux networks and individual sites, have long provided direct continuous observations, establishing flux measurements as a gold standard for quantifying carbon, greenhouse gas, water, and energy exchange. 

Despite their advantages, such as high temporal resolution, near-real-time availability, and mechanistic directness, direct flux methods have been slow to transition beyond academia. Key barriers include perceived methodological complexity, operational and cost constraints of traditional systems, insufficient geographic coverage, and the absence of a practitioner-oriented framework designed to deliver immediate societal and commercial value. 

This presentation introduces Harvesting Carbon 1.1, a station-level protocol for atmospheric measurement, reporting, and verifcation (aMRV) that provides simplified guidance and step-by-step workflows for carbon accounting. Beyond the original 1.0 version, it now includes a new critical component, FluxMapping, which transforms a single flux tower into tens of thousands of virtual sensors to generate spatially explicit, high-resolution flux maps across entire project domains. FluxMapping addresses the long-standing limitation of footprint-integrated fluxes by enabling source-sink attribution to specific areas, improved detection of small or borderline signals, and enhanced localization of individual sources. 

By strengthening sampling statistics and reducing uncertainty, FluxMapping order-of-magnitude enhances the scalability of direct flux measurements for aMRV and for integration with remote-sensing-based models. Applications include audit-ready datasets for carbon markets, fine-scale water-loss mapping, flux quantification for intermediate-size plots, and source localization for industrial sectors.  

The overarching aim of this presentation is to brainstorm advanced use of high-resolution flux measurements for actionable environmental decision-making and immediate societal benefits. 

14 Latest Tools for Using Direct Flux Measurement Outside Academia:Clear Guidance, Automation, Professional Services, and Weather Station-Inspired Approach

Poster

George Burba*

Daugherty Water for Food Global Institute, Lincoln, USA. LI-COR, Lincoln, USA. CarbonDew, Lincoln-Boulder, USA

Session

Session 31: Flux measurements for immediate societal benefits

Abstract text

Continental-scale research infrastructures and flux networks, alongside smaller networks and individual sites, directly measure evaporative water loss, heat, CO2, CH4, and other gas exchange between the surface and the atmosphere. Over four decades, such flux stations covered 2100+ stationary measurement points and various campaign sites.

Despite such major advantages as extremely high-resolution, real-time results, continuous and direct nature of flux measurements, their applications are only now and rather slowly entering fields beyond academia due to the perceived method complexity, actual complexity and cost of traditional instrumentation and site operation, lack of broad geographic data coverage, and absence of a comprehensive approach focused on direct flux measurements specifically tailored for bringing immediate societal benefits.

This presentation continues to address these challenges by simplifying explanations, offering detailed guides for practicing the method, presenting the latest lower-cost simple automated instrumentation and novel computing tools, facilitating peer-to-peer cross-sharing to increase data coverage and reduce station setup costs, and providing professional services for experiment design and execution. 

One example of the most recent developments is the 2025 publication of three new plain-language guides/protocols on direct dMRV/aMRV/MMRV. These aim to fundamentally change carbon markets by providing a direct, defensible, traceable, repeatable, real-time, evidence-based approach to quantify sequestration and emission in application beyond academia.

The ultimate goal of this presentation is to ignite discussions on utilizing flux measurements for practical decision-making applications to benefit society and to identify current needs, ideas, and examples for leveraging flux data in everyday decision contexts.

15 Meet the FLORES project: Mechanistically Tracking Forest Photosynthesis and Transpiration through Multi-scale Chlorophyll Fluorescence Signals

Poster

Quentin Beauclaire¹*, Bernard Longdoz¹, Ivan Janssens², Simon De Cannière², Jan De Pue³, Matthias Cuntz⁴, François Jonard⁵

¹BIODYNE Biosystems Dynamics and Exchanges, TERRA Teaching and Research Center, Gembloux Agro-Bio Tech, University of Liege, Liege, Belgium. ²Department of Biology, Research Group PLECO (Plant and Ecosystems), University of Antwerp, Wilrijk, Belgium. ³Royal Meteorological Institute of Belgium, Meteorological and Climatological Research,, Brussels, Belgium. ⁴Université de Lorraine, AgroParisTech, INRAE, Nancy, France. ⁵Earth Observation and Ecosystem Modelling Laboratory, SPHERES Research Unit, University of Liège, Liege, Belgium

Session

Session 16: Using sun-induced chlorophyll fluorescence to understand or scale EC fluxes

Abstract text

Recent advances in hyperspectral and microwave remote sensing products, particularly solar-induced chlorophyll fluorescence (SIF) and vegetation optical depth (VOD), provide complementary insights into photosynthetic activity and vegetation water status of forests, offering new opportunities for improved stress detection under climate change. To translate these observations into robust water and carbon flux estimates, mechanistic models that explicitly link remote sensing signals to underlying biological processes are needed. The FLORES research project (funded by BELSPO) addresses this objective by developing an innovative mechanistic approach through integration of multi-scale SIF and VOD observations, from leaf physiology to satellite monitoring. Anchored within the ICOS network, the project leverages continuous carbon and water flux measurements, top-of-canopy SIF, and canopy water dynamics from GNSS-T–derived VOD as key calibration and validation datasets for model development and evaluation. UAV-based hyperspectral (including SIF), thermal, and LiDAR sensors will be deployed to bridge leaf-scale processes and satellite observations. FLORES will further exploit FLEX and SMAP mission data to deliver robust estimates of forest carbon and water fluxes across temperate forests, strengthening the scientific basis for sustainable forest management under a changing climate. 

16 Uncertain soil organic carbon stock changes from eddy covariance compared with inventories and modelling in a Belgian cropland across five rotations

Oral

Quentin Beauclaire¹*, Bernard Longdoz¹, Laura Delhez¹, Neo Arquin¹, Marmar Sabetizadeh², Bruna Winck³, Benjamin Loubet³, Nicolas Saby⁴, Bernard Heinesch¹

¹BIODYNE Biosystems Dynamics and Exchanges, TERRA Teaching and Research Center, Gembloux Agro-Bio Tech, University of Liege, Liege, Belgium. ²Earth and Life Institute, Université Catholique de Louvain, Louvain-la-Neuve, Belgium. ³ECOSYS, University Paris-Saclay, INRAE, AgroParisTech, Palaiseau, France. ⁴Info&Sols, INRAE, Orléans, France

Session

Session 12: Comparing long-term eddy covariance measurements in terrestrial ecosystems with carbon stock variations: lessons and future challenges

Abstract text

Soils constitute the largest carbon pool in the terrestrial biosphere, storing two to four times more carbon than vegetation and the atmosphere combined. Consequently, a robust assessment of the methods available to quantify long-term  soil organic carbon (SOC) stock dynamics in croplands is essential to evaluate agriculture’s climate change mitigation potential. This study compares three independent approaches - decadal soil inventories, RothC model simulations, and carbon balance derived from eddy covariance (EC) fluxes, combined with explicit uncertainty analyses, to evaluate SOC stock changes at the BE-Lon ICOS cropland site (Belgium) over 21 years.

Soil inventories and RothC simulations indicated near-equilibrium SOC stocks. In contrast, EC-derived carbon budgets suggested mean annual SOC losses of 108 g C m⁻² yr⁻¹. If linearly extrapolated, such losses would deplete the initial SOC stock within approximately 55 years , which is unlikely due to constant management at BE-Lon over the last century. Concerning uncertainties, while analysis of the soil inventories showed no significant difference between the two sampling dates, uncertainty propagation revealed substantial dispersion in RothC simulations (±16 g C m⁻² yr⁻¹). For the flux-based approach, cumulative uncertainty in net biome productivity (NBP; ±118 g C m⁻² yr⁻¹) was comparable in magnitude to the estimated cumulative NBP, dominated by uncertainties related to spectral corrections and friction velocity threshold selection. Overall, integrating multiple independent approaches reveals both the potential and current methodological limitations of site-level SOC stock change estimation, emphasizing the need to reduce systematic uncertainties in EC-based carbon budegt to enable robust and reliable long-term assessments.

17 NEW IFS 125HR IN BOLOGNA AND FIRST COMPARISON OF RETRIEVED GHGs WITH A CO-LOCATED EM27/SUN

Poster

Andrè Achilli^1,2*, Elisa Castelli¹, Enzo Papandrea¹, Paolo Pettinari¹, Francescopiero Calzolari¹, Claudio Campenni¹

¹CNR-ISAC, Bologna, Italy. ²UNIBO, Bologna, Italy

Session

Session 23: Remote sensing and vertical profiling of atmospheric greenhouse gases for climate action

Abstract text

The city of Bologna is located in the Italian Po Valley, a region with high levels of anthropization and industrialization that make it one of the most polluted areas in Europe. To monitor the levels of greenhouse gases and pollutants in the region, in 2024 we installed a EM27/SUN Fourier Transform Spectrometer (FTS) in the Air Quality Observatory, located at the CNR-ISAC premises in Bologna. It acquires near infrared (NIR) interferograms of direct solar radiation, and these are analyzed via the PROFFAST software to retrieve total columnar content and dry air gas mole fractions (DMFs) of CO₂, CO, CH₄, and H₂O. Since this instrument is part of the COCCON network, we regularly send our products to them for publication on the ESA-EVDC webpage. In order to increase the detection capabilities of the observatory, in late 2025 we installed a new IFS 125HR FTS. This instrument measures solar interferograms in the mid infrared (MIR) and NIR ranges, and its high-resolution permits to retrieve lots of atmospheric components, including CO₂, CO, CH₄, H₂O, HDO, HF, HCl, HNO₃, HCN, ClONO₂, C₂H₆, and O₃. This instrument is compliant to the TCCON and NDACC networks, and the interferograms are analyzed following by using the GGG2020 software. 

Here we report an overview of the observatory, the first spectra acquired simultaneously by the two instruments, and a comparison of the retrieved DMFs of CO₂, CH₄ and CO.

18 A Framework for Ecosystem Flux Measurements and their Immediate Societal Benefits: A Case Study Across Nigerian Ecosystems

Poster

Prof John Stephen Kayode*

Federal University Lokoja, Kogi State, Nigeria

Session

Session 31: Flux measurements for immediate societal benefits

Abstract text

Ecosystem flux measurements — quantifying exchanges of carbon, water, and energy between terrestrial ecosystems and the atmosphere — provide critical data for environmental management, natural resource policy, climate adaptation planning, and sustainable development. Techniques such as eddy covariance and chamber methods enable continuous, high-resolution tracking of greenhouse gas exchanges and biophysical fluxes, linking ecosystem processes to drivers such as land use, water availability, and climate variability. Despite Nigeria’s varied ecosystems — including savannahs, forests, wetlands, and agroecosystems — there is limited systematic flux data to inform policy, agricultural decision-making, and carbon management strategies. This paper proposes establishing a network of regionally distributed flux measurement stations tailored to Nigeria’s key biomes to provide immediate societal benefits including improved climate risk assessments, agroecosystem productivity optimization, water stress monitoring, carbon budgeting support for national climate commitments, and urban air quality guidance. We outline a methodological framework for deploying state-of-the-art low-cost and standard flux measurement technologies, discuss integration with remote sensing and modeling approaches, and examine pathways for data utilization by stakeholders from farmers to policymakers. Based on global flux network experiences and recent technological advances, this case study demonstrates how actionable flux data can enhance ecosystem service valuation, climate mitigation planning, and sustainable land management in Nigeria. Strategic investments in flux measurement infrastructure will strengthen evidence-based environmental governance and contribute to socioeconomic resilience amid climate change.

19 Machine Learning-Enhanced LEAP Modelling For UK Net-Zero Pathways: An Integrated Framework for Anthropogenic CO₂ Emissions Forecasting to 2050

Poster

Francisca Oshim¹*, Ezekiel Nwibo¹, Adanna Akujiokwu¹, Marwa Waly¹, Vincent Uzomah², Yassin Osman¹

¹University of Greater Manchester, Bolton, United Kingdom. ²University of Salford, Salford, United Kingdom

Session

Session 18: Quantifying anthropogenic greenhouse gas emissions from continental to regional scales

Abstract text

Increase in CO₂ emissions from anthropogenic sources has been the primary driver of global warming, worsening the impact of climate change and making it one of the most important environmental problems globally. This study integrates Machine Learning (ML) evaluation and Low Emission Analysis Platform (LEAP) simulation to quantitatively evaluate the anthropogenic CO₂ emissions footprint from 1998 to 2050. The empirical findings from the Random Forest algorithm and SHapley Additive exPlanations (SHAP) analysis identified total energy demand (24.69%), energy intensity (21.81%), renewable share (19%), and EV penetration rate (12.77%) as the primary factors influencing CO₂ emission. Conversely, GDP per capita (10.44%), climate change levy (7.02%), and carbon price support (4.28%) were ranked as secondary drivers. CO₂ emissions forecasts using the exponential smoothing model indicated that, under business-as-usual (BAU), the UK would be unable to reach net-zero by 2050; however, it further indicated that the UK may not achieve net-zero by 2050 by only improving in the identified primary drivers, with a projected emission of 48 MtCO₂e. The LEAP scenario analysis indicated that net-zero is achievable but requires combined mitigation strategies and policy interventions with a projected net-zero value of -0.5 MtCO₂e. Improvements in energy efficiency, electrification, accelerated deployment of renewable energy, integration of thermal plants with CCS, and the use of hydrogen and bioenergy were identified as decarbonisation pathways feasible for achieving net-zero by 2050. The conclusion drawn from this research will benefit policymakers, industry professionals, and stakeholders by providing actionable pathways to achieve net-zero emissions in the UK by 2050.

20 Micrometeorological Assessment of CO₂ and H₂O Fluxes in Cotton Production of the Gediz Basin

Poster

Alican Eren¹*, Mehmet Ali Ul¹, Ali Levent Yağcı²

¹Ege University, İzmir, Turkey. ²Gebze technical university, İzmit, Turkey

Session

Session 6: Carbon cycle in the Mediterranean region: from the local to the regional scale

Abstract text

The Gediz Basin is one of Türkiye’s major cotton-producing regions. In recent years, declining water resources due to climate change have significantly restricted irrigation supplies, placing increasing pressure on agricultural production. Agriculture is both affected by and contributes to climate change. Cotton plants act as carbon sinks during the growing period by assimilating atmospheric CO₂ into biomass through photosynthesis, while releasing CO₂ through respiration. Therefore, continuous monitoring of carbon and water vapour fluxes is essential to understand plant–atmosphere interactions and support carbon and water balance modelling.

In this study, H₂O and CO₂ fluxes exchanged between the atmosphere and cotton were measured throughout the growing season using the Eddy Covariance method. Measurements were performed with a LI-COR LI-7500 gas analyser, recording sensible heat flux (H), latent heat flux (LE), and Net Ecosystem Exchange (NEE) at 30-minute intervals. Net radiation (Rn), soil heat flux (G), and meteorological variables including air temperature, relative humidity, precipitation, and atmospheric pressure were simultaneously monitored for energy balance assessment and data quality control.

Raw data were processed using EddyPro within the LI-COR SmartFlux system, and gap-filling and quality control procedures were applied using the REddyProc package. Results indicated that the cotton ecosystem functioned as a carbon sink during the growing season. Maximum daily carbon sequestration reached 8.3 g C m⁻² day⁻¹, while seasonal cumulative sequestration was −268.9 g C m⁻². Average daily evapotranspiration was 9.7 mm day⁻¹. These high-resolution flux data support regional carbon budgeting and sustainable water management.

21 ESTIMATING CARBON FLUXES FROM EDDY COVARIANCE MEASUREMENTS WITH SATELLITE-DERIVED VEGETATION INDICES IN A U.S. GRASSLAND: THE ARM- SOUTHERN GREAT PLAINS SITE

Poster

Obumneke Ohiaeri*

Texas Tech University, Lubbock, USA

Session

Session 22: Remote sensing and flux observations to link disturbance, ecosystem states, and carbon-water fluxes

Abstract text

Spatial upscaling of eddy covariance (EC) carbon fluxes remains limited by footprint heterogeneity and mismatch with satellite observations. This study evaluates the ability of footprint-aligned vegetation indices (VIs) derived from Sentinel-2 and Landsat-9 imagery to represent midday net ecosystem exchange (NEE) at the US-ARM Southern Great Plains grassland (2020–2024). Midday EC fluxes (12:00–14:00 local time) were paired with surface reflectance imagery clipped to the validated 90% cumulative flux footprint. Linear models (NEE = a*VI + b) were developed using NDVI, EVI, and NDMI and assessed under contrasting vapor pressure deficit (VPD) conditions. Across all midday observations (n = 293), EVI and NDMI explained substantially more variance in NEE (R² = 0.44 and 0.45; RMSE ≈ 1.5) than NDVI (R² = 0.23). Under dry atmospheric conditions (VPD ≥ 1.5 kPa), correlations remained strong for EVI (r = 0.66) and NDMI (r = 0.67) but were weaker for NDVI (r = 0.48), indicating that moisture-sensitive indices better capture physiological regulation of carbon uptake during atmospheric stress. Seasonal dynamics showed peak summer values of 0.58–0.65 (NDVI/EVI) and 0.45–0.50 (NDMI), consistent with tallgrass prairie phenology. Spatial mapping within the flux footprint showed strong heterogeneity in midday NEE under dry conditions (−8 to −2 µmol CO₂ m⁻² s⁻¹) and more uniform uptake during wet periods. These results demonstrate that footprint-aligned multispectral imagery can provide moderate and consistent representation of midday carbon exchange, while highlighting the importance of integrating atmospheric drivers for improved carbon flux modeling.

22 Vegetation Photosynthesis Model (VPM, v4.0): Integration of remote sensing and eddy flux tower sites for estimating daily gross primary production from sites to the globe

Poster

Xiangming Xiao¹*, li Pan², Cheng Meng¹, Baihong Pan¹

¹University of Oklahoma, Norman, USA. ²National University of Singapore, Singapore, Singapore

Session

Session 22: Remote sensing and flux observations to link disturbance, ecosystem states, and carbon-water fluxes

Abstract text

Accurate estimation of Gross Primary Production (GPP) of terrestrial vegetation is essential for understanding the carbon cycle. Here, we introduce Vegetation Photosynthesis Model (VPM v4.0) that estimates daily GPP with satellite images and daily climate data. We run the Vegetation Photosynthesis Model (VPM, v4.0) simulations with satellite images, atmospheric CO₂ concentration data, ERA5-Land climate, and annual land cover maps to estimate daily GPP of the global terrestrial ecosystems during 2000-2023. At the site scale, annual GPP and its interannual variability (IAV), and the trends from VPM-v4 have strong agreement with those from the eddy covariance tower sites that have >= 10 years of measurements, a total of 777 site‑years. At the global scale, global annual GPP estimate from VPM-v4 varies from 145 Pg C yr^-1 in 2000 to 166 Pg C yr^-1 in 2023, with an increasing trend of 0.78 Pg C yr^-2, which is higher than those from many terrestrial biospheric models (TBMs) and Earth System models (ESMs). The direct CO₂ effect on annual GPP varies among the biomes and is high in tropical and subtropical forests and contributes substantially to continued increase of global annual GPP during 2000-2023. The global GPP estimates from VPM-v4 falls within the range of annual GPP estimates from carbonyl sulfide (OCS)‑based method (157 Pg C yr^-1, 2000-2010), respiration‑based method (149 Pg C yr^-1), and ¹⁸O‑based method (150–175 Pg C yr^-1), which suggests that it could serve as useful data products for the studies of terrestrial carbon cycle.

24 Flow-Distortion Errors in Three-Component Ultrasonic Anemometers with Orthogonal and Non-Orthogonal Transducer Arrays: Implications for Surface-Layer Energy fluxes

Poster

Ivan Bogoev*, Brian Strickler

Campbell Scientific, Logan, USA

Session

Session 34: Manufacturers' session

Abstract text

Ultrasonic anemometers (UA) are frequently employed to measure wind, air temperature, and turbulent exchange of energy and matter in the atmospheric boundary layer. They are fast-response, linear, accurate, first-principle instruments.  A fundamental limitation of UA is the self-shadowing wake effect caused by the ultrasonic transducers and support structures, leading to underestimation of the wind measurement along the acoustic paths.  In a widely used non-orthogonal UA design each of the three acoustic paths is tilted 60 degrees from the horizontal plane and equally spaced 120 degrees around the vertical axes. The advantage of the non-orthogonal UA is that the transducers are taken out of the horizontal plane and the three sensing paths intersect forming a small measurement volume. Alternatively, in a less common orthogonal UA design, the acoustic paths are arranged perpendicular to each other and parallel to the axes of a Cartesian coordinate system, allowing the measurement of the vertical wind component by a single pair of transducers. A disadvantage of the orthogonal UA is the large separation between the wind components and the self-shadowing effects of the transducers in the horizontal plane.   This study is unique because the two UAs use the same ultrasonic transducers, have equal path length to transducer diameter ratios, utilize the same time-of-flight signal processing algorithm, sample rate and measurement bandwidth. The primary difference between the two UAs is the orientation of the six acoustic paths. We demonstrate the details of the design of the combined probe and present results from a field experiment.

25 Unlocking Flux Data: Enhancing Usability Through User-Centered Co-Design

Oral

Rachel Hollowgrass^1,2*, Sébastien Biraud¹, Christin Buechner¹, You-Wei Cheah¹, Danielle Christianson¹, Housen Chu¹, Trevor Keenan², Karla Leibowitz³, Sy-Toan Ngo¹, Fianna O’Brien¹, Dario Papale⁴, Gilberto Pastorello¹, Rosemary Ting¹, Brian Wang¹, Margaret Torn¹

¹Lawrence Berkeley National Laboratory, Berkeley, USA. ²University of California, Berkeley, Berkeley, USA. ³HyperArts, Oakland, USA. ⁴Università degli Studi della Tuscia, Viterbo, Italy

Session

Session 32: Unlocking climate research solutions through co-design

Abstract text

Regional flux networks have collected thousands of instrument-years of data essential for modeling Earth’s biogeochemical cycles. However, technical friction can limit the accessibility and utility of these datasets. This barrier can slow formal research, hinder flux-adjacent insights, and obstruct lay understanding and support. This can send an implicit message that flux data are only for the few.

This presentation explores how co-design and user experience (UX) methodologies can bridge this gap. By involving researchers and stakeholders directly in the design process, we can create tools that are more intuitive and usable. This presentation will include examples of engagement demonstrating how design thinking has improved data usability across diverse platforms, including graphical user interfaces (GUIs), submission quality controls, and generative AI integration. Attendees will learn how to identify usability bottlenecks and collaborate with design teams to build more impactful scientific services.

26 Improving Membrane-Based pCO₂ Measurements: Laboratory Evaluation and Redesign of the SubCtech OceanPack Race System

Poster

Kristin Kampen¹*, Tobias Steinhoff¹, Jana Fahning²

¹GEOMAR Helmholtz Centre for Ocean Research Kiel, Kiel, Germany. ²SubCtech GmbH, Kiel, Germany

Session

Session 5: Advancing marine CO₂ observations through next-generation sensors, integration and platform innovation

Abstract text

Accurate observations of surface ocean CO₂ are essential for reducing uncertainties in global ocean carbon‐uptake estimates, particularly in the Southern Ocean, where sparse data coverage contributes substantially to the persistent mismatch between model- and data-based uptake assessments. Within the TRICUSO project, we focus on improving membrane-based pCO₂ systems such as the SubCtech OceanPack Race, which offer unique opportunities for deployment on autonomous and citizen-science platforms but currently fall short of climate-quality requirements. Building on insights from the 2021 ICOS OTC pCO₂ instrument intercomparison, which highlights larger uncertainties in membrane equilibrator systems relative to classical air–water equilibrators, we aim to systematically characterize and reduce these uncertainties through controlled long-term laboratory experiments.

Our work focuses on quantifying the influence of key parameters—temperature stability, pressure effects, humidity control, and CO₂ detection precision—on overall pCO₂ accuracy. Using extended experiments in the GEOMAR test facility, we examine sensor response during abrupt environmental changes and evaluate long-term drift. Building on these diagnostics, we aim to innovate the SubCtech system and improve data quality by developing and testing targeted improvements, including enhanced thermal monitoring, optimized drying of carrier gases, and the integration of additional reference gases.

Here we present the current status of our analysis, outline the methodological framework for identifying critical uncertainty sources, and show initial results demonstrating where the OceanPack Race system requires redesign. These findings directly support the development of reliable, scalable pCO₂ measurements needed for emerging global frameworks such as the World Meteorological Organization Global Greenhouse Gas Watch.

27 Using carbonyl sulphide measurements for canopy carbon uptake estimates in needleleaf forests

Oral

Peter Bosman¹*, Maarten Krol^1,2

¹Meteorology and Air Quality Group, Wageningen University, Wageningen, Netherlands. ²Institute for Marine and Atmospheric Research, Utrecht University, Utrecht, Netherlands

Session

Session 13: Advancing approaches for quantifying greenhouse gas fluxes in terrestrial ecosystems

Abstract text

Carbonyl sulphide (COS) is an atmospheric trace gas that has been suggested as a proxy to estimate carbon uptake by plants. To this end, the concept of leaf relative uptake (LRU), the ratio of deposition velocities of COS and CO₂, has been introduced to obtain plant CO₂ uptake fluxes from COS flux measurements. In our study we use a coupled soil – canopy – atmospheric mixed layer model to simulate CO₂ and COS uptake by vegetation explicitly, and derive LRU. In this modelling framework, the exchange of COS is coupled to the exchange of H₂O and CO₂ via stomatal conductance. The latter is calculated using a photosynthesis model, accounting for separate exchange at sunlit and shaded leaves. The models are embedded in an inverse modelling framework, allowing for a structured model parameter estimation. We performed a parameter optimisation for a boreal forest in Finland (Hyytiälä), using observation data from July 2015. We took a holistic approach and aimed to obtain model parameters consistent with a large set of observations, including COS and CO₂ molar fractions (measured in and above the canopy) and fluxes. By optimising parameters, we obtained a good fit to many observation types simultaneously. We derived sunlit and shaded within-canopy LRU profiles, and identified drivers of LRU differences. Based on our results, we propose a new parameterisation of canopy-scale LRU based on absorbed PAR and vapour-pressure deficit of sunlit leaves near the canopy top. We applied the new parameterisation to two needleleaf forest locations.

28 Methodological harmonization of  Twenty Years of Carbon Flux Measurements at the FR-Gri ICOS Agricultural Site

Poster

Carmen Kalalian¹*, Pauline Buysse^1,2, Jérémie Depuydt¹, Anaïs Feron¹, Ahlam Mesmoudi¹, François Brabants¹, Nicolas P. Saby³, Bruna Winck¹, Florent Levavasseur¹, Pedro-Henrique Herig-Coimbra¹, Benjamin Loubet¹

¹UMR EcoSys, Université Paris-Saclay, INRAE, AgroParisTech, Palaiseau, France. ²UMR SAS, INRAE-Institut Agro, Rennes, France. ³UMR Info&Sols, INRAE, Orléans, France

Session

Session 14: Greenhouse gas fluxes in agroecosystems: processes, measurements and management implications

Abstract text

Long-term eddy covariance datasets are essential for quantifying agricultural carbon budgets and assessing soil carbon dynamics under changing climate and management conditions. However, methodological heterogeneity over time may introduce structural inconsistencies that affect trend detection and cumulative carbon balance estimates. 

We present a comprehensive reprocessing of 20 years (2005–2025) of eddy covariance measurements at the ICOS agricultural site FR-Gri (Grignon, France), located 40 km west of Paris. The entire dataset was processed using a fully harmonised workflow, spanning high-frequency raw data processing in EddyPro (including spectral corrections, time lag optimisation, quality control and filtering) through to consistent gap-filling and flux partitioning procedures. A unified methodological framework was applied across analyser generations and instrumental configurations to minimise artefacts arising from methodological evolution over time.

We investigate how harmonisation choices influence annual and multi-annual estimates of Net Ecosystem Exchange (NEE), Gross Primary Production (GPP), and ecosystem respiration (Reco). Particular attention is given to the sensitivity of inter-annual variability, long-term trends, and associated uncertainties to processing choices. The long-term carbon balance is then computed by including side-carbon flows and compared against soil carbon model simulations.

This work aims to improve the robustness and comparability of long-term carbon flux records from agricultural ICOS sites. By providing a consistent reanalysis of two decades of measurements, it strengthens the interpretation of agricultural carbon balances under evolving climate and management practices.

29 Seven-year continuous measurement of CO₂ and CH₄ at Atlas Mohammed V atmospheric research station in Northwest Africa

Poster

Ibrahim OUCHEN¹*, Michel RAMONET², Toure Dro TIEMOKO³, Morgan LOPEZ², Mohamed MASTERE^1,4, Abdelaziz MOTIAA⁵, Wahid Mellouki⁵

¹Department of Geomorphology and Geomatics, Scientific Institute, Mohammed V University in Rabat, Rabat 10106, Morocco, RABAT, Morocco. ²Laboratoire des Sciences du Climat et de l'Environnement (LSCE), IPSL, CEA-CNRS-UVSQ, Université Paris-Saclay, Gif-sur-Yvette, France, PARIS, France. ³Equipe de la Physique pour l'Environnement, Laboratoire de Physique Fondamentale et Appliquée, Université NANGUI ABROGOUA, 02 B.P. 801 Abidjan 02, Abidjan, Côte d'Ivoire., ABIDJAN, Côte d'Ivoire. ⁴School of Public Management, Governance and Public Policy, College of Business & Economics, University of Johannesburg, Johannesburg, South Africa. ⁵College of Chemical Sciences and Engineering (CCSE), African Research Center on Air Quality and Climate (ARC Air), Mohammed VI Polytechnic University (UM6P), Benguerir 43150, Morocco, Benguerir, Morocco

Session

Session 7: CO2 and CH4 cycle dynamics in Africa: observations, processes, and climate implications

Abstract text

Understanding GHG dynamics and improving carbon cycle assessments, particularly in under-sampled regions such as North Africa, requires continuous atmospheric measurements of carbon dioxide (CO₂) and methane (CH₄). The study reports on the first long-term in situ audits of CO₂ and CH₄ carried out by the Atlas Mohammed V (AMV) atmospheric research station, located at an altitude of 2,076 m in the Middle Atlas Mountains in Morocco (33.40° N, 5.10° W) between November 2018 and December 2024. The station, established as a member of the ICOS-France-Atmosphere network, offers an important data source for validating the satellite observations and defining regional carbon budgets. 

The annual growth rate of CO₂ and CH₄ where approximately 2.84 ± 0.36 ppm·year⁻¹ and 12.81 ± 2.95 ppb·year⁻¹, respectively, which corresponded to the regional trends. The use of HYSPLIT model for the analysis of back-trajectory clusters confirmed a significant seasonal variation in the origin of air masses. The seasonal ratios of ΔCH₄/ΔCO₂ were ranged from 12.69 ppb·ppm⁻¹ in winter (with r=0.74, indicating biomass/biofuel combustion), to 24.89 ppb·ppm⁻¹ in spring (r=0.36) 24.89 ppb·ppm⁻¹, and in autumn (r=0.52) by 22.40 ppb·ppm⁻¹. The seasonal cycle amplitudes of CO₂ and CH₄ were 6.1 ppm and 30 ppb, respectively, compared to those at ASK (8.3 ppm; 17 ppb) and IZO (8.1 ppm; 33 ppb). These results underscore the necessity of persistent ground-based monitoring in Africa for lessening the uncertainties related to global carbon budget estimations, and better understanding of the regional emission patterns and atmospheric transport processes.

30 Ud’A Trabocchi: a new ICOS-compliance atmospheric station in central Adriatic coast, Italy.

Poster

Piero Di Carlo*, Eleonora Aruffo, Alessandra Mascitelli, Piero Chiacchiaretta

University 'G. d'Annunzio' of Chieti-Pescara, Chieti, Italy

Session

Session 6: Carbon cycle in the Mediterranean region: from the local to the regional scale

Abstract text

The University ‘G. d’Annunzio’ of Chieti Pescara Atmospheric Observatory (Ud’A Trabocchi) aims to contribute to the ICOS network greenhouse gases measurements. The facility is in Central Italy on the Adriatic coast at 300m from the sea and, from model analysis (Villalobos, et al., 2025), is promising to cover the gap of observations between the ICOS stations Mt. Cimone and Tito Scalo, and should be representative of the GHG concentrations for Central Italy and Balkans. The measurement started in middle January of 2026. The station has a 15 m tall tower with seven sample points every 2 m. The instruments installed includes 3 Picarros for measurements of CO₂, CH₄, N₂O and isotopic ratio of CH₄ and CO₂; Meteorological station (Vaisala); Celiometer (Vaisala CL61) and Radon measurements (Mi.lan). The station includes also other measurements system such us: O₃, NO, NO₂, NO_x, CO and SO₂ (Thermo); NO₂ (CAPS, Teledyne); CO₂ flux (LI-COR); Aerosol concentration and size distribution (SMPS, TSI); column observation of O₃, NO₂, SO₂, HCHO (PANDORA 2S), Aerosol optical depth (CIMEL); GNSS reciver (Geoguard and Stonex); Meteorological station (Vaisala); Celiometer (Vaisala CL61). Finally, the facility includes a UAV system equipped with meteorological sensors and low-cost sensor to detect CO₂, CH₄, O₃ and NO_x.

Data of the first period of observation will be shown and discussed.

Reference

Villalobos, Y., et al., Towards improving top-down national CO2 estimation in Europe: potential from expanding the ICOS atmospheric network in Italy, ENVIRONMENTAL RESEARCH LETTERS, 20, 5, DOI10.1088/1748-9326/adc41e, 2025.

31 Large and increasing biospheric productivity of northern ecosystems

Oral

Matthias Cuntz¹*, Benjamin Smith^2,3, Josep G. Canadell⁴, Jürgen Knauer^2,4,5, Vanessa Haverd⁴

¹Université de Lorraine, AgroParisTech, INRAE, UMR Silva, Nancy, France. ²Hawkesbury Institute for the Environment, Western Sydney University, Penrith, NSW, Australia. ³Department of Physical Geography and Ecosystem Science, Lund University, Lund, Sweden. ⁴CSIRO Environment, Canberra, ACT, Australia. ⁵School of Life Sciences, Faculty of Science, University of Technology Sydney, Ultimo, NSW, Australia

Session

Session 19: Understanding feedbacks between greenhouse gas exchange processes and climate variability using in situ observations, remote sensing, and machine learning

Abstract text

Plants take up carbon dioxide (CO₂) through photosynthesis. How this will change with rising CO₂ concentrations in the atmosphere will strongly determine future climate change. Yet, this is a process critically unconstrained at the global scale. An increase in the seasonal variations of atmospheric CO₂ in recent decades indicates a positive trend in photosynthetic carbon uptake and a lengthening of the growing season in northern extra-tropical ecosystems. However, the biospheric characteristics behind these changes have not yet been fully explained.

We combined data‐driven seasonal cycles of plant productivity with carbon sinks across the range predicted by current biospheric process models to explain the seasonal variations of CO₂ at high and low northern latitudes over the past 40 years. We find that increases in seasonal variations can only be explained by a larger gross primary productivity (GPP) of northern ecosystems than most current estimates, equivalent to (51 ± 2) Pg(C) a⁻¹ around 2007, and by an increase of GPP about proportional (1.1 ± 0.3) to the increase in atmospheric CO₂, also larger than most current estimates. Our results highlight the importance of the interplay between vegetation productivity and its seasonal variations, providing an improved constraint to estimate the future behaviour of the terrestrial carbon sink.

32 Relations between european weather regimes and local air masses, greenhouse gases and air quality variability at OPE over the 2012-2025 period

Poster

Sebastien Conil¹*, Laurent Langrene¹, Morgan Lopez², Michel Ramonet², pascal Yiou²

¹Andra, Bure, France. ²LSCE, Gif Sur Yvette, France

Session

Session 19: Understanding feedbacks between greenhouse gas exchange processes and climate variability using in situ observations, remote sensing, and machine learning

Abstract text

OPE is a regional background atmospheric station dedicated to the long-term monitoring of in-situ atmospheric properties including weather parameters, air quality and aerosol properties, ambient air radioactivity and greenhouse gases.  Most of the measurements started between 2011 and 2015 and have been since operated by Andra with the support of other organisations such as Meteo France, Atmo Grand Est, ASNR and research labs such as Institut des Geosciences de environment and LSCE. OPE is a class 1 ICOS atmospheric station and GHG data are available on the carbon portal. Air quality data and aerosol data are available on the EMEP and GAW datacenter, EBAS, while meteorological data is reported on the French meteo.data.gouv.fr data service. We will use HYSPLIT air masses back trajectories as well as daily european weather regimes based on atmospheric circulations decompositions and classifications.

 After a brief presentation of the station set up, we will highlight the results of the time series decomposition of several air quality parameters and of the main GHG concentrations. We will first report on the main trends observed during that 2012-2025 period. Then we will assess the relationships between air masses, local wind regimes and GHG and air quality variability. A focus will be made on the relationships between European weather regimes occurrence and the local atmospheric variability. Intraseasonal variability of GHG and air quality properties will be related to European weather regimes in cold and warm seasons showing contrasting responses. 

33 Can carbon fluxes reveal ecosystem destabilization risk?

Oral

Filipe Manuel Andrade de Matos^1,2*, Marcos Fernández-Martinez¹, Roberto Molowny Horas¹

¹CREAF, Barcelona, Spain. ²UAB, Barcelona, Spain

Session

Session 22: Remote sensing and flux observations to link disturbance, ecosystem states, and carbon-water fluxes

Abstract text

Ecosystems face increasing risk of destabilization from global change, yet mechanisms driving increases in instability remain poorly understood. Eddy covariance towers provide critical high-frequency carbon flux measurements. We analyzed eddy-covariance fluxes (GPP, NEP, Re) from 65 FLUXNET sites spanning forests, shrublands, and grasslands globally to detect early warning signals using critical slowing down indicators (lag-1 autocorrelation, interannual variability) and flickering metrics (skewness, kurtosis). We examined climate variability, biodiversity, mycorrhizal associations, and vegetation type as potential drivers. Results indicate no widespread transition toward instability, but a subset of sites exhibits increasing destabilization risk linked to climate variability and climate extremes (droughts, heatwaves, fires). These findings demonstrate that high-frequency C data provide an essential baseline for detecting destabilization signals hidden in annual budgets. Integrating continuous in-situ flux measurements with hyperspectral, LiDAR/ALS, and SAR observations could enable spatially explicit predictions of carbon fluxes and evapotranspiration, improve disturbance attribution, and critically, quantify the role of biodiversity in buffering ecosystem instability. We highlight the urgent need for integrated observational frameworks that leverage eddy covariance infrastructure with satellite data to assess how biodiversity conservation can enhance ecosystem resilience and mitigate carbon-climate feedbacks under global change.

34 Continuous fossil fuel (ff) CO₂ fluxes based on ¹⁴C-calibrated proxy/ffCO₂ flux ratios and eddy-covariance flux measurements in Munich

Poster

Ann-Kristin Kunz^1,2*, Samuel Hammer³, Patrick Aigner⁴, Lars Borchardt⁵, Jia Chen⁴, Julian Della Coletta³, Markus Eritt⁵, Xochilt Gutiérrez⁵, Rainer Hilland^1,6, Christopher Holst⁷, Armin Jordan⁵, Natascha Kljun⁸, Richard Kneißl⁵, Daniel Kühbacher⁴, Changxing Lan⁷, Junwei Li⁴, Betty Molinier⁸, Susanne Preunkert³, Andreas Christen¹

¹University of Freiburg, Freiburg, Germany. ²Heidelberg University, Heidelberg, Germany. ³ICOS Central Radiocarbon Laboratory, Heidelberg University, Heidelberg, Germany. ⁴Technical University of Munich, Munich, Germany. ⁵Max Planck Institute for Biogeochemistry, Jena, Germany. ⁶TNO, Environmental Modelling, Sensing, and Analysis, Petten, Netherlands. ⁷Karlsruhe Institute of Technology, Garmisch-Partenkirchen, Germany. ⁸Lund University, Lund, Sweden

Session

Session 17: Urban greenhouse gas emissions and sinks: from advanced monitoring to source identification and impact

Abstract text

For an independent validation of urban fossil fuel (ff) CO₂ emission inventories, atmospheric measurements of CO and NO_x show great potential. Both gases are co-emitted with ffCO₂ during incomplete combustion and, unlike ffCO₂, can be measured directly and continuously using eddy covariance (EC). However, proxy-based estimation of continuous ffCO₂ fluxes requires knowledge of the local proxy/ffCO₂ flux ratios, which vary depending on the contributions of different source sectors and their respective emission ratios. The lack of direct, temporally-resolved observations and limited knowledge of the source sector contributions and their emission ratios are major sources of uncertainty.

To overcome this limitation, we combine continuous EC CO and NO_x fluxes with co-located, discrete ffCO₂ fluxes from ¹⁴CO₂ relaxed eddy accumulation measurements at a tall tower in Munich. The proxy/ffCO₂ flux ratios observed over 75 individual hours vary considerably. CO/ffCO₂ flux ratios show a dependence on mean air temperatures, consistent with an expected higher contribution of stationary combustion in winter. Due to the observed variability, the uncertainty of proxy-based ffCO₂ flux estimates is about 50 % larger than direct ¹⁴C-based estimates. Nevertheless, the unique data set enables the calculation of purely observation-based continuous ffCO₂ time series, which we compare with modeled estimates obtained from convolving emission inventories with flux footprints. In summer, NO_x-, CO-, and inventory-based ffCO₂ fluxes agree well, whereas in winter, the modeled fluxes are higher. We discuss similarities and differences between observed and modeled fluxes and flux ratios, and the implications for proxy-based partitioning of net CO₂ fluxes in cities.

35 New insight on the link between spatial and temporal dynamic of water flow and Carbon cycle in wetlands: a GIS-based approach

Poster

Nicolas DEVAU^1,2*, Mathias MAILLOT³, Damien RAMBOURG³, Zlatko NEDIC⁴, Danijela ZUNIC⁵, Natasa PRODANOVIC⁵, Laurent ANDRE^1,6

¹BRGM, ORLEANS, France. ²ISTO, UMR7327, Université d'Orléans, CNRS, BRGM, OSUC, Orléans, France. ³BRGM, Orléans, France. ⁴Competence Centre Ltd. for research and development, Vinkovci, Croatia. ⁵PSSS, Sombor, Serbia. ⁶ISTO, Orléans, France

Session

Session 13: Advancing approaches for quantifying greenhouse gas fluxes in terrestrial ecosystems

Abstract text

The carbon cycle in wetlands involves continuous carbon fluxes between the soil, water, and vegetation. However, water and carbon dynamics in wetlands remain poorly understood because monitoring networks are often absent or insufficient at many sites. This lack of data prevents the development of management strategies aimed at promoting carbon sequestration in wetlands. The objective of this work is to overcome this limitation by developing and validating an innovative framework to assess water dynamics and the carbon cycle in wetlands.

Regarding water dynamics, a gridded model simulating spatio-temporal variations in groundwater table depth was developed based on geostatistical approaches integrating topsoil elevation and surface water dynamics. For the carbon cycle, soil and biomass carbon pools were estimated from satellite data available at the European scale. Plant growth and diversity were explicitly incorporated into the estimation of biomass pools. Carbon fluxes from soil and biomass were assessed using proxy-based approaches. A surrogate model allowed linking groundwater table depth to aerobic respiration. Net primary production was used to estimate carbon fluxes from the atmosphere to biomass. For both fluxes, geostatistical models were applied to simulate their spatio-temporal patterns. 

The methodology was successfully applied to demonstration wetlands of the Danube Basin, and restoration scenarios are proposed to enhance their carbon sequestration capacity.

This work is a part of the Danube Wetlands and flood plains Restoration through systemic, community engaged and sustainable innovative actions (DaWetRest) project, which is co-funded by the European Union Horizon Europe programme (HORIZON-MISS-2022-OCEAN-O1-02), under Grant Agreement no. 101113015.

37 Assessing the ability of LPJ-GUESS-HYD to predict water stress responses in boreal forests using ICOS eddy-covariance flux data

Poster

Rose Brinkhoff*, Filipe Gomes de Almeida, Thomas Pugh, Cecilia Akselsson, Natascha Kljun

Lund University, Department of Earth and Environmental Sciences, Lund, Sweden

Session

Session 11: Climate feedbacks from ecosystem management and natural disturbances

Abstract text

Boreal forests are increasingly exposed to extreme heat and altered precipitation patterns, leading to periods of water stress that threaten their capacity to provide important ecosystem services. Management interventions can improve the resilience of forests to water stress, but our ability to implement such adaptation methods is contingent upon accurate identification of areas most susceptible to the adverse effects of this water stress. Recent advances in dynamic vegetation modelling have improved our ability to predict water stress responses in forests, including the integration of plant hydraulic processes into the ecosystem model LPJ-GUESS. Here, we evaluate the ability of this new adaptation, LPJ-GUESS-HYD, to detect water stress in three forests across Sweden. We identified periods of moderate, severe and extreme drought based on the Standardized Precipitation-Evapotranspiration Index (SPEI), and compared LPJ-GUESS-HYD carbon flux simulations with ICOS eddy-covariance flux data and satellite-based vegetation indices in drought and non-drought periods from 2015 to 2022. We found that LPJ-GUESS-HYD could accurately capture many water-stress-induced shifts in carbon fluxes and vegetation indices. However, its ability to detect these water stress responses varied largely between sites and years, and depended on the duration and intensity of the water stress. Our results provide insight into the factors determining the efficacy of LPJ-GUESS-HYD for predicting water stress responses, and highlight areas where improvement is needed. 

38 Constraining methane emissions in the EMME region using TROPOMI satellite data in an atmospheric inversion

Oral

Martin Vojta¹*, Rakesh Subramanian¹, Nikos Gialesakis², Rona L. Thompson³, Maria Kanakidou², Andreas Stohl¹

¹Department of Meteorology and Geophysics, University of Vienna, Vienna, Austria. ²Department of Chemistry, University of Crete, Crete, Greece. ³NILU, Kjeller, Norway

Session

Session 28: Combining data and models to support emissions estimation and policy at local to regional scales

Abstract text

Methane (CH₄) is the second most abundant anthropogenic greenhouse gas after CO₂. Anthropogenic sources account for roughly two-thirds of global CH₄ emissions, and approximately one-third of these anthropogenic emissions originate from fossil fuel production. The Eastern Mediterranean and Middle East (EMME) region is of particular importance, as it hosts a large share of global oil and gas extraction and therefore offers significant mitigation potential. However, methane emission estimates from bottom-up methods in this region remain highly uncertain due to the large number of emission sources.  The lack of atmospheric measurement stations in this area also challenges top-down approaches, however satellite observations offer an alternative constraint.

We use TROPOMI satellite observations together with the Lagrangian particle dispersion model (LPDM) FLEXPART and the inversion framework FLEXINVERT+ to estimate CH₄ emissions in the EMME region for the year 2020. Through an extensive set of sensitivity experiments, we investigate key challenges and parameter sensitivities when using satellite data in atmospheric inversions and identify the optimization of initial mole fractions as a dominant factor influencing inversion results. 

Our posterior emissions are significantly lower around the Persian Gulf compared to the prior estimate, based on the EDGAR v8.0 anthropogenic emissions inventory. Notably, similar reductions are reflected in the very recently released EDGAR_2025_GHG dataset, where updated activity statistics were used. This provides an interesting example in which independent refinements from a top-down approach converge with updates from a bottom-up inventory, illustrating how observational constraints and inventory development can jointly improve our understanding of regional emissions.

39 Stem CO₂ efflux and its relation to stem temperature and stem radial growth in three deciduous tree species over 4 years.

Poster

Manuel Acosta*, Eva Dařenová, Marko Stojanovic, Jan Krejza, Lukáš Kokrda, Marian Pavelka

Global Change Research Institute, CAS, Brno, Czech Republic

Session

Session 13: Advancing approaches for quantifying greenhouse gas fluxes in terrestrial ecosystems

Abstract text

It is well known that temperature is the most significant factor influencing stem CO2 efflux (S_CO2 ). However, some studies indicated that S_CO2 is also influenced by the dynamics of the stem radial growth. We measured the S_CO2 on three tree species (English Oak, Narrow-leaved Ash and Hornbeam) by an automated stem chamber system over 4 years to investigate the S_CO2 seasonal variation, the response of the S_CO2 to stem temperature (Q₁₀) by individual tree species and the relationship between S_CO2 and the stem radial growth for all studied tree species. Our results showed that the highest S_CO2 values were obtained in Ash trees, followed by Oak and Hornbeam trees, respectively. All investigated tree species showed different responses of S_CO2 to stem temperature over the investigated years (Q₁₀ ranged between 1.88 and 2.66). The stem radial growth was found to be a factor influencing the S_CO2, but only depending on the tree species and the period of the season. The strongest relationship between S_CO2 and stem radial growth was determined in the Hornbeam trees (R² = 0.59) in summer during the year 2020, while Oak and Ash showed a positive but weak relationship mainly during the period of stem growth. On the other hand, our results showed that stem radial growth is a complex process that, to some extent, influences stem CO² efflux, as it involves changes in ultrastructure and cell biochemistry that cause high metabolic activity, leading to higher rates of S_CO2.

40 Seeing through the canopy: COS and SIF constrained SCOPE modeling to investigate plant drought responses

Poster

Anna de Vries¹*, Felix M. Spielmann¹, Alexander Platter¹, Albin Hammerle¹, Egor Prikaziuk², Christiaan van der Tol², Georg Wohlfahrt¹

¹Department of Ecology, University of Innsbruck, Innsbruck, Austria. ²Faculty ITC, University of Twente, Enschede, Netherlands

Session

Session 16: Using sun-induced chlorophyll fluorescence to understand or scale EC fluxes

Abstract text

Understanding plant biophysical and biochemical responses to drought is critical for predicting ecosystem carbon–water exchange and gross primary productivity (GPP) under a changing climate. We investigate both biophysical (e.g. stomatal and non-stomatal limitations) and biochemical (e.g. photosynthetic efficiency and non-photochemical quenching) responses to natural drought events by integrating eddy covariance flux measurements of carbonyl sulfide (COS) and solar-induced fluorescence (SIF) as novel observational constraints into the Soil–Canopy Observation of Photosynthesis and Energy fluxes (SCOPE) model.

Because COS is taken up by plants similarly to CO₂ through shared diffusion pathways but is not re-emitted, COS flux measurements provide a robust proxy for GPP and stomatal conductance. While SIF offers insight into the light partitioning reactions at photosynthetic level. SCOPE is a physically based land-surface model that couples plant physiological processes with spectrally resolved radiative transfer within the canopy; incorporating COS and SIF introduces independent constraints on canopy conductance and carbon assimilation.

We leverage multiple years of concurrent COS and CO₂ flux measurements, together with fluorescence and hyperspectral reflectance observations, from a Scots pine–dominated montane forest in Austria that experienced naturally occurring drought periods. This integrated model–data framework has the potential to refine key ecosystem-scale processes, including stomatal regulation and water-stress responses. Eventually with the aim of enhancing our ability to detect and predict drought impacts on canopy function while tightening constraints on ecosystem carbon–water dynamics.

41 Exchange of CO₂, CH₄ and N₂O in a Sahelian semi-arid savanna – flux magnitudes and drivers

Poster

Aleksander Wieckowski¹*, Patrik Vestin¹, Jonas Ardo¹, Ousmane Diatta², Ousmane Ndiaye³, Michal Heliasz⁴, Tobias Biermann⁴, Peter Kornacher¹, Olivier Roupsard⁵, Torbern Tagesson¹

¹Lund University, Lund, Sweden. ²ISRA, Dakar, Senegal. ³ISRA, Dahra, Senegal. ⁴ICOS, Lund, Sweden. ⁵CIRAD, Dakar, Senegal

Session

Session 7: CO2 and CH4 cycle dynamics in Africa: observations, processes, and climate implications

Abstract text

The Sahel is one of the largest semi-arid regions in the world, serving as a biogeographic transition zone between the arid Sahara Desert and the humid Soudanian savannah. The region is experiencing rapid population growth, leading to increased demand for food and fuelwood. This has resulted in intensified land use, greater grazing pressure, and widespread conversion of natural vegetation to agricultural land. These land-use changes are altering the carbon and nitrogen cycles of Sahelian ecosystems, with potential implications for greenhouse gas (GHG) emissions. Despite this, there remains a lack of long-term, ecosystem-scale measurements of land-atmosphere GHG exchange across the African continent, making it difficult to assess the rates and drivers of change in GHG fluxes.

To address this gap, we present the first ecosystem-scale observations of CO_­2, CH₄ and N₂O fluxes in Africa using the flux-gradient method. A gradient measurement system was deployed on the eddy covariance tower at the Dahra site in Senegal, where continuous CO₂ flux measurements have been ongoing and will be used to validate the new observations. The system will operate continuously for one year, allowing us to assess seasonal variations in GHG fluxes, establish a greenhouse gas budget for the site, and identify key environmental drivers. This study provides a critical step toward understanding and quantifying GHG dynamics in one of the most vulnerable ecosystems in the world.

42 Estimating CO₂ emissions in the city of Zurich using building-resolved simulations and a Bayesian framework

Poster

Leonie Bernet*, Lionel Constantin, Lukas Emmenegger, Christoph Hüglin, Pascal Rubli, Dominik Brunner

Empa, Laboratory for Air Pollution / Environmental Technology, Dübendorf, Switzerland

Session

Session 17: Urban greenhouse gas emissions and sinks: from advanced monitoring to source identification and impact

Abstract text

Global greenhouse gas emissions originate largely from urban areas. Accordingly, cities such as Zurich, Switzerland, monitor their mitigation progress through bottom-up emission inventories. However, these inventories often carry large uncertainties. To address this, we use observation-based top-down approaches that combine atmospheric measurements with high-resolution models in an inversion framework. 

In Zurich, a city-wide CO₂ monitoring network was established in July 2022 as part of the European "ICOS Cities" project. The network consists of around 80 mid- or low-cost CO₂ sensors, providing observations for almost 4 years. In parallel, we simulate hourly CO₂ concentration fields using the building-resolved GRAMM-GRAL model system (Graz mesoscale and Lagrangian Model) at 10 m resolution. Due to its catalogue approach, the model can simulate urban emissions and dynamics for multiple emission sectors at low computational cost. We include 20 anthropogenic and biospheric emission sectors such as traffic, residential heating, industrial activities, human respiration or vegetation. 

By combining the observations with the simulated CO₂ fields using a Bayesian inversion, we trace the measured concentrations back to their emission sources. In the inversion process, we found best performances when using afternoon values only and when optimizing emissions for several years at once. We integrate mid- and low-cost sensor data into the system, assess their performance for estimating city-wide emissions and quantify discrepancies with the bottom-up inventories. By offering an independent top-down evaluation, the inversion framework helps verify and potentially improve the city's emission inventory, thereby supporting the pathway towards achieving net-zero emissions. 

43 INETFLUX - Innovative technologies cross scales to disentangle carbon dioxide and evapotranspiration fluxes of forests

Poster

Marian Pavelka¹*, Nina Buchmann², Roman Zweifel³, Lucie Homolová¹, Marko Stojanović¹, Jan Krejza¹, Štépánka Řehořková¹, Manuel Acosta¹

¹Global Change Research Institute, CAS, Brno, Czech Republic. ²ETH, Zurich, Switzerland. ³WSL, Birmensdorf, Switzerland

Session

Session 13: Advancing approaches for quantifying greenhouse gas fluxes in terrestrial ecosystems

Abstract text

A bilateral project in the framework of the Swiss-Czech Cooperation Programme was awarded to two ICOS national networks: ICOS Czech Republic and ICOS Switzerland. The INETFLUX project brings together one Czech institution (Global Change Research Institute CAS / GCRI  - CzechGlobe) and two Swiss research institutions (Eidgenössische Technische Hochschule Zürich / ETH Zurich, Eidgenössische Forschungsanstalt für Wald, Schnee und Landschaft / WSL) which all are part of ICOS. The aims of the project are to advance the understanding of forest net ecosystem exchange of CO₂ (NEE) and evapotranspiration (ET), specifically, to develop and validate new approaches for partitioning NEE and ET into their individual components and to identify their environmental drivers, also in the context of climate change. Stem growth measurements, stable isotope applications and remote sensing techniques will complement the flux measurements and provide additional constraints on the partitioned fluxes. The 3-year project boosts excellent scientific research and enhances bilateral cooperation between the two countries using the ICOS infrastructure. The research will take place at three ICOS stations: a subalpine spruce forest in Davos (Class 1 Ecosystem Station – CH-Dav) in Switzerland, a lower elevation spruce monoculture at Bílý Křiž (Class 2 Ecosystem Station – CZ- BK1) and a floodplain hardwood forest in Lanžhot (Class 1 Ecosystem Station – CZ- Lnz), both in the Czech Republic.

44 Does Forest Management Matter? Long‑Term Eddy‑Covariance Evidence from Continuous Cover and Rotation Systems in Sweden

Poster

Patrik Vestin*, Giulia Zudettich, Jutta Holst, Tobias Biermann

Lund University, Lund, Sweden

Session

Session 11: Climate feedbacks from ecosystem management and natural disturbances

Abstract text

Forests play a central role in the global carbon cycle and are key to climate change mitigation. Although forest ecosystems and management practices have been extensively studied, surprisingly little is known about how to manage forests sustainably while balancing environmental and societal demands. Scandinavian forests currently form a substantial, though seemingly declining, carbon sink. The dominant silvicultural system, rotation forestry (RF), is associated with large greenhouse gas (GHG) emissions following clear cutting. Continuous cover forestry (CCF) offers an alternative based on selective harvesting and natural regeneration, reducing ecosystem disturbance and potentially strengthening long term carbon sequestration. However, major knowledge gaps remain regarding the sink strength, GHG dynamics and climate sensitivity of CCF forests, creating substantial uncertainty in assessing their mitigation potential and determining optimal management strategies.

We present long term carbon flux measurements from the Rumperöd CCF forest in southern Sweden. Preliminary results for 2015–2023 show pronounced interannual variability in carbon exchange and a strong impact of the 2018 drought on ecosystem scale fluxes (Figure 1). The dataset will be extended to cover 2013–2025 and compared with observations from the nearby ICOS Hyltemossa RF forest and from a young RF stand in the close vicinity for the period 2015–2025, enabling an evaluation of carbon dynamics under contrasting management regimes.

45 Resolving In-Canopy Controls on VOC and Ozone Exchanges: a Multi-Layer Canopy Model Intercomparison Constrained by Above- and Below- Canopy Flux Observations

Poster

Bert Verreyken^1,2*, Laurens Ganzeveld³, Clément Dumont¹, Crist Amelynck^2,4, Bernard Heinesch¹

¹Gembloux Agro-Bio Tech, University of Liège, Gembloux, Belgium. ²Royal Belgian Institute of Space Aeronomy (BIRA-IASB)), Uccle, Belgium. ³Wageningen University, Wageningen, Netherlands. ⁴Department of Chemistry, Ghent University, Ghent, Belgium

Session

Session 24: Exchange of reactive gases and aerosols between the land surface and the atmosphere in natural and managed ecosystems

Abstract text

Forests are the largest global source of non-methane volatile organic compounds (NMVOCs) in the atmosphere. Recent research highlights the need for multi-layer canopies representation to simulate  ozone (O₃) concentrations and deposition in global/regional chemistry models. However, biophysicochemical processes along the soil—canopy—atmosphere continuum and their impact on ecosystem-level VOC exchanges remain uncertain. This is due to the system’s complexity and limited availability of real-world data.

To address this, we conducted three long-term campaigns (April to December) in 2022, 2023, and 2024 to measure VOC/O₃ concentrations at the ICOS Class 2 ecosystem site in Vielsalm (Belgium). Measurements were performed sequentially above the canopy, in the trunk space, and along the vertical gradient of the site’s 52 m flux tower. Furthermore, eddy covariance fluxes of VOCs and O₃ were calculated at the top of the tower (2022, 2023, and 2024) and in the trunk space (2023, 2024). 

We will present the application of 1-D multi-layer canopy exchange models to investigate specific mechanisms (biogenic emissions, stomatal and non-stomatal deposition, turbulent transport, and chemistry) and their control on bidirectional ecosystem-atmosphere exchange of VOCs and O₃ deposition. For this, we employ two 1-D multi-layer canopy exchange models (FORCAsTv2.0 and MLC-CHEM). While FORCAsTv2.0 uses a prognostic approach, MLC-CHEM is more observation-constrained in calculating exchanges of reactive trace gases. Complementary trunk-space and above-canopy fluxes, together with long-term vertical concentration profiles, provide a unique dataset to evaluate how canopy processes influence atmosphere–biosphere exchanges and improve process-level understanding beyond what is captured in large-scale chemistry models.

46 From Eddy Covariance to Regional Mapping: Physics-Informed Machine Learning for Peat Oxidation in Dutch Peatlands

Poster

Laurent Bataille¹*, Bart Kruijt¹, Wietse Franssen¹, Hong Zhao¹, Wilma Jans¹, Corine van Huissteden¹, Ignacio Andueza Kovacevic¹, Freek Engel¹, Ruchita Ingle¹, Tan Lippman¹, Jeferson Zerrudo¹, Jan Biermann¹, Hanne Berghuis¹, Laura van der Poel^2,1, Isabel Cabezas Dueñas¹, Niek Bosma³, Reinder Reinder³, Wiebe Borren⁴, Veronique Boon⁵, Ype van der Velde⁶, Ronald Hutjes¹

¹Wageningen University and Research, Earth Systems & Global Change, Wageningen, Netherlands. ²University of Copenhagen, Department of Geosciences and Natural Resource Management, Copenhagen, Denmark. ³Wetterskip Fryslân, Leeuwarden, Netherlands. ⁴Natuurmonumenten, Amersfoort, Netherlands. ⁵Free University , FacultyFaculty of Science, Earth and Climate, Amsterdam, Netherlands. ⁶Free University , Faculty of Science, Earth and Climate, Amsterdam, Netherlands

Session

Session 15: Peatlands in the global climate system: fluxes, feedbacks, and human pressures

Abstract text

Peat soil degradation in the Netherlands contributes an estimated 4.6–7 Mt CO₂ annually, accounting for approximately 3% of national greenhouse gas emissions. Under the Dutch Climate Agreement (2019), a mitigation target was set to reduce net CO₂ emissions from drained fen meadows by 1 Mt CO₂ yr⁻¹ by 2030. Achieving this objective requires robust attribution of peat-derived CO₂ emissions and reliable upscaling from site-level measurements to regional inventories.

Within the Dutch National Research Programme on Greenhouse Gases in Peatlands (NOBV), an extensive monitoring network of on-site and airborne eddy covariance (EC) observations across different peat types, drainage regimes, and grassland management practices. Direct measurement of peat oxidation at the ecosystem scale is not feasible using EC alone. Instead, peat-derived emissions are inferred from ecosystem-scale fluxes that integrate autotrophic respiration, heterotrophic decomposition, and vegetation turnover from mowing and regrowth.

To address this challenge, we develop a physically inspired machine learning that embeds hydrological and biogeochemical constraints within a data-driven architecture. Groundwater level dynamics, air-filled porosity, and soil physical properties are included as structured drivers to constrain model behaviour and limit empirical extrapolation. This hybrid strategy aims to enhance interpretability while maintaining the predictive flexibility of deep neural networks.

The trained models are designed for spatial upscaling by coupling flux-derived emission–response relationships with spatial datasets on drainage, peat properties, land use, and water management. By integrating ecosystem-scale observations, soil physics, and remote sensing, the framework links local EC measurements to regional carbon budgets.

47 First measurements of air‐sea CO2 fluxes in the central Mediterranean during the intense 2022–2023 marine heatwave

Oral

Mattia Pecci^1,2, Damiano Sferlazzo², Fabrizio Anello³, Silvia Becagli⁴, Simone Colella⁵, Lorenzo De Silvestri⁶, Tatiana Di Iorio⁶, Antonio Iaccarino⁶, Daniela Meloni⁶, Francesco Monteleone³, Salvatore Piacentino³, Elena Principato², Alcide di Sarra⁷*

¹INGV, Rome, Italy. ²ENEA, Lampedusa, Italy. ³ENEA, Palermo, Italy. ⁴University of Florence, Florence, Italy. ⁵CNR, Rome, Italy. ⁶ENEA, Rome, Italy. ⁷ENEA, Frascati, Italy

Session

Session 6: Carbon cycle in the Mediterranean region: from the local to the regional scale

Abstract text

Air-sea CO2 fluxes have beeen determined by combining continuous observations made at the Lampedusa Oceanoraphic Observatory (35.49°N, 12.47°E) and Atmospherc Observatory (35.52°N, 12.63°E), both contributing to ICOS. In this study, measurements obtained during the period December 2021 - June 2023 have been analysed.  This period includes an intense and prolonged marine heat wave, which developed throughout the period May 2022-April 2023. The data show that the Central Mediterranean on the annual scale currently acts as a CO2 sink, with an absorption phase in winter and an emission phase during summer. However, large differences exist between the two consecutive absorption periods included in the data set, with a 30% lower absorption during early 2023 compared to early 2022. The smaller CO2 absoprtion during the heat wave period appears to be linked to a strong reduction in intense wind episodes, which smaller air-sea exchange and vertical mixing in the ocean.

48 Improving estimates of the carbon flux – water loss relationship

Poster

Holger Lange¹*, Morgane Merlin¹, Junbin Zhao¹, Ryan Bright¹, Danielle Creek²

¹NIBIO, Ås, Norway. ²NMBU, Ås, Norway

Session

Session 13: Advancing approaches for quantifying greenhouse gas fluxes in terrestrial ecosystems

Abstract text

Carbon fluxes (GPP and NEE) are controlled by carbon-water coupling, as quantified e.g., by the underlying Water Use Efficiency (uWUE). This stand-level dynamic property involves VPD and evapotranspiration (ET), besides GPP. It is crucial to obtain reliable estimates for ET within the footprint of EC towers to understand the effect of increasing CO₂ concentrations on carbon storage and water losses. This is particularly relevant under extreme climatic conditions, such as prolonged drought episodes.

Traditionally, sapflow sensors are used for measuring the water demands for transpiration at the single-tree level. However, sapflow devices are not mandatory equipment at ICOS sites and expensive, limiting their deployment to a few trees in a stand, and they are not accessing physiological responses (radial growth, water storage) associated with stress (drought). Dendrometers, now required for ICOS sites, monitor stem size fluctuations from both radial growth and hydraulic dynamics of plant tissue. They have great potential to be more extensively deployed in forests.

We analyze the relationship between data from co-installed point dendrometers and sapflow sensors at the site NO-Hur. We link these relationships with tree physical attributes, meteorology and soil climate to assess whether a predictive model for sapflow can be built from measured diameter changes, tree properties and climate, to reduce the uncertainty of stand level tree transpiration estimation. It will help further inform the partitioning of water fluxes (ET) between the tree and understory layer from the EC water fluxes and provide a better estimate of uWUE, the carbon uptake – water loss relationship.

49 Beyond water solubility: physicochemical controls on VOC dry deposition in a mixed temperate forest

Poster

Clément Dumont¹*, Bert Verreyken^1,2, Niels Schoon², Benjamin Loubet³, Crist Amelynck^2,4, Bernard Heinesch¹

¹Gembloux Agro-Bio Tech, University of Liège, Liège, Belgium. ²Belgian Institute for Space Aeronomy, Brussels, Belgium. ³ECOSYS, Université Paris-Saclay, INRAE, AgroParisTech, Palaiseau, France. ⁴Department of Chemistry, Ghent University, Ghent, Belgium

Session

Session 24: Exchange of reactive gases and aerosols between the land surface and the atmosphere in natural and managed ecosystems

Abstract text

Volatile organic compounds (VOCs) strongly influence tropospheric ozone, secondary organic aerosol formation, and methane’s atmospheric lifetime. Although most VOCs are biogenically emitted, dry deposition to ecosystems represents an important but poorly constrained sink, particularly beyond a few highly water-soluble compounds.

We measured VOC concentrations and fluxes above and below the canopy at the mixed temperate ICOS forest site of Vielsalm (Belgium) over three growing seasons (2022–2024) using a PTR-ToF-MS instrument. Minimum deposition velocities were derived from negative net fluxes and interpreted with a two-layer (canopy–soil) big-leaf resistive model.

Significant deposition was observed for 47 VOC groups spanning a wide range of physicochemical properties. Median deposition velocities ranged from 0.4 cm s⁻¹ (formic acid) to 1.5 cm s⁻¹ (C₈H₈O₂H⁺). Below-canopy uptake accounted for ~10% of total deposition, indicating that canopy processes dominate VOC removal.

The widely used Wesely scheme reproduced deposition velocities only for highly water-soluble compounds (e.g. formaldehyde, formic and acetic acids), for which deposition increased with relative humidity and peaked in autumn. For many other VOCs, the model underestimated deposition by up to three orders of magnitude.

Under dry conditions, deposition velocities were positively correlated with the octanol–air partition coefficient, suggesting uptake into lipid phases such as the leaf cuticle. This complementary aqueous and lipid uptake was corroborated by an independent VOC flux dataset over a winter wheat field in the Paris region. Overall, our results call for deposition schemes that account for the deposition of both hydrophilic and hydrophobic VOCs to ecosystems.

50 Simultaneous optimisation of wetland methane processes and fluxes in a coupled data assimilation system

Oral

Rebecca Ward¹*, Aki Tsuruta¹, Anteneh Mengistu¹, Antti Leppänen¹, Tiina Markkanen¹, Antoine Berchet², Rajaran Vertiselvan², Jalisha Kallingal³, Marko Scholze³, Tuula Aalto¹

¹Climate System Research, Finnish Meteorological Institute, Helsinki, Finland. ²Laboratoire des Sciences du Climat et de l’Environnement, CEA-CNRS-UVSQ, Gif-sur-Yvette, France. ³Department of Physical Geography and Ecosystem Science, Lund University, Lund, Sweden

Session

Session 19: Understanding feedbacks between greenhouse gas exchange processes and climate variability using in situ observations, remote sensing, and machine learning

Abstract text

Improving methane (CH₄) emission estimates is critical for refining global and regional CH₄ budgets and informing climate policy. Wetland CH₄ emissions are typically quantified using either bottom-up process models or top-down atmospheric inversions. Process models can be calibrated against flux measurements to reduce parameter uncertainties, while inversions constrain regional-to-global fluxes using atmospheric observations within a Bayesian inversion framework. Parameter optimisation has been shown to improve agreement between process models and inversion results, suggesting that a combined, simultaneous application of these methods could further reduce uncertainties across spatial scales. 

As part of the Investigating Methane for Climate Action (IM4CA) project, this work will develop a coupled CH₄ data assimilation framework that simultaneously constrains wetland process parameters and large-scale fluxes by integrating eddy-covariance (EC) flux measurements and atmospheric CH₄ concentrations. The framework employs the JSBACH-HIMMELI wetland CH₄ model, targeting parameters governing respiration, methane production, oxidation, diffusion and plant transport. The system will be implemented within the Community Inversion Framework (CIF), with a focus on Arctic-boreal wetlands, a region characterised by large, climate-sensitive and uncertain CH₄ emissions.  

By jointly assimilating EC and atmospheric measurements, the framework will provide more robust parameter estimates and uncertainty quantification, enable identification of dominant wetland CH₄ processes, and produce flux estimates and budgets that better capture spatial and temporal variability. In the future, the framework could also be extended to assimilate auxiliary datasets such as satellite observations of inundation and vegetation, further strengthening constraints on wetland CH₄ emissions. 

51 Propagation of meteorological to physiological drought in observations

Poster

Chunhui Zhan*, René Orth

University of Freiburg, Freiburg, Germany

Session

Session 22: Remote sensing and flux observations to link disturbance, ecosystem states, and carbon-water fluxes

Abstract text

A meteorological drought is defined by extended periods of atmospheric water demand and precipitation deficit. It can lead to soil moisture depletion, i.e., hydrological drought. This can further trigger physiological drought, i.e., impaired vegetation functioning, posing risks to ecosystems through hydraulic failure, carbon limitation, and eventual plant mortality. The propagation of meteorological to physiological drought is a complex process, modulated by different vegetation, climatic, and edaphic factors. It is not yet well known at which time scale this propagation mostly occurs, and what the relevant drivers are.

This study aims to identify the time scales with the highest degree of propagation of meteorological to physiological drought. We use data from FLUXNET sites, including the Standardized Precipitation Evapotranspiration Index (SPEI) and eddy covariance data, along with additional databases related to plant traits and structural diversity. By analyzing the correlations between SPEI at various time scales and vegetation’s physiological responses—such as anomalies in gross primary productivity (GPP)—we identify relevant time scales for drought propagation.

Preliminary results from the 2018 European drought indicate that drought propagation occurs more rapidly (~15 days) in relatively dry regions, while it is slower (~30 days) in wet and arid areas where plants typically have adequate water access or are adapted to arid conditions. In the next step, we will analyse more droughts and other regions to enhance our understanding of drought propagation based on observations, aiming to inform the development of more accurate and reliable models.

52 Improved gap-filling of eddy covariance CO2 fluxes using remote sensing and environmental variables via  XGBoost

Poster

Simon De Canniere^1,2*, Thomas Servotte¹, Sebastian Wieneke³, Tim Verdonck¹, Ivan Janssens¹

¹University of Antwerp, Antwerpen, Belgium. ²Univeristy of Liege, Liege, Belgium. ³University of Leipzig, Leipzig, Germany

Session

Session 16: Using sun-induced chlorophyll fluorescence to understand or scale EC fluxes

Abstract text

Gap-filling of eddy covariance (EC) CO2 flux data is critical for quantifying ecosystem carbon balances, yet traditional methods like Marginal Distribution Sampling (MDS) do not adequately represent sub-daily carbon fluxes and it fails to leverage vegetation dynamics, which is especially problematic for filling in gaps longer than one week. This study evaluates the potential of eXtreme Gradient Boosting (XGBoost), a machine learning approach, to improve gap-filling of net ecosystem exchange (NEE) and gross primary production (GPP) by integrating remote sensing (RS) data and environmental data, both from in-situ measurements and from the ERA5 reanalysis model over a temperate pine forest (ICOS site BE-Bra). The results show XGBoost outperforms MDS for NEE gap-filling, with minimal performance degradation for gaps up to 56 days. Soil moisture and SIF improved predictions during warm periods (Air Temperature > 25 C), when these data were taken from in-situ sources. SHAP analysis revealed light-related drivers as dominant controls. During heatwaves, typically co-occurring with high-light conditions, soil water content became an important driver. Overall, the hybrid model achieved comparable model performance as the models with in-situ data, demonstrating the viability of satellite RS and reanalysis for operational gap-filling. However, in-situ irradiation turned out notably more useful compared to irradiation from a reanalysis. Our findings advocate for XGBoost as a robust tool to integrate multi-source data, advancing carbon flux quantification beyond traditional methods, espescially when it comes to modeling the sub-daily carbon fluxes, which is important when using EC data for evaluating remote sensing based carbon flux estimations.

53 Sward composition regulates carbon and Nitrous Oxide Emissions in Intensively Grazed Grasslands

Oral

Morad Mirzaei^1,2*, Rachael Murphy², Stuart Green³, Jesko Zimmermann³, Luis Lopez-Sangil², Karl Richards⁴, Matthew Saunders¹

¹School of Natural Sciences, Botany Discipline, Trinity College Dublin, Dublin, Ireland. ²Teagasc, Environment, Soils and Land Use Department, Johnstown Castle, Wexford, Ireland. ³Department of Agrifood Business and Spatial Analysis, Rural Economy and Development Programme, Teagasc, Ashtown Research Centre, Dublin, Ireland. ⁴Teagasc , Climate Centre, Johnstown Castle, Wexford, Ireland

Session

Session 14: Greenhouse gas fluxes in agroecosystems: processes, measurements and management implications

Abstract text

Grasslands dominate Ireland’s agricultural landscape and hold the country’s second largest stock of soil carbon. Nearly 90% of agricultural nitrous oxide (N₂O) emissions in Ireland arise from grazed grasslands, driven by nitrogen fertiliser and animal excreta deposition. Carbon dynamics in these systems are strongly influenced by grazing management, including stocking rate, timing, and defoliation intensity. However, the combined effects of sward composition and grazing regime, on carbon and N₂O dynamics remain poorly quantified. A two-year field study at the Teagasc Environmental Research Centre evaluated perennial ryegrass (PRG) monocultures and multispecies swards (MSS) under single and overlapping livestock urine deposition and protected urea fertilisation. N₂O fluxes were measured using static chambers, and net ecosystem exchange of CO₂ quantified via eddy covariance. UAV multispectral surveys mapped urine patches and upscaled plot-scale emissions to paddock scale. MSS consistently reduced cumulative N₂O emissions by 27–45% under overlapping urine and 39–71% under single applications relative to PRG, with further reductions of up to 60% under protected urea. These reductions coincided with enhanced plant N uptake (63%), improved NUE ( 38%), and lower yield-scaled emissions without compromising dry matter production. MSS also showed more stable CO₂ fluxes, lower carbon export, and improved soil–plant carbon retention. UAV surveys achieved 99% detection accuracy, revealing spatial heterogeneity and refining paddock-scale N₂O estimates. Overall, multispecies swards, integrated with appropriate grazing management, can regulate carbon and nitrogen fluxes, reduce N₂O emissions, and enhance ecosystem efficiency in intensively managed Irish grasslands.

54 Spatio-temporal reconstruction of coastal carbonate system dynamics along the land-ocean continuum using Deep Learning : the Gulf of Lion shelf

Poster

Romane Pollet¹*, Laurent Coppola^2,3, Raphaëlle Sauzède^4,5

¹LOV-SUR, Villegranche-sur-Mer, France. ²SU, Villegranche-sur-Mer, France. ³OSU-STAMAR, Paris, France. ⁴CNRS-INSU, Villefranche-sur-Mer, France. ⁵IMEV, Villefranche-sur-Mer, France

Session

Session 4: Carbon cycling in the land ocean aquatic continuum

Abstract text

Coastal zones along the land ocean continuum play a disproportionate role in the marine carbon cycle, yet their spatio-temporal variability remains poorly constrained due to strong environmental gradients, episodic river inputs and limited long-term observations. Within the ANR RiOMar project (grant number 22-POCE-0006), we develop a Deep Learning approach to reconstruct the dynamics of the coastal carbonate system in river influenced environments, with a focus on the Gulf of Lion shelf. 

The modelling strategy builds upon the CANYON-Med framework (“CArbonate system and Nutrients concentration from hYdrological properties and Oxygen using Neural network in the Mediterranean Sea”, Fourrier et al., 2020) and is adapted to coastal conditions. A deep learning model reconstructs the four dimensional variability of key carbonate system variables, including total inorganic carbon, total alkalinity and pH from a reference database of discrete bottle measurements. In parallel, a surface oriented data-driven approach estimates air-sea CO₂ fluxes from satellite observations. Both models rely on hydrological inputs including temperature and salinity, complemented by biogeochemical tracers such as dissolved oxygen or chlorophyll-a. 

By combining multi-scale in situ observations with a data driven framework, this approach enables reconstruction of carbonate system variability across the Gulf of Lion shelf over the past 15-20 years. By augmenting spatially and temporally dispersed in situ observations, the ANN provides a consistent representation of coastal carbonate chemistry and its response to riverine inputs, anthropogenic pressures and climate variability. This work improves understanding of coastal carbon cycling and supports quantification of air-sea CO₂ fluxes in dynamic coastal environments.

55 Quantifying the Spatiotemporal Patterns of Atmospheric Methane Emissions and Their Associations with Environmental Factors in Bihar, India

Poster

Avinash Dass*, Rajesh Kumar Ranjan

Central University of South Bihar, Gaya, India

Session

Session 19: Understanding feedbacks between greenhouse gas exchange processes and climate variability using in situ observations, remote sensing, and machine learning

Abstract text

Methane (CH₄) is a potent greenhouse gas with a global warming potential approximately 80 times greater than carbon dioxide over a 20-year timescale. Its atmospheric concentration has increased significantly due to anthropogenic activities such as agriculture, fossil fuel extraction, and waste management. Understanding the spatial and temporal variability of CH₄ emissions is crucial for effective mitigation strategies. This study examines the spatiotemporal distribution of atmospheric CH₄ over Bihar, India, using satellite observations from the Sentinel-5P TROPOspheric Monitoring Instrument (TROPOMI) between 2018 and 2023. We integrate CH₄ concentration data with meteorological variables and vegetation indices, including NDVI, EVI, and GPP, to assess emission drivers and identify regional hotspots and coldspots. Our findings reveal a 2.8% increase in annual mean CH₄ levels over the study period. Peak CH₄ emissions occur during the monsoon and post-monsoon seasons (August-October), coinciding with rice cultivation, wetland emissions, and flooding. Spatial analysis identifies hotspots over the mid-Gangetic plains, driven by intensive agriculture, livestock farming, and population density, while coldspots are found in forested and less urbanized southern and northern regions. Statistical analyses show positive correlations between CH₄ and vegetation indices and relative humidity (r ≈ 0.4, p < 0.01), while wind speed exhibits a negative correlation (r = -0.33, p < 0.01), highlighting the role of biomass productivity, moisture availability, and atmospheric mixing in CH₄ variability. Anomaly analyses confirm agricultural and hydrological factors as key drivers of CH₄ emissions. These findings call for targeted mitigation in flood-prone and agricultural regions to support climate and CH₄ reduction goals.

56 Monitoring global wetland methane emissions through TM5-4DAVR inversion system

Oral

Francesco Graziosi*

European Commission Joint Research Centre, Ispra, Italy

Session

Session 13: Advancing approaches for quantifying greenhouse gas fluxes in terrestrial ecosystems

Abstract text

Uncertainties in the balance of the methane (CH₄) sources and sinks complicate the interpretation of the observed interannual variability of atmospheric CH₄. Moreover, climate change may enhance natural CH₄ releases, prompting the need for monitoring global CH₄ emissions. To achieve this, we employ the TM5-4DVAR inverse model system driven by ECMWF-ERA5 meteorological data at a resolution. This four-dimensional inverse system generates monthly global fields of CH₄ fluxes across four source categories: wetlands, rice fields, biomass burning, and anthropogenic activities. The methane fluxes are optimized using surface-based measurements and satellite retrievals. The primary aim of this work is to identify the major geographical areas and source categories driving the interannual variability and trends of global CH₄ fluxes during the study period. Moreover, the temporal variability of natural methane fluxes is analysed in relation to physical parameters to investigate how natural CH₄ emissions respond to climate factors, focusing on wetland emissions. The production of methane in wetlands is due to microbial methanogenesis, which depends on temperature, water-table depth, and both the quality and quantity of organic matter. Climate change is expected to affect these three main drivers of methanogenesis. In this study, we focus on determining methane emissions from wetlands and their sensitivity to climate variables. We analysed the emission response to 2 m temperature (T) and total precipitation (P) based on ECMWF-ERA5 reanalysis, considering different climate zones. 

58 An Activity-Based Framework for Sectoral CO₂ Emission Estimation Using Multi-Pollutant Satellite Observations

Poster

Zhen Qu*

North Carolina State University, Raleigh, USA

Session

Session 18: Quantifying anthropogenic greenhouse gas emissions from continental to regional scales

Abstract text

Accurately tracking anthropogenic CO₂ emissions is essential for understanding and reducing their environmental and social impacts. We develop an activity-based framework to quantify sectoral anthropogenic carbon dioxide (CO₂) emissions using satellite observations. It leverages the ratios of co-emitted nitrogen oxides (NO_x), sulfur dioxide (SO₂), and carbon monoxide (CO) to attribute emission sources and optimize sectoral activity rates, thereby overcoming the limitations of using CO₂ observations alone to distinguish between anthropogenic and natural fluxes. Pseudo-observation tests with known solutions show that this framework reduces the normalized mean square error (NMSE) by 26-68% and normalized mean bias (NMB) by 10-63% relative to the prior emissions, when evaluated against the true sectoral CO₂ emissions, demonstrating that sectoral CO₂ emissions are recoverable when uncertainties are well characterized. Application to East Asia demonstrates that incorporating SO₂ and CO observations improves constraints on residential and industrial emissions that are weakly constrained by nitrogen dioxide (NO₂) under the traditional NO_x-proxy method. Posterior simulations show overall improved agreement with the Greenhouse Gases Observing Satellite (GOSAT) CO₂ dry column mixing ratios but also reveal that the activity-based framework is sensitive to CO-related uncertainties, highlighting the importance of accurate CO simulations, observations, and error characterization. 

59 Regenerative (low-input, diverse pasture) dairy farming reduces nitrous oxide emissions, but at the expense of the net carbon balance

Poster

Johannes Laubach¹*, John Hunt¹, Diego Carvalho¹, Scott Graham¹, Paul Mudge²

¹Bioeconomy Science Institute - Manaaki Whenua Landcare Research Group, Lincoln, New Zealand. ²Bioeconomy Science Institute - Manaaki Whenua Landcare Research Group, Hamilton, New Zealand

Session

Session 14: Greenhouse gas fluxes in agroecosystems: processes, measurements and management implications

Abstract text

On grazed grasslands, reducing nitrous oxide (N₂O) emissions and maintaining or increasing soil carbon stocks can make important contributions to mitigating greenhouse-gas emissions. In New Zealand, conventional dairy systems use predominantly ryegrass-clover pastures with mineral fertiliser applications to optimise growth. A synthesis of past measurements has shown that the net ecosystem carbon balances of such pastures are generally near-neutral. Here we investigate whether N₂O emissions can be reduced by "regenerative" dairy farm management, with much lower fertiliser input, highly diverse pastures and longer grazing rotations, and how such management affects the soil-pasture system's carbon balance. Measurements were made on a farm where half the area and herd are managed regeneratively and the other half conventionally. In late 2022 eddy covariance measurements of carbon dioxide (CO₂) exchange from both types of paddocks started and since early 2024 we have also measured N₂O. We have found N₂O emissions from the regenerative paddock about 40 % lower than from the conventional paddock, which means the emissions are roughly proportional to the rates of nitrogen inputs (including from the cows' excreta). However, in the first and third year of measurements, the regenerative paddock lost carbon, while in the second year its carbon balance was neutral and similar to that of the conventional paddock. On a CO₂-equivalent basis, the carbon losses overwhelmed the effect of the reduced N₂O emissions. These findings underscore the importance of whole-system greenhouse gas accounting when evaluating low-input dairy systems as mitigation strategies.

60 Multi-year land surface fluxes of energy and greenhouse gas (CO₂, CH₄)  fluxes from stationary and mobile eddy covariance systems in Dutch peatland

Poster

Bart Kruijt, Hong Zhao*, Wietse Franssen, Wilma Jans, Laurent Bataille, Ruchita Ingle, Freek Engel, Corine Van Huissteden, Ignacio Kovacevic Andueza, Jan Biermann, Ronald Hutjes

Wageningen University, Wageningen, Netherlands

Session

Session 15: Peatlands in the global climate system: fluxes, feedbacks, and human pressures

Abstract text

Peatlands cover approximately 9% of the Netherlands and constitute a critical carbon reservoir. However, intensive drainage for dairy farming has transformed these landscapes into greenhouse gas (GHG) sources, emitting approximately 6 Mton CO₂ yr−1. To meet the Dutch National Climate Law’s target of reducing emissions to 1.0 Mton CO₂ yr−1 by 2030, a suite of mitigation measures—including pressurized drainage, clay-to-peat soil coverings, and paludiculture—has been implemented under the National Research Programme on Greenhouse Gas Emissions from Fen Meadow Areas (NOBV).

Monitoring the efficacy of these interventions is challenging due to the potential trade-off between reduced CO₂ oxidation and increased CH₄ production under anaerobic conditions. Since 2019, a comprehensive observational framework has been established, featuring several stationary Eddy Covariance (EC) towers and a novel mobile EC platform. The mobile approach provides high-density spatial coverage, enabling direct comparisons between control and treatment plots across diverse management regimes without the prohibitive costs of multiple fixed installations.

This study introduces a synthesized, multiple-year (2022–2025) dataset encompassing flux measurements of CO₂, CH₄, latent heat, and sensible heat, and biometeorological drivers such as four-component radiation and soil heat flux. By aligning these Dutch records with international standards, we facilitate their integration into networks like ICOS and FLUXNET. The dataset is expected to serve as a robust foundation for annual GHG budgeting, process-based model development, and the objective evaluation of peatland restoration strategies in the Netherlands.

61 UFLUX: Bridging Knowledge-Based and Data-Driven Modelling for Precise, Granular, and Cost-Efficient Understanding of Ecosystem-Climate Change

Poster

Songyan Zhu¹*, Wenquan Dong², Jian Xu³

¹University of Southampton, Southampton, United Kingdom. ²Lund University, Lund, Sweden. ³Chinese Academy of Sciences, Beijing, China

Session

Session 10: Cross-scale responses of greenhouse gas variability to climate extremes

Abstract text

Global terrestrial ecosystems absorb roughly one-third of annual anthropogenic CO₂ emissions. This estimate is typically derived from ensembles of process-based ecosystem models, yet disagreement remains substantial: global carbon uptake estimates range from ~90 to 180 PgC (nearly 100% variation). Due to high computational demands, many models operate at coarse spatial resolutions (tens to hundreds of kilometres), limiting relevance for local climate action. This mismatch is acute in fragmented regions such as Europe, where 27% of land comprises croplands and grasslands in patches smaller than 0.02 km².

Data-driven approaches combine long-term Earth observations with flux tower measurements and show strong predictive skill, often converging near 140 PgC. However, satellite products (~0.25 km²) remain too coarse for heterogeneous landscapes and exceed typical flux tower footprints. Their black-box nature also limits ecological interpretability.

We introduce the Unified FLUXes (UFLUX) framework, integrating ecological theory with a deep learning architecture (SmartForest4D). UFLUX provides a unified, high-resolution representation of carbon, water, and energy fluxes. Unlike flux towers that measure only net exchange, UFLUX reconstructs the full carbon cycle, including biomass dynamics. It achieves strong global performance (cross-validated R² ≈ 0.9) and identifies systematic biases, including ~30% overestimation of carbon uptake in some data-driven products. Results highlight ecosystem respiration as a key constraint on net sequestration.

At fine scales (down to 0.0001 km²), UFLUX resolves grazing impacts and urban carbon dynamics. Robust to >1 month data gaps (>70% accuracy), it enables granular, cost-efficient monitoring to support climate resilience and sustainable land management.

62 A pebble in some scientist's shoes : 8 years of co-creation with a theater company on science mediation

Oral, as part of the programme in the event "An evening of science and art", see details here

Didier Voisin¹*, Jérémy Brunet², Claire Chappaz³, Emmanuel Cosme¹, Julie Cozic³, Julien Delahaye⁴, Guillaume Douady², Celine Goujon⁴, Jean-Louis Hodeau⁴, Gladys Mary³, Aurélien Masseboeuf⁵, Didier Mayou⁴, Geremy Panthou⁶, Camille Pasquier², Nicolas Prugniel², Virginie Jacquier-Roux⁷, Yvonne Soldo⁴, Nicolas Vigier³, Theo Vischel¹, Sylvie Zannier⁸, Barthélémy Champenois²

¹Université Grenoble Alpes, Institut des Géosciences de l'Environnement, Grenoble, France. ²Compagnie du Gravillon, Grenoble, France. ³Atmo Auvergne-Rhône-Alpes, Grenoble, France. ⁴Université Grenoble Alpes, CNRS, Institut Néel, Grenoble, France. ⁵Université Grenoble Alpes, CEA, CNRS, Spintec,, Grenoble, France. ⁶Université Grenoble Alpes, OSUG, Institut des Géosciences de l'Environnement, Grenoble, France. ⁷Université Grenoble Alpes, Centre de recherche en économie de Grenoble, Grenoble, France. ⁸Université Grenoble Alpes, PhITEM, Grenoble, France

Session

Session 33: Science and arts: How to communicate science?

Abstract text

Live theatrical performance is a powerful mediation tool, engaging audiences emotionally and intellectually. By weaving complex ideas into narrative, it makes abstract scientific concepts tangible and memorable, avoiding jargon to reach non-experts.

Since 2016, la Compagnie du Gravillon has developed an interdisciplinary collaboration with scientists from Grenoble University and other local institutions. Initially focused on physics, this partnership now spans environmental sciences and human/social sciences (SHS), recognizing their role in addressing "wicked problems". Uniquely, our collective of artists and scientists co-creates performances over the long term. artists join scientific discussions, while scientists participate in the writing and staging process.

Our lecture-performances blend improvisation with scientific rigor. After a 2018 play on light physics, we shifted to environmental themes: Point de Bascule (2022) on carbon offsetting; Zone Critique Eau/Air (2026) on water/air cycles; and Le paradoxe de la mobylette on mobility. Performances target diverse local audiences within 150 km of Grenoble, from rural farms to urban theaters, using bike tours to build lasting community connections.

Point de Bascule (300+ performances since 2022), revealed that processing the societal implications of the science required specific spaces for dialogue. Most performances are now accompanied by interactive workshops led by the scientists who co-wrote the play, as post-show discussions or standalone sessions with hands-on demonstrations.

While effective, challenges remain: reception varies by audience (age, background, context). This presentation explores key success factors—why this format works for complex issues like climate change—and how to address its limitations to broaden impact

63 CarbonWatch-Urban: Granular CO₂ emissions information for every town and city in New Zealand

Poster

Jocelyn Turnbull¹*, Laila Balzer¹, Gordon Brailsford¹, Perry Davy¹, Lydia DiCaprio¹, Lucas Domingues¹, Leigh Fleming¹, Sally Gray¹, Timothy Hilton¹, Liz Keller¹, Daemon Kennett¹, Haeyoung Lee¹, Erika Macapagal-Taluban¹, Sara Mikaloff Fletcher¹, Brayden Lewellen¹, Vanessa Monteiro¹, Rowena Moss¹, Stijn Naus¹, Peter Sperlich¹, Hayden Young¹, Scott Graham², John Hunt², Johannes Laubach², Katie Sewell¹, Adam Tipper³, Nancy Golubiewski⁴, Andrea Brandon⁵, Ephraim Irvine⁶, Mingjing Dong⁷, Selena Sheng⁷, Le Wen⁷

¹Earth Sciences New Zealand, Wellington, New Zealand. ²Bioeconomy Science Institute, Lincoln, New Zealand. ³StatsNZ, Wellington, New Zealand. ⁴Climate Change Commission, Wellington, New Zealand. ⁵Ministry for the Envrionment, Wellington, New Zealand. ⁶Ngāti Whātua Ōrākei, Auckland, New Zealand. ⁷University of Auckland, Auckland, New Zealand

Session

Session 17: Urban greenhouse gas emissions and sinks: from advanced monitoring to source identification and impact

Abstract text

Urban areas produce 70-80% of  fossil fuel CO₂ emissions, thus detailed information about urban emissions and offsetting potential is needed to focus efforts where they will have the most impact to deliver meaningful, measurable change.  CarbonWatch-Urban is a multi-tiered approach that provides fine-scale CO₂ emissions and sink information for all of New Zealand’s urban areas.  First, we established a high resolution bottom-up inventory of Auckland's fossil fuel CO₂ emissions , and optimised the UrbanVPRM land surface model to estimate the Auckland region biogenic CO₂ budget.  We use atmospheric observations of carbon dioxide, radiocarbon-in-carbon-dioxide, carbon monoxide, carbonyl sulphide and black carbon to partition Auckland’s CO₂ emission sources, and an atmospheric inversion to rigorously validate and improve the bottom-up flux estimates. We leverage the Auckland experience to expand the bottom-up emission modelling framework nationally, incorporating the improvements identified in Auckland, and validate the models using eddy covariance measurements in Christchurch and mobile laboratory observations from towns and cities spanning New Zealand’s climates, topographies and urban forms. This combined approach provides the most accurate estimates of New Zealand’s urban emissions. We utilise the granular emissions information in combination with spatially resolved socioeconomic information to identify drivers of emission patterns in New Zealand cities.

64 Source–Sink Optimization and Economic Assessment of CO₂ Capture, Transport, and Utilization from Indonesian Coal-Fired Power Plants for Enhanced Oil Recovery

Poster

Victor Siahaan*

PT PLN Energi Primer Indonesia, Jakarta, Indonesia

Session

Session 28: Combining data and models to support emissions estimation and policy at local to regional scales

Abstract text

This research presents a comprehensive assessment of carbon capture, transportation, and utilization potential from 104 coal-fired power plants (CFPPs) across Indonesia. The compiled database includes plant coordinates, annual coal consumption in 2024 and specified coal calorific value (GAR) ranges. The relationship between coal gross calorific value (GAR) and CO₂ emission factor was derived using IPCC default emission factors and calorific conversion parameters, enabling estimation of annual CO₂ emissions based on fuel consumption. Consistent with established emission estimation methodologies, higher coal calorific values correspond to higher emission factors, resulting in greater carbon release per unit of coal burned. Potential CO₂ utilization sites were limited to priority oil fields suitable for enhanced oil recovery (EOR), characterized by large remaining oil in place, mature reservoirs, API gravity above 25°, and reservoir pressures near minimum miscibility pressure (MMP), primarily located in South Sumatra, West Java, and East Java. Three region of oil fields were selected, and source–sink matching optimization was conducted using geodesic distance calculations, complemented by marine routing factors to estimate realistic transport distances and costs. Capture and transport costs were integrated to evaluate economic feasibility, followed by comparison with potential carbon pricing incentives. Results indicate that although CO₂-EOR utilization offers significant mitigation potential, financial gaps remain as the combined capture and transport cost of approximately 66.64 USD/tCO₂ exceeds commonly referenced carbon pricing incentives such as the United States 45Q benchmark, highlighting the importance of policy support and fiscal incentives to enable large-scale deployment in Indonesia.

65 Accounting for the urban biosphere for a Southern-Hemisphere sub-tropical city: alternatives to atmospheric inversions and eddy covariance

Poster

Timothy Hilton¹*, Elizabeth Keller¹, Leigh Fleming¹, Lucas Dominques^2,3, Jeremy Parry-Thompson^2,4, Hayden Young¹, Vanessa Monteiro^1,5, Sally Gray⁶, Beata Bukosa⁶, Stijn Naus⁶, Gordon Brailsford⁶, Rowena Moss⁶, Sara Mikaloff-Fletcher⁶, Lucy Hutyra⁷, Gurney Kevin⁸, Jocelyn Turnbull¹

¹Earth Sciences New Zealand, Lower Hutt, New Zealand. ²GNS Science, Lower Hutt, New Zealand. ³Nuclear and Energy Research Institute, São Paulo, Brazil. ⁴Greater Wellington Regional Council, Wellington, New Zealand. ⁵Universitat Autònoma de Barcelona, Barcelona, Spain. ⁶Earth Sciences New Zealand, Wellington, New Zealand. ⁷Boston University, Boston, USA. ⁸Northern Arizona University, Flagstaff, USA

Session

Session 17: Urban greenhouse gas emissions and sinks: from advanced monitoring to source identification and impact

Abstract text

The CarbonWatch-Urban project is tasked with estimating fossil fuel and biogenic carbon dioxide emissions for all of New Zealand's towns and cities at 500 metre spatial resolution and hourly temporal resolution. These communities range in size from Auckland (population 1.5 million; New Zealand's largest city) to numerous small towns. To provide emissions estimates in these diverse settings we employ separate models for fossil fuel emissions and biogenic (photosynthesis, respiration) fluxes, and we use a variety of atmospheric observations to constrain these models and support high-resolution atmospheric inversions.  The urban biosphere is challenging in the best of circumstances, and in New Zealand we must additionally allow for models that were developed and parameterised for cold-climate Northern Hemisphere cities whose vegetation is largely deciduous with long dormant periods during winter when foliage has been dropped.  These parameterisations are at best questionable for New Zealand cities as well as many other tropical and sub-tropical cities with year-round growing seasons and evergreen broadleaf vegetation.  Auckland, like many of these cities, also does not have a time series of eddy covariance data to use for model parameterisation.  Here we present an approach to tune a terrestrial biosphere model using flask-based atmospheric radiocarbon measurements from Auckland as an alternative to the parameterisation methods commonly employed.  We also demonstrate that these methods, in conjunction with a fossil fuel emissions inventory, can provide spatially and temporally granular emissions estimates without the intensity of resources (computational, observational, and scientific expertise) required to run a full atmospheric inversion.

66 Exploring the behaviour of the ocean and land carbon sinks with Jungfraujoch's observational record

Poster

Stuart Grange¹*, Peter Nyfeler¹, Vasileios Mandrakis¹, Filip Zürcher¹, Eliza Harris^1,2

¹University of Bern, Bern, Switzerland. ²High Altitude Research Stations Jungfraujoch and Gornergrat, Bern, Switzerland

Session

Session 18: Quantifying anthropogenic greenhouse gas emissions from continental to regional scales

Abstract text

The global carbon budget is a key system to understand with respect to global temperature change and the ocean and land carbon sinks have demonstrated high levels of resilience in the face of increasing anthropogenic emissions of carbon dioxide (CO₂) into the atmosphere. However, there are signals that these two critical carbon sinks’ abilities to sequester carbon are declining. 

Observations of CO₂ and oxygen (O₂; in the form of δ(O₂/N₂)) from both in situ analysers and flask sampling activities from the Jungfraujoch high-alpine observatory (3572 m above sea level) in the Swiss Alps have been used to constrain the carbon budget using the O₂-CO₂ partitioning method to illuminate the behaviour of the carbon sinks between 2005 and 2025. High-precision measurements of CO₂ and O₂ are useful for partitioning because these two species are intrinsically linked through various processes, however, the dissolution of CO₂ into the ocean does not involve O₂, thus allowing for the separation of the ocean and land sinks. The results indicate that the absolute carbon uptake of the ocean and land sinks has plateaued, and the carbon uptake of the land sink has declined between 2005 and 2025. At the same time, more CO₂ has remained in the atmosphere, thus increasing the atmospheric growth rate. These results provide more evidence from a purely observational record that the two carbon sinks are weakening. The complexities of the interpretation and possible downstream climate impacts will be discussed.

67 Extreme hydrological drought reduces CO₂ uptake in China’s largest floodplain lake

Oral

Xiaosong Zhao*, Lejun Zhao, Xingwang Fan

Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences, Nanjing, China

Session

Session 10: Cross-scale responses of greenhouse gas variability to climate extremes

Abstract text

Climate extremes are becoming more frequent and intense, yet how extreme hydrological events regulate CO₂ dynamics in floodplain lakes remains poorly understood. Poyang Lake, China's largest freshwater lake, has experienced frequent extreme drought events in recent years, raising concerns about the stability of its carbon sink function. Understanding the response of carbon dynamics to such extreme hydrological regimes is critically important for predicting climate feedbacks in these vulnerable ecosystems. Here we constructed a spatially explicit carbon dioxide (CO₂) flux dataset covering 2003−2022 for China’s largest floodplain lake (R² = 0.86, RMSE = 0.49 gC m⁻² d⁻¹). The annual fluxes varied from the peak annual flux in 2010 (52.57 ± 4.71 gC m⁻²) to the maximum carbon uptake in 2011 (−186.36 ± 7.27 gC m⁻²). Hydrological regime shifts dominated temporal variability, particularly the timing of water rise and recession. A 10-day delay in lake water rise enhanced CO₂ uptake by 19.20 gC m⁻², whereas a 10-day advance in lake water recession increased uptake by 11.63 gC m⁻². However, the enhancement of CO₂ sink can be impaired in case of excessively early or rapid lake water level decline. For example, the 2022 extreme drought reduced CO₂ uptake by over 20% compared to the normal drought years due to plant water stress and increased ecosystem respiration. Our findings reveal nonlinear responses of lake CO₂ fluxes to hydrological extremes, highlighting the need to integrate event-scale dynamics into carbon budgeting under changing climate.

68 Near Real-Time Optimization of Flask Sampling Strategy at ICOS Stations Using Model-Based Radiocarbon Contamination Assessment

Poster

Zhendong Wu^1,2*, Alex Vermeulen², Samuel Hammer³, Ute Karstens¹, Lynn Hazan⁴, Amara Abbaris⁴, Markus Eritt⁵, Timo Knaack³, Claudio D'onofrio^1,2

¹ICOS Carbon Portal, Department of Earth and Environmental Sciences, Lund University, Lund, Sweden. ²ICOS ERIC, Carbon Portal, Lund, Sweden. ³Institute of Environmental Physics, Heidelberg University, ICOS-CRL, Heidelberg, Germany. ⁴LSCE/IPSL, UMR CEA-CNRS-UVSQ, Gif-sur-Yvette, France. ⁵Max Planck Institute for Biogeochemistry,ICOS-FCL Lab, Jena, Germany

Session

Session 18: Quantifying anthropogenic greenhouse gas emissions from continental to regional scales

Abstract text

Radiocarbon (¹⁴C) measurements from atmospheric CO₂ flask samples are a key tool for quantifying fossil fuel CO₂ emissions. However, these measurements can be significantly influenced by ¹⁴C discharges from nuclear power plants. Such contamination reduces the usefulness of samples for fossil fuel source attribution and may increase analytical costs due to the high expense of radiocarbon analysis.

We developed the FlaskSampling App (https://flasksampling.app.icos-cp.eu/), a decision-support tool for optimizing flask sampling strategies at ICOS atmospheric stations. The application visualizes near real-time Δ¹⁴C simulation results and supports automated decision-making on whether to retain or discard collected flask samples based on a user-defined contamination threshold. The goal is to enable objective, station-level sample screening without manual intervention, ensuring that only uncontaminated samples proceed to costly laboratory analysis.

The backend of the App integrates backward simulations from the FLEXPART atmospheric transport model to estimate station-specific source footprints. FLEXPART is operated in both global (1° resolution) and European (0.25° resolution) domains, driven by 3-hourly ECMWF IFS operational meteorological data with a two-day latency. By combining simulated footprints with spatially resolved ¹⁴C emission inventories from European nuclear facilities, the tool quantifies the potential nuclear influence on individual flask samples.

This automated screening framework reduces unnecessary radiocarbon analyses, improves sampling efficiency, and enhances the reliability of fossil fuel CO₂ assessments within the ICOS network. By supporting consistent, data-driven sampling decisions, the FlaskSampling App contributes to more cost-effective and robust atmospheric radiocarbon monitoring across Europe.

69 Observation-based estimates of land-biosphere CO₂ fluxes in the northern high latitudes

Oral

Rona Thompson¹*, Gregoire Broquet², Audrey Fortems-Cheiney³, Philippe Peylin², Vladislav Bastrikov³

¹NILU, Kjeller, Norway. ²LSCE, Saclay, France. ³Science Partners, Paris, France

Session

Session 29: Using GHG measurement data to support national greenhouse gas inventories

Abstract text

The northern high latitudes are home to Boreal forests, which are thought to be an important sink of CO₂. In particular, Russia has the largest expanse of Boreal forest, covering about one quarter of its territory. However, estimates of the carbon sink over northern Boreal forests, and Russia in particular, are very uncertain. Russia has reported almost no change in forest biomass in its national inventories since the collapse of the Soviet Union, whereas satellite remote sensing indicates increased forest cover and biome productivity. 

In this study, we estimate land-biosphere fluxes of CO₂, specifically Net Ecosystem Exchange (NEE) and land-use change fluxes (together NEE + FLUC), using observations of CO₂ mole fractions in the FLEXINVERT inversion framework. The inversion was run for the entire northern hemisphere north of 41°N, but with a focus on northern Eurasia. NEE + FLUC are estimated for years 2012 to 2021, with 2021 being the last year of data available over Russia at the time of the study. The inversion followed a newly developed protocol in the EYE-CLIMA project in which the lateral transfer of carbon through harvest and river transport is accounted for in the prior. The results for NEE + FLUC, thereby, are more comparable to CO₂ LULUCF reported by national inventories. 

Results indicate that the Russian land biosphere is a weak sink of CO₂ after crop and wood harvest have been accounted for. However, large uncertainties remain owing to the sparse observational coverage in this region and uncertainties in the fossil and harvest emissions.

70 Optimal extension of the current CO₂ observation network using Incremental Optimisation

Oral

Prabir Patra*

JAMSTEC, Yokohama, Japan

Session

Session 26: Designing the ideal global greenhouse gas monitoring network

Abstract text

Development of Global Greenhouse Gases Watch (G3W) is envisaged by the World Meteorological Organisation (WMO) in order to support the policymakers and parties to enable better tracking of actions under the Paris Agreement and other emission mitigation efforts. The G3W aims to provide distributions of surface fluxes at 1x1 deg spatial resolution at monthly time intervals, reduced latency of month(s) and higher accuracy than present. To achieve this goal the current surface observation network is insufficient to make/provide measurements without any break due to environmental conditions (unlike the remote sensing instruments) because of non-uniform global coverage.

In the Task Team (TT)–Network of G3W, we aim to design an observation network that fill up the well-recognised gaps in an efficient way. Using the MIROC4-ACTM global transport model and the Incremental Optimisation (IO) approach, we have identified 50-odd optimal sites on top of the base observation networks consisting of 50 and 142 sites. The optimal sites are selected based on criteria of a posteriori flux uncertainty reduction (aFUR) in a 84-region inversion model (Chandra et al., 2022). Our results show that the optimal sites are placed in regions of Tropical and South America, Africa, Siberia/Central/South Asia, and Southeast Asia and Australia. By additional 50 sites over the land, the total land flux aFURs were 62% and 58% compared to the base cases using 50 and 142 sites, respectively. The respective aFURs were 17% and 33% compared to the total a priori flux uncertainty of 81 PgC yr^-1.

71 Quantifying net greenhouse gas exchange of urban green areas

Oral

Leena Järvi¹*, Veera Bilaletdin¹, Vertti Perttilä¹, Hermanni Aaltonen², Esko Karvinen², Jesse Soininen¹, Xiao Bai³, Mari Pihlatie¹, Liisa Kulmala²

¹University of Helsinki, Helsinki, Finland. ²Finnish Meteorological Institute, Helsinki, Finland. ³Aarhus University, Aarhus, Denmark

Session

Session 17: Urban greenhouse gas emissions and sinks: from advanced monitoring to source identification and impact

Abstract text

Cities are increasingly interested in harnessing urban vegetation to provide various ecosystem services including greenhouse gas (GHG) emission mitigation. The existing research has primarily centered around carbon sequestration and the exchange of other GHGs, and the role of soils have been widely neglected. Studies suggest that urban soils exhibit higher respiration rates, increased nitrous oxide (N₂O) emissions and reduced methane (CH₄) uptake compared to natural ecosystems. These alterations are driven by urban features such as fertilization, irrigation and nitrogen deposition, and changes in soil physical condition. To understand better the net GHG exchange of urban vegetation, we have conducted extensive observational campaign in Helsinki, Finland, from summer 2025 to summer 2026. The measurements cover eddy covariance (EC) measurements at the FI-Kmp station using MGA8 (MIROAnalytics) and Los Gatos gas analyzers, automatic soil chambers, mobile chamber platform and targeted manual chamber measurements focusing to quantify the impact of irrigation and biochar of the exchanges. The initial results show a significant reduction in CH₄ uptake under irrigation. CO₂ and N₂O showed no consistent response in summer 2025, which might be due to the relatively normal climatic conditions. Biochar appeared to suppress the largest N₂O flux events from soil, but no significant effects on CO₂ and CH₄ fluxes were seen. CH₄ fluxes showed pronounced spatial variability across the study site. The ecosystem level EC measurements reveal the surroundings to act as a source for both CH₄ and N₂O requiring more detailed analysis with footprint analysis to distinguish between anthropogenic and biogenic components. 

72 Lungs – visualizing urban carbon exchange though art and eddy covariance data

Oral, as part of the programme in the event "An evening of science and art", see details here

Leena Järvi¹*, Teemu Lehmusruusu², Ranja Hautamäki²

¹University of Helsinki, Helsinki, Finland. ²Aalto University, Espoo, Finland

Session

Session 33: Science and arts: How to communicate science?

Abstract text

Urban ecosystems are dynamic mosaics of anthropogenic emissions and biogenic carbon exchange. However, the patterns of carbon exchange between urban surfaces and the atmosphere remain largely invisible to citizens despite their fundamental role in the ongoing climate change. Here, we explore a novel approach to visualizing city-scale carbon dynamics by integrating high-frequency eddy covariance (EC) CO₂ flux measurements with three-dimensional laser scanning data to create an artistic representation of the city’s “lungs” – urban green areas taking up atmospheric carbon. These scientific data were transformed into a computer animation titled Lungs, presented as part of an urban light art festival, which renders invisible carbon flows into perceptible visual patterns. By translating EC data into a public art context, we aim to foster broader awareness of urban carbon exchange and the impact of green areas as carbon sinks and highlight the value of long-term, independent observations in understanding urban carbon budgets. This interdisciplinary approach demonstrates how ICOS data can contribute not only to scientific understanding but also to interpretations of how cities “breathe” - making complex biogeochemical processes tangible and accessible to society.

73 Low-Cost Electronic Noses for Methane Tracking

Poster

Guillem Domènech-Gil*

Department of Thematic Studies and Environmental Change (TEMA M), Linköping University, Linköping, Sweden

Session

Session 21: Emerging approaches for greenhouse gas flux measurements

Abstract text

Atmospheric methane (CH₄) concentrations are rising rapidly for partly unresolved reasons, and its 100-year global-warming potential is 28-34 times that of carbon dioxide by mass. Therefore, widely accessible means to quantify CH₄ are increasingly needed to understand variability, source attribution and support mitigation efforts. Low-cost sensor (LCS) approaches could offer an appealing complement to other measurement approaches. Enabling reliable measurements with LCSs could facilitate distributed networks for flux measurements and the identification of source-sink dynamics. However, robust calibration methods and interference compensation remain key barriers for LCSs.

We present a new strategy to address these challenges: an electronic-nose platform that combines multiple low-cost metal-oxide CH₄ sensors and with environmental sensors for temperature, humidity, and pressure. Using data from this platform and mechanistically informed machine‑learning models trained against reference equipment, we compensate for cross‑interference and retrieve CH₄ concentrations allowing to track environmental variability. 

We present results and differences between laboratory and field calibrations, and we outline limitations and pathways for improving our versatile approach, implemented to test data from both natural and anthropogenic sources. This strategy can help guide mitigation actions and pave the way for the widespread use of LCS networks in greenhouse gases flux measurements.

74 Current climatic effect of carbon cycling in boreal lakes

Poster

Janne Rinne¹*, Sari Juutinen², Pirkko Kortelainen³, Marja Heikkilä⁴, Tuula Larmola¹, Anne Ojala¹

¹Natural Resources Institute Finland (LUKE), Helsinki, Finland. ²Finnish Meteorological Institute, Helsinki, Finland. ³Finnish Environment Institute, Helsinki, Finland. ⁴Geological Survey of Finland, Helsinki, Finland

Session

Session 4: Carbon cycling in the land ocean aquatic continuum

Abstract text

The development and functioning of any given ecosystem creates a radiative forcing (RF) by causing perturbations to the atmospheric greenhouse gas levels. For example, mire ecosystems have collected carbon over millennia, creating a cooling RF, while methane (CH₄) emission creates a warming RF. Thus, their total climatic effect is a combination of these opposing effects¹.

Similarly, the current climatic effect of carbon cycling in the boreal lakes results from the accumulation of carbon into lake sediments over millennia, and the current CH₄ emission and carbon dioxide (CO₂) evasion from lake surface. These create opposing RF components. We are using data on carbon stored in the Finnish boreal lake sediments² with estimates of surface gas exchange^3,4 to estimate the RF of these systems using the newly developed ACME approach⁵.

The annual methane emissions from lake systems are generally considerably lower than the emissions from mire systems, similarly to the carbon storage densities. Boreal lakes can be a considerable source of CO₂ into the atmosphere. The range of estimates of current RF by the Finnish boreal lakes estimated by ACME approach will be presented and the role of the lakes to the current RF discussed.

References

1. Frolking et al, 2006, J. Geophys. Res., 101, G01008
2. Pajunen, 2000, Geological Survey Finland, Special Paper 29, 39-69
3. Kortelainen et al., 2006, Global Change Biol., 12, 1554-1567
4. Juutinen et al., 2009, Biogeosci., 6, 209-223
5. Rinne et al 2025, Environ. Res. Lett., 20, 064033

75 Simultaneous measurements of multiple greenhouse gases using a drone-based system with correction for gas sensor drift and improved source area identification

Oral

Magnus Gålfalk*, David Bastviken

Department of Thematic Studies—Environmental Change, Linköping University, Linköping, Sweden

Session

Session 20: Unmanned autonomous vehicles and proximal sensing in greenhouse gas research and monitoring

Abstract text

Two important challenges with UAS-based measurements of greenhouse gases (GHGs) are large baseline drifts of gas sensors and source identification when there are several sources distributed within a footprint. Ambient conditions, such as temperature and humidity, known to influence the accuracy of gas sensors, change fast during flight at different altitudes and speeds. Big such drifts limit UAS-based measurements to high and often anthropogenic emissions as they cause strong gradients in concentration levels. Emissions across landscapes often generate much weaker concentration gradients and are also more extended, making fluxes and source areas more challenging to constrain.

We present a newly developed approach to reduce this instrument drift significantly, enabling flux measurements in natural environments. We have also developed the approach further to allow the matching of gas structures on vertical wall flight paths to sources and sinks in the footprint using an independent tracer of air movements across the scene. The new method produces drift-corrected simultaneous measurements of multiple GHGs (CO₂, CH₄, N₂O) and wind data. We will present results using different flight strategies (e.g. single wall, two-wall, and rectangular walls) in both anthropogenic and natural environments.

76 Building a national atmospheric inversion framework for CO₂ and CH₄ in Belgium using CTDAS–WRF

Poster

Sieglinde Callewaert¹*, Jiaxin Wang^2,1,3, Friedemann Reum⁴, Julia Marshall⁴, Filip Desmet^1,5, Thilo Heinecke^5,6, Bart Dils¹, Mahesh Kumar Sha¹, Martine De Mazière¹, Bert Gielen⁵, Bernard Heinesch⁶

¹Royal Belgian Institute for Space Aeronomy (BIRA-IASB), Brussels, Belgium. ²Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China. ³University of Chinese Academy of Sciences, Beijing, China. ⁴German Aerospace Center (DLR), Institute of Atmospheric Physics, Oberpfaffenhofen, Germany. ⁵Plants and Ecosystems (PLECO), Department of Biology, University of Antwerp, Antwerp, Belgium. ⁶Gembloux Agro-Bio Tech, Université de Liège, Gembloux, Belgium

Session

Session 28: Combining data and models to support emissions estimation and policy at local to regional scales

Abstract text

National greenhouse gas (GHG) inventories reported to the UNFCCC (United Nations Framework Convention on Climate Change) are traditionally derived using bottom-up approaches based on activity data and emission factors. Complementary top-down approaches, combining atmospheric observations with transport modelling, provide an independent means to evaluate and improve national-scale emission estimates.

In this context, the VERBE project (Towards a GHG Monitoring and Verification System for Belgium, https://verbe.aeronomie.be) was initiated in 2022 to develop a national top-down GHG monitoring and verification system in support of Belgian climate policy and inventory reporting (MVS). A central objective of VERBE is the establishment of the methodological and observational framework (both in-situ and remote-sensing observations of GHG concentrations) required for atmospheric inverse modelling for MVS in Belgium, while building on existing expertise in ecosystem exchange modelling and monitoring of atmospheric GHG mole fractions. 

As part of this effort, the CarbonTracker Data Assimilation Shell coupled with WRF-GHG (CTDAS–WRF) has been implemented as the core inversion framework. The system is applied to CO₂ and CH₄ at the national scale, optimizing scaling factors for prior fluxes using an ensemble square root smoother. For CO₂, anthropogenic and biogenic flux components are treated separately. Atmospheric constraints are currently provided by in situ observations from the ICOS network. 

This presentation focuses on the current CTDAS–WRF configuration for Belgium, recent developments in its implementation, and insights from initial sensitivity experiments such as a performance assessment with respect to observations, in the context of supporting national GHG inventory verification and emission mitigation monitoring.

77 Laser based Δ¹⁴C-CO₂ monitoring for identifying urban fossil emissions

Oral

Patrick Siegwolf¹*, Andrew Whitehill¹, Stephan Henne¹, Dominik Brunner¹, Sangil Lee², Ruth Hill-Pearce³, Samuel Hammer⁴, Lukas Wacker⁵, Lukas Emmenegger¹, Béla Tuzson¹, Joachim Mohn¹

¹Empa, Laboratory for Air Pollution / Environmental Technology, Dübendorf, Switzerland. ²Korea Research Institute of Standards and Science, Daejeon, Korea, Republic of. ³National Physical Laboratory, Teddington, United Kingdom. ⁴Heidelberg University, Heidelberg, Germany. ⁵ETH, Zürich, Switzerland

Session

Session 17: Urban greenhouse gas emissions and sinks: from advanced monitoring to source identification and impact

Abstract text

Urban areas are major sources of anthropogenic CO₂ emissions. Yet, accurate quantification of local fossil emissions based solely on CO₂ mole fractions is difficult due to interfering biospheric fluxes. Radiocarbon in CO₂ (¹⁴C) is a direct tracer to distinguish fossil from biogenic CO₂. Currently, state-of-the art measurements are largely based on flask sampling and subsequent analysis by accelerator mass spectrometry. The resulting limited temporal coverage reduces effective footprint coverage, leading to episodic characterization of fossil fuel emissions.

Here, we present a novel laser-based analytical setup for continuous Δ¹⁴C-CO₂ measurements in a suburban area (Empa campus, Switzerland). The setup consists of a saturated-absorption cavity ring-down spectrometer (ppqSense), coupled with a pre-concentration unit (NC Technologies) and supplemented with a quantum cascade laser absorption spectrometer for CO₂ purity and δ¹³C-CO₂ analysis. The system achieves a 30-minute temporal resolution and a repeatability in the low per mille range in ambient air. We developed gas standards based on various biogenic and fossil CO₂ sources with varying isotopic composition to ensure consistency with reference laboratories. The data set is complemented by stable isotope measurements (δ¹⁷O-CO₂ and δ¹⁸O-CO₂).

We showcase first continuous Δ¹⁴C-CO₂ measurements and interpret the time series together with co-located CO₂ observations and high-resolution Lagragian transport simulations of fossil and biogenic CO₂ to quantify fossil enhancements at the ppm-level to attribute urban emission patterns at sub-hourly time scales. These measurements provide a pathway toward improved source apportionment, uncertainty reduction, and independent evaluation of urban emission inventories based on direct ¹⁴C-measurements.

78 Continuous CO₂ Stable Isotope Measurements at the High-Alpine Station Jungfraujoch

Poster

Patrick Siegwolf*, Lukas Emmenegger, Béla Tuzson

Empa, Laboratory for Air Pollution / Environmental Technology, Dübendorf, Switzerland

Session

Session 18: Quantifying anthropogenic greenhouse gas emissions from continental to regional scales

Abstract text

The stable isotopic composition of atmospheric carbon dioxide is characteristic of emission sources and processes shaping the atmospheric carbon cycle. However, sustained high-precision isotope measurements at remote sites remain technically challenging due to demanding infrastructure and maintenance requirements. Hence, long-term isotope time series are predominantly derived from discrete flask samples with limited temporal coverage and resolution.

At the high-alpine Jungfraujoch station, continuous CO₂ stable isotope measurements have been performed since 2008 using a quantum cascade laser absorption spectrometer with second time resolution, allowing the observation of both background trends and the characterization of short-term pollution events. After a major hardware upgrade in 2024, measurement performance improved substantially. The current setup achieves a precision better than 0.01 ppm for CO₂ and about 0.03 ‰ for δ¹³C-CO₂ and δ¹⁸O-CO₂, with respective accuracies of around 0.04 ppm and 0.05 ‰ on the WMO and VPDB scales. In parallel, quality control procedures were standardized, and the resulting data products are now routinely uploaded to the ICOS Carbon Portal.

This dataset provides a unique high-resolution long-term record of atmospheric CO₂ isotopic composition at Jungfraujoch, representative of the European background. The high temporal resolution enables robust detection of pollution events, supporting effective filtering for background analyses and characterization of isotopic signatures from major source regions, including episodic influence from various urban regions. Sustained operation and regular data submission are planned to strengthen Europe’s isotope observing capacity and support assessments of long-term trends in atmospheric CO₂ isotopic composition.

79 Title: Multi-Scale Greenhouse Gas Emission Monitoring and Assessment System in the ICOS Pilot City Munich

Oral

Jia Chen¹*, Josef Stauber¹, Daniel Kühbacher¹, Junwei Li¹, Patrick Aigner¹, Betty Molinier², Natascha Kljun², Ann-Kristin Kunz³, Andreas Christen³, Samuel Hammer⁴, Rainer Hilland⁵, Andreas Luther¹, Moritz Oliveira Makowski¹, Adrian Wenzel¹, Florian Dietrich¹, Stavros Stagakis⁶, Christopher Claus Holst⁷, Changxing Lan⁷, Dominik Brunner⁸

¹Professorship of Environmental Sensing and Modeling, Technical University of Munich, Munich, Germany. ²Department of Environmental and Geosciences, Lund University, Lund, Sweden. ³Faculty of Environment and Natural Resources, University of Freiburg, Freiburg, Germany. ⁴Heidelberg University, Heidelberg, Germany. ⁵Netherlands Organisation for Applied Scientific Research (TNO), The Hague, Netherlands. ⁶University of Basel, Basel, Switzerland. ⁷Karlsruhe Institute of Technology (KIT), Garmisch, Germany. ⁸Swiss Federal Laboratories for Materials Science and Technology (EMPA), Dübendorf, Switzerland

Session

Session 17: Urban greenhouse gas emissions and sinks: from advanced monitoring to source identification and impact

Abstract text

Cities are major sources of anthropogenic GHG emissions and, therefore, central to global climate-mitigation efforts. Robust, observation-driven monitoring systems are essential for evaluating progress toward the ambitious reduction targets set by many cities. Munich serves as a pilot city for the ICOS Cities initiative, which seeks to advance urban GHG measurement frameworks. Here, we report the first comprehensive GHG assessment for Munich derived from this multi-scale observational and modeling system.

We established a city-wide sensor network comprising 20 roof-top mounted mid-cost CO₂ systems, 52 low-cost CO₂ sites, and 30 low-cost air-quality units at street level, alongside our long-established MUCCnet with five solar-tracking spectrometers that measure total-column CO₂, CH₄, and CO. Further, we conducted tall-tower eddy covariance flux measurements of CO₂ and co-emitted species, relaxed eddy accumulation (REA) ¹⁴CO₂ measurements, biospheric field observations, and wind LiDAR profile measurements. We incorporated these data in self-developed high-resolution models, including emission inventories, transport modeling, flux footprint modeling, biospheric modeling, and various inverse modeling techniques. 

By integrating all observational and modelling components, we assessed Munich's GHG emissions for up to 6 years. We evaluated multiple urban emission estimation approaches, outlining the strengths and added value of each, and compared their outcomes with the city’s self-reported inventories. Within the uncertainty bounds, our estimates agree and can provide timely feedback for city climate action, whereas self-reported city inventories are typically delayed by 2-3 years. The resulting insights offer evidence-based guidance for shaping future mitigation strategies and provide recommendations to support robust emission accounting in other cities.

80 Characterizing anthropogenic emissions through continuous atmospheric CO₂ observations in Barcelona

Poster

Vanessa Monteiro¹*, Gara Villalba Mendez¹, Qing Luo¹, Roger Curcoll Masanes²

¹Universitat Autònoma de Barcelona, Barcelona, Spain. ²Universitat Politècnica de Catalunya, Barcelona, Spain

Session

Session 17: Urban greenhouse gas emissions and sinks: from advanced monitoring to source identification and impact

Abstract text

The urban greenhouse gas (GHG) monitoring network operating in the Barcelona Metropolitan Area (AMB) currently comprises five high-precision measurement sites equipped with Picarro instruments that continuously measure carbon dioxide (CO₂) and methane (CH₄). The network has been progressively expanded, with the earliest site installed in late 2021 and the most recent operational from 2026. The sites (Fabra, ICM, ICTA, IDAEA, and UPC-Agropolis) were strategically deployed to capture the diversity of urban landscapes across the AMB, including a natural forest, an urban coastal area, high-traffic infrastructure on the city outskirts, an urban park within a highly urbanized area, and peri-urban agricultural zones. This configuration enables the characterization of the heterogeneity of urban emissions and local atmospheric dynamics. Hourly averaged CO₂ mole fractions reveal clear spatial gradients between sites: the forested site (Fabra) consistently exhibits lower mole fractions, whereas ICTA, located near a major highway, shows the highest values and the greatest variability, agreeing with previous multi-site measurement campaigns in Barcelona.

These results demonstrate the value of sustained, high-resolution urban observations for capturing both spatial and temporal variability in GHG concentrations. Beyond improving the understanding of urban atmospheric processes and supporting model validation, the network data are intended to be used to inform urban-scale emission assessments and policy-relevant analyses. In particular, we aim to show how continuous atmospheric CO₂ observations can help identify emission hotspots and provide observational evidence to support the development and assessment of GHG mitigation strategies in a complex urban environment such as Barcelona.

81 When drought and heat in 2022 and 2023 turned the FR-Pue Mediterranean forest into a two-year net carbon source

Poster

Jean-Marc Limousin*, Jean Kempf, Jeanne Poughon, Serge Rambal, Jean-Marc Ourcival

CEFE, Univ Montpellier, CNRS, EPHE, IRD, Montpellier, France

Session

Session 10: Cross-scale responses of greenhouse gas variability to climate extremes

Abstract text

Southern France experienced record-breaking temperatures in 2022, associated with two consecutive years of low precipitation, severe drought episodes and recurrent heatwaves. Using the 20-year data series (2001-2021) of eddy-covariance carbon and water fluxes measured at the ICOS Mediterranean forest site FR-Pue (Puéchabon) prior to this event, we assess the impacts of these two exceptional years on the carbon budget of the forest. While the Puéchabon forest always behaved as a net carbon sink between 2001 and 2021, the carbon balance was reversed to a net annual carbon source in 2022 and 2023. This anomaly was caused by a deficit of photosynthesis, as leaf physiology was severely impacted by water stress and heat stress. Significantly lower photosynthetic rates were not restricted to the most stressful conditions but also manifested under most meteorological conditions even outside the summer period. This observation demonstrates that such extreme meteorological events have long lasting effects on tree physiology. However, these negative effects on photosynthesis were not observed during the following year when a complete recovery of photosynthetic rates was achieved with the production of new leaves. Nevertheless, the carbon budget in 2024 was also particularly low because of an excess of total ecosystem respiration compared to the long-term mean. This study is, yet, a rare example of an inversion of a forest carbon balance driven merely by meteorological conditions and it highlights the value of long-term observations to better interpret the consequences of extreme events on ecosystem functioning.

82 Inefficient Consumption of Natural Gas Drives Methane Emissions from a Megacity

Oral

Yuwei Zhao¹*, Andrew Hallward-Driemeier¹, Luke Schiferl², Trey Maddaleno³, Michael Vermeuel⁴, Dylan Millet³, Delphine Farmer⁵, Roisin Commane¹

¹Columbia University, New York City, USA. ²Lamont-Doherty Earth Observatory, Palisades, USA. ³University of Minnesota, St. Paul, USA. ⁴Purdue University, West Lafayette, USA. ⁵Colorado State University, Fort Collins, USA

Session

Session 17: Urban greenhouse gas emissions and sinks: from advanced monitoring to source identification and impact

Abstract text

Urban areas contain many anthropogenic methane sources that can be managed to mitigate climate change. The New York City metropolitan area (NYCMA) is one of the top 100 largest persistent sources of methane globally, and the largest urban methane emitter in the northeastern United States of America, yet these emissions are vastly underestimated. Using long-term tower data, we quantified monthly methane emissions and resolved the seasonal thermogenic fraction of these emissions using trace gas correlations. We observed a strong seasonal cycle in methane emissions with peaks during the heating (October to February) and cooling (June to August) seasons, with thermogenic sources dominating both seasons (98 +/- 9% to 86 +/- 13% , respectively). Incomplete combustion of natural gas (post-meter) was the dominant thermogenic signal observed, with limited influence from pre-meter natural gas. Thermogenic methane emissions were strongly correlated with natural gas deliveries over NYCMA. We calculated a natural gas methane loss rate of 1.7 +/- 0.6%, more than three times higher than expected, valued at approximately $300 million USD yr-1 in lost methane. Our results highlight the importance of accounting for year-round methane losses from inefficient combustion of natural gas and the critical need for appropriately targeted methane mitigation policies to improve combustion efficiency for end-use natural gas energy systems.

83 Changes in vegetation productivity and structure as a response to climate variability

Poster

Calyne Khamila¹*, Sadegh Jamali², Stéphanie Horion³, Tobias Biermann¹, Per-Ola Olsson¹, Torbern Tagesson¹

¹Department of Earth and Environmental Science, Lund University, Sölvegatan 12, SE-223 62, Lund, Sweden. ²Department of Technology and Society, Faculty of Engineering, Lund University, SE-221 00, Lund, Sweden. ³Department of Geosciences and Natural Resource Management (IGN), University of Copenhagen, Øster Voldgade 10, Copenhagen, Denmark

Session

Session 22: Remote sensing and flux observations to link disturbance, ecosystem states, and carbon-water fluxes

Abstract text

The increasing frequency and severity of extreme weather events over the past few decades, together with the societal demand for commodities threatens the Earth’s ecosystems. Sweden’s forest and wetland ecosystems are such an example, where the recent 2018 drought events have caused a loss in productivity and tree mortality. High demand of industrial raw materials is also driving deforestation. These factors make monitoring of forests and wetlands to track the state and conditions of ecosystem services a priority for Sweden. The main aim with this study is to analyse the response of forest and wetlands to the impacts of extreme weather events. We specifically explore the applicability of Earth observation (Sentinel-1 and 2) and a deep learning model (Bidirectional Encoder Representations from Transformers (BERT)) to map ecosystem disturbances across Sweden over the past 10 years. The maps will be evaluated against ICOS ground truth data, and thereafter be used for analysing drivers of ecosystem disturbances. 

84 Regional Inversion of New York City's CO₂ and CH₄ Emissions Using In Situ Sensors

Poster

Andrew Hallward-Driemeier¹*, Yuwei Zhao¹, Raghav Dhall², Roisin Commane^1,2

¹Columbia University, New York, USA. ²Lamont-Doherty Earth Observatory, Palisades, USA

Session

Session 18: Quantifying anthropogenic greenhouse gas emissions from continental to regional scales

Abstract text

New York City (NYC) is the largest city in the United States and ranks third among cities in the world in greenhouse gas emissions. Under Local Law 97, NYC committed to reduce greenhouse gas emissions from buildings by 40% in 2030, using a series of restrictions that took effect starting in 2024. NYC also implemented a congestion price around lower Manhattan at the start of 2025. NYC’s coastal location makes satellite retrievals difficult in the region, necessitating in situ measurements for MRV. Our group began a continuous monitoring network of CO₂ and CH₄ sensors around NYC in 2023, supplementing existing two existing sites from NOAA. Our network has grown to include 3 background sites and 9 sites within the NYC metro area, with a further 4 scheduled to come online in 2026. We sporadically deploy analyzers of CO, C2H6, and N2O to measure co-emitted species for sector analysis, and one of these sites measures ¹³CO₂ as well. We will use these measurements, along with the sparse satellite data, in regional inversions for CO₂ and CH₄ to evaluate the successes of NYC’s policy changes and identify sectors to target for further mitigation efforts.

85 Variabilité et voies de transport des gaz à effet de serre en Afrique de l'Ouest : l'étude de cas de Lamto

Poster

Touré Dro TIEMOKO*

University Nangui ABROGOUA, Abidjan, Côte d'Ivoire

Session

Session 7: CO2 and CH4 cycle dynamics in Africa: observations, processes, and climate implications

Abstract text

West Africa is particularly vulnerable to climate change, even though regional greenhouse gas (GHG) dynamics remain poorly documented. Understanding temporal variations and transport mechanisms of the main GHGs is a major scientific challenge for improving regional carbon balances and supporting climate policies. The Lamto geophysical station (Côte d'Ivoire) offers a unique opportunity to study atmospheric concentrations of CO₂, CH₄ and CO in West Africa.

This study aims to characterise the temporal variability (diurnal, seasonal and interannual) of these gases and to identify the source-sink relationships influencing their regional concentrations. Analyses are based on ten years of continuous high-precision measurements, coupled with Lagrangian air mass modelling using the FLEXPART model. A clustering approach applied to back-plumes (PES) has enabled the classification of atmospheric transport regimes and the assessment of the relative contribution of different source regions.

Results show significant seasonality, with peaks during dry season (December–February), related to biomass fires and the influence of Harmattan winds. Long-term increase rates for CO₂ (~2.24 ppm.year⁻¹) and CH₄ (~7 ppb.year⁻¹) are comparable to global trends. Cluster analysis highlights four main transport regimes, with continental air masses from the north and north-east accounting for nearly 40% of the variance in observed concentrations. The ΔCO/ΔCH₄ and ΔCO/ΔCO₂ ratios confirm the importance of combustion processes.

This work contributes to a better understanding of regional atmospheric dynamics in West Africa. Future prospects include strengthening the regional observation network to improve the quantification of GHG fluxes on an African scale.

86 Spatiotemporal dynamics of gross primary production across Europe 2014–2023 using PROBA-V and Sentinel-3 FAPAR

Poster

Mingyuan Zhang*, Lanhui Wang, Sadegh Jamali, Torbern Tagesson

Lund University, Lund, Sweden

Session

Session 22: Remote sensing and flux observations to link disturbance, ecosystem states, and carbon-water fluxes

Abstract text

Estimates of gross primary production (GPP) derived from satellite remote sensing are crucial for assessing how the terrestrial ecosystems respond to climate and other human-induced changes. While model comparison is a prerequisite for comprehensively understanding carbon dynamics, existing GPP products rely on a limited number of satellite sensors. In this study, the contemporary fraction of absorbed photosynthetically active radiation product derived from PROBA-V and Sentinel-3 was used to develop a novel GPP product (EU-GPP) based on a light use efficiency (LUE) model for the European continent. EU-GPP accounted for the distinct responses of biomes to temperature and water stresses by incorporating biome-specific environmental scalars. Evaluation against eddy covariance GPP and model comparison against other satellite-based GPP products demonstrated the applicability of EU-GPP for studying spatiotemporal carbon dynamics across Europe. The model also responded well to the drought-induced stress, indicating its ability to capture interannual variability and the impact of extreme weather events. EU-GPP highlighted increasing GPP trends in croplands during 2014–2023, underlining the recent advancements in agricultural management practices. The Mediterranean forest ecosystems exhibited weakening GPP, suggesting the strong adverse impact of summer droughts on these ecosystems. This study presents an LUE-based GPP product based on novel remote sensing data and provides an independent perspective for the monitoring of GPP across the European continent.

87 Cultivation of cattail on fen peat: A promising approach for greenhouse gas mitigation from peatlands formerly used for agriculture?

Poster

Sandra Koop*, Bärbel Tiemeyer, Philipp Köwitsch, Christian Brümmer

Thünen-Institut für Agrarklimaschutz, Braunschweig, Germany

Session

Session 15: Peatlands in the global climate system: fluxes, feedbacks, and human pressures

Abstract text

The project RoNNi (Sustainable Production and Utilisation of Cattail on Fen Sites in Lower Saxony; funding period 2023-2032) evaluates the potential of cattail (Typha spp.) paludiculture to mitigate greenhouse gas (GHG) emissions from currently drained fen peatlands in Northwestern Germany. Paludiculture is the wet cultivation of peatlands aimed at conserving the existing peat body while producing biomass. The project addresses the research gap concerning GHG emissions associated to paludiculture systems by quantifying the emissions of carbon dioxide (CO₂), and methane (CH₄) on cattail polders and CO₂, CH₄ and nitrous oxide (N₂O) emissions on drained agricultural reference sites within two different spatial clusters. For this, the eddy-covariance method and static closed chambers will be employed. High water levels in the cattail polder are expected to reduce or even eliminate CO₂ and N₂O emissions. However, water-saturated soil conditions and ponding are expected to increase CH₄ emissions. The impact of these CH₄ emissions on the overall GHG balance will be evaluated throughout the course of the project. The measurements of GHG emissions will commence in May 2026 within the first regional cluster. Measurements in the second cluster are scheduled to begin in spring 2027. The project activities further cover investigations on the effect of cattail paludiculture on the water balance, biodiversity, and nutrient dynamics, as well as economic and socioeconomic analysis of cattail production. By this, the RoNNi project comprehensively assesses the potential of cattail cultivation to mitigate GHG emissions from peatland soils, while also addressing broader environmental and (socio)economic dimensions.

88 Reducing Uncertainty in N₂O Emission Factors for Swedish Ley Systems Through Intensive Field Monitoring

Poster

Stig Sand¹*, Johannes Albertsson², Jonas Ardö¹, Tobias Biermann¹, Patrik Vestin¹

¹Lund University, Lund, Sweden. ²Swedish University of Agricultural Sciences, Alnarp, Sweden

Session

Session 14: Greenhouse gas fluxes in agroecosystems: processes, measurements and management implications

Abstract text

Approximately 38% of Swedish agricultural greenhouse‑gas emissions originate from soils, with nitrous oxide (N₂O) being the dominant contributor. The spatial and temporal variability of N₂O emissions is large, yet measurement data remain scarce and key processes remain insufficiently constrained. Swedish ley systems, consisting of grass- and legume-dominated forage rotations, play a central role in national livestock production. Within these systems, N₂O emissions are highly variable due to the strong interactions between management practices and soil biogeochemical processes. Variations in soil moisture and species composition reshape nitrogen mineralization and nitrification–denitrification, generating both short‑term N₂O pulses and longer‑term flux variability. These brief but intense events, often triggered by rainfall or soil thaw, are easily missed by low‑frequency monitoring, increasing uncertainty in national inventories and limiting the data needed to move beyond default IPCC Tier 1 emission factors. More representative measurements are therefore essential for accurately capturing the magnitude and timing of N₂O fluxes in these systems.

To address these gaps, we aim to better quantify and constrain soil‑derived N₂O emissions in ley systems in southern Sweden, thereby providing data that may improve Sweden’s national greenhouse‑gas inventory. Here, we combine manual static chamber measurements with a LI‑7820 N₂O analyzer to monitor fluxes across ley phases and management events at four field sites. Preliminary data show large spatial variability within and between sites, as well as pronounced emission pulses following fertilization and soil thaw.

89 Temporal complexity of ecosystem functioning: chaotic or periodic carbon fluxes?

Oral

Marcos Fernández-Martínez¹*, Ivan Janssens², Michael Obersteiner³, Peter Manning⁴, Filipe Andrade⁵, Eladio Rodríguez-Penedo¹, Josep Peñuelas⁶

¹CREAF, Bellaterra, Spain. ²Universiteit Antwerpen, Wilrijk, Belgium. ³Environmental Change Institute, University of Oxford, Oxford, United Kingdom. ⁴University of Bergen, Bergen, Norway. ⁵CREAF, Belatterra, Spain. ⁶CREAF-CSIC, Bellaterra, Spain

Session

Session 22: Remote sensing and flux observations to link disturbance, ecosystem states, and carbon-water fluxes

Abstract text

The development of non-linear dynamics and chaos theory elucidated that simple processes could lead to high complexity in the functioning of nature. Recent studies have shown that non-linear dynamics are common across populations of very different taxa. However, whether the energy and matter fluxes of entire ecosystems follow non-linear dynamics, such as ecosystem carbon (C) fluxes, and how complex these dynamics are, is still unknown. We here take a novel approach to looking at C fluxes from 57 ecosystems around the world to investigate whether they display periodic, low- or high-dimensional chaotic, or a random temporal behaviour by means of the correlation dimension. Hence, we assessed the temporal complexity of ecosystem functioning, and investigated its drivers and trends over time. Ecosystems with higher temporal complexity exhibited lower interannual variability and seasonality. Causal analyses indicated that ecosystems with larger C fluxes generally cause higher temporal complexity, and larger and temporally complex C fluxes contribute to reducing interannual variability, potentially indicating higher resistance to perturbations. Our results showed a positive trend in the complexity of GPP over time, which correlates with increasing annual GPP. Although the increase in GPP temporal complexity is on average very small (<0.5 degrees of freedom over 20 years), this result may indicate that ecosystems are increasingly responsive to endogenous or exogenous stimuli. Our results indicate non-linear dynamics are present in C fluxes, and that the short-term temporal complexity of ecosystem functioning can be informative about ecosystem properties otherwise missed by longer timescales.

90 CARIMED (CARbon, tracers, and ancillary data In the MEDiterranean Sea): A ship-based data synthesis product – overview and quality control procedures

Orals

Marta Álvarez¹*, Maribel García-Ibáñez², Nico Lange³, Alex Kozyr⁴, Antón Velo⁵, Toste Tanhua⁶

¹IEO-CSIC, A Coruña, Spain. ²IEO-CSIC, Palma de Mallorca, Spain. ³NORCE, Bergen, Norway. ⁴OCADS-NOAA, Silver Spring, USA. ⁵IIM-CSIC, Vigo, Spain. ⁶GEOMAR, Kiel, Germany

Session

Session 6: Carbon cycle in the Mediterranean region: from the local to the regional scale

Abstract text

The Mediterranean Sea (MedSea) is highly sensitive to climate-driven changes in temperature, oxygen, and pH, among other variables. To better assess these long-term trends, we developed CARIMED (CARbon, tracers, and ancillary data In the MEDiterranean Sea), the first comprehensive, harmonised data synthesis product for the MedSea. CARIMED integrates hydrographic, inorganic carbon, transient tracer, and ancillary measurements from 46 research cruises spanning the period from 1976 to 2018, containing observations for the entire water column across all MedSea sub-basins. A substantial component of the data was retrieved from fragmented or locally archived historical records, thus consolidating previously inaccessible measurements. Following global synthesis approaches, CARIMED applies a quality-controlled, and bias-adjusted framework. A key adaptation was the secondary quality control (2QC) procedure, specifically tailored to the MedSea's unique hydrography, utilising sub-basin divisions and supplementary checks (including statistical consistency assessments) to resolve complex, often contradictory, inter-cruise offsets. This rigorous process minimised systematic biases, yielding a dataset with improved consistency, and highlights the urgent need for adapted standard operating procedures and reference materials to address the MedSea biogeochemical particularities. CARIMED delivers two complementary, freely available products: the aggregated original cruise data product and the final bias-adjusted data synthesis product. This essential resource establishes a new benchmark for assessing long-term biogeochemical trends, validating regional ocean models, and supporting climate-change mitigation and adaptation strategies in this rapidly changing semi-enclosed basin.

Additional Authors: Civitarese, Cantoni, Belgacem, Schroeder, Acerbi, Coppola, Wagener, Fajar, Flecha, Giani, Giannoudi, Guallart, Hassoun, Huertas, Ibello, Keraghel, Louanchi, Luchetta, Pérez, Schirnick, Souvermezoglou, Urbini, Vidal & Ziveri

91 Modelling forest carbon dynamics under Climate Change across European forests

Poster

Gonzalo Oton¹*, Viorel Blujdea¹, Matteo Zampieri², Paul Rougieux³, Roberto Pilli², Erone Ghizoni Santos¹, Alessandro Cescatti¹, Mirco Migliavacca¹, Giacomo Grassi¹

¹European Commission, Joint Research Centre (JRC), Ispra, Italy. ²Consultant with the European Commission, Joint Research Center (JRC), Ispra, Italy. ³EIchen Lernen Data Science Consulting, Backnang, Germany

Session

Session 11: Climate feedbacks from ecosystem management and natural disturbances

Abstract text

The European Union aims to become the first carbon-neutral continent by 2050. Forest resources play a crucial role in EU climate, energy, environment, and bioeconomy policies. Achieving this goal requires robust technological and ecosystem-based carbon sinks, a challenge made increasingly complex by the accelerating effects of climate change. Reliable models are needed to evaluate pathways toward the 2050 target, capturing both short- and medium-term forestry processes and long-term climate patterns. 

The JRC Forest Carbon Model adapted for European forests as EU-CBM-HAT, accurately monitors carbon dynamics and reports on forest carbon balance. This model provides country- and regional-scale insights into forestry indicators and carbon dynamics under forest management scenarios; however, climate variability is not considered, limiting its applicability for long-term projections. To address long-term climate sensitivity in projections, a new climate module has been linked to the model, enabling scenarios that adjust biomass growth projections. Climate adjustments are introduced through interannual variability using Net Primary Production (NPP) scenarios from ISIMIP3b (Classic Model), under the assumption that NPP is strongly correlated with climate conditions. Model performance was assessed through intercomparison with TRENDY project outputs and MODIS data during the calibration period 2010-2020. The model was then calibrated over the historical period and used to evaluate forest dynamics and carbon balance across EU forests under projected climate conditions (SSP1-2.6 and SSP3-7.0) through 2100. 

High consistency was found across NPP models and between ISIMIP-ingested and simulated NPP values. This framework offers a robust foundation for including climate change impacts on NPP into CBM.

92 Surface Slicks Decouple the Air–Sea CO₂ Gradient and Bias Flux Estimates

Oral

Mariana Ribas Ribas*, Oliver Wurl, Edgar Fernando Cortés-Espinoza

Carl von Ossietzky Universität Oldenburg, Wilhelmshaven, Germany

Session

Session 5: Advancing marine CO₂ observations through next-generation sensors, integration and platform innovation

Abstract text

Accurate quantification of air–sea carbon dioxide (CO₂) exchange requires resolving small-scale variability at the ocean–atmosphere interface. Still, most flux calculations assume vertically well-mixed boundary layers and rely on CO₂ partial pressures (pCO₂) measured meters above and below the sea surface. Here, we present results from a July 2024 field campaign in the German North Sea, where we deployed two observing platforms to examine how near-surface stratification influences CO₂ flux estimates. We measured the difference between oceanic and atmospheric pCO₂ (ΔpCO₂) at 100 cm above and below the surface (meter-scale interface) and at 20 cm (centimeter-scale interface), allowing a direct comparison between traditional and high-resolution methods.

Slicks are surface phenomena where capillary waves are damped by accumulated organic carbon. We identified Non-slick and Slick Scenarios with clear differences: under Non-slick conditions, both ΔpCO₂ approaches were similar, producing comparable CO₂ fluxes. In contrast, Slick Scenarios caused a persistent increase of more than 100 µatm in pCO₂ at 20 cm above the surface, decoupling the near-surface atmosphere from the meter-scale gradient. As a result, near-surface ΔpCO₂ exceeded the meter-scale estimate by more than 100 µatm, leading to significant differences in calculated CO₂ fluxes (up to −35.6 mmol m⁻² d⁻¹), which influenced both magnitude and direction. These findings show that slick-induced stratification can bias meter-scale flux calculations and emphasize the importance of centimeter-scale, high-frequency CO₂ measurements using innovative, non-intrusive sensors and integrated observing platforms. Accurately resolving near-surface ΔpCO₂ is essential for improving flux estimates and advancing next-generation marine CO₂ monitoring systems.

93 A global 14C budget 1954-2023 based on atmospheric and surface-ocean data

Oral

Christian Rödenbeck¹*, Liam Blyth², Maksym Gachkivskyi³, Heather Graven², Thomas Laemmel⁴, Samuel Hammer³, Martin Heimann¹, Quan Hua⁵, Ralph F. Keeling⁶, Fabian Maier¹, Nabir Mamnun², Tobias Naegler³, Susanne Preunkert³, Carlos A. Sierra¹, Alexander Winkler¹, Sönke Zaehle¹

¹MPI BGC, Jena, Germany. ²Imperial, London, United Kingdom. ³UHEI, Heidelberg, Germany. ⁴UNIBE, Bern, Switzerland. ⁵ANSTO, Sydney, Australia. ⁶SIO, San Diego, USA

Session

Session 18: Quantifying anthropogenic greenhouse gas emissions from continental to regional scales

Abstract text

Measurements of radiocarbon (14C) in atmospheric CO2 or in dissolved inorganic carbon in the ocean provide various pieces of information about the global cycling of carbon. Such information includes the amount of CO2 emitted by anthropogenic burning of fossil fuels (exploiting the Suess effect from their being devoid of radiocarbon), or the turn-over of carbon in the terrestrial biosphere (exploiting the 14C pulse emitted into the atmosphere by nuclear bomb testing mainly in the 1960s). Due to the spatial and temporal sparsity of the atmospheric and oceanic 14C measurements, however, interpolation and budgeting approaches are needed to derive estimates resolved in space and time. Here we present a combination of (1) a global atmospheric 14C inversion and (2) a surface-ocean data interpolation, that yield data-based estimates of 14C fluxes and continuous atmospheric Delta14C fields over 1954-2023. As an application, these fields can be used as background or boundary fields in regional atmospheric inversions estimating fossil fuel CO2 emissions from atmospheric Delta14C data.

94 Constraining Urban Methane Sources in London using δ¹³C-CH₄ and Δ¹⁴C-CH₄ Measurements

Oral

Xinran Yang*, Heather Graven, Fang Liu

Imperial College London, London, United Kingdom

Session

Session 17: Urban greenhouse gas emissions and sinks: from advanced monitoring to source identification and impact

Abstract text

Urban areas contribute 21% of global anthropogenic methane emissions, and for waste, energy, and transport sectors, over 40% of global methane emissions originate from cities. According to the UK’s National Atmospheric Emissions Inventory (NAEI), London comprises only 0.65% of the UK’s land area, yet accounts for 2.7% of the UK’s total annual methane emissions and 9.1% of the UK’s fugitive methane emissions (e.g., natural gas leaks). However, the attribution of urban methane emissions to specific source sectors remains debated, and bottom-up inventory estimates often disagree with atmospheric observations. Multiple studies have suggested that fossil methane emissions in London are significantly underestimated in national inventories, highlighting persistent uncertainty in the attribution of fossil and biogenic methane sources.

Here, we present atmospheric CH₄ and δ¹³C-CH₄ measurements (2018–2026) and Δ¹⁴C-CH₄ measurements (2022–2026) from Imperial College London in central London to assess urban methane sources and their temporal evolution. We identify individual pollution events over the measurement period and apply Keeling plot analysis with δ¹³C-CH₄ to evaluate source contributions. Δ¹⁴C-CH₄ measurements are used to calculate the fossil fraction and are compared with simulated values from the Numerical Atmospheric-dispersion Modelling Environment (NAME) dispersion model coupled with the UK NAEI inventory and the Emissions Database for Global Atmospheric Research (EDGAR) global inventory. Our results provide updated constraints on methane source partitioning in London, evaluate the consistency of bottom-up estimates with atmospheric observations, and inform urban methane mitigation policy.

95 Future Methane Isotope Composition Changes as Indicators of Global Methane Source Changes Under Shared Socioeconomic Pathways

Poster

Xinran Yang¹*, Ryo Fujita², Heather Graven¹

¹Imperial College London, London, United Kingdom. ²Meteorological Research Institute, Tsukuba, Japan

Session

Session 26: Designing the ideal global greenhouse gas monitoring network

Abstract text

Methane (CH₄)—the second most important anthropogenic greenhouse gas—has strong potential for mitigating climate change through emission reductions due to its shorter atmospheric lifetime. The isotopic composition of CH₄ has changed over the industrial period as a result of CH₄ emissions from human activities. Scenarios of future human activities are widely used to explore potential climate and environmental changes, but future methane isotopic composition has not yet been simulated. Here, we conduct future methane isotope simulations for 2015–2050 using emission scenarios based on the Shared Socioeconomic Pathways (SSPs). We apply a simple box model whose parameters are calibrated to historical atmospheric CH₄ and isotopic observations and evaluate the model performance against atmospheric isotope measurements over the past decade. We examine the contributions of individual CH₄ source sectors to past and future atmospheric isotopic trends.

Our results show that atmospheric Δ¹⁴C-CH₄ and δD-CH₄ diverge across emission scenarios after 2020, whereas atmospheric δ¹³C-CH₄ shows limited sensitivity to scenario differences until around 2030. Modelled trends agree with observations over 2015–2024. Δ¹⁴C-CH₄ responds strongly to changes in fossil and nuclear power plant emissions, δD-CH₄ is sensitive to oxidation by OH sink, and δ¹³C-CH₄ primarily reflects biogenic versus fossil source contributions. Overall, our results suggest that long-term atmospheric monitoring of multi-isotopic CH₄ trends has strong potential to verify the influence of individual CH₄ source sectors on atmospheric methane growth. These results have implications for the long-term development of global methane monitoring strategies under changing future climates.

96 Improving the basis for assessing high-latitude forest carbon dynamics under climate change using forest inventory data in Sweden

Poster

Wenquan Dong¹*, Mengyuan Mu¹, Stefan Olin¹, Karl Piltz¹, Jonas Fridman², Thomas Pugh^1,3

¹Lund University, Lund, Sweden. ²Swedish University of Agricultural Sciences, Umeå, Sweden. ³University of Birmingham, Birmingham, United Kingdom

Session

Session 9: Carbon processes in high latitude/high altitude ecosystems under climate change

Abstract text

High-latitude forests store substantial amounts of carbon and constitute a key component of the global carbon cycle. However, assessing their vulnerability to climate change remains challenging, as large-scale estimates of forest carbon fluxes rely heavily on process-based models that often suffer from biases in key demographic processes, particularly tree growth and mortality. Such biases limit our ability to robustly quantify how boreal ecosystems may shift between carbon sink and source under changing climatic conditions.

Here, we present a data assimilation framework that integrates multi-census forest inventory data with the LPJ-GUESS dynamic vegetation model to reduce these biases and provide a stronger basis for analysing climate-driven changes in high-latitude forest carbon dynamics. This framework combines a state initialisation informed by observed forest structure from the earliest inventory census with rich observations at each grid, with a two-stage assimilation approach based on the Land Variational Ensemble Data Assimilation Framework (LAVENDAR) to bias-correct forest growth and mortality processes.

Applied to Swedish forests, a representative of the high-latitude forest, our framework significantly improves the simulated growth increments and mortality rates temporally and spatially. By enabling regionally and temporally adaptive parameterisation, our approach enhances the robustness of model-based assessments of forest carbon dynamics under a changing climate, particularly where growth and mortality processes may become more variable.

97 New insights on the production, biodegradation and recalcitrance of organic carbon from cultivated Saccharina latissima in Norway

Oral

Luiza Neves¹*, Sebastian Aker², David Aldridge¹, Deni Ribičić¹, Anders Brunsvik³, Ole Jacob Broch¹, Jorunn Skjermo¹, Murat Van Ardelan²

¹SINTEF Ocean, Trondheim, Norway. ²NTNU, Trondheim, Norway. ³SINTEF Industry, Trondheim, Norway

Session

Session 3: Blue carbon and seaweed: reforestation and cultivation

Abstract text

The production of recalcitrant dissolved organic carbon (RDOC) through bacterial transformation during kelp decomposition is a critical contributor to carbon sequestration from seaweed systems, and a major knowledge gap. We investigate the production and bioavailability of particulate (POC) and dissolved organic carbon (DOC) from Saccharina latissima, a primary candidate for cultivation in Europe, with a focus on RDOC analysis by FT-ICR MS.  DOC concentrations stabilise after 150 days and an average 58 % from initial levels remained as RDOC after 188 days of biodegradation. Photosynthates produced by the kelp and leaked in the DOC were predominantly aliphatic compounds (labile) containing nitrogen, sulphur or both (CHON, CHOS and CHONS), becoming available for consumption and transformation by bacteria. 1,805 new molecular formulas were produced by microbial transformation of the DOC during biodegradation (34 % of total). Aromatic and condensed aromatic compounds (RDOC), although low in quantities, were higher in kelp than control treatments. DOC and RDOC production lack seasonal data, as well as standardization in sampling protocols and analysis. Despite this, results presented may help inform carbon verification frameworks for marine CDR with seaweeds. Cultivated S. latissima provides an important role in the biological carbon pump through the continuous release of POC and DOC during grow-out at sea and “passive” CO₂ sequestration in the oceans via the production of RDOC. Besides the multiple uses for the harvested biomass, valuation of RDOC could help boost seaweed cultivation practices as a nature-based solution for climate change adaptation and mitigation, particularly at larger production scales. 

98 Assessing climatic and soil variables of bare soil albedo using comparison of ICOS cropland sites

Poster

Kyllian Duhoux*, Emmanuelle Vaudour, Carmen Kalalian, Ronny Lauerwald, Benjamin Loubet

UMR ECOSYS, INRAE AgroParisTech, Université Paris-Saclay, Palaiseau, France

Session

Session 11: Climate feedbacks from ecosystem management and natural disturbances

Abstract text

Albedo, is the solar reflectivity of a surface, i.e., its capacity to reflect incident shortwave radiation from the sun (visible and near and shortwave-infrared ranges). At ground level, any solar energy not absorbed is reflected. This reflected portion of the incident solar energy is called « albedo ». In agricultural systems, bare-soil albedo plays a vital role in the energy and water balance, and thus for the regulation of climate and micro-climate in the short-and medium term.

Albedo is influenced by numerous factors, including climatic (e.g., snow coverage), topographic (e.g., slope) or pedologic (e.g., soil type and soil composition, such as topsoil clay), and management (e.g., tillage) factors. Soil albedo is strongly controlled by the topsoil colour, which depends on mineral composition, (e.g. bright carbonates vs dark metal oxides) and organic matter content. It further depends on surface roughness which can be intentionally influenced by human activity, in particular tillage or no-tillage practices. Moreover, application of organic amendments like biochar, used for climate mitigation, significantly impact soil reflectance and albedo.

To better understand the influence of these driving factors, this study proposes a comparison between several cropland sites within the ICOS ecosystem network. Taking advantages of the dataset’s diversity in climatic and pedological properties, this study aims to determine the sensitivity of bare-soil albedo to a range of climatic and soil factors, focusing on bare-soil periods following tillage. We employed univariate and multivariate analyses to determine the albedo response to influential variables, and, finally, to evaluate multifactorial bare-soil albedo models.

99 Tundra in Flux as Borealization Reshapes Carbon Balance

Oral

Gangotri Chattopadhyay¹*, Anna Virkkala², Miska Luoto¹, Heidi Mod¹

¹University of Helsinki, Helsinki, Finland. ²Finnish Meteorological Institute, Helsinki, Finland

Session

Session 9: Carbon processes in high latitude/high altitude ecosystems under climate change

Abstract text

As boreal vegetation advances northward, Arctic tundra ecosystems are undergoing reorganization with major impacts on carbon cycling. Yet, mechanisms linking fine-scale vegetation changes to ecosystem CO₂ exchange remain poorly understood, limiting predictive capacity in carbon cycle models. Thus, here, we investigated how variation in tundra plant community structure regulates summertime fluxes across a borealizing landscape in northern Finland. We measured CO₂ exchange across 211 plots over three consecutive summers using chamber techniques to quantify productivity (GPP), ecosystem respiration (ER), and net ecosystem exchange (NEE). We then related these fluxes to vascular plant growth forms, cover and richness, and community-weighted functional traits while accounting for key abiotic drivers. Vascular plant growth form was the dominant control of flux variability (R² ≈ 0.25–0.34), with deciduous tall shrub communities showing the highest GPP and greatest NEE (i.e. carbon uptake) but also elevated ER, indicating accelerated carbon turnover with deciduous shrub expansion. Evergreen shrub plots exhibited lower and more variable NEE. Vegetation height and density (green fraction) also strongly explained fluxes, primarily enhancing photosynthetic capacity and associated respiration through vegetation-mediated effects on light and soil conditions. Among the functional traits, high leaf nitrogen content explained the greatest variability in fluxes (R² = 0.29 for GPP; 0.27 for NEE). Soil microclimate effects were comparatively weak, suggesting vegetation structure can outweigh local abiotic constraints during peak growing season. Our results indicate that tundra shrubification may increase carbon turnover, highlighting vegetation structure and functional traits as key regulators of high-latitude carbon dynamics under climate change.

100 The Dahra field site- long term monitoring of ecosystem characteristics of a semi-arid Savanna in Senegal

Poster

Torbern Tagesson¹*, Aleksander Wieckowski², Ousmane Ndiaye³, Ousmane Diatta³, Elhadji Mamadou SONKO⁴, Rasmus Fensholt⁵, Jonas Ardö²

¹Lund, Lund, Sweden. ²Lund University, Lund, Sweden. ³Centre Research Zootecnique, Dahra, Senegal. ⁴Universite de Cheikh Anta Diop de Dakar, Dakar, Senegal. ⁵University of Copenhagen, Copenhagen, Denmark

Session

Session 7: CO2 and CH4 cycle dynamics in Africa: observations, processes, and climate implications

Abstract text

The Sahel is a semi-arid biogeographic transition zone between the dry Sahara Desert in the North and the humid Soudanian Savannah in the south, and it is one of the world’s largest semi-arid regions. The region experiences a strong population growth, increasing the demand on the ecosystem services due to cropland expansion, increased pasture stocking rates and fuelwood extraction. In situ data are necessary to assess dynamic responses of ecosystems to changing environmental conditions, such as impact of anthropogenic management and climate changes. In addition, they are important for the parameterization and evaluation of remote sensing products and dynamic vegetation models. Currently, there are very few sites in the tropics, and in Africa in particular, with long-term measurements of ecosystem characteristics. Hence, there is a continuous lack of field data, and thereby a strong scientific need for calibration and validation data from this part of the world. The Dahra field site in Senegal is one of the sites with the longest ongoing time series of measurements of ecosystem characteristics across Africa, and here we are collecting data of herbaceous and tree species composition, tree structural parameters, edaphic conditions, herbaceous biomass, leaf are index, albedo, vegetation greenness, solar radiation absorption, hydrometeorological conditions, and land-atmosphere interactions (heat, water, carbon dioxide, methane and nitrous oxide fluxes). Here, we will present research conducted with these data. 

101 A Decade of Continuous Urban Greenhouse Gas Monitoring: CO₂, CO, and CH₄ Mole Fractions and Emissions Estimates from a Dense Tower-Based Observational Network in Indianapolis, USA

Oral

Zachary Barkley*, Kenneth Davis, Natasha Miles, Scott Richardson, James Marlow, Kevin Fletcher

Penn State University, University Park, USA

Session

Session 17: Urban greenhouse gas emissions and sinks: from advanced monitoring to source identification and impact

Abstract text

Urban greenhouse gas (GHG) monitoring is essential for achieving and enforcing climate policy objectives, particularly in cities where emissions are both highly concentrated and varied in source. Urban tower networks offer a robust approach for continuous, long-term monitoring of atmospheric composition and emissions. Indianapolis, USA, hosted one of the longest-running and densest urban observational tower networks as part of the Indianapolis Flux Experiment (INFLUX), with 15 sites measuring CO₂, CO, and CH₄ from 2011 to 2025. Here we present an analysis of this extensive dataset, deriving atmospheric enhancements and citywide emission rate estimates for all three gases from 2014-2023. CO₂ emissions show a clear decline beginning in 2015, coincident with the conversion of the Harding Street powerplant from coal to natural gas. A second, transient decrease in CO₂ is observed during the onset of the COVID-19 pandemic in 2020. Variations in CO during the pandemic closely track those in CO₂, supporting attribution of the reductions primarily to decreased vehicle emissions. In contrast, CH₄ emissions remain consistent throughout the record and agree with independently developed bottom-up inventories, though sectoral attributions vary greatly compared to national inventory estimates. CH4 emissions from leaks in the natural gas distribution sector are found to be equal to ~2% of throughput despite pre-meter distribution infrastructure being composed of plastic, leak-tight material. The results from this project demonstrate the capability of dense urban tower networks to detect sector-specific emission changes, evaluate bottom-up inventories, and provide an independent, observation-based framework for monitoring citywide GHG emissions and policy-relevant trends.

102 Incorporating Management and Canopy Information into Machine‑Learning Gap‑Filling of Eddy‑Covariance NEE in Irish Grasslands

Oral

Enola Barvaeva¹*, Katarina Domijan¹, Andrew Parnell², Rachael Murphy³

¹Maynooth University, Maynooth, Ireland. ²University College Dublin, Dublin, Ireland. ³Teagasc, Wexford, Ireland

Session

Session 14: Greenhouse gas fluxes in agroecosystems: processes, measurements and management implications

Abstract text

In intensively managed grasslands, eddy-covariance records of net ecosystem exchange (NEE) often contain gaps filled with models using only weather and time. We test whether adding grassland management and canopy information to machine-learning gap-filling models improves half-hourly NEE estimates. Half-hourly NEE data from two eddy-covariance towers in rotationally grazed dairy systems at Johnstown Castle, Ireland, are analysed. After quality control, two predictor sets are defined: (1) a baseline with photosynthetic photon flux density (PPFD), meteorological variables (air temperature, humidity, rainfall, global radiation, vapour pressure deficit) and time encodings; and (2) an augmented set that also includes repeated sward height and dry-matter biomass measurements, a nitrogen (N) application variable, days-since-management indicators, and a phytomass index from day–night NEE differences, interpolated across the artificial gaps. Random Forest, XGBoost and multilayer perceptron models are tested in an artificial-gap experiment with non-overlapping missing blocks from hours to several weeks, training on data outside each block and predicting NEE within it. Performance is evaluated using mean absolute error (MAE), root mean square error (RMSE) and coefficient of determination (R²). Across both sites, gap lengths and models, the augmented predictors improve gap filling over the weather-only baseline. Improvements are small for short gaps but grow for multi-day and multi-week gaps, especially within 30 days after grazing, when management and canopy effects on NEE are strongest. These findings argue for routinely including management-aware predictors in flux networks and modelling to reduce gap-filling uncertainty and better capture short-term management impacts on ecosystem–atmosphere carbon exchange.

103 Potential soil humidity control on sub-alpine grassland annual net ecosystem exchange - a preliminary comparison of FR-Clt with similar ecosystem stations

Poster

Alise Robert¹, Didier Voisin¹*, Catherine Coulaud², Hélène Barral³, Jean-Martial Cohard¹, Nina Buchmann⁴, Kukka-Maaria Kohonen⁵, Marta Galvagno⁶, Michele Eugenio d'Amico⁷, Nicolas Bonfanti⁸, Philippe Choler⁹

¹Université Grenoble Alpes, Institut des Géosciences de l'Environnement, Grenoble, France. ²Université Grenoble Alpes, CNRS, Institut des Géosciences de l'Environnement, Grenoble, France. ³Université Grenoble Alpes, IRD, Institut des Géosciences de l'Environnement, Grenoble, France. ⁴ETH Zurich, Department of Environmental Systems Science, Zurich, Switzerland. ⁵ETH Zurich, Department of Environmental Systems Science, Zurich, France. ⁶Environmental Protection Agency of Aosta Valley, Climate Change Unit, Aosta, Italy. ⁷Milano University, Department of Agricultural and Environmental Sciences, Milano, Italy. ⁸Université Savoie Mont-Blanc, Edytem, Chambéry, France. ⁹Université Grenoble Alpes, CNRS, Laboratoire d'Ecologie Alpine, Grenoble, France

Session

Session 10: Cross-scale responses of greenhouse gas variability to climate extremes

Abstract text

Low temperatures in sub/alpine and polar regions limit carbon mineralization, favoring soil organic matter accumulation. Climate warming threatens this balance, particularly in the Alps’ subalpine-to-alpine transition, where soil carbon destabilization risks are highest. Yet, flux measurement stations in this critical altitude band (around 2000 m a.s.l.) remain scarce, limiting our understanding of ecosystem-atmosphere exchanges.

This study compares annual net ecosystem CO₂ exchange (NEE) and its drivers at three sub-alpine grassland stations (FR-Clt, IT-Tor, CH-Aws) using ICOS/FluxNet data (2020–2025). Despite similar climatic conditions—radiation, temperature, and snow cover duration—FR-Clt uniquely exhibits consistent annual net CO₂ emissions, while IT-Tor and CH-Aws act as CO₂ sinks. The key difference lies in summer soil water content (SWC): at FR-Clt, SWC drops below a 30% volumetric threshold every July, triggering a shift from uptake to emission. This pattern is absent at the other stations, where SWC remains above the threshold and CO₂ uptake persists until August. The earlier onset of water stress at FR-Clt curtails its CO₂ sink period, likely explaining its distinct annual NEE.

These findings highlight the potential vulnerability of sub-alpine grasslands to water-limited conditions under climate change.

This presentation will (1) contrast the stations’ NEE budgets, (2) stress SWC’s pivotal role, and (3) discuss implications for carbon cycling in warming mountain ecosystems.

104 Methane emissions from coal mines in New South Wales, Australia

Poster

Ida Jandl^1,2*, Nasimeh Shahrokhi², Cathy Trudinger², Peter Rayner³, Nicholas Deutscher⁴, Matt Garthwaite⁵, Hassan Nawaz⁴, Nicholas Jones⁴, Robyn Schofield¹

¹School of Geography, Earth, and Atmospheric Sciences, University of Melbourne, Melbourne, Australia. ²CSIRO Environment, Aspendale, Australia. ³The Superpower Institute, Melbourne, Australia. ⁴School of Science, and Environmental Futures Research Centre, University of Wollongong, Wollongong, Australia. ⁵CSIRO Space and Astronomy, Canberra, Australia

Session

Session 18: Quantifying anthropogenic greenhouse gas emissions from continental to regional scales

Abstract text

Coal mines are responsible for 10% of the global anthropogenic methane emissions, yet their emission estimates remain significantly uncertain. Therefore, the Intergovernmental Panel on Climate Change (IPCC) highlights the importance of independent validation of methane inventories estimated by coal mine facilities.

Coal mining regions usually exhibit complex topography with a cluster of multiple coal mining pits. For example, in the Hunter Valley of New South Wales (NSW), Australia, there are more than 20 coal mines within an area of about 3600 km2, with the largest open pit mine measuring more than 9 km in width. Currently available gridded inventories (EDGAR, Open Methane) are based on emissions from point locations, but they do not accurately reflect the spatial characteristics of the emissions of the open pit coal mines. Here, we develop a gridded 2 km x 2 km inventory for NSW based on Australia’s national reporting for 2024 using polygon shapes to identify the accurate locations of the emissions from the mine pits.

Our gridded inventory is used as forcing for our offline coupled WRF-CMAQ transport model to simulate methane mole fractions in the atmosphere. We compare the simulated methane mole fractions to satellite (TROPOMI) and ground-based (EM27/SUN) observations. This study lays the foundation for a 4D-Var inversion framework to quantify methane emissions in the New South Wales coal mining region.

105 Photoprotection dominates over water stress in delaying spring photosynthesis resumption in cold-region evergreen forests

Poster

Yunpeng Luo*

Swiss Federal Institute for Forest, Snow and Landscape Research WSL, Birmensdorf, Switzerland

Session

Session 9: Carbon processes in high latitude/high altitude ecosystems under climate change

Abstract text

Spring photosynthetic recovery in cold‐region evergreen needleleaf forests is often delayed relative to model predictions, but the physiological drivers of this mismatch remain unclear, particularly the relative importance of cold‐induced photoprotection versus hydraulic limitation for constraining carbon uptake. We integrated coordinated leaf-, canopy-, and ecosystem‐scale measurements with modeling analyses at two Picea abies forests in Germany and Switzerland that experienced contrasting winter–spring climate conditions to diagnose controls on gross primary productivity (GPP) recovery and light‐use efficiency (LUE). Spring GPP recovery was strongly delayed at the colder Swiss site but not at the warmer German site. The delay was primarily explained by enhanced photoprotective pigment accumulation and reduced LUE, while hydraulic constraints were minor. Photochemical reflectance index (PRI) captured cold‐induced LUE reductions and reduced model overestimation of spring GPP by ~42%, although PRI–LUE coupling weakened below −5 °C. Photoprotection is the dominant physiological constraint on spring photosynthetic recovery in cold evergreen forests. Incorporating mechanistic, observation‐based representations of photoprotection into ecosystem and Earth system models can substantially improve predictions of forest carbon uptake under a warming and increasingly variable climate.

106 Feasibility of virtual tall towers approaches for estimating CO₂ mixing ratio in the atmospheric boundary layer

Poster

Lediane Marcon-Henge¹*, Alexander Graf¹, Matthias Peichl²

¹Institute of Bio- and Geosciences: Agrosphere (IBG-3), Research Centre Jülich, Jülich, Germany. ²Swedish University of Agricultural Sciences, Department of Forest Ecology and Management, Umea, Sweden

Session

Session 29: Using GHG measurement data to support national greenhouse gas inventories

Abstract text

Applications such as atmospheric modelling, carbon-budget estimation, quantification of greenhouse gas (GHG) sources and sinks, and understanding the processes driving these gases in the atmosphere rely on a comprehensive network of atmospheric observations. Virtual Tall Tower (VTT) approaches can expand GHG observations by scaling existing ecosystem measurements to tall-tower heights (from 50 m to >100 m). We evaluated the feasibility of two VTT approaches for estimating boundary-layer CO₂ concentrations from eddy covariance (EC) measurements. The main dataset was the combined ecosystem–atmosphere station Svartberget in northern Sweden (SE-Svb, SVB), a part of the ICOS (Integrated Carbon Observation System) network; other ICOS sites served as supplementary examples. Well-calibrated  EC measurements appeared to be the most crucial and challenging requirement. The VTT method described by Haszpra et al. (2015) performed slightly better than our newly proposed VTT, but produced about three times fewer valid data points for Svartberget. In a sensitivity analysis of the former method, we identified planetary boundary-layer height as one of the most sensitive variables affecting mixing-ratio estimates. Furthermore, the largest biases when comparing VTT estimates to tall-tower measurements occurred in the morning. Our newly proposed VTT is based on direct inspection of the turbulence-resolving high-frequncy measurements from the  EC system. In initial tests across different set-ups, we observed that vertical wind information contributes to improved CO₂ estimates. Although further evaluation of both VTT methods is required, integrating calibration routines into existing  EC stations is a first step toward implementing VTT to expand the network of atmospheric observations.

107 Comparison of top-down and bottom-up greenhouse gas emission estimates with national inventory submissions: highlights from five user stories (AVENGERS)

Poster

Roxana Petrescu¹*, Sander Houweling², Marko Scholze³, Christian Mielke⁴, Hugo Denier van der Gon⁵, Margreet van Zanten⁶, Dominik Brunner⁷, Anna-Maria Jönsson³, Mattias Lundblad⁸, Angela Fiore⁹, Thomas Kaminski¹⁰, Andre Butz¹¹, Lara Abou Chehade¹², Alex Vermeulen¹³, Ioannis Cheliotis², Joël Thanwerdas⁷, Arjo Segers⁵, Yohanna Villalobos³, and all members of the AVENGERS Consortium³

¹CMCC, Viterbo, Italy. ²VU Amsterdam, Amsterdam, Netherlands. ³Lund University, Lund, Sweden. ⁴UBA, Dessau, Germany. ⁵TNO, Utrecht, Netherlands. ⁶RIVM, Bilthoven, Netherlands. ⁷EMPA, Dubendorf, Switzerland. ⁸SLU, Uppsala, Sweden. ⁹ISPRA, Rome, Italy. ¹⁰iLab, Hamburg, Germany. ¹¹Uni Heidelberg, Heidelberg, Germany. ¹²CMCC, Sassari, Italy. ¹³ICOS ERIC Carbon Portal, Lund, Sweden

Session

Session 28: Combining data and models to support emissions estimation and policy at local to regional scales

Abstract text

Methodological development of greenhouse gas (GHG) verification frameworks is essential for effective application in Europe and for broader international cooperation under UNEP, IPCC, and UNFCCC in support of the Paris Agreement. Building on the VERIFY project, which pioneered reconciliation of top-down (TD) atmospheric inversions and bottom-up (BU) inventory estimates for EU Member States, subsequent initiatives such as CoCO₂ and AVENGERS have further advanced this effort.

AVENGERS identifies best practices for integrating TD and BU approaches, addressing methodological inconsistencies, emission factors, and uncertainties through strengthened collaboration between scientific and inventory communities, which led to tangible progress and targeted reconciliation exercises in five countries.

Preliminary CO₂ results indicate that atmospheric inversions estimate a stronger biospheric sink than previous models, particularly in Germany, with notable trends also the Netherlands, and Switzerland. While fossil fuel emissions remain dominant, net land carbon sources have declined since 2010 due to reduced emissions and an intensifying sink. Inversions generally align with inventories for large emitters but are less robust for smaller countries.

For CH₄, intercomparisons reveal consistent positive anomalies over Benelux, suggesting inventory underestimation. For N₂O, inversions show shifting seasonality, recent emission declines, and high sensitivity to extreme climate events.

Bottom-up DGVM simulations confirm their value for EF estimation but highlight substantial structural and parameter uncertainties, emphasizing the need for improved calibration, validation, and methodological refinement.

This presentation summarizes preliminary results comparing inventory-based and top-down GHG estimates across five user stories, highlighting sectoral trends, spatio-temporal patterns, process-specific insights, and associated uncertainties.

108 Sources and Burial of Sedimentary Organic Carbon of the Radial Sand Ridge Field in the South Yellow Sea (east China) on a Centennial Scale

Poster

Xiuqiang Peng^1,2*, Minjing Wang^1,2, Yuru Yan^1,2, Qun Liu¹, Kai Ouyang¹, Ping Zuo^3,2,4

¹Jiangsu Geological Bureau, Nanjing, China. ²Key Laboratory of Coastal Salt Marsh Ecosystems and Resources, Ministry of Natural Resources, Nanjing, China. ³Key Laboratory of Coast and Island Development of the Ministry of Education, Nanjing University, Nanjing, China. ⁴School of Geography and Ocean Science, Nanjing University, Nanjing, China

Session

Session 4: Carbon cycling in the land ocean aquatic continuum

Abstract text

The Radial Sand Ridge Field in the South Yellow Sea, a large-scale tidal sand ridge geomorphological complex in shallow continental shelf waters, preserves critical environmental evolution information and serves as a key archive for marine carbon cycling research. Analysis of total carbon, organic carbon (OC), inorganic carbon (IC), total nitrogen, stable carbon isotopes (δ¹³C), and 210Pb radionuclides in 30 sediment cores from this system reveals significant carbon storage. In the upper meter of sediment, OC stocks are estimated at 14.82 million tons, while IC stocks reach 29.60 million tons, with IC density approximately 1.7 ~ 5.7 times that of OC per unit area. The carbon burial rate here is relatively high compared to other sedimentary settings, slightly exceeding global averages for muddy tidal flats (147.7 g C m⁻²a⁻¹) and seagrass meadows (138 ± 38 g C m⁻²a⁻¹), though lower than global salt marshes (244.7 ± 26.1 g C m⁻² a⁻¹) and mangroves (230.9 ± 26.0 g C m⁻² a⁻¹). The system annually buries approximately 0.48 Tg C, comparable to the entire Chinese salt marsh ecosystem (0.263–0.749 Tg C a⁻¹). Spatially, δ¹³C values increase offshore, indicating greater marine organic matter contribution. Temporally, the proportion of marine-sourced OC in nearshore areas has risen over the past century. These findings highlight the globally significant role of the Radial Sand Ridge System as an efficient carbon burial and storage site on continental shelves, exceeding previous assessments.

109 Warming effect on carbon allocation in Laminaria hyperborea is controlled by seasonal temperature

Poster

Kiara Franke^1,2*, Concepción Iñiguez³, Ulf Karsten², Inka Bartsch⁴, Angelika Graiff²

¹University of Aarhus, Aarhus, Denmark. ²University of Rostock, Rostock, Germany. ³University of Malaga, Malaga, Spain. ⁴Alfred Wegener Institute, Bremerhaven, Germany

Session

Session 3: Blue carbon and seaweed: reforestation and cultivation

Abstract text

The North-East Atlantic kelp species Laminaria hyperborea has a significant carbon capture potential. Under global warming, this kelp forest forming seaweed is already affected by decreasing abundance at its southern distribution edge. However, the physiological consequences of year-round warming on natural populations remain understudied. The present study focused on the net primary production (NPP) and organic carbon release in L. hyperborea in the German Bight (Helgoland). In all four seasons, meristematic blade discs were subjected to seasonal ambient temperatures (spring: 7°C, summer: 16°C, autumn: 14°C, winter: 6°C) and elevated temperatures (delta +4°C) under simulated in situ irradiances. Under warming, annual NPP was 14% higher (347 g C m^–2 yr^–1) than at ambient temperature (303 g C m^–2 yr^–1). In addition, warming exacerbated seasonal differences, resulting in two times higher NPP rates in spring and three times lower NPP rates in autumn than under ambient conditions. The release of carbon as either particulate organic carbon or dissolved organic carbon (DOC) was six times and three times higher, respectively, in high-temperature seasons (summer and autumn) than in spring and winter. Annually, warming resulted in 217% higher DOC release rates. Our study demonstrated the importance of investigating the effects of year-round warming on carbon uptake and allocation in kelp species. We therefore recommend modelling NPP rates under different global change scenarios and considering the release of carbon by kelp species when estimating their role in coastal carbon cycles. 

110 Improving Accessibility of CO₂ Reference Gases for Ocean Observation Platforms

Poster 

Carsten Spisla, Clemens von Scheffer*, Kristin Kampen, Melf Paulsen, Tobias Steinhoff

GEOMAR Helmholtz Centre for Ocean Research Kiel, Kiel, Germany

Session

Session 5: Advancing marine CO₂ observations through next-generation sensors, integration and platform innovation

Abstract text

High-quality pCO₂ measurements of the surface ocean are essential for marine carbon cycle research. Many instruments use traceable reference gases to ensure they fulfil the fundamental requirement for ensuring measurement accuracy and comparability across platforms. Currently, many marine observing platforms rely either on reference gases provided by the ICOS CAL laboratory or on a limited number of commercial suppliers. There, typically available gas cylinder sizes are 10, 20 or 50 L, which can be impractical for smaller platforms such as sailing vessels, volunteer observing ships, coastal stations, and short-term deployments.

To support the broader carbon observation community, we established the Ocean Observation Services (OOS) facility at GEOMAR to prepare and distribute CO₂ reference gases in basically any, but primarily smaller cylinder volumes (2–10 L), traceable to international calibration scales. The goal is not to replace existing calibration services, but to complement them by providing flexible supply options, reducing logistical constraints, and alleviating demand pressure on central calibration laboratories.

This initiative particularly targets mobile and non-ICOS platforms contributing data to the Surface Ocean CO₂ Atlas (SOCAT) and its synthesis products, where access to appropriately sized reference gases remains a bottleneck. By expanding availability and flexibility, we aim to strengthen data quality assurance across the marine carbon observation network.

We will present the technical setup, quality control procedures, traceability strategy, and first experiences with distribution to external users, and discuss how this service can be integrated within the existing ICOS calibration frameworks.

111 Observation of Atmospheric Deposition Flux and Its Eco-environmental Effects in Terrestrial Ecosystems

Poster

Qiufeng Wang^1,2*, Jianxing Zhu³, Yanlong Jia⁴, Nianpeng He⁵, Qiongyu Zhang⁶, Haili Yu⁵, Yue Xi^1,2, Yanran Chen^1,2, Zihan Tai^1,2, Guirui Yu^1,2

¹Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing, China. ²College of Resources and Environment, University of Chinese Academy of Sciences, Beijing, China. ³Hebei Agricultural University, Baoding, China. ⁴Department of Forestry, Hebei Agricultural University, Baoding, China. ⁵Department of Ecology, Northeast Forestry University, Harbin, China. ⁶Tianjin University, Tianjin, China

Session

Session 31: Flux measurements for immediate societal benefits

Abstract text

The atmospheric deposition of nitrogen, phosphorus, and acid is increasing due to human activities. This may not only bring negative effects on ecosystem, such as biodiversity damage, soil acidification, reducing soil buffer capacity, but also bring positive effects on ecosystems, such as providing nutrient element, increasing crop yields, promoting forest growing, and so on. Therefore, it will be helpful for evaluating the impact of atmospheric nitrogen, phosphorus, and acid deposition to ecosystem processes and functioning to explore their dynamics, patterns, and the influencing factors. Based on Chinese Ecosystem Research Network (CERN) and other stations, we established the Observation Network of Atmospheric Chemistry Deposition in typical terrestrial ecosystems (China_WD) which consists of more than 50 ecosystems covering forest, grassland, cropland, desert, lake, wetland, and city. The network observes atmospheric nitrogen, phosphorus, acid, and other deposition simultaneously. Based on the observation data, we investigated the spatio-temporal evolution of atmospheric nitrogen deposition, its dynamic formation mechanisms, the spatial and temporal patterns of multi-component atmospheric deposition and their interrelationships, and further elucidated the multiple resource and environmental effects induced by atmospheric deposition.

112 Ecosystem sensitivity to atmospheric and soil drought, a comparative analysis. 

Poster

Mahum Naseer*, Fran Lauriks, Marilyn Roland, Matteo Campioli, Ivan Janssens

University of Antwerp, Antwerp, Belgium

Session

Session 10: Cross-scale responses of greenhouse gas variability to climate extremes

Abstract text

In a changing climate, drought has transitioned from climate anomaly to a more persistent environmental state across terrestrial ecosystems. As a result, ecosystems are increasingly exposed to compound drought events, i.e. periods in time during which soil moisture declines and atmospheric water demand increases simultaneously. Although many studies examine ecosystem responses to compound drought, the individual drought stressors and their interaction remains less studied. This study investigates the individual and compounding impacts of soil and atmospheric drought on the carbon and water fluxes of a forest and heathland ecosystem during the growing season (April – September) for four consecutive years (2021 – 2024). The two ICOS study sites, a dry heathland (Maasmechelen) and a Scots pine dominated forest (Brasschaat), are located in Flanders (Belgium) and share similar climatic conditions. Over the study period, ecosystem functioning was evaluated by assessing the impacts of relative extractable water (REW) and vapor pressure deficit (VPD) on the net ecosystem exchange (NEE), gross primary production (GPP) and evapotranspiration (ET), obtained from eddy covariance measurements. We identify the critical thresholds at which soil and atmospheric drought alter functioning of the two ecosystems. This provide insights in (i) the resilience of the two different ecosystems to the drought type, (ii) how drought resistance evolves during the growing season, (iii) and how and when these thresholds are breached and alter the diurnal patterns. Our findings offer insights into the plant physiological responses and highlight the critical need to decouple drought-associated drivers to accurately assess ecosystem resilience under a changing climate.

113 Development of an Integrated Urban Greenhouse Gas Monitoring System in Gliwice (Poland): The MONCO2 Approach

Poster

Barbara Sensuła*, Alicja Ustrzycka

Silesian University of Technology, Institute of Physics- CSE/Division of Geochronology and Environmental Isotopes, Gliwice, Poland

Session

Session 17: Urban greenhouse gas emissions and sinks: from advanced monitoring to source identification and impact

Abstract text

Monitoring atmospheric greenhouse gases in urban environments is essential for understanding carbon cycle dynamics and identifying anthropogenic emission sources. The MONCO2 project, conducted at the Institute of Physics – CSE of the Silesian University of Technology, aims to develop and implement high-resolution methods for measuring CO₂ and CH₄ concentrations together with their isotopic composition.

The measurement infrastructure includes a GMP343 probe installed on the roof of the Institute building in Gliwice, a dedicated CO₂ sensor, and a biweekly air sampling system. Flask samples are analyzed spectrometrically to determine radiocarbon (¹⁴C) content and compared with background air. A significant depletion of atmospheric ¹⁴C has been observed in the Gliwice area, reflecting the Suess effect and confirming a strong contribution of fossil fuel–derived CO₂.

In parallel, measurements are performed using Cavity Ring-Down Spectroscopy (CRDS), enabling simultaneous, high-frequency determination of CO₂, CH₄, water vapor concentrations, and carbon isotopic composition. The analyzer, currently undergoing testing and calibration, provides continuous datasets.

The integration of concentration and isotopic measurements supports source attribution and improves understanding of urban carbon cycle processes. The approach enables differentiation between fossil and biogenic CO₂ contributions and assessment of spatial contrasts between vegetated and industrial zones. MONCO2 strengthens regional monitoring capacity and contributes to the development of modern methodologies for urban greenhouse gas observation.

This work is supported by the programme European Funds for Silesia 2021–2027: The modern methods of the monitoring of the level and isotopic composition of atmospheric CO₂ (project no. FESL.10.25-IZ.01-06C9/23-00, PI Barbara Sensuła).

114 Reducing Spatial Bias at Heterogenous Sites: A Novel Eddy Covariance Filtering Approach Integrating Footprint Modelling and Remote Sensing

Oral

Torben Oliver Callesen¹*, Anna Candotti¹, Sarah Treby², Samantha Grover³, Damiano Zanotelli¹, Massimo Tagliavini¹, Leonardo Montagnani¹

¹Free University of Bozen-Bolzano, Bozen-Bolzano, Italy. ²University of Melbourne, Melbourne, Australia. ³RMIT University, Melbourne, Australia

Session

Session 22: Remote sensing and flux observations to link disturbance, ecosystem states, and carbon-water fluxes

Abstract text

Eddy covariance (EC) measurements of net ecosystem exchange (NEE) over sites with multiple land cover classes may suffer from significant uncertainty due to heterogeneity in biological source/sink distribution and flux footprint variation, masking the “true” signal of the target system. We present a filtering approach based on the footprint-weighted contribution of non-target areas (non-target footprint fraction; NTFF) to 30-minute flux measurements using remote sensing-derived land cover classification at 10 m resolution.

Our findings, based on the analysis of 3-4 months of data across three heterogenous sites, indicated an apparent reduction in source/sink distribution-related bias relative to unfiltered data. The apparent total NEE over the whole measurement period increased slightly at a tall forest site likely due in part to NTFF acting as a weak advection indicator. Total NEE became more negative by up to 189.32 g C m^-2 at a mountain vineyard affected by nighttime anthropogenic emissions but was relatively unaffected at an Alpine peatland site where the non-target vegetation was difficult to differentiate using satellite information. Energy balance closure was improved at all sites.

This filtering method can be implemented at EC sites using existing measurements and open access remote sensing information, providing a potential tool for improving the representativeness and inter-comparability of carbon budgets measured at heterogeneous sites. This approach can easily be integrated into existing processing pipelines, complementing standard quality control methods. Further research could develop an objective NTFF threshold and validate results at other sites against independent methods.

115 Spatially-constrained carbon modelling identifies precipitation–fire functional controls on biomass stocks and carbon dynamics across the Southern African Woodlands ecoregion

Oral

Mathew Williams¹*, David Milodowski¹, Luke Smallman¹, Casey Ryan¹, Gabi Hegerl¹, Kyle Dexter¹, Carla Roesch¹, Stephen Sitch², Mike O'Sullivan², Aude Valada³, Iain McNicol¹

¹University of Edinburgh, Edinburgh, United Kingdom. ²University of Exeter, Exeter, United Kingdom. ³UMR, Montpellier, France

Session

Session 7: CO2 and CH4 cycle dynamics in Africa: observations, processes, and climate implications

Abstract text

Southern African woodlands (SAW) are the world's largest savanna, covering ∼ 3M km², but their carbon balance and its interactions with climate and disturbance are poorly understood. Here we produce a C budget of the SAW region consistent with observations; and diagnose its functional variation and interactions with climate and fire. Using 1506 independent 0.5° pixel model calibrations we produce a regional C analysis (2006–2017) constrained with Earth observations of biomass (ALOS-PALSAR 2007-10)) and leaf area index (MODIS 2006-17), forced with local climate (CRU-JRAv2.1), burned area (MODIS) and forest loss (GFW) estimates. The estimated regional net biome production is neutral, −0.08 Mg C ha⁻¹ yr⁻¹ (95 % uncertainty interval −1.67/1.66), with fire emissions contributing ∼ 0.88 Mg C ha⁻¹ yr⁻¹ (0.36–2.51). The emergent spatial variation in biogenic fluxes and C pools is strongly correlated with mean annual precipitation and burned area. However, there are multiple, potentially confounding, causal pathways through which variation in environmental drivers impacts the spatial distribution of C stocks and fluxes, mediated by spatial variations in functional parameters like allocation, wood lifespan, and fire resilience. Patterns of biomass and C cycling across the region are the outcome of climate controls on production and vegetation–fire interactions which determine residence times, linked to spatial variations in key ecosystem functional characteristics. This model-data fusion approach provides routes for delivering national estimates of C balance across the region to support UNFCCC processes, and insights for sustainable forest management at national scales linked to localised traits and processes.

116 Establishment-Year Carbon, Energy and Water Fluxes in Perennial Grain versus Annual Cropping System in Southern Sweden

Oral

Veronika Widengren¹*, Jonas Ardö¹, Jutta Holst¹, Tobias Biermann¹, Peter Kornacher¹, Johannes Albertsson², Patrik Vestin¹

¹Lund University, Lund, Sweden. ²Swedish University of Agricultural Sciences, Alnarp, Sweden

Session

Session 14: Greenhouse gas fluxes in agroecosystems: processes, measurements and management implications

Abstract text

Agriculture is a major source of global greenhouse gas emissions, and faces challenges such as erosion, nutrient leakage in addition to a strong reliance on external inputs such as energy, seeds, pesticides and fertilizers. These challenges are largely driven by the cultivation of annual crops with shallow roots and frequent soil disturbance. Perennial crops, such as intermediate wheatgrass (IWG, Thinopyrum intermedium), could offer a sustainable alternative through its deep and extensive root systems and minimal tillage. This study on IWG establishment evaluates carbon, energy, and water fluxes in comparison with winter oilseed rape (Brassica napus) during the first crop cycle.

A 10-ha IWG field was established next to an equal-sized conventional annual crop reference field in southern Sweden. Using eddy covariance, the first crop cycle showed that IWG had 50% higher net carbon uptake than winter oilseed rape (net ecosystem exchange (NEE) -212 vs. -141 g C m⁻², respectively). The net ecosystem carbon balance (NECB), including the harvested above-ground biomass, showed a small net uptake in the IWG compared to a small net emission in winter oilseed rape (-49 and 31 g C m^-2, respectively). Comparison of energy balance and water use efficiency (WUE) showed that IWG exhibited a more stable temporal pattern, suggesting a conservative growth strategy and resource‑use behaviour. Despite a challenging start with flooding of the IWG field, the results suggest that perennial crops have a greater uptake of carbon in comparison to conventional annual crop systems already in the establishment year.

117 Interannual variability in net ecosystem exchange and evapotranspiration in a semi-arid rainfed barley –legume rotation with summer fallow in Cyprus

Oral

Hakan Djuma¹*, Niovi Christodoulou¹, Ioannis Sofokleous¹, Andreas Savvides¹, Christos Zoumides¹, Minas Arnaoutis², Adriana Bruggeman¹

¹The Cyprus Institute, Nicosia, Cyprus. ²Department of Agriculture, Ministry of Agriculture, Rural Development and the Environment, Nicosia, Cyprus

Session

Session 6: Carbon cycle in the Mediterranean region: from the local to the regional scale

Abstract text

The future of cereal cultivation at the dry margins of semi-arid Mediterranean environments is increasingly threatened by climate change. This study quantifies the seasonal and interannual variability of net ecosystem exchange (NEE) and evapotranspiration (ET) in a barley (Hordeum vulgare L. var. achna)–pea (Pisello proteico var. astronaut) rotation using eddy covariance over a five-year period (2020–2025). The site is a 12-ha rainfed field in the Mesaoria Plain in Cyprus with mean annual rainfall of 315 mm and temperature of 19.6 °C. The analysis covers nine seasons: three barley winter seasons, two pea winter seasons, and four intervening summer fallow seasons. Notably, the driest barley season (171.2 mm) showed the greatest net CO₂ uptake (NEE: −2.25 kg m⁻²) compared with the two wetter barley seasons (258.1 mm and −1.54 kg m⁻²; 315.7 mm and −1.98 kg m⁻²), and also produced the highest grain yield (2.49 t ha⁻¹ vs. 1.37 and 0.94 t ha⁻¹). Total growing degree days from sowing to senescence for this driest season were between those of the wetter seasons (1784 vs. 1675 and 2161). The ET was 50% of precipitation (85.6 mm) in the driest season, compared with 35% in the wetter seasons (91.1 and 109.5 mm). The NEE of the four fallow seasons ranged from −0.37 to 0.28 kg m⁻² and showed no clear relationship with precipitation. The results highlight the importance of continuous eddy covariance measurements for improving our understanding of carbon and water fluxes in marginal Mediterranean agroecosystems.

118 German CH₄ emissions derived from in-situ observations using ensemble-enhanced ICON–ART model

Poster

Valentin Bruch*, Niklas Becker, Andrea Kaiser-Weiss

Deutscher Wetterdienst, Offenbach, Germany

Session

Session 28: Combining data and models to support emissions estimation and policy at local to regional scales

Abstract text

Inverse modeling can validate national GHG emissions inventories by combining atmospheric observations and transport simulations. But uncertainties and errors in these simulations present a major challenge. We address this challenge by employing the numerical weather prediction model ICON at 6.5 km resolution and estimating the uncertainties using a transport ensemble. We use a priori emission data from national inventory compilers and in-situ observations from ICOS to compute national-scale CH4 emissions with a focus on Germany for the years 2018 to 2024.

Our inversion method is an extended synthesis inversion that scales initially defined flux categories while self-consistently adjusting uncertainties to the estimated fluxes. Implemented in the new dynamic uncertainty inversion tool DUBFI, this method is optimized to minimize biases in the emission estimates. We compare the inversion results to a second method that uses a very similar transport setup but estimates fluxes using a local ensemble transform Kalman filter (LETKF) that optimizes emissions and concentrations of CH₄ simultaneously. This method is more suitable when the spatial distribution of emissions is unknown or for near-real-time monitoring of greenhouse gas emissions.

Preliminary results show good overall agreement of the emission estimates from both methods in all well-observed regions. For northwestern Germany and the Netherlands, the posterior emissions are significantly higher than in the national inventories. These results are part of the national ITMS project that develops an integrated greenhouse gas monitoring system for Germany.

119 The current status of the Oceans in the Global Greenhouse Gas Watch.

Poster

Richard Sanders¹*, Adrienne Sutton², Galen McKinley³, Peter Landschutzer⁴, Tobias Steinhoff⁵, Siv Lauvset¹, Dorothee Bakker⁶, Siyabula Hamnca⁷, Louise Delaigue⁸, Maciej Telszewski⁹

¹NORCE, Bergen, Norway. ²NOAA Pacific Marine Environmental Laboratory, Seattle, USA. ³Columbia University and Lamont-Doherty Earth Observatory, New York, USA. ⁴VLIZ, Ostend, Belgium. ⁵Geomar, Kiel, Germany. ⁶University of East Anglia, Norwich, United Kingdom. ⁷CSIR, Cape Town, South Africa. ⁸Laboratoire d’Océanographie de Villefranche, Villefranche-sur-mer,, France. ⁹International Ocean Carbon Coordination Project, Sopot, Poland

Session

Session 26: Designing the ideal global greenhouse gas monitoring network

Abstract text

The Ocean is a major CO₂ sink taking up 30% of anthropogenic carbon emissions, slowing climate change and giving us  time for mitigation and adaptation actions, and potentially,  carbon dioxide removal (CDR). We know this from the value chain that links surface CO₂ observations from multiple groups, to SOCAT, the surface ocean CO₂ Atlas, synthesized global air-sea CO₂ flux estimates from the Surface Ocean CO₂ Mapping Intercomparison (SOCOM), and then the Global Carbon Budget (GCB) which reports to the UNFCCC Conference of Parties (COP). In the near future, this value chain will be extended to include a WMO sponsored Global Greenhouse Gas Watch which will require  surface flux fields  from data-based products similar to those produced by  SOCOM. These would be used as the ocean prior for a series of inversion calculations aimed at producing GHG fluxes across the Earth’s surface (both ocean and land) on a monthly basis at a spatial resolution of 1 degree. Ocean carbon cycle scientists have been strongly engaged in the planning for this via a series of actions including setting up a Surface Ocean CO₂ Observing Network (SOCONET) as a GOOS observing network and supporting Observing System Simulation Experiments (OSSEs) to identify key areas where additional  observations will be required. Here we review the history and development of these activities, and describe the set of priority actions identified for implementation in the near future that will be considered by the WMO congress in late 2026.  

120 The Metadata Advantage: Elevating Interoperability of Ocean Climate Science

Poster

Charlotte Miskin-Hymas¹*, Steve Jones²

¹British Oceanographic Data Centre, National Oceanography Centre, Southampton, United Kingdom. ²Flanders Marine Institute (VLIZ), Ostend, Belgium

Session

Session 1: How big is the open ocean carbon sink?

Abstract text

Surface ocean partial pressure of carbon dioxide (pCO₂) is a fundamental variable for understanding air–sea CO₂ exchange, a key process governing Earth’s climate system. The value of pCO₂ observations depends critically on the completeness and quality of the accompanying metadata. Rich metadata—such as sensor type, serial number, sampling depth, environmental context and platform characteristics—provided as machine‑readable content enables robust interpretation, fairer uncertainty assessment, and reliable intercomparison of datasets.  

High‑quality, interoperable pCO₂ data are also increasingly important for greenhouse gas monitoring frameworks, including the emerging World Meteorological Organization (WMO) Greenhouse Gas Watch initiative, which aims to provide an authoritative, integrated view of atmospheric and oceanic carbon fluxes.  

ICOS shares surface ocean pCO₂ data with the Surface Ocean CO₂ Atlas (SOCAT). Such coordinated data stewardship supports global efforts across the entire ocean observation landscape such as the Global Ocean Observing System (GOOS), the Ocean Data and Information System (ODIS) and in turn, the UN Ocean Decade, which rely on consistent and high‑quality data flows. These datasets are strengthening the evidence base required for climate assessments by bodies such as the Intergovernmental Panel on Climate Change (IPCC) and for producing robust global carbon budgets. 

Ultimately, rigorous metadata and strong data management frameworks underpin our ability to monitor the changing ocean, evaluate climate mitigation strategies, and inform policy decisions in a rapidly warming world. This presentation will describe how enhancing ICOS ocean data with rich and machine-readable metadata is directly contributing to a more interoperable global data infrastructure. 

121 Evaluation of Precipitable Water Vapor (PWV) Reference Standards at High Altitude: Systematic Biases in the Andes

Poster

Fernando Velarde¹*, Wara Carvajal¹, Alexander Cede^2,3, Morgan Lopez⁴, Laura Mirabal¹, Manuel Roca², Michel Ramonet⁴, Laura Ticona⁴, Marcos F. Andrade¹

¹Laboratory for Atmospheric Physics, Universidad Mayor de San Andrés, La Paz, Bolivia, Plurinational State of. ²LuftBlick, Innsbruck, Austria. ³NASA Goddard Space Flight Center, Greenbelt, MD, USA. ⁴LSCE, Gif-sur-Yvette, France

Session

Session 23: Remote sensing and vertical profiling of atmospheric greenhouse gases for climate action

Abstract text

Ground-based networks like COCCON and the Pandonia Global Network (PGN) are primary reference standards for the geophysical validation of satellite-derived Precipitable Water Vapor (PWV) and the establishment of Monitoring and Verification Support (MVS) capacities. In high-altitude urban environments like La Paz (3420 m a.s.l.), precise PWV data are critical for validating missions like Sentinel-5P and CO2M. However, these standards rely on an assumption of pressure-independent consistency within retrieval algorithms—a premise largely untested in low-pressure regimes.

To address this, we evaluated the performance of three instruments part of known infrastructures—EM-27/SUN (Obs4Clim), Pandora (PGN), and CIMEL (AERONET)—during an intensive campaign in La Paz. Results show systematic biases at high altitude conditions compromising atmospheric product traceability. While instruments exhibited high correlation (R²≥0.98), significant offsets were identified: the EM-27/SUN showed a 13% wet bias relative to Pandora, while the CIMEL displayed a dry bias of 19–23%. Preliminary validation against radiosondes suggests Pandora retrievals align most closely with the "true" vertically integrated PWV.

To assess global scalability, we conducted a network-wide intercomparison of 55 co-located Pandora/CIMEL stations across diverse latitudes and altitudes. While correlation remains robust (R²>0.98), retrieval slopes deviate from the ideal 1.0 at sea level to ~0.75 above 3 km. This confirms a fundamental breakdown of retrieval assumptions in low-pressure regimes. To maintain the integrity of global reference networks and ensure the reliability of MVS capacities, further investigation is required to harmonize retrieval results across the full range of terrestrial atmospheric pressures.

122 Improving and validating the Clifton’s model for cuticular and soil ozone deposition in a broadleaf forest

Oral

Giacomo A. Gerosa¹*, Davide Plebani^1,2, Angelo Finco¹, Riccardo Marzuoli¹

¹Catholic University of the S.H., Brescia, Italy. ²Catholic University (KU Leuven), Leuven, Belgium

Session

Session 24: Exchange of reactive gases and aerosols between the land surface and the atmosphere in natural and managed ecosystems

Abstract text

Ozone dry deposition to vegetated surfaces occurs through stomatal uptake and non-stomatal processes involving external plant surfaces, such as leaf cuticles, and soils. While stomatal deposition has been extensively investigated, non-stomatal pathways remain less understood. In 2020, Clifton et al. proposed a mechanistic model to estimate non-stomatal ozone deposition to dry and wet cuticles and soils. However, the model was developed on a purely theoretical basis and has not been validated against field measurements.

This study presents a calibration and validation of Clifton’s non-stomatal resistance model for a broadleaf forest. Total ozone deposition resistance simulated with the model was compared with values derived from eddy covariance flux measurements collected in 2021 and 2022 at Bosco Fontana (Italy). Partitioning of total resistance enabled a more detailed assessment of the modeled cuticular component.
In addition, modeled soil deposition resistance (Rsoil) was evaluated against measurements obtained using automatic soil chambers at the same site in 2024 and 2025. 

Results indicate that Clifton’s model reasonably reproduces cuticular deposition (Rcut) and its seasonal dynamics. However, it fails to capture Rsoil under increasing soil water content, as it explicitly neglects ozone dissolution and reaction within water films in soil pores.
To address this limitation, we introduced a simplified representation of pore water and ozone–water reactions into the Rsoil formulation. This modification substantially improved agreement with measured soil resistance.

Overall, the results are promising, although validation at additional sites is required before general conclusions can be drawn.

123 First Aircore and total column measurement campaigns in South America carried out in Toledo, Bolivia (2023-2025)

Oral

Michel Ramonet¹*, Laura Ticona Ticona¹, Alan Callau², Maixent Cassagne¹, Olivier Laurent¹, Dylan Lopez¹, Morgan Lopez¹, Antoine Parent¹, Louis-Jeremy Rigouleau¹, Zarela Tuco², Fernando Valerde², Marcos Andrade²

¹LSCE, Gif-sur-Yvette, France. ²UMSA, La Paz, Bolivia, Plurinational State of

Session

Session 23: Remote sensing and vertical profiling of atmospheric greenhouse gases for climate action

Abstract text

The installation of an atmospheric composition observatory in Chacaltaya, now a global station in the WMO network, has enabled Bolivia to develop strong expertise in the analysis of atmospheric composition, including reactive gases, aerosols, and greenhouse gases. The ICOS 2026 conference is an opportunity to present several scientific results based on this expertise. We have also drawn on this program to set up the first two campaigns to measure vertical GHG profiles using AIRCORE, combined with co-located total column measurements, intended to provide the first evaluation dataset for spatial measurements in South America. Two intensive campaigns were carried out by a Franco-Bolivian team in the Oruro region in August 2023 and July 2025. The first campaign enabled the measurement of six vertical profiles of CO₂, CH₄, CO, and H₂O (from the surface up to an altitude of 25 km), and 11 profiles were measured during the second campaign. At the same time, an FTIR spectrometer measured the total columns of the same gases, and in 2025 in particular, special effort was made to synchronize the vertical profiles with satellite overpasses (TROPOMI, OCO2/3). During this presentation, we will detail the vertical profiles and total columns measured from the ground, as well as their comparison with CAMS simulations and satellite measurements.

124 Reducing uncertainties on tropical methane emissions from wetlands in the Congo Basin using satellite observations

Poster

Sophie Wittig¹*, Martin Vojta¹, Rakesh Subramanian¹, Hartmut Bösch², Manuel Gloor³, Anjumol Raju¹, Seyed Omid Nabavi¹, Andreas Stohl¹

¹University of Vienna, Vienna, Austria. ²University of Bremen, Bremen, Germany. ³University Of Leeds, Leeds, United Kingdom

Session

Session 7: CO2 and CH4 cycle dynamics in Africa: observations, processes, and climate implications

Abstract text

Atmospheric methane (CH₄) is a significant greenhouse gas with both anthropogenic and natural sources which is responsible for around a quarter of the global warming that has occurred since pre-industrial times. Natural CH₄ sources have become an increasingly important driver of the CH₄ growth rate in the atmosphere due to climate feedbacks caused by global warming, however they remain particularly difficult to quantify. The largest natural source is biogenic CH₄ from wetland regions. Estimates of emissions from tropical wetlands are subject to particularly high levels of uncertainty due to the limited availability of stationary and mobile measurements in the in the southern hemisphere of the Earth.

Here, we aim at improving estimates of tropical CH₄ emissions in the Congo Basin using satellite observations from the TROPOMI (Tropospheric Monitoring) instrument in a Bayesian inverse modelling framework in order to optimize CH₄ fluxes from multiple emission scenarios. We use the inverse modelling tool FLEXINVERT, which is based on backward simulations of the Lagrangian particle dispersion model FLEXPART (FLEXible PARTicle). In an attempt to estimate background concentrations as accurately as possible, we use global 3D concentration fields obtained with the linear transport model FLEXPART-LCM.

This work is part of the Horizon-Europe project IM⁴CA, with the aim of preparing the inverse modelling set-up for novel total column measurements that will be installed in the Congo during the course of the project.

125 Estimating fossil fuel CO₂ signals using atmospheric O₂ measurements at Cabauw, The Netherlands

Poster

Loïs de Beijl¹*, Bert Kers¹, Jordi Vilà-Guerau de Arellano², Harro Meijer¹, Ingrid Luijkx²

¹University of Groningen, Groningen, Netherlands. ²Wageningen University & Research, Wageningen, Netherlands

Session

Session 18: Quantifying anthropogenic greenhouse gas emissions from continental to regional scales

Abstract text

Reliable quantification of fossil fuel CO₂ emissions is essential for tracking progress towards emission reduction targets. Atmospheric oxygen (O₂) measurements offer additional information on anthropogenic CO₂ emissions through the coupling of CO₂ and O₂ in carbon cycle processes. O₂ therefor allows to distinguish fossil fuel signals from biospheric influences. However, continuous high precision O₂ observations are still limited within existing atmospheric monitoring networks, particularly at tall tower stations.

As part of the CORSO project, a high precision atmospheric O₂ measurement system was installed at the Cabauw observatory, which is part of ICOS. Measurements were first carried out at 207 m and later expanded to three heights (27 m, 67 m, and 207 m), allowing the investigation of vertical gradients and derive combined CO₂ and O₂ flux estimates. The first six months of data were collected during CORSO and compared with existing ICOS measurements of CO₂ and meteorological variables.

The current work focuses on the use of these observations to derive O₂ fluxes and exchange ratios, with the aim of separately identifying rural and anthropogenic sources from the region. In addition, O₂ is explored as a tracer to identify and quantify fossil fuel contributions in the atmospheric time series. Following the recent renovation of the tower, measurements will be resumed in 2026, allowing the extension of the time series. First results from the continued observations will be presented, together with an outlook on the potential of atmospheric O₂ measurements within ICOS to improve regional fossil fuel emission estimates.

126 Long-term C sinks of two unmanaged, partly disturbed mountain forest ecosystems in Austria

Poster

Johannes Kobler, Johannes Peterseil, Thomas Dirnböck*

Environment Agency Austria, Vienna, Austria

Session

Session 11: Climate feedbacks from ecosystem management and natural disturbances

Abstract text

Temperate mountain forests play a critical role in the global carbon (C) cycle. Under ongoing climate change, however, the forest C sink (i.e. NEP) is increasingly at risk due to forest disturbance by intensified climatic extremes and biotic disturbance agents. The cessation of forest management, often proposed as a climate-protection strategy, has been debated because it may increase forest susceptibility to disturbance.

This study quantified long-term NEP of two widespread forest ecosystems at the LTER site Zöbelboden within the Northern Limestone Alps. These forests have remained unmanaged since 1993 and have experienced varying forest disturbances. 

Disturbances reduced living tree biomass in two plots, whereas one plot remained undisturbed. While the undisturbed stand maintained a strong persistent C sink, disturbance shifted one disturbed plot into a net C source. Contrasting NEP responses were driven by differences in site productivity, C losses from legacy coarse deadwood pools, and related tree species–specific decomposition rates.

Our results highlight the central role of coarse deadwood dynamics in controlling long-term NEP in unmanaged, disturbed forests. These findings provide important insights for climate-oriented forest management and improve understanding of the resilience of carbon sinks in widespread temperate mountain forest ecosystems. All R-based analyses were released as an open data product within eLTER, enabling transferability of the NEP calculation to other forest sites.

127 Remineralisation of terrestrial dissolved organic carbon in the Bothnian Bay inferred from stable carbon isotopes

Oral

Abdirahman Omar*

NORCE Research, Bergen, Norway

Session

Session 4: Carbon cycling in the land ocean aquatic continuum

Abstract text

The processing of terrestrial dissolved organic carbon (tDOC) within coastal and shelf seas plays a crucial role in the global carbon cycle, yet its fate remains poorly understood. In Nordic coastal regions, rivers deliver substantial quantities of tDOC due to the surrounding landscapes, which are rich in organic carbon. The consequences for coastal carbon cycling and local ecosystems are strongly influenced by how this tDOC is transformed and behaves in the marine environment. 

In this presentation, we will examine the proportion of the initial tDOC input that is remineralised to dissolved inorganic carbon (DIC) in the Bothnian Bay, as well as the potential implications for seasonal acidification and air-sea CO ₂ exchange. To address these questions, we employ multi-cruise biogeochemical datasets, incorporating both riverine and marine measurements of inorganic and organic carbon, alongside their stable carbon isotopic signatures (δ¹³C). A conservative mixing model is constructed for biogeochemical variables to predict hypothetical concentrations resulting from the mixing of river and marine waters. Deviations between measured data and these predicted values are then analysed using isotope mass balance calculations, enabling quantification of tDOC remineralisation. 

128 Agroecological practices and carbon turnover in Mediterranean soils

Oral

Evangelia Koukianaki¹*, Maria Lilli¹, Roberta Farina², Silvia Vanino², Carlos Alberto Torres Guerrero³, Eduardo Garcia Braga³, Nikolaos Nikolaidis¹

¹Technical University of Crete, Chania, Greece. ²Council for Agricultural Research and Economics, Research Centre for Agriculture and Environment, Rome, Italy. ³Technological Centre in Biodiversity, Ecology, Environmental and Food Technology, University of Vic, Catalonia, Spain

Session

Session 6: Carbon cycle in the Mediterranean region: from the local to the regional scale

Abstract text

Mediterranean agroecosystems are highly sensitive to carbon depletion due to climatic constraints, intensive land use (Aguilera et al., 2013) and low soil organic carbon content (Grilli et al., 2021). Agroecological practices restore soil carbon stocks, improve soil structure and enhance ecosystem resilience (Somashekar et al., 2024; Blaix et al., 2025). However, their influence on carbon stabilization processes and the persistence of carbon in the different soil fractions remains poorly quantified.

The objective of this study was to quantify carbon inputs, identify the soil fraction with the greatest carbon sequestration and evaluate carbon turnover among different Mediterranean soils and management practices.

The Carbon, Aggregation and Structure Turnover (CAST) model, which was used in this study, simulates the macro-aggregate formation (around particulate organic matter, POM) and disruption to form micro-aggregates and silt-clay sized micro-aggregates (Stamati et al., 2013). Soil samples from three farms in Greece, Italy and Spain were separated into water-stable aggregate fractions to measure their carbon content, which was used to initialize and calibrate the model. 

The carbon input, in Greek Avocado 1 and 2 plantations was 10.7 and 13.0 tC/ha/yr, for the Italian olive groves was 3.4 tC/ha/yr while for the Spanish pasture and agroforestry simulations was 2.2 and 4.4 tC/ha/yr respectively. The majority of SOC was found in the POM fraction ranging from 58% in the olives to 79% in Avocado 1. The turnover time was found to be extremely long in the decomposition of macro-aggregates which indicate soil stability and fertility.

129 Measuring air-sea CO₂ fluxes using an innovative instrumented ocean gliders approach

Oral

Paco STIL¹, Louise DELAIGUE¹, Felix MARGIRIER², Rémi EMMETIERE², Tobias STEINHOFF³, Socratis LOUCAIDES⁴, Susan HARTMAN⁴, Anita FLOHR⁴, Andrew GATES⁴, Laurent COPPOLA¹*

¹LOV, Villefranche/Mer, France. ²ALSEAMAR, Rousset, France. ³GEOMAR, Kiel, Germany. ⁴NOC, Southampton, United Kingdom

Session

Session 5: Advancing marine CO₂ observations through next-generation sensors, integration and platform innovation

Abstract text

Air-sea carbon dioxide (CO₂) exchange is a key variable of the Earth’s climate system. Over 2014-2023, the ocean is estimated to have absorbed 3.2 ± 0.4 GtC yr⁻¹, around one third of anthropogenic CO₂ emissions. Yet, substantial gaps persist between estimates based on global ocean biogeochemistry models and observation-based surface fCO₂ products, largely because in situ CO₂ measurements remain sparse and seasonally biased. These limitations are especially important at high latitudes, where fCO₂ reconstructions may overstate variability by up to ~30%. Expanding autonomous observations is therefore critical to reduce uncertainty in the ocean carbon sink. For the first time, wz deployed a Sea-Explorer glider (ALSEAMAR) integrating a membrane-based Pro-Oceanus Mini CO₂ sensor together with an acoustic wind sensor (Purpoise). This configuration enabled co-located measurements of the key variables (fCO₂ and wind speed) needed to estimate air-sea CO₂ fluxes while simultaneously profiling the upper ocean to 1000 m depth. Field trials at the DYFAMED site (2024/2025) assessed Purpoise wind measurements against the DYFAMED buoy and evaluated Mini CO₂ observations through intercomparison with CTD-calibrated water-column measurements. A second deployment at the Porcupine Abyssal Plain site (June 2025) tested the near-surface performance of the Mini CO₂ sensor against the GO CO₂ system operated aboard RRS James Cook. Results demonstrate the technical feasibility of deriving glider-based air-sea CO₂ flux estimates and highlight the value of multi-sensor autonomous platforms to complement existing observing networks, improve spatiotemporal coverage, and ultimately reduce uncertainties in the global ocean carbon budget.

130 Multi-decadal evolution of seawater carbonate chemistry and air-sea CO₂ fluxes in a warming northwestern Mediterranean Sea

Poster

Samir Alliouane¹, Steeve Comeau¹, Jean-Pierre Gattuso^1,2, Laurent Coppola¹*, Jean-Michel Grisoni³, Laure Mousseau¹, Frédéric Gazeau¹

¹Laboratoire d'Océanographie de Villefranche (CNRS/SU), Villefranche-sur-Mer, France. ²Institute for Sustainable Development and International Relations (Iddri-SciencesPo), Paris, France. ³Institut de la Mer de Villefranche (CNRS/SU), Villefranche-sur-Mer, France

Session

Session 6: Carbon cycle in the Mediterranean region: from the local to the regional scale

Abstract text

This study examines long-term environmental changes in the Bay of Villefranche-sur-Mer (northwestern Mediterranean Sea) using two complementary observing systems: the SOMLIT-Point B coastal time series (weekly sampling since 1995 at 1 and 50 m depth) and the high-frequency Environment Observable Littoral (EOL) buoy (30-minute measurements since 2018). Combining multi-decadal low-frequency observations with high-frequency monitoring allows the separation of long-term environmental trends from short-term variability and extreme events, while providing a comprehensive assessment of changes in physical and carbonate chemistry parameters as well as air–sea CO₂ fluxes. 

Surface waters at Point B warmed significantly between 1995 and 2023 at a rate of 0.039 ± 0.004 °C yr^-1. The strongest warming occurs during summer, with July trends reaching 0.1 °C yr^-1, reflecting increasing thermal extremes. Over the same period, salinity shows no significant trend at the surface but increases slightly at 50 m. 

Between 2007 and 2024, the carbonate system shows pronounced changes. Dissolved inorganic carbon increases (2.47 ± 0.10 µmol kg^-1 yr^-1), seawater partial pressure of CO₂ (pCO₂) rises (5.22 ± 0.18 µatm yr^-1), and pH declines rapidly (-0.0043 ± 0.0001 yr^-1), indicating accelerated ocean acidification. Driver decomposition reveals that dissolved inorganic carbon accumulation explains around 3/4 of the pH decrease, while warming contributes ~25%. Because seawater pCO₂ increases faster than atmospheric pCO₂, the bay shifts from a net CO₂ sink to a source around 2014. High-frequency observations further highlight the importance of sampling resolution for detecting short-term variability and improving the interpretation of long-term coastal environmental trends.

131 Evapotranspiration of the subalpine spruce forest Davos: 27 years of fluxes, a handful of drivers, one main water source

Plenary

Nina Buchmann¹*, Marius Floriancic¹, Iris Feigenwinter¹, Lukas Hörtnagl¹, Harsh Beria¹, Ankit Shekhar², Mana Gharun³

¹ETH Zurich, Zurich, Switzerland. ²Indian Institute of Technology Kharagpur, Kharagpur, India. ³University of Munster, Munster, Germany

Session

Session 13: Advancing approaches for quantifying greenhouse gas fluxes in terrestrial ecosystems

Abstract text

Norway spruce forests are widespread in Europe, yet their long-term evapotranspiration (ET) dynamics and seasonal water sources remain insufficiently constrained. Here we analysed 27 years (1997-2023) of ET measured with eddy covariance at the subalpine spruce forest Davos-Seehornwald, an ICOS-CH Class 1 Ecosystem station (CH-Dav). We determined ET flux drivers using advanced machine learning and identified tree water uptake patterns using stable water isotopes (2020–2022). While annual precipitation amounts remained roughly constant, potential ET increased, but actual ET declined throughout the observation period. Concurrently, water use efficiency (i.e., gross primary production GPP/ET) stayed stable, consistent with a tight physiological control. Driver attribution revealed pronounced seasonality: ET of the Norway spruce forest was mainly temperature-limited, while vapor pressure deficit (VPD) and soil moisture were less important. However, under very warm and dry conditions (based on z-scores), high VPD affected ET more than low soil moisture. Isotope-based water source partitioning showed that Norway spruce trees in this mono-specific natural stand relied substantially on soil water from deeper layers (50–70 cm), much in contrast to results from spruce trees in mixed forests (water uptake from top 30 cm). Deep water uptake often accounted for >50% of total tree water uptake, reaching >80% during the year 2020. In addition, spruce shifted from cold-season-recharged soil water early in the growing-season to warm-season precipitation. Together, these results suggest an increasing vulnerability of mono-specific spruce forests to future intensifying drought and heat conditions, when summer precipitation cannot sustain transpiration any longer.

132 Atmospheric dryness effects on canopy chlorophyll fluorescence and Gross Primary Production (GPP) in a deciduous forest during heat waves

Oral

Zhaohui Li¹*, Gabriel Hmimina², Gwendal Latouche¹, Daniel Berveiller¹, Abderrahmane Ounis², Yves Goulas², Kamel Soudani¹

¹Ecologie Société Evolution (ESE), Université Paris-Saclay, CNRS, AgroParisTech, 91190, Gif-sur-Yvette, France. ²Laboratoire de Météorologie Dynamique (LMD), Sorbonne Université, IPSL, CNRS, École polytechnique, 91128, Palaiseau Cedex, France

Session

Session 16: Using sun-induced chlorophyll fluorescence to understand or scale EC fluxes

Abstract text

Sun-Induced chlorophyll Fluorescence (SIF) is a key proxy of Gross Primary Production (GPP), yet its reliability weakens when plants experience stress. This study investigates the feasibility of canopy-level active chlorophyll fluorescence (LED-Induced chlorophyll Fluorescence, LIF), which directly measures the apparent fluorescence yield (F_yieldLIF), in detecting plant responses to abiotic stress. Conducted at the French Fontainebleau-Barbeau (ICOS FR-Fon) station during the 2022 European heatwaves, continuous eddy covariance fluxes, SIF, F_yieldLIF, and environmental variables were acquired. We investigated how high Vapor Pressure Deficit (VPD) affected fluorescence (SIF and F_yieldLIF), GPP, and their relationships. At the half-hourly scale, high atmospheric dryness led to a weakened SIF-GPP correlation (R² dropped from 0.49 to 0.17), while the F_yieldLIF-GPP correlation increased significantly (R² rose from 0.07 to 0.43). Daily scale analysis showed overall improved correlations for both, indicating time-scale-dependent responses. Furthermore, under high VPD and clear skies, F_yieldLIF proved to be a more robust proxy for the maximum photosynthetic rate (Amax) than SIFy (Photosynthetically Active Radiation (PAR)-normalized SIF), with R² reaching 0.85 compared to SIFy’s 0.56. This study highlighted F_yieldLIF's advantage in detecting plant responses to dryness stress. The advantage of F_yieldLIF stems from its stable excitation source and fixed geometry, allowing it to directly capture physiological regulation like Non-Photochemical Quenching (NPQ) without the structural and radiative transfer interference that confounds SIF signals. We concluded that integrating canopy-level active fluorescence measurements with passive SIF is essential for accurate mechanistic interpretation and physiological validation of SIF signals, especially under future, more frequent extreme climatic events.

133 CO₂-carbon sink strength in a warm-temperate Southern Hemisphere peat bog

Poster

David Campbell¹, Aaron Wall¹*, Jordan Goodrich², Jack Pronger², Louis Schipper¹

¹University of Waikato, Hamilton, New Zealand. ²Bioeconomy Science Institute, Hamilton, New Zealand

Session

Session 15: Peatlands in the global climate system: fluxes, feedbacks, and human pressures

Abstract text

Kopuatai natural wetland in Aotearoa New Zealand is the largest remaining intact raised peat bog covering around 100 km². Between 2012 and 2025, eddy covariance measurements of carbon and water fluxes were made near the centre of Kopuatai in an area dominated by the peat-forming species Empodism robustum. Kopuatai sits in a warm temperate maritime climatic zone conducive for year-round growing conditions. The hydrological regime is characterised by a shallow, stable water table (mean water table depth of −92 mm) enabled by low evaporation rates (average of 579 mm y⁻¹). During the 14-years of measurements, net ecosystem production (NEP) was a total net uptake of 41.4 t C ha⁻¹. Annual NEP during these 14 years increased from ~2 t C ha⁻¹ y⁻¹ in 2012 to ~4 t C ha⁻¹ y⁻¹ by 2025. This increase in CO₂ exchange coincided with warming in the adjacent marine coastal area that led to air and peat temperatures increasing by 1 °C during the study period. There was a significant (P < .05) positive relationship between annual air temperature with both annual NEP and gross primary production, but not ecosystem respiration. This 14-year period has provided a “natural” regional-scale warming experiment whereby Kopuatai has responded with increased carbon sequestration. While it is unclear whether carbon sequestration would occur with ongoing warming (and at what rate), it is encouraging that the ecosystem has been resilient to the moderate regional climatic warming that has observed over the past 14 years.

134 Soil sampling vs carbon balance: detecting management-related soil C change on New Zealand dairy farms

Oral

Aaron Wall¹*, David Campbell¹, Seager Ray¹, Jordan Goodrich², Louis Schipper¹

¹University of Waikato, Hamilton, New Zealand. ²Bioeconomy Science Institute, Hamilton, New Zealand

Session

Session 12: Comparing long-term eddy covariance measurements in terrestrial ecosystems with carbon stock variations: lessons and future challenges

Abstract text

Understanding the impact agricultural management practices have on soil carbon changes has been a persistent theme of New Zealand’s research agenda for the past several decades. Using eddy covariance measurements of CO₂ exchange coupled with quantification of carbon imports and exports, more than 70 site-years of net ecosystem carbon balances (NECB) have been collected across a range of management interventions on dairy farms. We showed that the NECB of altering sward composition and diversity was carbon (C) neutral to a small loss of C, pasture renewal and transitioning between pasture and crop led to moderate C losses, while periodic cropping (one-off crops in long-term pastures) led to large C losses. Using periodic cropping as an example, we examined whether direct soil C sampling could confirm the measured NECB changes (i.e. a change of ~5 t C ha⁻¹ before vs. after the crop). By analysing the variability from 30 cores taken per paddock per sampling, power analysis suggested that the change in soil C stock would have needed to be between 14 and 24 t C ha⁻¹ to be detectable using soil sampling. In contrast, the NECB approach allowed for the detection of changes in C of typically between 1 and 2 t C ha⁻¹ yr⁻¹. Thus, we conclude that the NECB method is more suitable than direct soil sampling for measuring management-related changes in soil C of the magnitude expected under New Zealand dairy farming (typically ±3 t C ha⁻¹, but larger for cropping events).

136 Seasonal Soil Temperature Dynamics and Permafrost Sensitivity in High-Latitude Regions

Poster

YEONJIN SON*, Jonghun Kam

POSTECH, POHANG-SI, Korea, Republic of

Session

Session 9: Carbon processes in high latitude/high altitude ecosystems under climate change

Abstract text

Soil temperature dynamics play a critical role in regulating premfrost stability and ecosystem processes in high-latitude regions. Understanding how seasonal thermal dynamics propagate through the soil profiles is therefore essential for interpreting ecosystem responses to climate change. This study investigates long-term, depth-resolved soil temperature observations across northern Eurasia to examine seasonal asymmetry and vertical responses in the subsurface. By analyzing multi-decadal observations from multiple soil layers, the research focuses on identifying how seasonal warming and cooling phases vary with depth and how these changes relate to surface climate conditions. 

In addition, soil temperature variations are examined together with near-surface air temperature and snow conditions in order to better understand the processes linking surface climate forcing and subsurface thermal dynamics. These analyses aim to identify observational indicators of subsurface thermal change that can help evaluate how well the Earth system models represent seasonal thermal processes in permafrost regions. 

By focusing on seasonal dynamics and vertical thermal structure, this study contributes to improving our understanding of how subsurface thermal processes influence permafrost sensitivity and ecosystem responses in high-latitude environments under ongoing climate warming. 

 By integrating soil temperature with near-surface air temperature and snow conditions, the research identifies key indicators of subsurface thermal changes relevant to climate model evaluation. The results aim to improve understanding of how seasonal processes control permafrost sensitivity under ongoing climate change. 

137 Optimal extension of the current CO₂ observation network using Incremental Optimisation

Poster

Aparna Aparajita¹*, Prabir K. Patra², Ravi K. Kunchala¹

¹Centre for Atmospheric Sciences, IIT Delhi, New Delhi, India. ²Research Institute for Global Change, JAMSTEC, Yokohama, Japan

Session

Session 26: Designing the ideal global greenhouse gas monitoring network

Abstract text

The current greenhouse gas (GHG) measurement network remains very sparse (~200 sites) compared to global meteorological observation sites. The Global Greenhouse Gases Watch (G3W), an initiative by the World Meteorological Organisation (WMO), aims to produce distributions of surface fluxes at 1×1 degree spatial resolution and monthly intervals in near real-time. High-precision GHG measurements are expensive, given the stringent accuracy and long-term stability goals set by the WMO of ~0.2 ppm (Rolling Review of Requirements, OSCAR database). Thus, expanding the measurement network, particularly into regions currently inaccessible to established groups, requires careful spatial planning. In this study, we identify the most suitable observational sites by analysing characteristics derived from remote sensing observations of surface parameters and infrastructure availability. More than 300 CTBTO International Monitoring System stations were evaluated for their suitability as GHG observational sites. Approximately 50 optimal site locations were identified on top of base observation networks of 50 and 142 sites, achieving flux uncertainty reductions (FUR) of 62% and 58% in 84-region inversions, respectively. A FLEXPART-based analysis ranks candidate sites using footprint analysis and locational details to avoid proximity to high-emission facilities. We recognise that "background" conditions vary across observation parameters; for instance, agricultural lands and forest ecosystems exhibit distinct sensitivities for CH₄, N₂O, and CO₂. Emission maps convolved with these flux fields identify relatively background sites for all three major GHGs. These results provide a robust framework for prioritising future GHG network expansion under the WMO G3W initiative, improving global flux estimates and supporting climate policy.

138 Regional atmospheric inversions for the monitoring of continental to national scale CO2 LULUCF fluxes and fossil fuel emissions in Europe

Oral

Audrey Fortems-Cheiney¹*, Elise Potier¹, Hannah Allen², Antoine Berchet², Isabelle Pison², Philippe Peylin², Vladislav Bastrikov¹, Philippe Ciais², Tianqi Shi², Frédéric Chevallier², Yilong Wang³, Rona Thompson⁴, Samuel Hammer^5,6, Grégoire Broquet²

¹Science Partners, Paris, France. ²Laboratoire des Sciences du Climat et de l’Environnement, Gif-sur-Yvette, France. ³State Key Laboratory of Tibetan Plateau Earth System, Environment and Resources (TPESER), Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing, China. ⁴NILU – Norsk Institutt for Luftforskning, Kjeller, Norway. ⁵Institute of Environmental Physics, Heidelberg University, Heidelberg, Germany. ⁶ICOS Central Radiocarbon Laboratory, Heidelberg University, Heidelberg, Germany

Session

Session 18: Quantifying anthropogenic greenhouse gas emissions from continental to regional scales

Abstract text

Continental to national scale CO₂ atmospheric inversions generally target land ecosystem fluxes, whose uncertainties dominate those in the fossil fuel emissions with the current observation networks. The fossil fuel emissions are monitored locally with observations and models dedicated to specific hotspots. Furthermore, the inversions usually ignore the carbon lateral transfer and emissions associated to wood and crop harvesting, and leaching from soils to river, lakes and seas. This hampers comparisons to the LULUCF emission inventories. Developments are thus needed to ensure regional inversions can support the national inventories of anthropogenic CO₂ emissions.

We present a synthesis of CO₂ regional inversion studies over Europe aiming to provide land ecosystem fluxes matching the scope of the LULUCF inventories, and fossil fuel emissions. They are based on CO₂ observations from the European/ICOS surface network and from the NASA/JPL OCO-2 satellite mission, and on the coupling between the Community Inversion Framework (CIF) and the CHIMERE atmospheric chemistry transport model. Combining estimates of sinks and sources associated to harvests and river discharge with flux components from the ORCHIDEE vegetation model in the pre and post processing of the prior and posterior inversion estimates increases the consistency with LULUCF emission inventories. The use of the intensive sampling of radiocarbon of the ICOS network, and of satellite observations of co-emitted pollutants, supports the derivation of fossil fuel emissions. The analysis highlights current limitations (uncertainties in annual budgets of the LULUCF sink, in the multi-tracer approaches), but also provide positive perspectives for the modeling and observations.

139 Accelerating CO₂ and CH₄ source changes in European coastal observatories revealed by STILT and in situ observations

Poster 

Jose Adame¹*, Ute Karstens², Maurizio Busetto³, Giorgio Tranchida⁴, Margarita Yela⁵, Tuula Aalto⁶, Paolo Cristofanelli³

¹Atmospheric Sounding Station. El Arenosillo observatory. National Institute for Aerospace Technology (INTA)., Mazagon. Huelva, Spain. ²ICOS Carbon Portal. Department of Earth and Environmental Sciences. Lund University., Lund, Sweden. ³National Research Council of Italy. Institute for Atmospheric Science and Climate (CNR-ISAC)., Bologna, Italy. ⁴National Research Council of Italy. Institute for the study of anthropogenic impacts and sustainability in the marine environment (CNR-IAS)., Granitola, Italy. ⁵Atmospheric Research and Instrumentation Branch. National Institute for Aerospace Technology (INTA)., Torrejón de Ardoz. Madrid., Spain. ⁶Finnish Meteorological Institute., Helsinki., Finland

Session

Session 6: Carbon cycle in the Mediterranean region: from the local to the regional scale

Abstract text

Regional carbon dioxide (CO₂) and methane (CH₄) sources were analyzed at two European coastal observatories, El Arenosillo (ARN, Spain) and Capo Granitola (CGR, Italy), using in situ measurements combined with STILT model simulations over the period 2006–2023. Agreement between observations and model outputs was robust, with correlation coefficients ranging from 0.599 to 0.900 for CO₂ and 0.693 to 0.859 for CH₄. Footprint analysis indicated that ARN was predominantly influenced by Atlantic maritime airflows, the Iberian inland, and the western Mediterranean, whereas CGR was affected by central Mediterranean circulation, southern Italy, Sicily, and North Africa. Background contributions dominated the total signal (~98%), while regional sources accounted for biogenic CO₂ (0.81–0.91%), fossil fuel CO₂ (0.47–0.56%), anthropogenic CH₄ (~1.4%), and natural CH₄ (0.25–0.29%). Although total CO₂ and CH₄ concentrations increased in accordance with global trends, source-specific analyses revealed significant changes between 2006–2010 and 2019–2023. Overall concentration trends accelerated by factors of 1.6 for CO₂ and 2.9–3.6 for CH₄. Biogenic CO₂ at ARN increased 22-fold, from 0.14 ± 1.08 to 3.03 ± 1.07 ppm dec⁻¹, whereas at CGR biogenic contributions shifted from negative to positive trends. Natural CH₄ trends weakened by 46% at ARN and reversed sign at CGR, indicating a transition toward reduced sink strength or enhanced emissions. Anthropogenic CH₄ emissions were stabilized or slightly reduced. These results highlight rapidly changing greenhouse gas source dynamics in European coastal regions, reflecting the influence of climate feedbacks, and emphasize the need for broader-scale observational studies.

140 Tropical Urban Green Area fluxes on Local climate in Tropical City, Bangkok, Thailand

Poster

Parkin Maskulrath*

Kasetsart University, Bangkok, Thailand

Session

Session 17: Urban greenhouse gas emissions and sinks: from advanced monitoring to source identification and impact

Abstract text

Tropical urban flux measurements are applied to support the use of green spaces to mitigate UHI and regulate the urban carbon cycle. This study quantifies energy and carbon dioxide (CO₂) fluxes across multiple urban green spaces in Bangkok, Thailand, with particular focus on the Bang Kachao mega urban green area (16 km²), where newly established eddy covariance (EC) systems were deployed in 2026. The flux data were used alongside a long-term EC system at the urban Kasetsart University (KU-Tower) site and smaller parks. Flux data were combined with surface temperature and mobile surveys to assess urban flux dynamics.

The results showed that larger parks exhibited higher latent heat (LE) and reduced sensible heat (H) fluxes. Chatuchak Park recorded the highest 24-hour average LE of 60.06 ± 18.66 W/m², contributing to localized cooling of 0.33 ± 0.02 °C. Conversely, smaller parks surrounded by high aspect-ratio street canyons, such as Santiphap Park, showed minimal cooling effects, with nearly equal H and LE fluxes (46.35 ± 10.81 W/m²). Restricted air circulation and limited vegetation produced warming of 2.64 ± 1.15 °C.

Initial analysis of the diurnal CO₂ flux patterns at Bang Kachao reveals positive nighttime fluxes (3–9 µmol m⁻² s⁻¹), indicating ecosystem respiration, while negative midday fluxes (−5 to −6 µmol m⁻² s⁻¹) demonstrate active photosynthetic carbon uptake, resulting in a net daily carbon sink when compared to smaller urban parks. These findings emphasize critical role of large urban green spaces in enhancing thermal regulation and carbon mitigation in tropical megacities.

141 An AI-Driven Framework for Efficient Plume Analysis of Open-Path N₂O, CH₄, and NH₃ Measurements

Poster

Weihao Shen¹*, Ruisheng Jiang¹, Ting-Jung Lin², YIn Wang¹

¹HealthyPhoton Technology, Ningbo, China. ²Ningbo Institute of Digital Twin, Ningbo, China

Session

Session 34: Manufacturers' session

Abstract text

Accurate characterization of greenhouse gas (CH₄, N₂O) and atmospheric pollutant (NH₃) plumes is essential for quantifying point-source emissions and understanding regional carbon–nitrogen cycling. However, current plume analysis workflows remain limited by manual plume identification, subjective background correction, and signal desynchronization in closed-path systems caused by sorption effects of polar molecules such as NH₃.

Here we present an automated plume analysis framework driven by generative artificial intelligence for multi-species open-path gas measurements. The framework performs end-to-end transformation of raw observation sequences into structured analytical outputs, integrating plume detection, background removal, feature extraction, and emission rate inversion. High-fidelity input data are provided by a self-developed open-path quantum cascade laser spectrometer, which delivers inherently synchronized 10 Hz measurements of CH₄, N₂O, and NH₃ and eliminates sampling-line artifacts.

Field validation demonstrates that the AI-driven workflow reduces data interpretation time from hours to minutes while maintaining high analytical robustness. In industrial and wastewater treatment environments, transient plumes of CH₄ (up to 7539 ppb) and NH₃ (background enhancement of ~37 ppb) were detected. The system further identified coherent multi-species signatures (r = 0.62, p < 0.01; lag ±1 s) and extracted representative source fingerprints (CH₄/NH₃ ≈ 10), with emission estimates validated through controlled release experiments. Real-time analytical feedback also enables adaptive mobile sampling strategies to capture stochastic emission events.

This integrated framework provides an efficient digital approach for real-time plume characterization and supports improved monitoring of industrial emissions and greenhouse gas sources.

142 Regional variability and enhanced trends of atmospheric methane in the Atlantic–Mediterranean transition region from TROPOMI and GOSAT observations

Poster

Jose Adame¹*, Robert Parker^2,3, Pablo Hidalgo⁴, Margarita Yela⁵

¹Atmospheric Sounding Station. El Arenosillo observatory. National Institute for Aerospace Technology (INTA)., Mazagon. Huelva, Spain. ²National Centre for Earth Observation, University of Leicester, Leicester, United Kingdom. ³Earth Observation Science, School of Physics and Astronomy, University of Leicester., Leicester, United Kingdom. ⁴Department of Integrated Sciences, Center for Natural Resources, Health and Environment (RENSMA), University of Huelva,., Huelva, Spain. ⁵Atmospheric Research and Instrumentation Branch. National Institute for Aerospace Technology (INTA)., Torrejón de Ardoz. Madrid, Spain

Session

Session 6: Carbon cycle in the Mediterranean region: from the local to the regional scale

Abstract text

The spatial distribution and temporal evolution of atmospheric methane (CH₄) in the Atlantic–Mediterranean transition region of southwestern Europe were investigated using TROPOMI and GOSAT satellite observations. TROPOMI XCH₄ data revealed marked spatial variability across the Iberian Peninsula, with enhanced mixing ratios over major river valleys and the southern plateau, areas strongly influenced by agricultural and livestock activities. A marine–continental transition zone in southwestern Iberia was identified as a regional CH₄ hotspot. Seasonal variability and regional trends were analysed within this area. The lowest XCH₄ levels were observed over a semi-natural mountainous region (1871.9 ± 18.5 nmol mol⁻¹), whereas the highest growth rate (10.9 ± 0.7 nmol mol⁻¹ yr⁻¹, 2019–2023) was detected there. In contrast, an inland wetland showed the highest mean XCH₄ (1888.7 ± 15.8 nmol mol⁻¹) but a lower regional trend (6.4 ± 0.9 nmol mol⁻¹ yr⁻¹). Long-term GOSAT observations (2009–2023) indicated a mean regional trend of 8.7 ± 0.3 nmol mol⁻¹ yr⁻¹. A comparison between the first and last five years of the record revealed a marked acceleration in XCH₄ growth, with trends increasing from 6.5 ± 0.9 to 14.9 ± 1.2 nmol mol⁻¹ yr⁻¹, corresponding to an acceleration ratio of 2.3. Similar enhancement patterns were identified over the adjacent Atlantic Ocean and central Mediterranean, with corresponding ratios of 2.7 and 2.4, respectively. These findings highlight the complex regional dynamics of methane and the importance of satellite observations to understand its evolving budget. Future studies will focus on identifying potential sources in Western Europe based on land-use mapping and TROPOMI observations.

143 Working towards an independent benchmark of EC water fluxes with a water bucket model

Poster

Anne van Poecke*, Simon De Cannière, Ivan Janssens

University of Antwerp, Antwerp, Belgium

Session

Session 22: Remote sensing and flux observations to link disturbance, ecosystem states, and carbon-water fluxes

Abstract text

Here we present an analysis of the water balance at the Brasschaat ICOS site. The standard suite of ICOS observations (evapotranspiration, precipitation, groundwater depth and soil moisture profiles) will be combined with sap flow-, dendrometer- and GNSS-VOD data to constrain the role of transpiration. Jointly, these data will enable to parameterize a water bucket model to estimate water drainage. We also present a new type of instrument that will be installed at the site in Autumn to measure horizontal groundwater flows. If these instruments work well, the wealth of data collected at the site would enable a strong and independent benchmarking of the eddy covariance-derived latent heat flux.

144 Forest–atmosphere exchange of O₃ and NO: multi-level flux evidence from a temperate deciduous forest of the Po valley

Oral

Davide Plebani^1,2*, Angelo Finco¹, Riccardo Marzuoli¹, Bart Muys², Giacomo Alessandro Gerosa¹

¹Department of Mathematics and Physics, Università Cattolica del Sacro Cuore, Brescia, Italy. ²Department of Earth and Environmental Sciences, KU Leuven, Leuven, Belgium

Session

Session 24: Exchange of reactive gases and aerosols between the land surface and the atmosphere in natural and managed ecosystems

Abstract text

Ozone (O₃) and nitric oxide (NO) are key reactive trace gases regulating atmospheric chemistry and air quality. O₃ is a major pollutant that reduces forest productivity, particularly in Mediterranean regions, thereby weakening the forest carbon sequestration. Accurate quantification of O₃ dry deposition is therefore essential for vegetation risk assessment. Forest soils are among the largest natural sources of NO, although size and distribution of emissions remain uncertain. The reaction between soil-emitted NO and O₃ represents an efficient but often overlooked non-stomatal sink, potentially contributing up to ~20% of total O₃ deposition. Simultaneous measurements of O₃ and NO are thus crucial to disentangle chemical, depositional, and emission processes within and above forest canopies.

This study examines O₃ and NO dynamics in a temperate mixed oak–hornbeam forest at the ICOS site of Bosco Fontana (IT-BFt, Italy). In 2025, measurements combined eddy-covariance fluxes of O₃ and NO above and below the canopy, vertical concentration profiles (2, 8, 32, and 40 m), and soil fluxes using dynamic chambers.

Results highlight the role of soil-derived NO in forest–atmosphere O₃ exchange. Following summer rainfall, elevated soil NO emissions increased near-surface NO concentrations (0.1–2 m), with no comparable increase aloft, indicating rapid O₃ titration within the lowest meters. During high-emission periods, nighttime low-O₃ conditions occasionally produced positive NO fluxes at 8 m, suggesting temporary accumulation and upward transport.

Overall, findings emphasize the importance of soil-emitted NO in enhancing non-stomatal O₃ removal in forests, highlighting the need to integrate this process into dry deposition parameterizations.

145 Methane emissions in northern peatlands: Integration of process-based model and multi-site observations

Oral

Mousong Wu^1,2*, Wenzhuo Duan¹, Matthias Peichl³, Hongxing He⁴, Nigel Roulet⁴, Sara Knox^4,5, Per-Erik Jansson⁶

¹International Institute for Earth System Science, Nanjing University, Nanjing, China. ²Frontiers Science Center for Critical Earth Material Cycling, Nanjing, China. ³Department of Forest Ecology and Management, Swedish University of Agricultural Sciences, Umeå, Sweden. ⁴Department of Geography, McGill University, Montreal, Canada. ⁵Department of Geography, The University of British Columbia, Vancouver, Canada. ⁶KTH Royal Institute of Technology, Stockholm, Sweden

Session

Session 15: Peatlands in the global climate system: fluxes, feedbacks, and human pressures

Abstract text

Northern peatlands are crucial carbon reservoirs (~1/3 of global soil C) and the largest natural CH₄ source. Projecting their CH₄ emissions remains challenging due to complex environmental controls, limited observations, and model simplifications, yet the contribution of non-growing season emissions remains poorly quantified. These gaps limit accurate peatland-climate feedback projections.

To address this, we enhanced the CoupModel for comprehensive gas process representation in soil and plants. We simulated O₂, CO₂, and CH₄ fluxes alongside energy/water fluxes at pristine northern peatlands along a Boreal-Arctic gradient. This study: (1) identified key parameters controlling CH₄ emissions, evaluated model performance, and revealed hysteresis in CH₄-temperature responses; (2) explored shoulder-season emission mechanisms across a freeze-thaw gradient; (3) compared active versus passive restoration impacts on CH₄ fluxes.

The calibrated model reproduced hourly CH₄ fluxes well (R²: 0.18–0.60). CH₄-temperature dependence showed seasonal hysteresis, with 40% higher temperature sensitivity (Q10=5.2) during below-average water table depths. Diurnally, nighttime fluxes exceeded daytime fluxes by 14–23%. Shoulder-season (autumn freeze & spring thaw) emissions contributed up to 24.8% of annual totals, with autumn emissions up to 15 times higher. While phase duration varied along the gradient, emission intensity was consistently higher during autumn freeze. Model projections indicate active restoration enhances net CO₂ uptake (-137 g C m⁻² yr⁻¹ by 2100), whereas passive rewetting increases CO₂ emissions (90 g C m⁻² yr⁻¹). Both sites show increased future CH₄ emissions, with the actively-restored site emitting more (26 vs. 10.8 g C m⁻² yr⁻¹).

146 Introducing Physics-Informed Transpiration Estimates: using physics-informed machine learning to partition eddy covariance-measured evapotranspiration

Oral

Emma Cochran*, Elke Eichelmann

University College Dublin, Dublin, Ireland

Session

Session 13: Advancing approaches for quantifying greenhouse gas fluxes in terrestrial ecosystems

Abstract text

As a key component of gas exchange between terrestrial ecosystems and the atmosphere, transpiration (T) connects global water, carbon, and climate patterns. While evapotranspiration (ET) measurements are widely available, largely thanks to global flux networks, ground-truth T data collection is time and labor intensive, and thus relatively few datasets exist. This disparity highlights the need to estimate ecosystem T from eddy covariance-measured ET to better understand ecosystem water dynamics. Physics-informed machine learning (PI-ML) constrains models with physical laws and ecosystem assumptions, presenting a solution for producing these estimates without extensive T data for model training and validation. Here, Physics-Informed Transpiration Estimates (PITE), a PI-ML model, is introduced to estimate T/ET at a soybean agricultural site in Ontario, Canada. Using eddy covariance data, PITE assumes T follows a sinusoidal curve over a diel cycle and is bound by the assumptions that T is negligible overnight and that evaporation is negligible when the ecosystem experiences its hottest and driest conditions. PITE estimates were validated against leaf-level T and soil evaporation lysimeter measurements from the 2025 growing season and benchmarked against other ET partitioning methods. Seasonal trends were evaluated alongside leaf area index, photosynthetic activity, and relevant meteorological and flux variables, including differences in seasonal T dynamics between a conventional and diverse cropping rotation and between corn and soybean systems. The physically realistic T estimates produced by PITE reflect the model's ability to capture the underlying physical processes driving ET partitioning across varied agricultural settings, supporting better water management and food security.

147 Carbon capture and using modern biotechnology to make products in demand by society which can replace fossil carbon sources and help reduce fossil carbon emisisons.

Poster

Fleur de Haas¹*, Theo Verleun²

¹GOA-Ventures, Ede, Netherlands. ²GOA-ventures, Ede, Netherlands

Session

Session 34: Manufacturers' session

Abstract text

Seaweed bioproccesing leads to a complete new coastal economy. 

Climate change effects will make seaweed more abundantly available and as such a new biomass for bio-processing comes available. Mother earth is putting algae, like seaweeds, in the race to battle CO₂ increase to stop further acidification of the oceans. The growth of seaweed goes together with capturing of CO₂ and as such is a corrective action in the acidifying of the oceans. 

If we can now use this seaweed to make products which are in demand by society and which can replace products which are notorious for the cause of releaseing fossil CO₂ into the atmosphere we are able to put the breaks on CO₂ emisisons.

Hence, we want to present the biopreocessing of seaweeds and make out of each kilo of seaweed both food-grade functional protein (supporting protein tranition from animal proteins to plantbased proteins) as well as a renewable energy source; Biogas (supporting the energy transition from fossil fuels towards renewable energy). The bioprocess itself can be operated with part of the produced renewable erengy. 

Being able to master such technology offers a 3 fold benefit in carbon management; a) captuing by seaweed growth b) protein traniston & c) energy transition. This mulitplier will substantially contribute to putting the breaks on fossil CO₂ emisisons.

in addition to these effects on CO₂, such developments will create new and alternative jobs in coastal area's and more prosperity for coastal communities, who are suffering from decline in fish-stocks and its associated jobs.

148 Advancing Grassland Process Representation in the dynamic vegetation model LPJ-GUESS: Evaluation Against ICOS Observations

Oral

Anna-Kristina Voss¹*, Stefan Olin¹, David Wårlind¹, Torbern Tagesson¹, Jürgen Knauer², Jonas Ardö¹

¹Lund University, Lund, Sweden. ²University of technology Sydney, Sydney, Australia

Session

Session 14: Greenhouse gas fluxes in agroecosystems: processes, measurements and management implications

Abstract text

Grassland ecosystems dominated by herbaceous plants provide essential ecosystem services, most prominently food provision, and play a key role in the global carbon cycle and in sustaining biodiversity. They are therefore central to both human well-being, and Earth system functioning. Despite their ecological importance, grassland ecosystems and herbaceous species are insufficiently represented in dynamic global vegetation models, limiting our ability to project their responses to climate and land-use change. For example, a realistic representation of phenology of grassland ecosystems is needed for analysing management practices including grazing, cutting and ploughing. However, many vegetation models lack key grass-specific processes that determine vegetation persistence and regrowth both in unmanaged and managed grasslands. As a consequence, it remains difficult to analyse feedbacks between grassland vegetation, soils and the atmosphere and the effect management has on them.

   In this study, we compare the performance of the dynamic global vegetation model LPJ-GUESS against ICOS sites. We advanced the representation of herbaceous species in the model by implementing a daily carbon allocation scheme, grass-specific traits for i.e. allometry and life-cycle dynamics to better capture growth and seasonal development of grassland vegetation. We evaluated the effects of the new model representation on biomass accumulation, carbon stocks and fluxes, both in absence of management and under different management scenarios.  These developments allow for a more mechanistic simulation of grassland responses to different management regimes.

149 The contemporary global terrestrial carbon cycle - a systemic model-data fusion analysis

Poster

Luke Smallman, David Milodowski, Mathew Williams*

University of Edinburgh, Edinburgh, United Kingdom

Session

Session 19: Understanding feedbacks between greenhouse gas exchange processes and climate variability using in situ observations, remote sensing, and machine learning

Abstract text

The terrestrial carbon (C) cycle is critical to climate regulation, provisioning ecosystem services and biogeochemical cycles. Development of climate mitigation strategies requires rigorous, uncertainty bounded, systemic information on C-dynamics and their underpinning ecological processes. Significant effort has been invested to quantify contemporary C-cycling through use of specific observations of land surface properties. These are often incomplete in space and time, contain poorly characterised errors, and lack internal consistency checks, making the construction of a reliable global C budget a significant challenge. 

We present a rigorous, global (0.5×0.5 deg), multi-decadal (2003-2024), analysis of the terrestrial C-cycle at a monthly time step. We use a state-of-the-art Bayesian model-data fusion framework (CARDAMOM) to calibrate a model of terrestrial ecosystems (DALEC) using ecologically relevant spatio-temporal observations (e.g. leaf area, biomass, soil C), and forcing (meteorology, CO2, disturbance). The resulting systemic analysis is a pixel-scale calibration. CARDAMOM uniquely propagates observational uncertainties through to DALEC's parameters (i.e. ecosystem properties) and simulates C pools and fluxes for each pixel across the vegetated land surface. 

The key scientific result is that current global multi-decadal datasets are inadequate to provide >95% confidence on the sign of net exchange across 80% of the vegetated land surface. However, our main objective is to enhance accessibility of this open-access dataset through a thorough description of its key features: calibration and evaluation metrics, photosynthate allocation fractions and tissue residence times, and their uncertainties. These outputs can inform the evaluation of land surface models, and identify key uncertainties in C-cycling to target them. 

150 Seasonal variability and quantification of methane emissions across Africa

Oral

Ines Kamoun¹*, Aurélien Sicsik-Paré¹, Eldho Elias^1,2, Marielle Saunois¹, Adrien Martinez¹, Isabelle Pison¹, Grégoire Broquet¹, Antoine Berchet¹

¹Laboratoire des Sciences du Climat et de l’Environnement, LSCE/IPSL, CEA-CNRS-UVSQ, Université Paris-Saclay, Gif-sur-Yvette, France. ²Max Planck Institute for Biogeochemistry, Jena, Germany

Session

Session 7: CO2 and CH4 cycle dynamics in Africa: observations, processes, and climate implications

Abstract text

The African continent contributed approximately 14% to the total annual methane (CH₄) emissions between 2010 and 2019 (Saunois et al., 2025) with a high increase in emissions between 2019 and 2024 (He et al, 2025). Moreover, CH₄ emissions and concentrations over Africa are subject to pronounced seasonal variability. The largest amplitudes occur in the equatorial region and parts of Eastern Africa, particularly near South Sudan. This variability is mainly driven by wetland emissions, which are the main source of methane in Africa, accounting for approximately 40% of total emissions. However, current CH₄ emission estimates over Africa remain highly uncertain due to sparse observational constraints and the complexity of several emission processes. For instance, bottom-up inventories for equatorial Africa emission estimates range from 28 to 83 Tg/year according to the Global Methane Budget (Saunois et al., 2025).

In this study, we use TROPOMI satellite observations to assess the changes in CH₄ mixing ratios across different African regions between 2019 and 2024. We quantify the increase rate of CH₄ mixing ratios and focus on its spatial heterogeneity at the regional scale. Our results reveal that the tropical region shows the most pronounced growth in methane concentrations. Moreover, using the Community Inversion Framework (CIF) coupled with the chemistry transport model CHIMERE at a spatial resolution of 0.3°×0.3°, we quantify CH₄ emissions for the continent and investigate the drivers underlying the seasonal variability of CH₄ mixing ratios. We focus mainly on the seasonal emissions pattern of agriculture, wetlands and biomass burning.

151 Assessing PROSAIL model performance to estimate Carbon-Based Constituents in alfalfa

Poster

Giacomo Panza^1,2*, José Luis Pancorbo³, Aldo Dal Prà³, Lorenzo Brilli³, Carlos Camino⁴, Cinzia De Benedictis^1,2, Camilla Chieco¹, Federico Carotenuto¹, Giandomenico De Luca³, Simona Maccherini², Beniamino Gioli³

¹CNR-IBE, Bologna, Italy. ²University of Siena, Siena, Italy. ³CNR-IBE, Firenze, Italy. ⁴Wageningen University and Research, Wageningen, Netherlands

Session

Session 22: Remote sensing and flux observations to link disturbance, ecosystem states, and carbon-water fluxes

Abstract text

Alfalfa (Medicago sativa L.) is a key fodder crop valued for its high nutritional quality. Its forage quality largely depends on the balance between protein content and structural carbohydrates such as lignin, cellulose, and hemicellulose. These compounds belong to the broader group of Carbon-Based Constituents (CBCs), which also includes sugars and starch and represents an important plant biophysical variable. Because CBC content varies across growth stages in response to metabolic activity and climatic conditions, their monitoring is essential to improve estimates from modelling approaches and to support management decisions.

In this study, a methodological framework combining Sentinel-2 satellite imagery with a hybrid Artificial Neural Network (ANN)– Radiative Transfer Model (PROSAIL) inversion approach, implemented through the ToolsRTM R package, was developed to estimate CBC concentrations across different alfalfa growth stages in two fields in Italy over the 2023–2025 period.

Results indicate that PROSAIL-derived CBC estimates from Sentinel-2 reflectance spectra were consistent with ground-truth laboratory measurements collected in 2025 (R² = 0.75). The CBC estimated across years showed that the lowest values occurred at the early vegetative stage (~2.46 mg cm⁻²), followed by the late bud stage (~3.32 mg cm⁻²). The highest concentrations were observed from mid-vegetative to early bud stages (~4.42 mg cm⁻²), while early flowering (pre-cut) showed slightly lower but still elevated values (~3.57 mg cm⁻²). These results confirm the reliability of the hybrid ANN–PROSAIL approach for CBC retrieval in alfalfa.

152 Assessing CO₂ fluxes in an Irish grassland ley and a continuous tillage system

Poster

Lyubov Bragina¹*, Mary Bridget Lynch¹, Michele McCormack¹, George Gleasure¹, Ryan Burger², Mark Boland¹, Rebecca Hall¹, Rachael M. Murphy²

¹Teagasc, Agricultural Catchments Programme, Wexford, Ireland. ²Teagasc Climate Centre, Wexford, Ireland

Session

Session 14: Greenhouse gas fluxes in agroecosystems: processes, measurements and management implications

Abstract text

Under the 2030 Climate Framework, Ireland must reduce LULUCF emissions by 37–58% from 2018 levels, with up to 6% offset by carbon sinks. Grasslands and arable land cover approximately 59% and 12% of the country, respectively, making them key targets for carbon removal. As soil carbon dynamics vary with climate, soil type, and management, agricultural soils can act as either a source or sink of CO₂.  This study quantified CO₂ NEE and NECB at cropland and grassland sites in South-East Ireland over two years. The cropland site was managed for spring barley, while the grassland site had been converted from arable production three years before the study and has since been managed under a zero-grazing (ZG) system. CO₂ fluxes were measured at both sites using the eddy covariance technique, where soil and climatic conditions were comparable. Detailed field management data included dates of key soil management practices, fertiliser types and rates, and grain, straw, and grass yields for both sites. Preliminary results showed that grassland under the ZG system had higher photosynthetic carbon uptake and ecosystem respiration than the cropland system. Both sites were net carbon sinks based on NEE alone, with values of −613.55 g C m^-2 y^-1 and −520.42 g C m^-2 y^-1 in 2023 and 2024 for grassland, and −322.90 g C m^-2 y^-1 and −321.61 g C m^-2 y^-1 for the arable site. However, carbon export through biomass removal offset the NEE sink, indicating that only ZG grassland functioned as a long-term carbon sink.

153 GHG observations in the Bologna urban area (northern Italy)

Poster

Paolo Cristofanelli¹*, Elisa Castelli¹, Andrè Achilli^1,2, Maurizio Busetto¹, Francescopiero Calzolari¹, Simonetta Montaguti¹, Enzo Papandrea¹, Paolo Pettinari¹, Matteo Rinaldi¹

¹Consiglio Nazionale delle Ricerche - Istituto di Scienze dell'Atmosfera e del Clima, Bologna, Italy. ²Bologna University, Bologna, Italy

Session

Session 17: Urban greenhouse gas emissions and sinks: from advanced monitoring to source identification and impact

Abstract text

Northern Italy is among the most polluted and densely populated regions in Europe. The diverse land use of the Po Basin makes it a major contributor to greenhouse gas emissions from multiple sectors. Medium and large cities, together with highly industrialized areas, contribute substantially to emissions from industrial processes, fuel combustion, waste management, and natural gas distribution.

Bologna is a medium‑sized city of 180 km², with a population of 391,810, located at the foothills of the Apennines. Urban emissions from official bottom‑up inventories remain poorly characterized. 

From April 2024 to July 2025, continuous in‑situ measurements of CO₂ and CH₄ were carried out in Bologna at 20 m a.g.l. on the roof of the CNR‑ISAC building in the northern suburbs using a CRDS analyser. At the same time, vertical columns of CO₂, CH₄, and CO were measured with an EM27‑SUN spectrophotometer within the COCCON network. Local and basin‑scale wind circulations strongly influenced the observations: nighttime conditions were generally affected by air masses from the city centre, whereas during daytime a substantial contribution from the wider Po Basin was observed.

This study characterizes the diel and seasonal variability of atmospheric CO₂ and CH₄ in Bologna’s urban environment. We analyse differences associated with air masses influenced by the city centre versus those originating from the Po Basin plain. Finally, we compare temporal patterns captured by in‑situ and column measurements, discussing their complementary value and limitations for improving constraints on urban greenhouse gas emissions.

154 Can crop management change albedo in Swedish Agriculture?

Poster

Jonas Ardö, Veronika Widengren*, Patrik Vestin

Earth and Envoronemntal Science, Lund University, Lund, Sweden

Session

Session 22: Remote sensing and flux observations to link disturbance, ecosystem states, and carbon-water fluxes

Abstract text

Albedo impacts surface energy balance and varies with agricultural practices. Management choices of crop type and coverage period influence soil exposure and erosion, soil carbon sequestration, nutrient leakage as well as surface energy balance via albedo. Hence, albedo‑related co‑benefits may also arise when using cover crops to reduce nutrient losses by the use of cover crops.

     We calculate broadband albedo from time series of Copernicus Sentinel-2 bi-directional reflectance distribution function  corrected top-of-canopy reflectance to minimize disturbances from the atmosphere and snow. We derive daily blue-sky albedo from MODIS combined with diffuse radiation data from SMHI’s STRÅNG dataset. Using these time series of Sentinel-2 and MODIS albedo we compare albedo of annual crops versus albedo of perennial crops in order to quantify management effects on albedo in south Swedish agriculture.

     In situ data from ICOS agricultural sites are used for calibration and validation. Special consideration is given to our intensively monitored site in Alnarp, where we evaluate annual and perennial crops in terms of climate mitigation potential (carbon sequestration and albedo).

     To assess variations over time we calculate trends of annually integrated albedo (Sentinel-2) of individual agricultural fields for the study period in southern Sweden. We relate detected trends to land management to investigate the climate mitigation potential related to albedo. Methodological problems and observational uncertainties are discussed.

155 From Disturbance to Restoration: Tracking Ecosystem Recovery in a Mediterranean Forest

Poster

Anjali Thapa¹*, Gabriele Guidolotti², Giacomo Nicolini³, Simone Sabbatini³, Leonardo Ancillotto⁴, Dario Capizzi⁵, Daniele Cecca⁶, Dora Cimini⁶, Olivia Dondina⁷, Adriana Mariotti³, Michele Mattioni², Francesco Mazzenga⁸, Emiliano Mori⁴, Riccardo Salvati⁶, Paolo Sconocchia², Giulia Bonella⁶, Carlo Calfapietra², Giorgio Matteucci⁹, Giuseppe Scarascia Mugnozza¹, Dario Papale¹

¹Department for Innovation in Biological, Agro-Food and Forest System (DIBAF), University of Tuscia, Viterbo, Italy. ²National Research Council of Italy, Research Institute on Terrestrial Ecosystems (CNR-IRET), Porano, Italy. ³Euro-Mediterranean Center on Climate Change (CMCC), Lecce, Italy. ⁴National Research Council of Italy, Research Institute on Terrestrial Ecosystems (CNR-IRET), Sesto Fiorentino, Italy. ⁵Regione Lazio, Rome, Italy. ⁶Segretariato Generale della Presidenza della Repubblica – Servizio Tenuta di Castelporziano, Rome, Italy. ⁷University of Milano-Bicocca, Milan, Italy. ⁸National Research Council of Italy, Institute of BioEconomy (CNR-IBE), Rome, Italy. ⁹National Research Council of Italy, Institute of BioEconomy (CNR-IBE), Sesto Fiorentino, Italy

Session

Session 6: Carbon cycle in the Mediterranean region: from the local to the regional scale

Abstract text

At the Castelporziano Presidential Natural Reserve, located 25 km southwest of Rome, Italy, the combined outbreak of two invasive pests, Toumeyella parvicornis and Tomicus destruens, has caused almost the complete destruction of about 1000 ha of Pinus pinea forest, of which nearly 600 ha were mostly monospecific, in less than six years. While the loss is devastating, this widespread dieback and subsequent felling have created a rare opportunity to establish a supersite under Mediterranean climate. 

The supersite integrates three new monitoring plots along with long term ICOS ecosystem station, in collaboration with major European research infrastructures (ICOS, eLTER, LifeWatch). Among the four plots, two represent post pine reforestation areas, one following active plantation approach with some untouched patches and the other natural evolution. The remaining two are reference plots: a mature deciduous oak forest and a mature evergreen oak forest monitored by the existing ICOS station. Each plot is equipped with an eddy covariance system for continuous measurements of CO₂, water and energy exchanges, with the new towers operational since August 2024. The primary objective is to compare the pace and functional trajectory of carbon recovery, as reflected by net ecosystem carbon exchange and its component fluxes between the restoration areas, using mature forests as functional references. Flux measurements are integrated with continuous meteorological monitoring and vegetation assessments, enabling the concurrent recovery of water and energy fluxes under different pathways. Biodiversity is also monitored across plots. This comprehensive approach establishes a practical framework for monitoring recovery following extreme disturbance. 

156 Bridging Tower Observations and Satellite Estimates: Evaluating MODIS GPP in a Tibetan Alpine Grassland

Poster

Nithin D Pillai^1,2*, Christian Wille¹, Manuel Helbig¹, Torsten Sachs^1,3

¹GFZ Helmholtz Centre for Geosciences, Potsdam, Germany. ²Institute of Geosystems and Bioindication, Technische Universität Braunschweig, Braunschweig, Germany. ³Institute of Geoecology, Technische Universität Braunschweig, Braunschweig, Germany

Session

Session 13: Advancing approaches for quantifying greenhouse gas fluxes in terrestrial ecosystems

Abstract text

One of the potential causes of the discrepancies between satellite-based and eddy covariance (EC) -derived gross primary productivity (GPP) estimates lies in the spatial scale mismatch between time-varying EC flux footprints and fixed satellite pixels, but the importance of this mismatch for satellite-EC comparisons has rarely been tested quantitatively. In this study, we quantified the role of spatial representativeness in explaining discrepancies between moderate resolution imaging spectroradiometer (MODIS) GPP and EC-derived GPP in an alpine steppe ecosystem on the Tibetan Plateau (TP), a region of high ecological and climatic importance. MODIS GPP underestimated EC-derived GPP at this site, with a bias and RMSE of −8.73 and 9.19 g C m⁻² 8 d⁻¹, respectively. During periods of active vegetation, MODIS GPP showed moderate temporal agreement with EC-derived GPP (R² = 0.60) but a large relative bias (−65.3 %). Restricting EC observations to the footprints exhibiting high vegetation similarity to the MODIS pixel improved the temporal agreement (R² = 0.69), but the substantial negative bias persisted (−9.16 g C m⁻² 8 d⁻¹; −66.3 %). These results indicate that the spatial mismatch between EC footprints and MODIS pixels is unlikely to be the cause of the underestimation. Instead, algorithmic biases in MODIS GPP and temporal scaling differences between satellite and tower measurements are likely the main contributors to the lower MODIS GPP at this site. This underestimation highlights the need for improved, high-resolution satellite-based carbon flux products for Tibetan alpine grasslands.

157 Communicating ecosystem-atmosphere exchanges: An illustrated exploration of eddy covariance

Poster

Susanne Wiesner*

Technical University of Denmark, Kongens Lyngby, Denmark

Session

Session 33: Science and arts: How to communicate science?

Abstract text

Understanding how ecosystems exchange carbon, water, and energy with the atmosphere is central to climate research, yet these ideas often remain abstract for wider audiences, as eddy‑covariance concepts can be difficult to grasp. I am creating a comic that uses visual storytelling to introduce ecosystem processes through the lens of land-atmosphere exchanges and the fluxes they generate. The comic seeks to explain the eddy covariance method in a way that makes a complex measurement approach approachable and interesting. It shows why the technique is essential for understanding ecosystem processes, detecting climate‑driven changes, and supporting our ability to adapt to shifting environmental conditions. Rather than focusing on technical detail, the narrative uses the method as a framework for helping readers build an intuitive sense of how ecosystems function and how flux patterns reflect environmental change. By combining clear explanations with a story‑driven format, the comic aims to make the science behind ecosystem-atmosphere exchanges more accessible and memorable.

In an era when AI‑generated imagery is increasingly common, the project also reflects on the value of intentionally crafted scientific illustration for communicating nuance, maintaining trust, and giving science communication a distinctive voice. Although I am not a professional illustrator, the comic is hand‑drawn to emphasize clarity, intentionality, and a personal scientific perspective. By presenting this work, I hope to gather feedback both on how illustrated narratives can support learning about ecosystem-atmosphere interactions and on how the comic itself can be improved as a piece of science communication.

158 Tuning the turbulence parameterisation of a NWP model for atmospheric transport: a case study using ICOS stations

Poster

Diego Jiménez de la Cuesta*, Navid Mouji, Jochen Förstner, Andrea Kaiser-Weiss

Deutscher Wetterdienst, Offenbach am Main, Germany

Session

Session 28: Combining data and models to support emissions estimation and policy at local to regional scales

Abstract text

Given that greenhouse-gas emissions are usually located at surface, turbulence has an important role in the vertical distribution of these emissions and their subsequent large-scale transport, determining the concentrations of the emitted gas and its vertical profile at observational sites.

ICOS stations measure at different levels the concentration of greenhouse gases, as well as basic meteorological variables such as wind speed, temperature and humidity. Usually, the measurement heights lie within or near the boundary layer, depending on the time of the day and season.

In numerical weather prediction (NWP) models, like the German Weather Service's ICON-ART, and at resolutions finer than 10 km, turbulent flows in complex terrain become important when we look at the atmospheric transport of tracers. In this study, we focus on a set of ICOS stations that are representative of (i) different topographic complexity and (ii) the prevailing direction of the large-scale flow. We compare with model equivalents of these ICOS stations in ICON-ART in Limited Area Mode at a ~6.5 km resolution grid over Europe.

We analyse how the variation of relevant tuning parameters of the turbulent parameterisation of ICON-ART improve or worsen the representation of meteorological variables. We find a range of parameters that (i) physically makes sense, (ii) could reduce the model-observation mismatch, and (iii) provide a basis for an ensemble with perturbed turbulence settings that provide a better constraint of the transport error. Additionally, we characterise regions that need higher resolutions for representing turbulent flows as atmospheric transport needs.

159 NEE in relation to climate change and extreme events in a Danish beech forest during three decades.

Poster

Kim Pilegaard*, Susanne Wiesner, Andreas Ibrom

Technical University of Denmark, Lyngby, Denmark

Session

Session 11: Climate feedbacks from ecosystem management and natural disturbances

Abstract text

We  present a detailed analysis of 30 years of continuous CO₂ fluxes measured over a deciduous forest in Denmark.  We show that there is a highly significant trend of increased carbon uptake period and increased net ecosystem exchange (NEE) of CO₂. 

We relate the findings to climate change, especially increasing temperature and CO_2-concentrations in the atmosphere. We also study the effects of extreme weather events, such as summer droughts and heat waves on NEE. In the most severe case the annual NEE was reduced by 25%, compared to a predicted normal year.

When we started our measurements in 1996, the forest was very close to neutral with respect to CO₂ exchange. Using past local climate data such as rainfall, temperature and radiation, supplemented by remote sensing vegetation indices and land surface temperature, we try to make an estimate of the NEE in the years preceding our measurements. The results show that in the years before we started our measurements, the atmospheric climate characteristics were less favourable for carbon sequestration in the forest.

160 Characterising sub-daily vegetation water dynamics across European forest ecosystems using a GNSS transmissometry network

Poster

Nathan Van der Borght¹*, Susan Steele-Dunne¹, Emma Tronquo^1,2, Anna Selina Neyer¹, Paco Frantzen¹, Rob Mackenzie¹, Hans van der Marel¹

¹TUDelft, Delft, Netherlands. ²Ghent University, Ghent, Belgium

Session

Session 20: Unmanned autonomous vehicles and proximal sensing in greenhouse gas research and monitoring

Abstract text

Sub-daily vegetation water dynamics are critical indicators of ecosystem resilience to heat and drought, modulating carbon-water coupling from hours to seasons. Yet these dynamics remain poorly constrained at ecosystem scale. Microwave remote sensing offers a promising pathway, but current satellite missions lack the spatiotemporal resolution to resolve diurnal cycles at the site level.

To bridge this gap between in-situ monitoring and remote sensing, we are using GNSS transmissometry (GNSS-T): a proximal sensing approach to estimate vegetation optical depth (VOD), a microwave proxy for canopy water content, from the attenuation of GNSS signals between above- and below-canopy receivers. The resulting high-resolution time series can track the daily filling and draining of canopy water storage.

We deployed 21 GNSS receivers across nine European forest sites (May–September 2025), primarily co-located with ICOS Ecosystem stations, spanning a climatological and ecological gradient from northern Finland to central Italy. Pairing GNSS-T with eddy-covariance fluxes, meteorology, and soil moisture allows us to relate canopy water changes to water availability and evaporative demand.

Early results reveal diurnal VOD dynamics with site-specific responses to vapour pressure deficit and latent heat flux, which may offer new insights into ecosystem-scale hydraulic capacitance. Looking ahead, continuous GNSS-T records can guide future spaceborne microwave missions by quantifying the diurnal cycle of a microwave-derived proxy for vegetation water content, informing sampling strategies, and characterising sensitivity across forest types. By integrating novel sensors within the ICOS network, this study demonstrates how proximal sensing can advance our understanding of ecosystem carbon-water coupling.

161 Determining the role of the South Baltic in the atmosphere-ocean CO₂ balance: effect of the EF(SA) on the EC CO₂ fluxes

Poster

Iwona Wrobel-Niedzwiecka*, Małgorzata Kitowska, Violetta Drozdowska, Przemysław Makuch, Karol Kulinski, Fernando Aguado Gonzalo, Jacek Piskozub

Institute of Oceanology, Polish Academy of Sciences, Sopot, Poland

Session

Session 1: How big is the open ocean carbon sink?

Abstract text

CO₂ is a natural constituent of Earth’s atmosphere, accounting for approximately 0.04%, yet it remains one of the principal regulators of the global climate system. Global estimates of air–sea CO₂ fluxes rare still associated with substantial uncertainty, largely because they inadequately represent the pronounced spatial heterogeneity of coastal seas, the complexity of coupled physical–biogeochemical interactions, and the contribution of episodic extreme events.

To address this  gap, we conducted air–sea gas-exchange observations during four research cruises in the central and southern Baltic Sea (November 2022; May 2023; May 2024; October 2024). The campaigns comprised high-frequency atmospheric and oceanic observations required for both the eddy covariance (EC) and bulk flux approaches, complemented by water sampling to quantify the enrichment factor (EF) of surface-active substances in the sea-surface microlayer.

A seasonal comparison between spring (typically lower wind speeds) and autumn (typically higher wind speeds) indicates that the gas transfer velocity (k) is more robustly parameterised as a function of friction velocity u_* under well-developed turbulence (R²= 0.44–0.63 in autumn) than as a function of neutral 10 m wind speed U_10N alone (R² = 0.13–0.18), as u_* can catch variability than U_10N alone can not see (surface stress, wave state, wave breaking). Furthermore, the results demonstrate a systematic damping effect of EF(SA) on k, with reductions of up to 40%.

162 Quantifying nitric oxide (NO) fluxes from vegetable cropland: High-frequency eddy covariance measurements with chemical loss correction

Poster

Yuting Zhang¹*, Dong Wang², Kai Wang¹, Xunhua Zheng¹

¹Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China. ²National Marine Data and Information Service, Tianjin, China

Session

Session 24: Exchange of reactive gases and aerosols between the land surface and the atmosphere in natural and managed ecosystems

Abstract text

Agricultural soils are an important but highly uncertain source of reactive nitrogen oxides (NO_x). Quantifying soil nitric oxide (NO) emissions from intensively managed croplands remains challenging due to the high spatial and temporal variability. 

In this study, continuous NO flux measurements were conducted over a subtropical vegetable cropland in the middle reaches of the Yangtze River, China, using the eddy covariance (EC) method during a two-month period (August 4–October 3, 2016). Parallel static chamber measurements were also performed. A simplified chemical loss correction approach was developed to account for rapid NO-O₃-NO₂ reactions between the soil surface and the measurement height using concentrations of NO, NO₂, O₃ and footprint-based transport time estimates.

NO fluxes exhibited strong temporal variability, characterized by a pronounced fertilization-induced emission pulse and a single-peak diurnal pattern. Chemical reaction correction increased EC fluxes by approximately 30%. Chamber measurements captured the temporal dynamics but produced substantially lower cumulative emissions than EC estimates. The cumulative NO emission measured by EC during the observation period corresponded to 0.54% of the applied nitrogen (~330 kg N ha^-1), whereas chamber measurements estimated only 0.17%.

These results demonstrate the large emission pulses following fertilization in intensive vegetable systems. It helps improve emission estimates from agricultural soils and provide new insights into nitrogen cycling under high nitrogen input management.

163 High greenhouse gas emissions after grassland destruction

Poster

Iris Feigenwinter¹*, Yi Wang^2,1, Lukas Hörtnagl¹, Fabio Turco¹, Regine Maier³, Nina Buchmann¹

¹ETH Zürich, Zürich, Switzerland. ²Lund University, Lund, Sweden. ³Swiss Federal Institute for Forest, Snow and Landscape Research WSL, Birmensdorf, Switzerland

Session

Session 11: Climate feedbacks from ecosystem management and natural disturbances

Abstract text

Grassland destruction, here defined as the renewal or renovation of permanent grassland or the termination of temporary grassland within a crop rotation, is a common agricultural management practice in Europe. While typically employed to improve sward quality or prepare the soil for subsequent crops, grassland destruction can also substantially disturb the ecosystem’s carbon and nitrogen cycle, leading to increased carbon dioxide (CO₂) losses and nitrous oxide (N₂O) emissions. Yet the magnitude of these emissions at ecosystem scale and the role of management timing and intensity remain poorly understood. Using eddy covariance measurements, we aimed to identify the effects of grassland destruction timing and intensity on N₂O fluxes, and to quantify the ecosystem GHG budgets (CO₂, N₂O, methane (CH₄)) for three grassland destruction events. N₂O fluxes during grassland destruction accounted for 18 to 66% of annual N₂O emissions. They increased with fertilizer N inputs, but also with the Soil Tillage Intensity Rating (STIR), an indicator for soil disturbance. Annual GHG budgets in the year of grassland destruction were -109, 807, and 2122 g CO₂ eq m^-2 yr^-1, representing a wide range from a small GHG sink to a large GHG source. Thus, despite beneficial impacts on vegetation composition and yields, grassland destruction can temporarily transform agroecosystems into large GHG sources. These findings underscore the need to consider these events when assessing the climate change mitigation potential of grasslands. In addition, they need to be incorporated into GHG models to ultimately support the development of sustainable agricultural management and related policies.

164 A New Tall Tower to Fill The Observational Gap In UK Greenhouse Gas Emissions Verification

Poster

Chris Rennick¹, Adam Wisher², Cameron Yeo^1,3, Sara Defratyka^4,2, Emmal Safi¹, Dafina Kikaj¹, Tom Gardiner¹, Jagadeesh Yeluripati⁵, Tim Arnold^6,2, Mat Williams²*

¹NPL, Teddington, United Kingdom. ²University of Edinburgh, Edinburgh, United Kingdom. ³University of Bristol, Bristol, United Kingdom. ⁴LSCE - UVSQ, Gif-sur-Yvette, France. ⁵James Hutton Institute, Aberdeen, United Kingdom. ⁶Lund University, Lund, Sweden

Session

Session 18: Quantifying anthropogenic greenhouse gas emissions from continental to regional scales

Abstract text

Continuous measurements of greenhouse gases (GHG) and tracer species in the well-mixed troposphere are essential for providing the high-quality observations required for inversion modelling used in regional scale emissions verification. Scotland has represented an observational gap in the UK greenhouse gas monitoring network. The existing UK tall tower network is located at telecommunications masts, established as the Deriving Emissions Linked to Climate Change (UK-DECC) network. However, the location of these sites offers limited sensitivity to emissions from Scotland.  

A new purpose-built 100 m tall tower has been commissioned as the first dedicated GHG observatory in the UK, located on the Balruddery Research farm, Invergowrie. The tower is equipped with pumped inlets at 100 m and 50 m above ground level enabling continuous measurements of key GHGs (methane, carbon dioxide, carbon monoxide and nitrous oxide) and isotopic tracers to help identify emissions sources. Meteorological sensors for wind speed at the inlet heights, and temperature at 10 m intervals, support interpretation of atmospheric transport modelling. The vertical sampling provides information on the boundary layer dynamics and enable the resolution of Scottish GHG sources and sinks. Measurements are traceable to WMO scales and follow measurement and quality control procedures that ensure compatibility across the UK network and ICOS atmospheric stations. 

This poster will present the scientific motivation, and preliminary results from the first year of operation, demonstrating added sensitivity this observatory provides to Scottish regional emissions and contribution to UK emissions verification. 

165 From leaves to canopies, chambers to EC: revisiting scaling  up photosynthesis assuming optimality

Poster

Bart Kruijt¹*, Wilma Jans¹, Gijs Hogeboom¹, Lyanne Van den Berg², Rich Norby³, Corine van Huissteden¹, Laurent Bataille¹, Ruchita Ingle¹, Tan Lippman¹, Oscar Hartogensis¹, Patrick Meir⁴, Tomas Domingues⁵, Sabrina Garcia⁶

¹Wageningen University, Wageningen, Netherlands. ²Louis Bolk Instituut, Bunnik, Netherlands. ³University of Tennessee, Knoxville, USA. ⁴University of Edinburgh, Edinburgh, United Kingdom. ⁵University of Sao Paulo, Riberao Preto, Brazil. ⁶Instituto Nacional de Pesquisas da Amazonia, Manaus, Brazil

Session

Session 13: Advancing approaches for quantifying greenhouse gas fluxes in terrestrial ecosystems

Abstract text

Modelling whole-canopy GPP as well as quantifying GPP in complex or small vegetation stands simplifies greatly if and where theories on optimal use of light by photosynthesis apply, through acclimation of photosynthetic capacity to the average local light environment. In these cases, leaf photosynthetic capacity scales linearly with absorbed PPFD, such that integrated canopy GPP scales linearly with canopy FAPAR. GPP would then linearly relate to photosynthesis of fully sunlit leaves. This elegant optimality hypothesis has been an object of study for many decades. Here we will review and revisit the idea, using both older and new data sets from diverse ecosystems.

Despite its simplicity, testing the optimality hypothesis in real vegetation has proven challenging. Apart from technical difficulties such as determining the mean light environment, other factors seem to also play a role in optimal response of photosynthesis to the environment. Where verification against eddy covariance fluxes is tried, the representative footprints are hard to match and respiration has to be scaled up as well.

We will analyse some relatively recent data sets that allow scaling up and testing optimality, including data from Amazon rain forest, peatlands and swamp forest but also from a long series of annual student field courses in grasslands and pine forests in Europe. Finally, we will consider practical approaches to estimate canopy GPP by scaling up.

166 Linking Eddy Covariance Flux Measurements with Satellite Observations to Characterize Coastal CO₂ Exchange in the Southern Baltic Sea

Poster

Małgorzata Kitowska*, Iwona Niedźwiecka

Institute of Oceanology, Polish Academy of Sciences, Sopot, Poland

Session

Session 22: Remote sensing and flux observations to link disturbance, ecosystem states, and carbon-water fluxes

Abstract text

In the southern part of the Baltic Sea - a region characterized by limited observational accessibility, strong seasonal variability, and strong influence of coastal processes - consistent, long-term direct measurements of gas exchange between the ocean and atmosphere remain scarce. This observational gap limits our ability to quantify regional carbon budgets and to evaluate satellite-derived estimates of carbon fluxes in coastal environments.

To address this challenge, a pilot field campaign was conducted between February and May 2025 to assess the feasibility of continuous CO₂ flux monitoring over the south Baltic Sea and to explore the potential of integrating in-situ flux observations with remote sensing products. The eddy covariance (EC) CO₂ station was installed on the rooftop of the Institute of Oceanology of the Polish Academy of Sciences in Sopot. The EC dataset was complemented with independent meteorological observations, including wind speed, wind direction, and air temperature, enabling the characterization of atmospheric transport conditions and flux footprint variability. To place the in-situ observations in a broader spatial context, the measurements were evaluated alongside satellite-derived products from the SatBałtyk system. This integration as a first step toward building a long-term observational framework allows preliminary assessment of how satellite observations can support the interpretation of coastal flux dynamics and improve the spatial representativeness of site-level measurements. The results highlight the complexity of carbon exchange in coastal zones, where marine processes, atmospheric transport, and urban influences interact.

167 Comparative Analysis of Carbon Dioxide (CO₂), Methane (CH₄) and Ozone (O₃) Concentrations over two urban cities in Northern Nigeria: A case study of Kano and Katsina

Poster

Muawiya Sani¹*, Maryam Idris², Rabia Sa'id Salihu²

¹Centre for Atmospheric Research, Anyigba, Nigeria. ²Bayero University Kano, Kano, Nigeria

Session

Session 7: CO2 and CH4 cycle dynamics in Africa: observations, processes, and climate implications

Abstract text

Urban centers in developing regions are increasingly important contributors to atmospheric greenhouse gas emissions and ozone-related air pollution, yet long-term observational analyses remain limited across sub-Saharan Africa. This study presents a comparative assessment of carbon dioxide (CO₂), methane (CH₄), and ozone (O₃) over two major northern Nigerian cities; Kano and Katsina using datasets from GEOS-Chem (CO₂), the Atmospheric Infrared Sounder (AIRS; CH₄), and MERRA-2 (O₃) for the period 2015–2021. Monthly concentrations were analyzed using descriptive statistics, seasonal climatologies, and the Modified Mann-Kendall (MMK) trend test combined with Sen’s slope estimator. CO₂ exhibited statistically significant increasing trends in both Kano (Z = 8.421, p < 0.000001) and Katsina (Z = 8.376, p < 0.000001), with Sen’s slope magnitudes of 2.46 ppm yr⁻¹ and 2.45 ppm yr⁻¹, respectively. Mean CO₂ concentrations increased from 399 ppm in 2015 to 417 ppm by 2021, consistent with global atmospheric growth rates. Methane displayed statistically significant upward trends in Kano (Z = 6.315, β = 0.0114 ppm yr⁻¹) and Katsina (Z = 7.022, β = 0.0127 ppm yr⁻¹), with stronger seasonal variability than CO₂. In contrast, ozone showed weak positive slopes that were not statistically significant (p > 0.05), although clear seasonal cycles were observed with dry-season maxima and wet-season minima. Kano consistently exhibited higher CO₂ and CH₄ concentrations than Katsina, reflecting differences in urbanization, industrial activity, and traffic emissions. These findings highlight the growing influence of anthropogenic emissions in northern Nigeria and emphasize the importance of methane mitigation for climate stabilization and regional air-quality management.

168 The Internet of the Environment (IoE) Ecosystem – From Eddy Covariance Greenhouse Gas Flux Measurements to Cloud-Based Post-Processing and Flux Spatialization

Poster

Frank Griessbaum*, Taylor Thomas, James Kathilankal, Sasha Ivans, George Burba, Tyler Barker, Daniel Singer, Andrew Parr, Israel Begashaw, Jason Hupp

LI-COR Environmental, Lincoln NE, USA

Session

Session 34: Manufacturers' session

Abstract text

The Internet of the Environment (IoE) is a subset of the Internet of Things (IoT) in which environmental sensors with embedded power and communications, continuously monitor and report on the state of the natural environment. Choosing an IoE approach over a legacy infrastructure approach enables higher densities of measurements in space, which allows for better representation of the spatial heterogeneity across the Earth system. 

This approach is supported by low-cost Carbon and Water Nodes, based on Eddy Covariance (EC) sensors, including the LI-720 (CO₂, H₂O, ETa, LE, H) and LI-710 (ETa, LE, H), complemented by a suite of onboard and wirelessly connected biometeorological sensors. Traditional Eddy Covariance systems based on LI-COR Smartflux will also be IoE-enabled, reporting to the same IoE cloud infrastructure.

The IoE Ecosystem combines edge and cloud processing tools enabling a data pipeline that transforms raw sensor data into analysis-ready-data (ARD). This transformation occurs on the sensors themselves (within the sensing-layer) and in the cloud where computing capacity is not limited. Derived data products, from agricultural and natural ecosystems, provide insights into the quality of the flux data, estimates for gap-filled data during periods of time when the sensors were not operational, flux accumulation, flux footprints, and spatialized fluxes to extend the estimate of fluxes to a broader spatial area.

This presentation provides an updated overview of the IoE Ecosystem, covering its sensing-layer hardware, cloud integration, and online flux post-processing data pipeline.

169 Hypertemporal LiDAR Monitoring of Typha Biomass Dynamics at the Rewetted Peatland Site Zarnekow

Poster

Christoph Lotz*, Christian Wille, Manuel Helbig, Tatjana Zivkovic, Torsten Sachs, Martin Herold, Benjamin Brede

GFZ Helmholtz Centre for Geosciences, Potsdam, Germany

Session

Session 20: Unmanned autonomous vehicles and proximal sensing in greenhouse gas research and monitoring

Abstract text

Rewetted peatlands can play a key role in climate mitigation strategies and integral parts of the Integrated Carbon Observation System (ICOS) network. While eddy covariance systems provide continuous measurements of ecosystem-scale carbon fluxes, complementary temporally resolved three-dimensional observations of vegetation structure in peatlands are extremely rare. Camera-based approaches are limited to chromatic information and surface proxies, whereas LiDAR captures direct geometric and volumetric properties of vegetation. High-frequency structural LiDAR data are largely missing, limiting the ability to link vegetation structural development to observed flux variability.

We present a hypertemporal (sub-hourly) proximal sensing system deployed at an ICOS-associated rewetted peatland site dominated by Typha spp., designed to capture canopy structural dynamics and derive aboveground biomass proxies. An automotive-grade LiDAR sensor was adapted for autonomous ecological monitoring and configured to repeatedly scan a defined vegetation sector. The resulting three-dimensional point clouds allow extraction of canopy height distributions, vertical density profiles, and volumetric metrics at sub-daily resolution, providing unprecedented temporal detail of vegetation growth.

Analyses focus on capturing the green-up phase, when rapid structural changes occur and biomass accumulation is most pronounced. LiDAR-derived biomass estimates will be compared to eddy covariance measurements to explore their potential (i) to contextualize observed flux dynamics and (ii) to serve as structural predictors for carbon exchange. By providing temporally resolved biomass information, the system offers a pathway to interpret flux variability within the peatland and to explore upscaling from the eddy footprint to the broader ecosystem.

170 Reach out with your outreach!

Poster

Rolf Niemann*

Vattenhallen Science Center, Lund, Sweden

Session

Session 33: Science and arts: How to communicate science?

Abstract text

Vattenhallen Science Center at Lund University is a meeting place for science, educators, schools, and the general public.

Researchers play an important role in the development of new exhibitions and visitor activities. We will present some of the collaborations we have had in recent years.

We offer various ways to communicate your research, for example by participating in events such as European Researchers’ Night, the Young Scientists Competition, joining a school visit, or opening your research station to visitors.

171 Temporal variability and transport patterns of greenhouse gases (CO₂, CH₄) and CO in West Africa: the case of the Lamto geophysical station

Poster

Touré Dro TIEMOKO^1,2*, Michel RAMONET^3,4, Fidèle YOROBA^5,2, Lopez Morgan³, Kobenan Benjamin KOUASSI^5,2, Marc DELMOTTE³, Adama DIAWARA^5,2

¹Université Nangui ABROGOUA, Abidjan, Côte d'Ivoire. ²Station Géophysique de Lamto, N'Douci, Côte d'Ivoire. ³Laboratoire des Sciences du Climat et de l'Environnement, CEA-CNRS-UVSQ, Université Paris-Saclay, Gif-Sur-Yvette, France. ⁴CNRS, Gif-Sur-Yvette, France. ⁵Université Félix Houphouët Boigny, Abidjan, Côte d'Ivoire

Session

Session 7: CO2 and CH4 cycle dynamics in Africa: observations, processes, and climate implications

Abstract text

West Africa is particularly vulnerable to climate change, even though regional greenhouse gas (GHG) dynamics remain poorly documented. Understanding temporal variations and transport mechanisms of the main GHGs is a major scientific challenge for improving regional carbon balances and supporting climate policies. The Lamto geophysical station (Côte d'Ivoire) offers a unique opportunity to study atmospheric concentrations of CO₂, CH₄ and CO in West Africa.

This study aims to characterise the temporal variability (diurnal, seasonal and interannual) of these gases and to identify the source-sink relationships influencing their regional concentrations. Analyses are based on ten years of continuous high-precision measurements, coupled with Lagrangian air mass modelling using the FLEXPART model. A clustering approach applied to back-plumes (PES) has enabled the classification of atmospheric transport regimes and the assessment of the relative contribution of different source regions.

Results show significant seasonality, with peaks during dry season (December–February), related to biomass fires and the influence of Harmattan winds. Long-term increase rates for CO₂ (~2.24 ppm.year⁻¹) and CH₄ (~7 ppb.year⁻¹) are comparable to global trends. Cluster analysis highlights four main transport regimes, with continental air masses from the north and north-east accounting for nearly 40% of the variance in observed concentrations. The ΔCO/ΔCH₄ and ΔCO/ΔCO₂ ratios confirm the importance of combustion processes.

This work contributes to a better understanding of regional atmospheric dynamics in West Africa. Future prospects include strengthening the regional observation network to improve the quantification of GHG fluxes on an African scale.

172 An Automated Forest Monitoring LiDAR (FML) for High-Temporal Resolution Forest Structure Observation

Poster

Christoph Lotz¹*, Clemens Nothegger², Peter Dorninger², Martin Herold¹, Benjamin Brede¹

¹GFZ Helmholtz Centre for Geosciences, Potsdam, Germany. ²4D-IT, Wien, Austria

Session

Session 20: Unmanned autonomous vehicles and proximal sensing in greenhouse gas research and monitoring

Abstract text

High-temporal-resolution observations of forest canopy structure remain rare, as conventional terrestrial laser scanning (TLS) campaigns are episodic and labor-intensive. To address this gap, we developed an automated monitoring LiDAR system, designed for continuous operation and acquisition of near-TLS-quality three-dimensional scans under field conditions.

The system integrates a scanning LiDAR unit with automated control and environmental protection, enabling repeated structural surveys without manual intervention. Installed permanently in a forest stand, it generates multi-temporal point clouds at predefined intervals, producing consistent and reproducible structural time series. In contrast to most currently deployed monitoring LiDAR scanners, the system provides higher spatial resolution while maintaining autonomous operation and compatibility with established protocols. The scanner is meant to be deployed as part of StrucNet, which aims to establish a network of ground-based laser scanners to monitor vegetation structure dynamics across ecosystems through time. The presented system contributes to this framework by delivering structurally detailed observations suitable for integration into such networks.

Hypertemporal observations enable quantification of canopy phenology, leaf area index dynamics, and fine-scale changes in canopy architecture. Multi-temporal LiDAR captures detailed three-dimensional structural development that is not accessible from episodic TLS campaigns alone, allowing investigation of dynamic canopy processes at timescales relevant for carbon and water exchange.

By combining autonomous operation with high-resolution structural data, this monitoring LiDAR provides a robust tool for long-term forest structure observation and supports the integration of structural information into ecosystem research infrastructures such as StrucNet and ICOS.

173 Connecting neighborhood canopy coverage to biogenic CO2 uptake across Paris, France

Poster

Anni Karvonen¹*, Minttu Havu¹, Laura Bignotti², Benjamin Loubet², Leena Järvi^1,3

¹Institute for Atmospheric and Earth System Research/Physics, Faculty of Science, University of Helsinki, Helsinki, Finland. ²ECOSYS, INRAE, AgroParisTech, Université Paris-Saclay, Palaiseau, France. ³Helsinki Institute of Sustainability Science, Faculty of Science, University of Helsinki, Helsinki, Finland

Session

Session 17: Urban greenhouse gas emissions and sinks: from advanced monitoring to source identification and impact

Abstract text

The majority of global greenhouse gases are emitted in urban areas. In addition to reducing the anthropogenic sources, biogenic sinks are important in mitigating emissions. Urban green areas take up carbon dioxide (CO₂) in photosynthesis, and they also cool the microclimate. These benefits at neighborhood level are dependent on the amount of canopy coverage, but the needed extent remains yet uncertain. 

This study utilized the Surface Urban Energy and Water balance Scheme (SUEWS) model to analyze biogenic CO₂ flux variability across the greater Paris area. SUEWS simulates energy, water, and CO₂ exchanges at a local neighborhood scale. The research was conducted from March 2024 to June 2025, using the ICOS Cities eddy covariance measurements from an urban forest in Vincennes to validate the model. Initially, we examined the effects of various biogenic CO₂ parameters (such as park trees, street trees, and forests) on modeling CO₂ and heat fluxes. We scaled our findings by modeling CO₂ dynamics across the greater Paris area at a resolution of 500 m x 500 m, comparing CO₂ uptake with canopy coverage to evaluate how urban vegetation mitigates emissions at the neighborhood scale. 

174 Quantifying GHG exchange in the floodplain forest

Poster

Natalia Kowalska¹*, Adam Bednařík¹, Eva Darenova¹, Jan Vavřina^1,2

¹Global Change Research Institute CAS, Brno, Czech Republic. ²Department of Forest Ecology, Faculty of Forestry and Wood Technology, Mendel University in Brno, Brno, Czech Republic

Session

Session 13: Advancing approaches for quantifying greenhouse gas fluxes in terrestrial ecosystems

Abstract text

Floodplain forests play a significant role in regulating methane (CH₄) exchange; however, the contribution of aquatic surfaces to ecosystem-scale CH₄ fluxes remains insufficiently constrained. Climate change and anthropogenic pressures are altering key environmental drivers, particularly groundwater table dynamics and soil temperature, potentially shifting the balance between CH₄ sources and sinks. We investigated greenhouse gas (GHG) exchange in a temperate floodplain forest near Lanžhot, Czech Republic, currently characterized by relatively dry conditions and spatially distributed aquatic ecosystems (streams and ponds).

Ecosystem-scale CH₄ and N₂O fluxes were continuously quantified using the eddy covariance (EC) method and analysed in relation to emissions from aquatic ecosystems, soils and tree stems within the EC footprint. Component-specific fluxes of CH₄, CO₂, and N₂O were measured during approximately two-week intensive campaigns in each season (spring, summer, autumn and winter) from streams, ponds, adjacent soils, and tree stems using chamber-based approaches. Floating chambers and bubble traps were used to quantify diffusive and ebullitive emissions from water bodies, respectively. Soil and stem fluxes were assessed along transects extending from water bodies into the forest interior using chamber methods.

We hypothesized that the studied floodplain forest is a source of CH₄ and N₂O and that ecosystem-scale CH₄ and N₂O fluxes are largely driven by emissions from aquatic ecosystems, whereas episodic increases in groundwater level enhance the relative contribution of terrestrial components. Preliminary results support this hypothesis and improve the interpretation of EC-derived CH₄ fluxes, contributing to a more comprehensive GHG budget of temperate floodplain ecosystems.

175 Thermal imaging of forest canopy temperatures: Retrieving high fidelity and spatio-temporally resolved information for ecosystem stress monitoring

Poster

Jennifer Susan Adams¹*, Alexander Damm^1,2, Michael Niederberger¹, Kathrin Naegeli¹

¹University of Zurich, Zurich, Switzerland. ²Swiss Federal Institute of Aquatic Science and Technology (Eawag), Dubendorf, Switzerland

Session

Session 20: Unmanned autonomous vehicles and proximal sensing in greenhouse gas research and monitoring

Abstract text

Canopy temperature is a critical property of forest ecosystems, driven by processes of evapotranspiration and water- and heat transfer between soil, atmosphere and vegetation. Highly resolved spatio-temporal observations of canopy temperatures can shed light on these processes, their relationship with micro- and macro-climate, and are sensitive proxies for forest resilience to climate extremes. Thermal imagery can particularly help observe canopy temperatures at fine spatio-temporal scales, providing insight on ecosystem- and species-specific thermal niches and responses to extreme events and/or climate change.

This contribution presents a new ground observation set-up to monitor canopy temperatures over two Swiss forests (mixed and coniferous), which have been priority sites (ICOS and FLUXNET) for measuring forest gas exchange, environmental stress responses and biodiversity over many years. The instrumentation derives high-precision spatially resolved canopy and species-specific temperatures from thermal imagery complemented by pointing infrared radiometers, since its installation in 2024. Particular emphasis is placed on cascading levels of correction to ensure high-fidelity observations, including laboratory correction using blackbody calibration facilities, atmospheric and in-situ bias correction using thermal reference calibration plates. 

We show example use cases of this unique dataset for understanding a) species-specific responses to extreme summer temperatures and their relationship with abiotic factors (meteorological and eddy flux observations), b) ecosystem and species-specific diurnal temperature modelling to provide insights on canopy thermoregulation, c) ecosystem differences of Norway Spruce canopy temperatures between the two different ecosystems as well as d) their use as a thematic validation site for upcoming high-resolution satellite missions (e.g., TRISHNA, LSTM).

176 Net greenhouse gas balance in Mediterranean vineyard and walnut agroecosystems: a three-year field study

Poster

Lorenzo Brilli¹*, Marianna Nardino², Camilla Chieco², Christian Massari³, Paolo Benettin⁴

¹National Research Council of Italy - Institute of Bioeconomy, Firenze, Italy. ²National Research Council of Italy - Institute of Bioeconomy, Bologna, Italy. ³National Research Council of Italy – Research Institute for Geo-Hydrological Protection, Perugia, Italy. ⁴Department of Earth Surface Dynamics - University of Lausanne, Lausanne, Switzerland

Session

Session 14: Greenhouse gas fluxes in agroecosystems: processes, measurements and management implications

Abstract text

Agriculture, forestry and other land use contribute to about one fifth of global anthropogenic greenhouse gas (GHG) emissions. Reducing emissions from these systems therefore requires a better understanding of the environmental and management drivers controlling GHG exchange. 

Here we quantified the three-year (2023–2025) GHG balance of two Mediterranean perennial cropping systems in Italy: (i) a 1-ha organically managed vineyard in the Chianti area (43°31′N, 11°17′E) and (ii) intensive walnut (Juglans regia L.) orchard in the Po Plain (44°51′N, 11°23′E). 

The vineyard is a 50-year-old Vitis vinifera L. cv. Sangiovese plantation under organic management with Vicia faba cover cropping and green manure, whereas the walnut system represents a juvenile, fertilized intensive plantation. At both sites, net ecosystem CO₂ exchange (NEE) was continuously measured using eddy covariance (EC), while soil N₂O fluxes were quantified using static chambers.

The vineyard acted as a carbon sink, with annual net C balances of 123.5, 454.5 and 286.8 g C m⁻² yr⁻¹ (2023–2025) and low N₂O emissions (0.068 g N m⁻² yr⁻¹). The walnut orchard showed stronger C uptake (−240, −150 and −223 g C m⁻² yr⁻¹) but higher N₂O emissions associated with fertilization (0.39 g N m⁻² yr⁻¹). Overall, the three-year mean GHG balance was −1017.5 g CO₂-eq m⁻² for the vineyard and −518 g CO₂-eq m⁻² for the walnut orchard, indicating that both systems acted as net GHG sinks, although fertilization reduced the mitigation potential of the walnut system.

177 Flux processing pipeline as a source of uncertainty in ecosystem CO2 budget estimates

Oral

Santeri Satalahti¹*, Mika Aurela², Alexander Buzacott³, Pasi Kolari³, Mika Korkiakoski², Asta Laasonen³, Samuli Launiainen¹, Hannakaisa Lindqvist², Annalea Lohila^2,3, Ivan Mammarella³, Tuukka Petäjä³, Janne Rinne¹, Timo Vesala³, Aki Vähä³, Olli Peltola¹

¹Natural Resources Institute Finland (Luke), Helsinki, Finland. ²Finnish Meteorological Institute (FMI), Helsinki, Finland. ³University of Helsinki, Helsinki, Finland

Session

Session 29: Using GHG measurement data to support national greenhouse gas inventories

Abstract text

Measured carbon dioxide (CO2) balances depend on the processing of raw eddy covariance (EC) data into the final data product. The processing pipeline covers routines such as spectral corrections, quality filtering and gapfilling, that may be implemented differently in different pipelines. Hence, flux datasets processed through different pipelines are not necessarily equivalent. While individual processing choices (e.g. different gapfilling algorithms) have been cross compared, the uncertainties stemming from different processing pipelines have been less studied. 

To address this, we processed one year of raw CO2 flux data from three EC sites with three pipelines developed in Finnish research institutes and compared the resulting datasets. At all sites, annual CO2 balances differed by up to 7.2%...12.2% between the pipelines and their mean, with similar differences in partitioned fluxes. We found differences in spectral corrections, friction velocity thresholds, filtered data coverage, and between gapfilled and measured fluxes, and identified gapfilling as a key contributor to the annual balance mismatch. At ICOS site FI-Hyy, the mean of our CO2 balance estimates matched the balance calculated by ICOS ETC within 1%, but with discrepancy in filtered data coverage in ICOS data (21%) and our datasets (43%...50%).

Highlighting the differences, best practices and sources of uncertainty across pipelines supports harmonising flux processing between different observation networks. This is crucial, as the fluxes provide the ground truth for land-surface model development, carbon-cycle remote sensing, and increasingly for AI-based flux upscaling. Moreover, this allows utilising the existing research infrastructure more efficiently for e.g. national greenhouse gas inventories.

178 Integrated Proximal Sensing and Ground-Based Monitoring at Majadas de Tiétar: SIF and TIR to improve Energy Balance models

Oral

Luis Alonso Chorda¹*, Arnaud Carrara¹, Javier Pacheco-Labrador², Vicente Burchard², M. Pilar Martin²

¹CEAM: Mediterranean Center of Environmental Studies, Valencia, Spain. ²Environmental Remote Sensing and Spectroscopy Laboratory (SpecLab), Spanish National Research Council (CSIC), Madrid, Spain

Session

Session 16: Using sun-induced chlorophyll fluorescence to understand or scale EC fluxes

Abstract text

Observation capabilities of the ICOS ES-LMa ecosystem station are enhanced by implementing continuous proximal sensing and tree scale measurements to build an integrated multiscale, high-frequency (10-30 min) observational system addressing critical knowledge gaps regarding how Mediterranean tree species respond and recover from climate extremes such as compound heat and water stress. This issue is particularly relevant for Quercus ilex, the dominant tree species in the Mediterranean area, which presents high drought tolerance but currently limited understanding of its threshold responses to stress.

The main objective is to explore the potential of these combined observations to understand the processes that regulate the functional response of Mediterranean savanna to heat and water stresses, where structural heterogeneity and fast stress responses remain poorly captured by current Earth Observation systems and standard monitoring approaches, insufficient to capture short-term adaptive physiological responses occurring at hourly to minute timescales, such as stomatal closure, water transport regulation, photoprotective mechanisms, and xanthophyll cycle dynamics.

The overall infrastructure delivers continuous ecosystem-scale measurements of turbulent fluxes, meteorological variables, comprehensive soil water observations and continuous tree-scale information on tree water status and water transport dynamics. An innovative integration of state-of-the-art proximal remote sensing techniques: Thermal Imaging (TIR) to resolve spatial patterns of vegetation surface temperature variations; Sun-Induced Fluorescence (SIF) and VIS-NIR Reflectance to distinguish active photosynthesis from photoprotective responses.

The integrated dataset will also be used for: (i) improving TSEB/3SEB evapotranspiration models; (ii) contributing to upscaling strategies for the Cal/Val of ESA's FLEX mission. 

Preliminary data and results are presented.

179 Seasonal Carbon Forecasts Predicted the 2022 European Drought's Impact Months in Advance

Poster

Ruben van 't Loo¹*, Auke van der Woude¹, Manouk Vermeulen¹, Ingrid Luijkx¹, Wouter Peters^1,2

¹Wageningen University & Research, Wageningen, Netherlands. ²Rijksuniversiteit Groningen, Groningen, Netherlands

Session

Session 10: Cross-scale responses of greenhouse gas variability to climate extremes

Abstract text

To improve our understanding of the future strength of the land carbon sink we need to better understand the impacts of extreme events on carbon uptake. Current low-latency monitoring systems have identified the driving role of droughts for developing these anomalies, emphasizing the need for a better understanding of the development of droughts that occur on a seasonal timescale. Increasingly better seasonal weather forecasts now provide the opportunity to predict the impacts of extreme weather on the carbon cycle before events occur, offering a true test of our current understanding of these processes. Here, we demonstrate the novel approach of forecasting the impacts of droughts several months in advance by developing a seasonal ensemble carbon forecast using ECMWF’s seasonal forecast (SEAS5) and the SiB4 land surface model. We demonstrate the system’s power for a well-studied hindcasting case: the extreme European summer drought of 2022. We show that this system can predict drought impacts 1-3 months in advance, using ICOS flux observations and satellite data to score the prediction. Furthermore, we identify the dominant drivers and main sources of predictability from meteorological forcing, climate trends and initial conditions. These results highlight the potential of seasonal forecasting to understand the drought impacts on the carbon cycle. This study provides a stepping stone towards integrated weather and carbon forecasting on a seasonal timescale, and can act as an early warning system for land managers.

180 Recent advances in δ¹³C-CO₂ and δ¹³C-CH₄ measurements in Italy: perspectives and challenges of an established national collaborative network

Poster

Francesco D'Amico¹*, Ivano Ammoscato¹, Alcide G. di Sarra², Tatiana Di Iorio³, Salvatore Piacentino⁴, Damiano M. Sferlazzo⁵, Paolo Cristofanelli⁶, Simonetta Montaguti⁶, Isabella Zaccardo^7,8, Antonella Buono^7,8, Lucia Mona⁷, Giulia Zazzeri⁹, David Lowry¹⁰, Rebecca E. Fisher¹⁰, Thomas Röckmann¹¹, Carina van der Veen¹¹, Teresa Lo Feudo¹, Claudia R. Calidonna¹

¹Consiglio Nazionale delle Ricerche - Istituto di Scienze dell'Atmosfera e del Clima, Lamezia Terme, Italy. ²Agenzia nazionale per le nuove tecnologie, l'energia e lo sviluppo economico sostenibile, Frascati, Italy. ³Agenzia nazionale per le nuove tecnologie, l'energia e lo sviluppo economico sostenibile, Santa Maria di Galeria, Italy. ⁴Agenzia nazionale per le nuove tecnologie, l'energia e lo sviluppo economico sostenibile, Palermo, Italy. ⁵Agenzia nazionale per le nuove tecnologie, l'energia e lo sviluppo economico sostenibile, Lampedusa, Italy. ⁶Consiglio Nazionale delle Ricerche - Istituto di Scienze dell'Atmosfera e del Clima, Bologna, Italy. ⁷Consiglio Nazionale delle Ricerche - Istituto di Metodologie per l'Analisi Ambientale, Tito, Italy. ⁸Università della Basilicata - Dipartimento di Ingegneria, Potenza, Italy. ⁹Ricerca sul Sistema Energetico, Milano, Italy. ¹⁰Royal Holloway University of London - Department of Earth Sciences, Egham, United Kingdom. ¹¹Universiteit Utrecht - Institute for Marine and Atmospheric Research, Utrecht, Netherlands

Session

Session 6: Carbon cycle in the Mediterranean region: from the local to the regional scale

Abstract text

In the past few years, four Italian atmospheric stations started continuous measurements of δ¹³C-CO₂ and δ¹³C-CH₄ based on Cavity Ring Down Spectroscopy. These stations are located in four regions, and have peculiar characteristics: Lampedusa (code: LMP) is an island site south of the Strait of Sicily; Lamezia Terme (LMT) is a coastal station in Calabria, southern Italy; Monte Cimone (CMN) is a mountain site (2165 m a.s.l.) in the northern region of Emilia-Romagna; Potenza (POT) is a continental, 100 m tall tower site in the southern region of Basilicata. Preliminary studies have investigated the influence of emissions linked to fossil fuel use and a possible interaction between mineral particles originated from the Sahara desert, on the fractionation of carbon in atmospheric CH₄. In these years, activities have been performed to better characterize the analytical performance of the instrumentation and investigate their usability for continuous monitoring of δ¹³C-CO₂ and δ¹³C-CH₄. The network is currently working in conjunction with foreign institutions with the objective of introducing common reference standards, calibration, and round-robin procedures. These procedures are challenged by the logistics of shipping cylinders to locations such as the mountain site of CMN, and the island of LMP. Despite these challenges, the network is expected to introduce continuous calibrated measurements of δ¹³C-CO₂ and δ¹³C-CH₄ in 2026, effectively contributing to a more detailed understanding of carbon cycle dynamics in the central Mediterranean Basin. In this work, we will also provide an overview of δ¹³C-CO₂ and δ¹³C-CH₄ variability at these four measurement sites in Italy.

181 Divergent CO₂ emission patterns and thermal response in two “climate-neutral by 2030” Italian cities. Investigating the role of thermal breakpoints and anthropic pressure

Poster

Simone Putzolu*, Tommaso Giordano, Lorenzo Brilli, Valentina Marchi, Alessandro Zaldei, Carolina Vagnoli, Giovanni Gualtieri, Beniamino Gioli

CNR-IBE, Firenze, Italy

Session

Session 17: Urban greenhouse gas emissions and sinks: from advanced monitoring to source identification and impact

Abstract text

This study presents a four-year analysis (2021–2025) of urban eddy covariance CO₂ fluxes measured in Florence and Prato, two adjacent Italian cities participating in the EU “100 Climate-Neutral Cities” mission. Seasonal mean diurnal cycles show comparable nighttime baseline emissions (within ~5 µmol m⁻² s⁻¹), but consistently higher daytime fluxes in Florence. The intercity hourly difference peaks in winter (up to ~25 µmol m⁻² s⁻¹), remains substantial in spring and autumn (~20 µmol m⁻² s⁻¹), and persists in summer (~15 µmol m⁻² s⁻¹).

Cold-season cycles are highly synchronized (r ≈ 0.9), reflecting similar traffic and heating patterns under shared climatic forcing. In contrast, summer dynamics diverge markedly. Florence exhibits a sustained midday emission plateau, indicating how urban function shapes metabolic patterns: its historic, service-oriented center maintains higher daytime activity compared to Prato’s more residential and industrial profile. This signal is further reinforced by Florence’s strong tourism-driven transient population, which amplifies daytime emissions beyond those associated with resident inhabitants alone.

Thermal breakpoint analysis, the temperature below which heating induces a linear rise in emissions, reveals distinct city-specific signatures despite similar climate. Prato shows a lower breakpoint (16.7°C), whereas Florence responds at a higher threshold (20.4°C) and with a steeper temperature sensitivity. Above these thresholds, the weak temperature-flux relationship (R² < 0.15) indicates that summer emissions are largely decoupled from meteorological control and driven by stochastic urban activities.

These results highlight that achieving carbon neutrality requires integrating city-specific thermal responses, urban functional structure, and tourism dynamics into tailored emission mitigation strategies.

182 CH4 inverse modelling using the ICOS tall tower network and implications for emissions from the Benelux for the period 2010-2021

Poster

Ioannis Cheliotis¹*, Sander Houweling¹, Friedemann Reum², Eleftherios Ioannidis³, Dominik Brunner⁴, Joel Thanwerdas⁴, Marko Scholze⁵, Yohanna Villalobos⁵, Vishnu Thilakan Manaparambil⁵

¹Department of Earth Sciences, Vrije Universiteit, Amsterdam, Netherlands. ²Deutsches Zentrum für Luft- und Raumfahrt e.V., Institut für Physik der Atmosphäre, Oberpfaffenhofen, Germany. ³Research & Development Satellite Observations, Royal Netherlands Meteorological Institute (KNMI), De Bilt, Netherlands. ⁴EMPA, Swiss Federal Laboratories for Materials Science and Technology, Dübendorf, Switzerland. ⁵Department of Physical Geography and Ecosystem Science, Lund University, Lund, Sweden

Session

Session 29: Using GHG measurement data to support national greenhouse gas inventories

Abstract text

In the Horizon Europe project Avengers, three atmospheric inversion models have been used to study CH4 emissions over Europe, namely the WRF-Chem coupled with the CarbonTracker Data Assimilation Shell (CTDAS), ICON-ART Community Inversion Framework (CIF) and Lund University Modular Inversion Algorithm (LUMIA). Inversion results for the period 2010-2021 have been used to study emission variations and trends in Western European countries that are well constrained by the measurement network. The analysis of the results revealed that the ICON-ART CIF model showcases the best performance out of the three with regards to the optimization of the CH4 fluxes throughout the period under study. The ICON-ART CIF results showed the largest decrease in the root mean square error as well as the bias. Regarding the WRF-CTDAS, the RMSE comparison for the validation stations showed an increase for the posterior mixing ratios of the validation stations, indicating that the emission improvement due to the optimization is limited to the regions around the stations used in the inversions. The LUMIA model, showed an increase in the BIAS for the posterior mixing ratios. Focusing on the Benelux region we notice some positive patterns in the flux differences between the posterior and the prior runs. The largest upward emission adjustment is found for the ICON-ART CIF inversion that agrees best with the atmospheric measurements. WRF-Chem CTDAS is used to study the emission adjustments over the Benelux in further detail and the role to ICOS sites Cabauw and Lutjewad located in The Netherlands.

183 Estimating urban CO₂ fluxes in Auckland, New Zealand with an atmospheric transport model inversion: a synthetic data study

Oral

Stijn Naus^1,2,3*, Sara Mikaloff-Fletcher¹, Beata Bukosa¹, Jocelyn Turnbull^1,4, Timothy Hilton¹, Elizabeth Keller^1,5, Stuart Moore¹, Vanessa Monteiro^1,6, Daemon Kennett¹, Sally Gray¹, Leigh Fleming¹, Hayden Young¹, Martina Franz¹, Gordon Brailsford¹

¹Earth Sciences New Zealand, Wellington, New Zealand. ²Wageningen University & Research, Wageningen, Netherlands. ³ICOS Carbon Portal, Lund, Sweden. ⁴University of Colorado, Boulder, USA. ⁵Victoria University of Wellington, Wellington, New Zealand. ⁶Universitat Autònoma de Barcelona, Barcelona, Spain

Session

Session 17: Urban greenhouse gas emissions and sinks: from advanced monitoring to source identification and impact

Abstract text

Accurate quantification of urban CO₂ emissions is essential for effective climate mitigation, yet separating anthropogenic and biospheric contributions remains challenging. We present the first urban-scale inversion framework for Auckland (New Zealand), integrating high-resolution flux estimates, atmospheric transport modelling, and a new urban observation network. Using an Observing System Simulation Experiment, we assess system performance and sensitivities. In the baseline configuration, the inversion reduces daily flux errors by 28% for both anthropogenic and biospheric components. We demonstrate that, especially in an urban setting, wind-filtered non-afternoon observations are essential for correcting non-afternoon fluxes, such as night-time respiration and morning rush hour. For Auckland specifically, transport model resolution plays an important role, with correct topographic representation substantially influencing flux estimates. While the framework successfully separates flux components under realistic conditions, the solution relies on accurate prior flux spatial structures and uncertainty representation. This work establishes a methodological foundation for urban CO₂ monitoring systems designed to inform emission mitigation policy in Auckland, with lessons learnt translatable to other cities.

184 Revisiting Energy Balance Non-Closure through Multivariate Atmospheric Regime Analysis

Oral

Neo Arquin*, Bernard Longdoz, Bernard Heinesch

University of Liege, Gembloux Agro-Bio Tech, Gembloux, Belgium

Session

Session 13: Advancing approaches for quantifying greenhouse gas fluxes in terrestrial ecosystems

Abstract text

To better understand the global carbon budget, terrestrial greenhouse gas (GHG) emissions can be quantified using the eddy covariance (EC) technique. However, this method sill suffers from the well-known energy balance closure (EBC) problem observed at most EC stations. Mesoscale organized atmospheric structures have been proposed as the main hypothesis to explain this persistent lack of EBC. Several studies suggest that, since turbulent transport mechanisms affect both energy and scalar fluxes, systematic energy imbalance may indicate regime-dependent biases in GHG emissions estimates. This could lead to underestimated long-term budgets, which is why improving our understanding of the mechanisms underlying energy balance non-closure is essential.

The ICOS station in Lonzée (BE-Lon, cropland) has one of the longest and most complete data series on cropland in Europe, offering more than 20 years fluxes measurements. Our study proposes a new approach based on multivariate analysis and temporal segmentation to identify and characterize coherent atmospheric episodes. These episodes are subsequently grouped using clustering techniques into statistically distinct regimes, whose EBC characteristics are then quantified and compared. First results show that, within each atmospheric stability condition, several regimes are identified, displaying contrasted mean EBC levels. However, transitions between these configurations remain gradual, suggesting a continuum rather than sharply separated states. In addition, remote sensing products are used to quantify landscape heterogeneity and assess its influence on the occurrence of the identified atmospheric regimes.

185 Sources of Urban CO2 Modeling Uncertainty: An Ensemble-Based Sensitivity Test of Beijing CO2 Concentrations

Poster

Kevin Fletcher¹*, Kenneth Davis¹, Zachary Barkley¹, Tomohiro Oda^2,3, Hiroshi Suto⁴

¹The Pennsylvania State University, University Park, USA. ²Universities Space Research Association, Washington, USA. ³The University of Maryland, College Park, USA. ⁴Japan Aerospace Exploration Agency, Tsukuba-city, Japan

Session

Session 28: Combining data and models to support emissions estimation and policy at local to regional scales

Abstract text

There are three primary sources of uncertainty in the forward modeling of CO₂: atmospheric transport, emissions within the domain, and inflow into the model domain. We investigate the contribution of each source of uncertainty to the overall uncertainty using a WRF-CHEM ensemble in a domain centered on Beijing with 18 transport members, 3 anthropogenic flux priors, and 2 background CO₂ members. The transport members focus on variability within the planetary boundary layer, where transport most affects CO₂ tracers emitted at the surface. Each flux prior is at a different spatial and temporal resolution, providing a spread of realistic possibilities. The ensemble is run across four separate winters (January 2020-March 2023) to minimize effects from biogenic fluxes. We validate the ensemble’s CO₂ performance against partial column CO₂ concentration measurements from the Greenhouse Gases Observing Satellite and the wind speed and direction from soundings in and around Beijing. We evaluate each of the sources of uncertainty for their contribution to the overall uncertainty. In addition, we evaluate how much each source of uncertainty can be minimized, allowing us to estimate the anthropogenic emissions compared to the priors.

186 Estimating European CH₄ fluxes using the CarboScope Regional atmospheric inversion system (CSR)

Poster

Frank-Thomas Koch^1,2*, Saqr Munassar², Luana Basso², Christian Rödenbeck², Dagmar Kubistin¹, Christoph Gerbig²

¹Deutscher Wetterdienst, Hohenpeissenberg, Germany. ²Max Planck Institute for Biogeochemistry, Jena, Germany

Session

Session 28: Combining data and models to support emissions estimation and policy at local to regional scales

Abstract text

With then extension of atmospheric stations network delivering continuous observations, top-down inverse transport modelling of greenhouse gases (GHGs) can be implemented in a quasi-operational framework.  Within the Intergrated GHG System for Germany (ITMS), we applied the CarboScope-Regional inversion system (CSR), which embeds the regional inversion within a global inversion using the two-step approach [1]. The regional CH4 inversion uses Lagrangian mesoscale transport from STILT and prior fluxes for peatlands, mineral soils, biomass burning, termites, anthropogenic emissions from EDGAR v6.0, and ocean fluxes. In addition an inversion with the latest release of the EDGAR inventory (EDGAR 2025 report) was used.  

The protocol for the inversion follows the methane regional inversion intercomparison project for Europe [2].  Latest European Obspack (2025) compilation of atmospheric methane, including  ICOS and non-ICOS stations was used to perform inversions for  2006-2024. Far field contributions were derived from a global CarboScope methane inversion. The model domain covers most of Europe (33°N – 73°N, 15°W – 35°E) with a spatial resolution of 0.25 degree for fluxes and 0.5 degree for flux corrections inferred by the inversion. 

Posterior methane fluxes and uncertainties for period 2006-2024 are presented on annual and monthly temporal resolution for the European domain and on national spatial scale for Germany.   

1. Rödenbeck C., et. al, A two-step scheme for high-resolution regional atmospheric trace gas inversions based on 1independent models, Atmos. Chem. Phys, 9, 5331-5342, 2009  
2. Ioannidis E., et. al, An inter-comparison of inverse models for Estimating European CH₄ emissions, ESSD, 18, 167-19, 2026 

187 Quantifying  coverage of aquatic vegetation in two contrast lakes combining UAV and Satellite    imagery

Poster

Zhengran Yin*, Martin Karlson

linköping university, linköping, Sweden

Session

Session 13: Advancing approaches for quantifying greenhouse gas fluxes in terrestrial ecosystems

Abstract text

In the vegetated littoral zone, the interface between terrestrial and aquatic environments, aquatic macrophyte productivity can rival that of rainforests. Neglecting this zone in lake carbon modelling can seriously bias system-scale budgets by overestimating terrestrial carbon inputs to pelagic waters. However, characterizing macrophytes in the littoral zone is practically challenging and time-consuming. To enable scalable quantification of the extent of macrophyte types (floating, emergent and submerged) and coverage, we propose a remote-sensing workflow that combines ultra-high-resolution UAV imagery with high-resolution satellite data and demonstrate it for two contrasting lakes in Sweden (Skogaryd and Erken). Multispectral UAV imagery acquired at peak vegetation provide detailed, spatially explicit reference data (“ground truth”) for macrophyte mapping, while satellite imagery (Planet Scope) extends these observations to broader spatial coverage. We quantify percent coverage of macrophyte types and use UAV-derived products to calibrate and validate satellite-based estimates, reducing uncertainty in shoreline-scale mapping of the littoral zone. Targeting the peak vegetation period also supports seasonal interpretation by capturing maximum extent and facilitating subsequent phenological comparisons. In future work, we plan to link percent cover to biomass production, providing an input for revising littoral carbon flux parameterizations in existing models. This integrated UAV–satellite approach offers a practical pathway for rapid, large-area assessment of vegetated littoral zones, helping constrain lake carbon budgets at scales relevant to regional and global synthesis. 

188 COS in cities: Exploring possibilities and limitations

Oral

Jesse Soininen¹*, Kukka-Maaria Kohonen², Stavros Stagakis³, Ara Cho⁴, Betty Molinier⁵, Pascal Rubli⁶, Rainer Hilland^7,8, Natascha Kljun⁵, Liisa Kulmala², Leena Järvi^1,9

¹Institute for Atmospheric and Earth System Research/Physics, University of Helsinki, Helsinki, Finland. ²Finnish Meteorological Institute, Helsinki, Finland. ³Department of Environmental Sciences, University of Basel, Basel, Switzerland. ⁴Meteorology and Air Quality, Wageningen University & Research, Wageningen, Netherlands. ⁵Department of Earth and Environmental Sciences, Lund University, Lund, Sweden. ⁶Empa, Swiss Federal Laboratories for Materials Science amd Technology, Dübendorf, Switzerland. ⁷nvironmental Meteorology, Institute of Earth and Environmental Sciences, University of Freiburg, Freiburg, Germany. ⁸The Netherlands Organisation for Applied Scientific Research (TNO), Petten, Netherlands. ⁹Helsinki Institute of Sustainability Science, University of Helsinki, Helsinki, Finland

Session

Session 17: Urban greenhouse gas emissions and sinks: from advanced monitoring to source identification and impact

Abstract text

Cities are significant sources of carbon dioxide (CO₂) emissions. Understanding the complex behavior of the net source requires a better understanding of biogenic CO₂ uptake, or gross primary productivity (GPP). Carbonyl sulfide (COS) serves as a tracer for GPP due to its similar stomatal uptake pathway with CO₂, without being respired back into the atmosphere. This study employs the eddy covariance (EC) technique to calculate vertical fluxes of COS in Helsinki, Finland (FI-Kmp), and Zurich, Switzerland (CH-Har). In Helsinki, COS-based GPP estimates were compared with solar-induced fluorescence (SIF), chamber measurements, and conventional EC partitioning of GPP and respiration. In Zurich, EC measurements were combined with flux footprints, and analysis focused on identifying and quantifying anthropogenic emissions of COS. To separate biospheric and anthropogenic COS contributions, biospheric COS fluxes were estimated using the Simple Biosphere Model version 4 (SiB4).

Results indicate that COS flux measurements can reliably estimate GPP in areas dominated by vegetation, yielding values comparable to other EC estimates in Helsinki. In Zurich, anthropogenic emissions from natural gas combustion interfered with COS signals, preventing its use as a GPP tracer. An approximate emission factor of 0.5 ± 0.2 ppt COS ppm CO₂^-1 was derived from this source. Additionally, drought stress during heatwaves diminished the relationship between COS and GPP.

This research underscores the potential and challenges of using COS as a GPP tracer in urban settings, situating the findings within the broader context of global carbon cycling and the need for accurate urban carbon accounting.

189 Isolating Temperature Effects in Eddy Covariance Data Using Explainable Machine Learning

Oral

Flavian Tschurr*, Nina Buchmann

ETH, Zurich, Switzerland

Session

Session 14: Greenhouse gas fluxes in agroecosystems: processes, measurements and management implications

Abstract text

The response of ecosystems to temperature is central to carbon flux or crop modelling. Yet, its interpretation under field conditions remains challenging because temperature co-varies strongly with radiation, vapour pressure deficit, water availability, and other environmental drivers. This raises a key question for eddy covariance (EC) flux studies: How can temperature response curves be derived, and to what extent do they reflect intrinsic physiological constraints rather than contextual environmental interactions?

Using partitioned gross primary production (GPP) across 14 cropland sites in Europe and the USA, spanning more than 100 growing seasons of winter crops and maize, we applied explainable machine learning (XGBoost combined with SHAP) to estimate isolated temperature response curves while explicitly accounting for co-varying drivers. Across crop types and environments, isolated temperature responses converged robustly on the widely used, process-based Wang-Engel function, with identifiable minimum (T_min), optimum (T_opt), and maximum (T_max) temperature values. T_opt was consistently the most robustly constrained threshold, whereas T_min and T_max were less tightly defined due to reduced data density at temperature extremes.

Despite strong environmental heterogeneity, the functional form of the temperature response curve was preserved across sites, while parameter values varied systematically, indicating environmental modulation without alteration of the underlying physiological structure. These findings demonstrate that EC data can recover intrinsic temperature constraints on GPP, consistent with established process-based functions, while explicitly addressing non-linear driver interactions. This framework allows improving parameterization of ecosystem-scale carbon flux models and to derive response-function-based characterizations of ecosystem functioning across environmental gradients and variables.

190 Integrating surface and satellite measurements for a more detailed understanding of CO and CH₄ variability in Calabria, Southern Italy

Poster

Teresa Lo Feudo*, Francesco D’Amico, Ivano Ammoscato, Daniel Gullì, Mariafrancesca De Pino, Claudia R. Calidonna

Consiglio Nazionale delle Ricerche - Istituto di Scienze dell'Atmosfera e del Clima, Lamezia Terme, Italy

Session

Session 6: Carbon cycle in the Mediterranean region: from the local to the regional scale

Abstract text

This study analyzes the variability of carbon monoxide (CO) and methane (CH₄) in 2020–2025 by integrating two complementary observational datasets: in-situ surface measurements from the CNR ISAC World Meteorological Organization – Global Atmosphere Watch (WMO/GAW) station located in Lamezia Terme (Calabria, Italy), and satellite observations from the Copernicus Sentinel-5P (SP5) mission focusing on the same area. The analysis reveals recurring seasonal patterns in both datasets. CO exhibits pronounced winter maxima, primarily associated with increased combustion emissions and enhanced atmospheric stability that limits vertical mixing. CH₄ also shows higher winter concentrations, likely linked to biogenic activity and seasonal cycles. Superimposed on these regular seasonal variations are significant interannual fluctuations. The year 2021 emerges as particularly anomalous for both gases, with marked increases detected in the time series; these anomalies are consistent with widespread wildfires that affected the Mediterranean region, whose impact is evident as elevated surface CO concentrations and enhanced total column values were retrieved. Additional CH₄ buildups are observed at the time of the 2020 COVID-19 lockdown, indicating a progressive increase in atmospheric CH₄ concentrations that may be associated with changes in global source/sink balances, contributions from biogenic sources, and on a local scale, inputs from livestock and landfills. The comparison between ground-based and satellite measurements highlights periods of strong agreement as well as intervals of divergence, mainly driven by large-scale atmospheric transport processes. This integrated observational approach proves essential for identifying emission anomalies, improving the understanding of meteorological controls on GHGs, and strengthening regional environmental monitoring capabilities.

191 Svalbard seawaters as a significant source of CH₄ to the atmosphere during summer 2024: from Arctic amplification to glacial meltwater

Poster

Coraline Leseurre¹*, Bruno Delille², Axelle Brusselman², Hannelore Theetaert¹, Michiel T'Jampens¹, Ozan Efe², Thanos Gritzalis¹

¹Flanders Marine Institute (VLIZ), Ostende, Belgium. ²University of Liège (ULiège), Liege, Belgium

Session

Session 10: Cross-scale responses of greenhouse gas variability to climate extremes

Abstract text

The Arctic is undergoing unprecedented warming, with temperatures rising more than twice the global average, a phenomenon known as Arctic amplification. This extreme climate change is significantly impacting ecosystems and greenhouse gas (GHG) dynamics in this region. CH₄, the second most important anthropogenic GHG after CO₂, is becoming a significant contributor to atmospheric emissions in the Arctic. While open-ocean CH₄ remains near atmospheric equilibrium (~2% of the global CH₄ budget), emerging evidence indicates that the Arctic shelf can be a major source of CH₄. 

To better constrain the processes controlling CH₄ air-sea exchange in these regions, we investigated the sea surface CH₄ concentrations around Svalbard Archipelago (an area affected by glacial runoff and complex geology) during a summer cruise conducted on board the RV Skagerak (University of Gothenburg,) in August 2024. We observed systematically elevated CH₄ concentrations (mean 11 nmol L^-1) with spatial patterns linked to bathymetry and known hydrocarbon seep locations. The highest values (up to 105 nmol L^-1) occurred near marine-terminating glaciers and were not associated with strong salinity anomalies. This indicates that subglacial inputs from the glacier bedrock may dominate over the simple freshwater dilution from the glacial meltwater plume. 

Our observations and analysis suggest that Svalbard’s surface waters acted as a pronounced source of CH₄ to the atmosphere during summer 2024. We will discuss the relative roles of geological seepage and subglacial pathways (especially in the context of the transition from marine to land-terminating glaciers) and their implications for present and future Arctic shelf CH₄ emissions. 

192 Improved Aboveground Biomass Estimates at ICOS Forest Sites Using Terrestrial Laser Scanning to Reduce Uncertainty on Carbon Stock Changes.

Poster

Barbara D'hont¹*, Geike Desloover^1,2, Kim Calders², Bert Gielen¹

¹University of Antwerp, Antwerpen, Belgium. ²Ghent University, Ghent, Belgium

Session

Session 12: Comparing long-term eddy covariance measurements in terrestrial ecosystems with carbon stock variations: lessons and future challenges

Abstract text

Quantifying carbon stocks at ICOS forest sites is necessary to interpret eddy-covariance flux measurements in the context of underlying biomass dynamics. A substantial share of the forest carbon stock is stored in trees, especially within their above‑ground biomass (AGB). Trees at ICOS stations are routinely inventoried for diameter at breast height (DBH), height, and species, from which AGB is typically derived using existing allometric equations. These allometries are often not site-specific and are frequently based on limited calibration datasets, which can introduce substantial uncertainty and bias. Here, we evaluate the use of terrestrial laser scanning (TLS) to improve AGB estimation at ICOS forest stations. TLS captures highly detailed 3D point clouds of individual trees, allowing tree volume to be estimated directly and converted to AGB using wood density. We therefore developed site-specific allometric relationships using TLS measurements collected in the continuous sample plots (4 x 2000 m² per ICOS station)  to estimate AGB of the entire ICOS flux tower target area using the repeated forest inventories. These TLS-derived allometries were compared with estimates obtained from conventional allometries as well as from species-specific TLS-based target areas. By reducing uncertainty in forest carbon stock estimates, this approach supports a more reliable interpretation of ecosystem carbon fluxes at ICOS sites.

193 Continuous atmospheric CO₂ and O₂ observations for constraining regional fossil fuel CO₂ emissions in the United Kingdom

Poster

Karina Adcock¹*, Penelope Pickers^1,2, Grant Forster^1,2, Andrew Manning³, Chris Rennick⁴, Tim Arnold^5,6, Adam Wisher⁶

¹Centre for Ocean and Atmospheric Sciences, School of Environmental Sciences, University of East Anglia, Norwich, United Kingdom. ²National Centre for Atmospheric Science, University of East Anglia, Norwich, United Kingdom. ³Department of Science and Mathematics, National Technical Institute for the Deaf, Rochester Institute of Technology, New York, USA. ⁴National Physical Laboratory, Teddington, United Kingdom. ⁵Department of Earth and Environmental Sciences, Lund University, Lund, Sweden. ⁶School of GeoSciences, University of Edinburgh, Edinburgh, United Kingdom

Session

Session 18: Quantifying anthropogenic greenhouse gas emissions from continental to regional scales

Abstract text

Measurements of atmospheric oxygen (O₂) provide valuable insight into carbon cycle processes, particularly for distinguishing whether variations in atmospheric carbon dioxide (CO₂) arise from land, ocean or anthropogenic fluxes. Since land biospheric exchanges of O₂ and CO₂ are strongly anticorrelated, while ocean exchanges are not, combined measurements of O₂ and CO₂ are used to calculate Atmospheric Potential Oxygen (APO), a tracer that is conservative with respect to land biospheric processes. Fossil fuel combustion also produces anticorrelated O₂ and CO₂ fluxes, but with different exchange ratios than land; therefore, short‑term variations in APO can be used to quantify local to regional fossil fuel emissions.

Here, we present three atmospheric time series of O₂, CO₂, and APO from the United Kingdom: (1) a 16‑year record from the Weybourne Atmospheric Observatory in Norfolk; (2) a 4‑year record from the Heathfield Tower in Sussex; and (3) preliminary observations from the Invergowrie Tower in Scotland, started earlier this year. We also introduce the design of OXYMOBILE, a new mobile atmospheric O₂ measurement laboratory housed within an electric van and currently under development as a community resource. OXYMOBILE will be available for use to investigate fossil fuel CO₂ emissions, biospheric carbon uptake, carbon capture and storage systems, and terrestrial ecosystem fluxes. Together, these datasets and measurement capabilities provide new opportunities for constraining fossil fuel CO₂ emissions in the United Kingdom.

194 Strategic design of a global optimal network for atmospheric CO2 monitoring

Oral

David Matajira-Rueda¹*, Charbel Abdallah¹, Thomas Lauvaux^1,2

¹Université de Reims Champagne-Ardenne, REIMS, France. ²Laboratoire des Sciences du Climat et de l’Environnement, Paris, France

Session

Session 26: Designing the ideal global greenhouse gas monitoring network

Abstract text

To contribute to the Global Greenhouse Gas Watch (G3W) initiative, optimal designs for atmospheric monitoring networks, specifically for CO2, are presented for several representative regions worldwide. These designs are obtained using the CRO²A algorithm (Concepteur de Réseaux Optimaux d’Observations Atmosphériques), which, through a direct approach, minimizes the number of ground-based observation stations (cost-effectiveness) and maximizes the quality of the collected data (reliability).

The optimal designs proposed by the CRO²A also allow for the integration or assimilation of already installed ground-based stations and the corresponding assessment of the existing networks that contain them. The results are based on the processing of the anthropogenic CO2 mixing ratios from the Copernicus Atmosphere Monitoring Service (CAMS) modelling system. The results are compatible with those obtained using other transport models and approaches, such as the Lagrangian Particle Dispersion Model and inverse modeling, respectively, but with lower computational cost and without using uncertainty reduction as an objective function.

Through the CRO²A, it is possible to define a set of simple but essential rules to create the necessary references for an appropriate global monitoring network that can be integrated into the WMO and by the GHG community. The analysis of monitoring networks using CRO²A also enables the gradual, phased deployment of additional stations that may be required later.

To demonstrate the benefits of the designs proposed by CRO²A, a comparison of results is intended, considering various transportation models across different regions globally, with particular attention to the impact of the approach used.

195 Managing User Requirements within the Integrated Greenhouse Gas Monitoring System for Germany (ITMS)

Poster

Buhalqem Mamtimin¹*, Elena Zwerschke², Jennifer Mueller-Williams², Dagmar Kubistin², Christian Mielke³, Roland Fuß⁴, Andrea Kaiser-Weiss¹

¹Deutscher Wetterdienst (DWD), Offenbach am Main, Germany. ²Deutscher Wetterdienst (DWD), Hohenpeißenberg, Germany. ³Umweltbundesamt (UBA), Dessau-Roßlau, Germany. ⁴Thünen Institute (TI), Braunschweig, Germany

Session

Session 32: Unlocking climate research solutions through co-design

Abstract text

The Integrated Greenhouse Gas Monitoring System for Germany (ITMS) is a national system combining atmospheric observations with bottom-up and inverse modelling to support greenhouse gas monitoring and national reporting. Within this system, a structured approach for managing user requirements is developed in Module V (Utilisation).

Inputs from key stakeholders (e.g. German Federal Environment Agency UBA and Thünen Institute) are systematically collected and translated into user requirements, which are then categorized and prioritized using a structured framework based on manually defined requirement IDs, a three-level prioritization scheme, and stakeholder mapping. These inputs are gathered through multiple channels, including stakeholder meetings, direct exchanges, and targeted user surveys.

This approach will link user requirements to ITMS data products generated by other modules and ensure transparent traceability between requirements and outputs, including sectoral and regional emission information, standardized formats, visualization, and high-level summaries to support policy decisions and public communication.

Continuous stakeholder engagement and annual updates ensure that requirements remain aligned with user needs. Close collaboration with the UBA and the Thünen Institute is an integral part of system development. Both partners are actively involved in coordination, scientific activities, and the co-design of services to ensure that outputs meet the requirements for national reporting and are suitable for the intended purpose.

196 A Novel Approach to Quantifying Methane Emissions and Its Corresponding Strategic Scheme for Ensuring Low Uncertainty

Poster

David Matajira-Rueda¹*, Charbel Abdallah¹, Thomas Lauvaux^1,2

¹Université de Reims Champagne-Ardenne, REIMS, France. ²Laboratoire des Sciences du Climat et de l’Environnement, Paris, France

Session

Session 21: Emerging approaches for greenhouse gas flux measurements

Abstract text

Among the various techniques for quantifying trace gas emissions across diverse environments, the one based on instrumented unmanned aerial vehicles (UAVs) is flexible and cost-effective but presents significant uncertainty due to the rapid motion of gas plumes and unpredictable weather, among other factors.

Therefore, we propose an algorithm that automatically and sequentially structures the necessary procedures to obtain an accurate estimate of methane emissions: the "8lind Date" strategic scheme.

This approach, in addition to leveraging the fact that UAVs have become an ideal platform for monitoring emissions in restricted or often difficult-to-access areas, temporally and spatially characterizes the signals captured by the instrument. The mentioned procedures create the conditions for applying a novel, low-uncertainty approach to emissions quantification. Furthermore, each procedure is optimally configured to account for characteristics such as the instrument's sampling frequency, battery life, and drone speed, thereby providing greater versatility to the monitoring process.

"8lind Date," among other features, provides optimal flight paths for UAVs based on energy criteria and spatiotemporal coverage of sampling windows. These paths provide much more efficient mapping and measurements than those obtained through traditional flight paths, without requiring exhaustive sampling, capturing only the key information.

Furthermore, a comparison process using methane columns is presented and proposed, employed as a performance test to evaluate flight paths for sampling according to the climatic conditions of the measurement environment.

197 CO₂ and H₂O balance of temperate agroforestry and open cropland or grassland systems

Oral

José Ángel Callejas Rodelas^1,2,3*, Justus van Ramshorst^1,4, Anas Emad¹, Alexander Knohl^1,5, Sarah Choe⁶, Maren Langhof⁷, Christian Markwitz¹

¹Bioclimatology, Faculty of Forest Sciences and Forest Ecology, University of Göttingen, Büsgenweg 2, 37077 Göttingen, Germany. ²Andalusian Institute for Earth System Research (CEAMA-IISTA), University of Granada, Granada 18006, Spain. ³Department of Applied Physics, University of Granada, Granada 18071, Spain. ⁴Quanterra Systems Ltd., Centenary House, Peninsula Park, Exeter EX2 7XE, United Kingdom. ⁵Centre for Biodiversity and Land Use, Faculty of Forest Sciences and Forest Ecology, University of Goettingen, Büsgenweg 2, 37077 Göttingen, Germany. ⁶Soil Science of Tropical and Subtropical Ecosystems, Faculty of Forest Sciences and Forest Ecology, University of Göttingen, Büsgenweg 2, 37077 Göttingen, Germany. ⁷{Institute for Crop and Soil Science, Julius Kühn-Institut (JKI), Federal Research Center for Cultivated Plants, Braunschweig, Germany

Session

Session 14: Greenhouse gas fluxes in agroecosystems: processes, measurements and management implications

Abstract text

Agroforestry (AF) systems are hypothesized to have a higher CO₂ sequestration potential but similar evapotranspiration (ET) compared to open cropland (OC) or grassland (OG) systems. Empirical evidence from field measurements is, however, still missing. In this study, CO₂ and H₂O flux densities were measured using lower-cost eddy covariance systems for 3 to 5 years over 4 pairs of AF and OC or OG systems. Multi-annual sums of net ecosystem CO₂ exchange (NEE) indicated a CO₂ uptake higher by 596 gC m^-2 at the AF on average. Two of the sites were a CO2 sink at the AF, but a CO₂ source at the OC and OG. ET was higher at the AF by 122 mm on average, with relative differences between OC and OG lower than for NEE.

The analysis of environmental drivers showed that AF systems were less sensitive to higher temperatures and vapor pressure deficit, providing a stronger CO₂ uptake, due to the enhanced physiological activity of the trees and the extended growing season. Nevertheless, there were differences between years, explained by the different crops and varying meteorological conditions. Winter crops showed a higher CO₂ uptake compared to summer crops, as shorter periods of bare soil decrease soil respiration. Water use efficiency was similar between AF and OC or OG at the beginning of the growing season, while it got higher at the AF after crops’ ripening. These results indicate the potential of AF systems to enhance climate change mitigation in agriculture if appropriate management practices are applied.

198 Towards Improved  Flux Tower Footprint Integration with Remote Sensing Observations

Poster

Linara Arslanova¹*, Martin Jung¹, Sophia Walther¹, Jake Nelson¹, Qi Yang¹, Pedro-Henrique Herig-Coimbra², Gregory Duveiller¹

¹Max Planck Institute for Biogeochemistry, Jena, Germany. ²NRAE, Palaiseau, France

Session

Session 19: Understanding feedbacks between greenhouse gas exchange processes and climate variability using in situ observations, remote sensing, and machine learning

Abstract text

Eddy covariance flux towers provide ground-truth observations for calibrating and validating machine-learning models that use satellite remote-sensing observations as input variables to upscale ecosystem fluxes. At the same time, satellite remote sensing is characterized by a fundamental trade-off between spatial and temporal resolution: high-spatial-resolution imagery better resolves sub-footprint spatial heterogeneity whereas coarse-spatial-resolution sensors offer frequent observations but often fail to capture fine-scale variability in heterogeneous landscapes.
An open-access remote-sensing dataset that is globally consistent across space and time and provides high spatial and temporal resolution for eddy covariance flux tower sites is currently lacking. Therefore, this study aims to enhance remote-sensing information content and strengthen cross-scale integration of flux tower footprints to support predictive upscaling.
To this end, a multi-sensor, multi-scale data fusion framework is under development to integrate complementary optical multi-spectral satellite observations (Sentinel-2 and VIIRS) into a harmonized, analysis-ready product. The current development status and first results of the framework are presented.

Key words: Eddy covariance; flux footprint; multi-sensor data fusion; analysis-ready data
product

199 The characterisation and use of a cost-effective in situ methane sensor for facility-scale emission quantification using uncrewed aerial systems

Oral

Noni van Ettinger¹*, Steven van Heuven¹, Huilin Chen^1,2

¹Rijksuniversiteit Groningen, Groningen, Netherlands. ²Nanjing University, Nanjing, China

Session

Session 21: Emerging approaches for greenhouse gas flux measurements

Abstract text

Methane (CH₄) is a promising target for climate change mitigation due to its relatively short atmospheric lifetime, high global warming potential compared to CO₂, and its emissions typically being inadvertent rather than required. Reliable detection of CH₄ emissions is essential for effectively guiding such mitigation efforts, but cost-effective sensors and platforms have traditionally been lacking. However, recent advances in technology have improved the availability of sensors that potentially are "good enough". In this study, we present the characterisation and field deployment of a cost-effective, in situ methane sensor (LGD-compact A CH₄; Axetris AG, Kaegiswil, Switzerland), which can be used for facility-scale emission quantification by integration with an uncrewed aerial system (UAS). The sensor underwent a comprehensive laboratory evaluation to characterise sensor sensitivities and quality control requirements. Subsequently, the sensor was deployed near a variety of CH₄ sources to demonstrate its performance under varying environmental conditions and source types. Deployment of the sensor across diverse emission sources, including dairy farms, a landfill and ship exhaust plumes, demonstrates the broad applicability of the technique. The performance of the in situ sensor was evaluated against a high-precision reference instrument (active AirCore) to assess sensor accuracy, efficiency and applicability. These results highlight the potential of this sensor as part of a cost-effective quantification platform, thereby increasing the sensor network to aid in minimising uncertainties in UAS-based monitoring. 

200 Measuring dry deposition fluxes of NO, NO₂ and O₃ in a coastal dune area

Poster

Kim Vendel¹*, Kevin Felter¹, Mark Eijkelboom¹, Marty Haaima¹, Margreet van Zanten^1,2

¹National Institute for Public Health and the Environment (RIVM), Bilthoven, Netherlands. ²Wageningen University & Research (WUR), Wageningen, Netherlands

Session

Session 24: Exchange of reactive gases and aerosols between the land surface and the atmosphere in natural and managed ecosystems

Abstract text

We measure dry deposition fluxes of nitrogen oxides (NO and NO₂) and ozone (O₃) during an ongoing campaign at coastal dune area Solleveld. This Dutch Natura 2000 site is surrounded by a large city, industrial areas and extensive horticultural greenhouses, all major sources of nitrogen oxides. Deposition of nitrogen compounds (N_r) leads to acidification and eutrophication of sensitive ecosystems such as coastal dunes, disrupting ecological balance and reducing biodiversity. Quantifying nitrogen deposition into natural areas remains an important challenge for atmospheric deposition models. Accurate measurements of exchange fluxes between vegetation and atmosphere are essential for validation and calibration of those models. Such measurements are still scarce, and previous Dutch campaigns focused solely on main contributor ammonia, while nitrogen oxides also account for 25% of the nitrogen dry deposition. To get a more complete picture, the Solleveld campaign is a unique combination of dry deposition flux measurements of NH₃ as well as the complete O₃-NO-NO₂ triad. In this presentation, we focus on NO, NO₂ and O₃ fluxes, while another presentation in this session by Rutledge-Jonker covers ammonia. 

We present first results of half-hourly measurements of dry deposition fluxes, obtained with the aerodynamic flux-gradient method. The aim of this study is to evaluate the current parametrization of NO, NO₂ and O₃ in DEPAC-1D for a dune vegetation. Data analysis will therefore focus on diurnal cycles and seasonal variations of the deposition fluxes, correlations with meteorological and vegetation parameters and vertical flux divergence due to chemical reactions between the three compounds.

201 First estimates of dissolved organic carbon transport from land to the Baltic and Western Seas through a modelled coastal zone

Poster

Moa Edman*, Simon Pliscovaz, Iréne Wåhlström, Emilie Breviere

SMHI, VÄSTRA FRÖLUNDA, Sweden

Session

Session 4: Carbon cycling in the land ocean aquatic continuum

Abstract text

The waters in Sweden's coastal zone are affected by runoff from land and by the open sea outside the coastal zone. The runoff from land supplies large amounts of organic material to the coastal zone, including dissolved organic carbon. This supply colours the water brown, a phenomenon known as brownification. Consequently, the water transparency represented by the Secchi depth is affected. In order to simulate the effects and processes related to organic material, calculations of organic carbon were implemented into the Swedish coastal zone model (SCM). SCM consists of vertically high-resolved coupled basins in which the equation solver PROBE is used. The basins are based on how the Water Framework Directives divides the coastal zone and the model calculate biogeochemical sinks and sources with SCOBI. The sinks and sources for dissolved organic carbon has been implemented in the latter. Preliminary model results for the Baltic and Western Seas show a seasonal variation in autochthonous (produced in situ) dissolved organic carbon, while concentration of allochthonous (transported, in this case; transported from land) dissolved organic carbon is more stable throughout the year. The simulated Secchi depths show more seasonal variation but are slightly overestimated, i.e. too deep. The new capability of SCM was applied to make a first estimation of dissolved organic carbon transport through the Swedish coastal zone and to the Baltic and Western Seas.

202 Enhancing Ecosystem Carbon and Water Flux Simulations through VOD Assimilation: Insights from the 2025 Los Angeles Wildfires

Poster

Mousong Wu¹*, Lu Hu¹, Huajie Zhu¹, Weimin Ju¹, Jingming Chen²

¹Nanjing University, Nanjing, China. ²University of Toronto, Toronto, Canada

Session

Session 19: Understanding feedbacks between greenhouse gas exchange processes and climate variability using in situ observations, remote sensing, and machine learning

Abstract text

Vegetation optical depth (VOD) derived from microwave satellites provides critical information on vegetation water content and has emerged as a valuable remote sensing metric for quantifying water stress impacts on vegetation carbon uptake. As ground validation of VOD remains challenging, terrestrial biosphere models (TBMs) serve as essential tools for its comprehensive validation and broader application. By incorporating plant hydraulic schemes, TBMs can simulate key vegetation water status variables, such as leaf water potential, to facilitate VOD simulation. In this study, we developed a plant hydraulics module within the adjoint-based Nanjing University Carbon Assimilation System (NUCAS). This module explicitly considers water supply driven by the soil-leaf water potential gradient and water demand due to potential transpiration and plant water storage, simulating water movement along the soil-plant-atmosphere continuum (SPAC). Parameter sensitivity analysis identified saturated hydraulic conductivity and minimum leaf water potential as key parameters influencing VOD. We evaluated model performance and subsequently assimilated satellite-derived VOD observations into the hydraulically enhanced NUCAS framework over the Los Angeles region, which experienced extensive wildfires in spring 2025. Results demonstrated that the assimilation of VOD effectively constrained vegetation hydraulic processes, leading to improved model representation of ecosystem carbon and water fluxes. Furthermore, our results revealed a potential link between simulated vegetation water stress and wildfire occurrence. By integrating remote sensing data with process-based modeling frameworks, this study advances our understanding of carbon-water-climate feedback processes, underscoring the value of VOD in improving the prediction of greenhouse gas exchange under climate variability. 

203 Constraining methane emissions and their variability over two decades using atmospheric inverse modelling

Poster

Luana S. Basso*, Christian Rödenbeck, Christoph Gerbig

Max Planck Institute for Biogeochemistry, Jena, Germany

Session

Session 28: Combining data and models to support emissions estimation and policy at local to regional scales

Abstract text

Methane (CH₄) is the second most important greenhouse gas. Its atmospheric concentrations have more than doubled since pre-industrial times, and its growth rates have been highly variable in recent decades. Despite extensive efforts using both bottom-up and top-down approaches, significant uncertainties remain in the global methane budget, particularly regarding the contributions of natural and anthropogenic sources. 

In this study, we use the Jena CarboScope global inversion system to estimate global methane surface fluxes for the period from 2000 to 2024. The Bayesian atmospheric inversion framework optimizes surface emissions by combining atmospheric transport modelling with long-term methane observations. Inversions were performed at a horizontal resolution of 3.8° × 5°. Prior emissions include wetland fluxes (ORCHIDEE), anthropogenic emissions (EDGAR), biomass burning (GFEDv4s), and additional minor sources such as termites, freshwater, geological seepage, and the ocean. We performed two complementary inversions. In the first setup, we assimilated only stations that have provided measurements since 2000, ensuring temporal consistency and enabling a robust assessment of long-term trends. In the second setup, we expanded the network by including stations with at least ten years of data, substantially increasing spatial coverage for the most recent decade and improving constraints on regional emission patterns. We present the temporal evolution of methane emissions over the last two decades, analyze their spatial distribution and regional emission patterns, and evaluate the consistency of the inversion results using independent observational datasets across different regions. We also assess how network configuration influences inferred trends and spatial variability.

204 Mechanistic Modeling of Forest Carbon Flux Responses to the 2018 Drought

Oral

Filipe Gomes de Almeida*, Rose Brinkhoff, Cecilia Akselsson, Natascha Kljun, Thomas A. M. Pugh

Lund University, Lund, Sweden

Session

Session 10: Cross-scale responses of greenhouse gas variability to climate extremes

Abstract text

Extreme climate events increasingly disrupt ecosystem carbon dynamics, yet mechanistic understanding of their impacts on greenhouse gas fluxes remains limited. The extreme 2018 drought in Sweden provides a striking example, causing severe reductions in forest productivity. We applied recent developments in the process-based dynamic vegetation model LPJ-GUESS to simulate carbon flux responses to this drought at two Swedish ICOS sites: Norunda (a mixed stand of Norway spruce and Scots pine) and Hyltemossa (a monoculture of Norway spruce). Two model versions were combined: a European-optimized parameterization of key tree traits and structure, and a version incorporating an enhanced mechanistic plant hydraulic scheme that can represent contrasting stomatal regulation strategies. Model outputs were evaluated against high-resolution carbon and water flux measurements from the Swedish ICOS sites. Results indicate that inclusion of hydraulic processes, particularly the representation of isohydricity, plays an important role in capturing drought-driven reductions in carbon fluxes, especially during the extreme 2018 event. This work demonstrates the potential of process-based models to clarify the mechanisms driving carbon flux variability during extreme droughts and provides a foundation for more accurate predictions of how drought affects forest carbon dynamics and ecosystem responses under future climate extremes.

205 Methane sources in cities re-interpretated using new laboratory results

Poster

Roisin Commane*, Raghav Dhall, Andrew Hallward-Driemeier, Luke Schiferl, Yuwei Zhao

Columbia University, New York, USA

Session

Session 17: Urban greenhouse gas emissions and sinks: from advanced monitoring to source identification and impact

Abstract text

Using a network of long-term methane observation sites (roof-top and towers) located around New York City (pop. 20M), we quantified the city-scale emission rate of methane (Schiferl et al., ACP, 2025) and carbon monoxide (CO) (Schiferl et al., ACP, 2024) for six years at an urban core site. These long-term studies did not identify any annual trend but observed a strong seasonal cycle in both methane and CO emissions that were correlated across the winter-to-spring transition. We also observed a depleted ethane:methane ratio relative to the pipeline in plumes associated with combustion of natural gas. We undertook a laboratory study to sample the in-stack exhaust of a natural gas boiler under different operating conditions. We observed large methane and ethane emissions during boiler cycles (known as methane slip). Badly operated boilers with insufficient air flow released large amounts of CH₄, C₂H₆ and CO during the incomplete combustion of natural gas during the duty cycle. This low combustion efficiency burning also depleted C₂H₆relative to the C₂H₆:CH₄ ratio of the incoming pipeline, which is consistent with our observations at the urban core rooftop site. Previous studies using ethane:methane ratios would have underestimated the natural gas contribution of methane emission. Instead, we can use the 1Hz variability (correlation coefficients) in methane, ethane and CO to identify natural gas and combustion related sources within the urban plume. We can also identify methane emissions after the combustion process, which helps distinguish whether mitigation policies should target natural gas suppliers or consumers. 

206 CO₂ emission reduction scenarios and atmospheric trend indicators for climate policy in the Aix-Marseille metropolitan area (France): a synthesis of the interdisciplinary COoL-AMmetropolis project.

Poster

Irène Xueref-Remy^1,2*, Marine Claeys^1,3, Michelle Leydet¹, Valéry Masson³, Frédérique Hernandez^4,5, Ludovic Lelandais¹, Alexandre Armengaud⁶, Sonia Oppo⁶, Jérôme Dubois⁵, Marie-Laure Lambert⁵, Coralie Demazeux⁵, Jean-Philippe Mévy¹, Damien Bouchard⁶, Jean Wurtz³, Aurélie Riandet¹, Pierre-Eric Blanc², Benoît Carré², Fitahiana Andriamahafaly⁴, Zoé Dubreuil Szymanski⁴

¹Aix Marseille Univ, Avignon Université, CNRS, IRD, IMBE (Institut Méditerranéen de Biodiversité et d’Ecologie marine et continentale), Marseille, France. ²Observatoire de Haute Provence, CNRS, Saint-Michel l'Observatoire, France. ³Centre National de Recherche Météorologique, Toulouse, France. ⁴Laboratoire Mécanismes d'Accidents (LMA), Université Gustave Eiffel, Salon-de-Provence, France. ⁵Aix-Marseille Université, Laboratoire Interdisciplinaire Environnement Urbanisme, Marseille, France. ⁶ATMOSUD, Marseille, France

Session

Session 6: Carbon cycle in the Mediterranean region: from the local to the regional scale

Abstract text

The Aix-Marseille-Provence (AMP) metropolis is the second-largest urbanized area in France and accounts for about 10% of national CO₂ emissions. This coastal region hosts one of the country’s largest industrial and port complexes, making it a major emission hotspot. According to the regional inventory, the industrial and energy production sectors contribute about one third of AMP’s CO₂ emissions, comparable to the maritime sector, while road traffic accounts for roughly 25%.

Within the COoL-AMmetropolis project, atmospheric field campaigns were conducted to independently assess and improve the regional CO₂ emissions inventory. At the same time, the AMP territory is implementing multiple mitigation strategies aimed at achieving carbon neutrality, although the coherence and potential effectiveness of these policies remain largely unexplored.

In this study, we analyze sectoral mitigation plans and identify several contradictory actions. We quantify CO₂ emission trends over the past 15 years and project their evolution to 2030 and 2050 based on currently planned measures, defining a “trend scenario” representing the likely emissions trajectory of the metropolis.

Using the MESO-NH mesoscale atmospheric model, we simulate this scenario to quantify changes in atmospheric CO₂ concentrations over the AMP area in 2030 and 2050. We also construct a “sobriety scenario” consistent with carbon neutrality by 2050 and estimate the associated reductions in atmospheric CO₂ concentrations. These simulations provide benchmarks for interpreting long-term observations from our reference station in central Marseille and support the future development of an atmospheric indicator to assess the effectiveness of urban CO₂ mitigation policies.

207 The effect of land management in Irish agricultural systems on grassland carbon sequestration

Poster

Eleanor Lampard^1,2*, Matthew Saunders¹, Rachael Murphy², Jack Bishop², James Rambaud², Kate Devereux², Eoin Dunne³, Brendan Horan⁴, Aine Murray⁴, Lucia Gil⁴

¹Trinity College Dublin, Dublin, Ireland. ²Teagasc, Johnstown, Ireland. ³Teagasc, Athenry, Ireland. ⁴Teagasc, Moorepark, Ireland

Session

Session 14: Greenhouse gas fluxes in agroecosystems: processes, measurements and management implications

Abstract text

Ireland currently relies on the IPCC’s default Tier 1 land emission factors to determine soil organic carbon (SOC) stock changes in managed grasslands on mineral soils. While grasslands store significant carbon, they are also the largest net emitter in the LULUCF sector. Our study aims to support the development of country-specific, Tier 2.5 land management factors for Irish soils. Current literature shows mixed results on whether multi-species swards (MSS) sequester more carbon than perennial ryegrass and white clover (PR+WC) systems. To address this, we evaluated sheep and dairy enterprises utiliSing both systems across multiple sites and years to quantify inter-annual variability in soil C stocks within different agri-climatic zones.We are currently analySing 2–3 years of data from eddy covariance towers across four sites with varied management: Co. Galway: Sheep-grazed PR+WC (220 kg N ha⁻¹), Co. Cork: Two dairy farms (MSS at 272 kg N ha⁻¹ and PR+WC at 175 kg N ha⁻¹), and Co. Tipperary: Dairy PR+WC (122 kg N ha⁻¹). By coupling eddy covariance results with farm management data, we quantified total C imports and exports to calculate the Net Ecosystem Carbon Balance (NECB) for 2024/25. Preliminary results indicate these grazing systems are currently weak carbon sinks. Drawing concrete conclusions on the comparative effect of MSS versus PR+WC requires further multi-year analysis. This study will refine national Irish inventories, providing the data needed to accurately track agricultural emissions and support national climate targets

208 Quantification of greenhouse gases fluxes in rice fields under different straw management practices

Poster

Cinzia De Benedictis^1,2*, Camilla Chieco¹, Marianna Nardino¹, Daniela Famulari¹, Federico Carotenuto¹, Giacomo Panza^1,2, Lorenzo Brilli³, Luisa Neri¹, Giuditta Meloni⁴, Maddalena Campi⁴, Cinzia Panigada⁵, Simona Maccherini², Beniamino Gioli³

¹National Research Council - Institute of BioEconomy (CNR-IBE), Bologna, Italy. ²University of Siena - Department of Life Science, Siena, Italy. ³National Research Council - Institute of BioEconomy (CNR-IBE), Firenze, Italy. ⁴Diagram S.p.A, Jolanda di Savoia (FE), Italy. ⁵Rurall, Jolanda di Savoia (FE), Italy

Session

Session 14: Greenhouse gas fluxes in agroecosystems: processes, measurements and management implications

Abstract text

Rice paddies account for approximately 48% of anthropogenic greenhouse gas (GHG) emissions from croplands and contribute 22% and 11% of total agricultural methane (CH₄) and nitrous oxide (N₂O) emissions, respectively. Optimizing field management practices, such as water and straw management, is critical to understand GHG budgets and develop effective climate change mitigation strategies.

The aim of this study is to quantify the CO₂, CH₄, and N₂O fluxes, over a two years period, in two paddy fields located in the Po Valley (Italy) under different rice straw management practices: soil incorporation versus open field burning. 

CO₂ and CH₄ fluxes are monitored continuously at ecosystem level using two eddy covariance (EC) towers (each equipped with Li-7500D and Li-7700 IRGA). Soil N₂O emissions are measured using a non-steady-state flow-through chamber (Li-8210-s) coupled to an infrared gas detector (Li-7820). 

When considering CO₂ alone, the cumulative emissions from the first growing season indicate that both rice fields acted as CO₂ sinks, each absorbing approximately 5.5 t CO₂ ha^-1. 

CH₄ emissions are comparable between fields and are estimated at 1.2 t CO₂e_100y ha^-1 in the straw burning field and 1.0 t CO₂e_100y ha^-1 in the straw incorporated field. N₂O emissions are higher in the straw burned field (10 t CO₂e_100y ha^-1 vs. 5.15 t CO₂e_100y ha^-1), likely due to topography-driven flooding differences.

The knowledge gathered from this project will contribute to the optimization of good agricultural management practices in rice cultivation. In addition, it will also represent an important contribution to the Italian GHG emissions inventory.

209 Drivers and Long-term Trend of Carbon Uptake and Acidification in Norwegian Fjords

Poster

Abdirahman Omar¹*, Lars Asplin², Svein Rune Erga³, Ingunn Skjelvan¹

¹NORCE Research, Bergen, Norway. ²Institute of Marine Research and Bjerknes Centre for Climate Research, Bergen, Norway. ³University of Bergen, Bergen, Norway

Session

Session 4: Carbon cycling in the land ocean aquatic continuum

Abstract text

Norwegian fjords function as natural sinks for atmospheric CO2, providing an essential

ecosystem service—but this sink also acts as a driver of ocean acidification within these

systems. In this study, we present, for the first time, a quantification of the long-term trend in

fugacity of CO2 (fCO2) and associated acidification variables, documenting how CO2

dynamics in the fjords have evolved over the last decades.

Our findings reveal that freshwater input plays an important role in modulating CO2

undersaturation, thereby enhancing the fjords’ capacity for atmospheric CO2 uptake. Seasonal

pH fluctuations are primarily governed by variations in Dissolved Inorganic Carbon (DIC),

which are linked to biological processes such as photosynthesis and respiration. Crucially, our

analysis shows that the sustained uptake of atmospheric CO2 has led to a measurable increase

in acidification within Norwegian fjords. Moreover, the findings highlight that changes in

freshwater input—whether due to climate change or human activities—can have significant

consequences for both CO2 uptake and acidification in these environments. These insights

underscore both the climate mitigation value of fjord CO2 uptake and highlight that the

biogeochemical consequences of this process must be carefully considered when developing

Carbon Dioxide Removal (CDR) methods aimed at enhancing oceanic carbon uptake.

210 The CAPASOS instrument: ∆pCO₂ measurements from USVs

Oral

Ute Schuster^1,2, Witold Tatkiewicz¹*, Luke Home¹

¹University of Exeter, Exeter, United Kingdom. ²ICOS OTC, Exeter, United Kingdom

Session

Session 5: Advancing marine CO₂ observations through next-generation sensors, integration and platform innovation

Abstract text

The CAPASOS instrument was designed and built at the University of Exeter (UoE, UK) to address the scientific community’s need for a low-cost, reliable system capable of collecting high-quality ∆pCO₂ data in remote ocean regions. The instrument consists of two main components: a dry compartment (the Core Unit) housing custom electronic PCBs, sensors, valves, pumps, etc.; and a wet section (the Equilibrator) where air–sea gas exchange occurs. A key design principle was minimisation of size and power consumption (3W average, up to 12W during start-up) to facilitate integration into Uncrewed Surface Vehicles (USVs). The result is a compact, flexible, and energy-efficient device suitable for autonomous deployments.

Within the framework of the HORIZON EUROPE GEORGE project, UoE and Offshore Sensing AS (Norway) collaborate to integrate the instrument into the SailBuoy surface uncrewed vehicle (USV). The integrated system was deployed from, and recovered again at, the Irish coast in July and August 2025, completing a 60-day mission, travelling over 1,000 km to and from the Porcupine Abyssal Plain – Sustained Observatory (PAP-SO). Post-recovery assessment of hardware and data analysis are continuing, with improvements being identified.

Current efforts focus on resolving identified issues and preparing for three planned deployments: 1) Northern Mediterranean Sea, summer 2026 (2^nd demo in the GEORGE project), 2) Southern Ocean, autumn 2026 (TRICUSO project), and 3) Southern Ocean, autumn 2027 (TRICUSO project).

We will present details of the completed deployments and experiences of the development and improvement process.

211 Heat and CO₂ fluxes above ground-based photovoltaic sites: measurements and simulations

Poster

Emma Lopez^1,2*, Virginie Moreaux¹, Jean-Christophe Domec³, Christophe Chipeaux², Cyriane Garrigou², Denis Loustau²

¹Ginger Burgeap, Lyon, France. ²INRAE ISPA, Villenave d’Ornon, France. ³Bordeaux Sciences Agro, Gradignan, France

Session

Session 12: Comparing long-term eddy covariance measurements in terrestrial ecosystems with carbon stock variations: lessons and future challenges

Abstract text

The expansion of ground-mounted photovoltaic parks is transforming land use, often replacing forested, semi-natural ecosystems or industrial brownfield and potentially altering local biophysical functioning. Understanding how these infrastructures modify soil-vegetation-atmosphere interactions is essential to assess and mitigate their environmental impacts.

In this context, we measured CO₂, sensible heat and latent heat fluxes over a three-year period using eddy covariance at two solar parks in rural and urban areas in France. In parallel, the mechanistic biophysical model GO+ (Moreaux et al., 2020), originally developed for forest ecosystems like the ICOS reference site near the rural solar park, was adapted for solar park ecosystems, including vegetation and panels interactions. In this study, we performed a comparison between in situ flux measurements and model results. Field observations of fluxes, and surface variables were used to evaluate the model's performance. We explored the effect of vegetation cover and management practices around panels on energy, evapotranspiration, carbon exchanges, and microclimate regulation. Both the model and the measurements indicate a net loss of carbon from the solar park following deforestation. Unexpectedly, the sensible heat flux from the solar park did not exceed that of the forest, whereas the latent heat flux from vegetation and soil beneath the panels, although lower than that of the forest, was maintained throughout the year. 

This work contributes to a better understanding of soil-plant-atmosphere interactions in emerging landscapes integrating renewable energy deployment. It also helps to identify pathways for maintaining or improving biophysical and ecological performance relative to previous land use.

212 Influence of open vegetation fires on atmospheric methane at two European GAW/WMO and ICOS measurement sites.

Poster

Rabia Ali Hundal^1,2,3, Michela Maione², Davide Putero⁴, Ute Karstens⁵, Dagmar Kubistin⁶, Matthias Lindauer⁶, Jennifer Müller-Williams⁷, Paolo Cristofanelli³*

¹Scuola Universitaria Superiore Pavia, Pavia, Italy. ²Urbino University, Urbino, Italy. ³Consiglio Nazionale delle Ricerche - Istituto di Scienze dell'Atmosfera e del Clima, Bologna, Italy. ⁴Consiglio Nazionale delle Ricerche - Istituto di Scienze dell'Atmosfera e del Clima, Torino, Italy. ⁵ICOS Carbon Portal, Department of Earth and Environmental Sciences, Lund University, Lund, Sweden. ⁶Deutscher Wetterdienst, Hohenpeissenberg Meteorological Observatory, Hohenpeissenberg, Germany. ⁷Deutscher Wetterdienst, Hohenpeissenberg Meteorological Observatory, Hohenpeissenberg, Italy

Session

Session 6: Carbon cycle in the Mediterranean region: from the local to the regional scale

Abstract text

Open vegetation fires are a significant source of atmospheric methane (CH₄), with plumes that can be transported over long distances under favorable meteorological conditions. 

This study evaluates how air masses affected by fires in Europe and the Mediterranean influenced short‑term CH₄ enhancements at Mt. Cimone (CMN, Italy) and Hohenpeissenberg (HPB, Germany) from 2015 to 2020. Fire‑influenced days were identified by combining 5‑day LAGRANTO back‑trajectory with GFED5 fire data. CH₄ residuals obtained by time‑series decomposition were analysed, with additional sensitivity tests based on increasing CO‑residual thresholds to isolate combustion‑rich plumes.

Both sites exhibited clear but episodic CH₄ enhancements during fire‑affected periods. At CMN, fire days produced marked increases in the upper tail of the CH₄ distribution: the 95^th percentile increased from 28.3 to 54.5 ppb and the 99^th from 47.3 to 79.2 ppb. The frequency of detected events showed a clear seasonal cycle, peaking in summer–autumn, with average CH₄ increases of +15 ppb in the warm season and +20 ppb in the cold season.

At HPB, fire‑influenced days were less frequent but associated with stronger CH₄ anomalies, with the 95^th percentile of CH₄ residuals increasing from 50.1 to 74.8 ppb. The detected events were concentrated in late summer–early autumn and were largely characterized by intense episodic plumes.

Preliminary comparisons with STILT+GFAS simulations indicates that LAGRANTO+GFED identifies stronger CH₄ signals from fire activity at both the measurement sites. These differences will also be discussed. 

213 Evaluating neural radiance fields as a proximal sensing tool for tree structure: A comparison with terrestrial laser scanning across open to closed stands

Poster

Sruthi Moorthy Krishna Moorthy Parvathi*, Maren Lechner, Henry Cerbone, Bhavya Palugudi, Graham Taylor, Roberto Salguero-Gomez

University of Oxford, Oxford, United Kingdom

Session

Session 20: Unmanned autonomous vehicles and proximal sensing in greenhouse gas research and monitoring

Abstract text

Accurate quantification of woody vegetation structure is fundamental for estimating above-ground biomass and carbon stocks, yet current high-precision approaches such as terrestrial laser scanning (TLS) remain costly and logistically demanding. Neural Radiance Fields (NeRF) offer a novel alternative by reconstructing three-dimensional structure from conventional imagery while simultaneously producing photorealistic renderings of vegetation scenes. However, their structural reliability across varying levels of canopy complexity remains poorly understood.

We compared NeRF-derived reconstructions with co-located TLS data within an urban setting (a public park in Oxford, UK), spanning a gradient of tree densities and architectural complexity from open parkland configurations to locally dense canopy clusters. Structural metrics including diameter, tree height, crown area and volume, and branching complexity were extracted to assess reconstruction completeness and geometric accuracy across tree size classes. Results indicate that NeRF captures small (<2 m height) with high completeness, outperforming TLS where occlusion and beam divergence limit returns on fine branches. In relatively open urban environments, large trees are reconstructed with strong geometric agreement. However, performance declines in densely clustered areas where canopy occlusion restricts multi-view coverage, leading to incomplete representation of upper crowns and interior branching structure.

Overall, NeRF performs best in structurally simple and semi-open environments such as urban green spaces and savannas. Given its low equipment requirements and realistic rendering capability, NeRF shows strong potential as a scalable proximal sensing tool for woody vegetation structure, supporting improved biomass estimation and carbon monitoring frameworks in ecosystems where traditional lidar deployment is constrained.

214 Integrating ecological knowledge and machine learning for partitioning net ecosystem exchange

Poster

Henriikka Vekuri*, Yu Zhou, Lukas Hörtnagl, Nina Buchmann

ETH Zurich, Department of Environmental Systems Science D-USYS, Institute of Agricultural Sciences, Zurich, Switzerland

Session

Session 13: Advancing approaches for quantifying greenhouse gas fluxes in terrestrial ecosystems

Abstract text

Eddy covariance measurements of net ecosystem exchange (NEE) are often partitioned into its component fluxes, gross primary productivity (GPP) and total ecosystem respiration (TER). This is done to gain process understanding, advance ecosystem model development and assess ecosystem responses to environmental change. The most widely used partitioning methods are the nighttime and daytime approaches, which fit simple nonlinear models to estimate GPP and TER using only a few environmental drivers. However, both GPP and TER are influenced by many additional factors, such as soil water content and temperature, beyond the commonly used drivers air temperature, vapor pressure deficit, and shortwave radiation. Moreover, traditional approaches assume that the temperature sensitivity and reference rate of respiration are identical during daytime and nighttime, and their parameter estimates may suffer from equifinality. To address these limitations, we develop a knowledge-guided machine learning (KGML) model for NEE partitioning. Our approach incorporates process understanding into the model structure and training procedure while maintaining the flexibility of ML to capture complex nonlinear responses. We compare the outputs of our model with standard nighttime and daytime approaches using ICOS Level 2 data from grassland and forest ecosystems, as well as with partitioned fluxes derived from the flux variance similarity method using raw, high-frequency data from a selected subset of sites. In our analysis, we focus on (1) consistency of partitioning estimates across approaches, (2) differences in partitioning estimates across ecosystem types, and (3) the ability of the methods to capture rapid changes, for example following precipitation events.

215 High-resolution reconstruction of the air-sea flux of CO₂ from the coast to the open ocean in the Western Tropical Atlantic (12^oS-4^oS)

Poster

Nathalie Lefèvre¹*, Manuel de Jesus Flores Montes², Doris Regina Aires Veleda², Denis Diverrès³

¹IRD LOCEAN, Paris, France. ²UFPE, Recife, Brazil. ³IRD, Plouzané, France

Session

Session 4: Carbon cycling in the land ocean aquatic continuum

Abstract text

The northeastern Brazilian coast is a poorly documented area despite its potential role in regional carbon budgets. Here we investigate the air–sea CO₂ exchange between 12°S and 4°S. 

Surface ocean and atmospheric CO₂ fugacity (fCO₂) have been measured continuously from an underway system aboard a merchant ship sailing between Europe and South America from 2008 to 2024. 

We use these repeated in situ observations to train a random forest model and reconstruct air–sea CO₂ fluxes at 1/12° resolution in the study region. The results indicate that the region acts overall as a source of CO₂ throughout the year averaging 5.4 TgC/yr. However, seasonal CO₂ sinks appear in the southern part of the domain (<8°S) from July to September in the 2008–2024 climatology, with a spatial extent that varies from year to year. Because the region is oligotrophic, these sink patterns are primarily associated with variations in sea surface temperature and water masses rather than biological activity. The strongest CO₂ uptake occurs in July 2012 (0.16 TgC) covering 28% of the area. It is linked to the large-scale cooling observed in the South Atlantic that year.

A narrow coastal band exhibits elevated chlorophyll concentrations and organic matter, suggesting enhanced biological influence. Additional observations are needed in this area to better constrain and validate the model. Nevertheless, as the regional flux is dominated by offshore waters, uncertainties in this coastal band are unlikely to significantly affect the overall CO₂ flux estimates.

216 Closing the gap: towards local-scale greenhouse gas flux quantification with UAV- and tripod-borne hyperspectral remote sensing

Poster

Sophie Vidal*, Daniel Hernan Orozco, Jesper Nørlem Kamp, Christoffer Karoff

Aarhus University, Aarhus, Denmark

Session

Session 20: Unmanned autonomous vehicles and proximal sensing in greenhouse gas research and monitoring

Abstract text

Accurate, spatially resolved quantification of greenhouse gas fluxes remains a major challenge in climate monitoring, emission verification, and the study of ecosystem processes. Established methods such as eddy covariance towers, static chambers, and satellite retrievals have important limitations in spatial resolution, temporal coverage, or observational scale. Proximal hyperspectral sensing may help bridge this gap.

This project investigates UAV- and tripod-borne hyperspectral remote sensing in the short-wave infrared range (970-2500 nm) for detection and quantification of CO₂ and CH₄ using the HySpex Mjolnir S-620 camera. Method development and testing will be carried out across different application domains, starting with: (1) controlled release experiments to determine detection limits and establish retrieval workflows; (2) restored wetland sites in central and northern Jutland, where flux mapping combined with soil process analysis may improve understanding of emission drivers; and (3) industrial and point sources such as biogas facilities and landfills supporting emission reporting and verification.

Column concentrations will be retrieved using a matched filter approach. Flux quantification methodology will be evaluated and adapted; including the Integrated Mass Enhancement method combined with wind data e.g. from a portable LiDAR wind scanner, and machine learning approaches. Initial wetland deployments will be supported by static chamber and eddy covariance measurements as independent reference data.

By September 2026, we expect to present detection limits from controlled release trials and first wetland observations. These results will provide a basis for discussion of methodological challenges, potential integration with ICOS-related monitoring activities, and future upscaling to regional modelling frameworks.

217 From flux tower to art work: art–science collaboration for communicating sustainability research in tropical landscapes

Poster

Alexander Knohl^1,2*, Christian Stiegler³, Ajeng Nurul Aini⁴, Fabian Brambach⁵, Adhari Donora⁶, Michael Fürst^7,8, Jana Juhrbandt⁹, Nina-Maria Knohl⁷, Immanuel Manurung⁹, Aiyen Tjoa¹⁰

¹Bioclimatology, University of Goettingen, Goettingen, Germany. ²Center for Sustainable Land Use and Biodiversity, University of Goettingen, Goettingen, Germany. ³Earth Sciences New Zealand, Wellington, New Zealand. ⁴ruangrupa, Jakarta, Indonesia. ⁵Forest Botanical Garden and Arboretum, University of Goettingen, Goettingen, Germany. ⁶Rumah Budaya Sikukeluang, Riau, Indonesia. ⁷Forum Wissen, University of Goettingen, Goettingen, Germany. ⁸Film Museum Potsdam, Potsdam, Germany. ⁹Environmental and Resource Economics, University of Goettingen, Goettingen, Germany. ¹⁰Faculty of Agriculture, Tadulako University, Palu, Indonesia

Session

Session 33: Science and arts: How to communicate science?

Abstract text

Effective science communication extends beyond disseminating research findings; it also highlights the practices, perspectives, and collaborations that shape knowledge production. Integrating artistic collaborations into communication strategies can open complementary pathways for engaging diverse audiences with scientific research.

Here, we present a collaboration between the University of Göttingen and Indonesian universities within the EFForTs project, which investigates ecological and socioeconomic functions of tropical rainforest transformation systems in Sumatra, Indonesia. The project operates a long-term eddy covariance flux tower in an oil palm plantation and examines how land-use change affects ecosystem functioning, greenhouse gas fluxes, biodiversity, and socioeconomic dynamics. To expand communication and engagement, researchers collaborated with documenta fifteen in Kassel and the Indonesian artist collective Rumah Budaya Sikukeluang as part of Lumbung Indonesia to develop transdisciplinary formats connecting science, art, and local communities.

A central component was the Semah Bumi festival in Jambi, a region undergoing rapid transformation from tropical rainforest to oil palm and rubber plantations. The festival served as dialogical platform where researchers, artists, and community members exchanged perspectives on ecological change, local knowledge, and cultural practices. The event also launched the “Sustainable Village” transformation lab, a long-term initiative focused on social learning, experimentation, and co-creation of sustainable land-use pathways. Outcomes were presented with corresponding objects from the University’s scientific collections in the Saujana Membumi exhibition at Forum Wissen in Göttingen and through collaborations with German schools.

This case study illustrates how artistic practice can expand the communicative reach of sustainability science and foster co-creation, reflexivity, and transformative learning.

218 Leveraging Ships of Opportunity to Improve Southern Ocean Carbon Sink Estimates

Oral

Jacqueline Behncke¹*, Toste Tanhua¹, Lucie Knor¹, Kevin O’Brien², Richard Sanders³, Peter Landschützer⁴

¹GEOMAR Helmholtz Centre for Ocean Research Kiel, Kiel, Germany. ²OceanOPS, Plouzané, France. ³NORCE Norwegian Research Centre & Bjerknes Centre for Climate Research, Bergen, Norway. ⁴Flanders Marine Institute (VLIZ), Ostend, Belgium

Session

Session 1: How big is the open ocean carbon sink?

Abstract text

The Southern Ocean carbon sink moderates Earth’s climate by absorbing about 40% of anthropogenic CO₂ emissions that end up in the ocean; yet the uncertainty in its estimate is substantial, on the same order of magnitude as EU fossil fuel emissions. Recent efforts to reduce these uncertainties include citizen-science initiatives using sailboats in professional races. Repeated measurements from these races have been shown to improve estimates of the ocean carbon sink. However, sailing races typically follow similar routes every 2–4 years, leaving vast areas of the Southern Ocean undersampled.

Equipping additional vessels, such as cruise ships, could strengthen observational coverage and improve Southern Ocean carbon sink estimates.

Here, we assess this potential using observing system simulation experiments (OSSEs). We treat output from the biogeochemical HAMOCC model as ground truth and apply real-world sampling masks. Using the neural network reconstruction method (SOMFFN), we quantify the impact of distinct observing strategies on Southern Ocean carbon sink estimates.

We evaluate an economical OSSE scenario in which vessels already operating in the Southern Ocean are equipped with autonomous CO₂ sensors. Specifically, we quantify the added value of a single tourist expedition cruise near the Antarctic Peninsula providing CO₂ measurements, and assess the potential impact of regularly operating Southern Ocean cruise ships that currently lack CO₂ observations but could be instrumented with autonomous sensors.

Our results demonstrate that expanding autonomous CO₂ measurements on existing vessels can substantially improve Southern Ocean carbon sink estimates in a cost-effective, scalable, and actionable way. 

219 Designing a Global Δ¹⁴CO₂ Network: Conceptual Framework for Observing System Simulation Experiments

Oral

Timo Knaack¹*, Christian Rödenbeck², Fabian Maier², Sanam N. Vardag³, Samuel Hammer¹

¹ICOS CRL, Institute of Environmental Physics, Heidelberg University, Heidelberg, Germany. ²Max Planck Institute for Biogeochemistry, Jena, Germany. ³Institute of Environmental Physics, Heidelberg University, Heidelberg, Germany

Session

Session 26: Designing the ideal global greenhouse gas monitoring network

Abstract text

Radiocarbon (¹⁴CO₂) offers a unique possibility to differentiate anthropogenic and biogenic sources of CO₂, as CO₂ from fossil fuel combustion is completely devoid of the ¹⁴C isotope. Despite its diagnostic strength, ¹⁴CO₂ remains underutilised in global monitoring systems due to analytical complexity and limited sampling coverage. To enable scalable deployment, autonomous sampling technologies are currently being developed to support unattended operation under remote supervision. This development motivates a realistic assessment of the scientific benefits of an expanded global ¹⁴CO₂ monitoring network, including its potential role within the Global Greenhouse Gas Watch (G3W) initiative.

Here we present a conceptual framework for a set of Observing System Simulation Experiments (OSSEs) designed to evaluate the benefit of a global Δ¹⁴CO₂ network for estimating, e.g. fossil fuel CO₂ emissions. In this study, we focus on the experimental design for conducting these OSSEs. The global ¹⁴CO₂ networks will be evaluated against various scientific objectives, including estimating continental fossil-fuel CO₂ emissions. The framework assesses the sensitivity of network design parameters, such as station density, site locations, and temporal sampling patterns, with respect to these objectives. Furthermore, it evaluates the impact of reducing measurement uncertainties through replicate sample analysis at strategic sites.

This work establishes a roadmap of necessary OSSEs and evaluation metrics to efficiently assess the information gain provided by an extended global Δ¹⁴CO₂ network, linking posterior uncertainty reduction and detection thresholds to spatial attribution performance and cost–efficiency, thereby offering a tool for optimal network design.

220 Integrated observations of oceanic CO2 in the North-Western Mediterranean Sea : new insights from the MOOSE network.

Poster

Thibaut Wagener¹*, Dominique Lefevre¹, Laurent Coppola²

¹MIO/CNRS/AMU, Marseille, France. ²LOV/CNRS/SU, Villefranche sur Mer, France

Session

Session 6: Carbon cycle in the Mediterranean region: from the local to the regional scale

Abstract text

The north-western Mediterranean Sea (NWMS), within a rather small geographical area, encompasses several key oceanographic processes: a convection area allowing ventilation of intermediate and deep water masses during winter, strong primary production variability related to nutrients input during winter and interaction between coastal and open sea waters. During the last decade, high sea surface temperatures have been reported in summer whereas less intense winter seems to reduce ventilation of intermediate and deep water masses. Physical and chemical properties of this basin have been monitored for the last 30 years, with a significant increase of these observations during the last decade in the frame of the french national sustained observing network (in particular MOOSE and SOMLIT). This makes the NWMS one of the most studied oceanographic areas of the globe. In this study, we will focuss on the main results provided by this integrated observing system. Long-term trends recorded over the last 20 years in term of inorganic carbon storage and ocean acidification will be presented with a focus on the seasonal variability and mechanisms involved in the CO₂ transfer between the surface and the deep ocean. Recent observations during high sea surface temperatures events in summer will be analyzed. The particular effect of the Rhone river discharge on the inorganic carbon cycling of this sub-bassin will also be considered. In conclusion, future directions on how this integrated effort could provide added value to the ICOS European research infrastructure will be discussed.

221 Greenhouse gas fluxes from a managed forest chronosequence in southern Sweden

Poster

Erica Jaakkola*, Tobias Biermann, Lena Ström, Patrik Vestin, Anders Lindroth

Lund University, Lund, Sweden

Session

Session 11: Climate feedbacks from ecosystem management and natural disturbances

Abstract text

Terrestrial ecosystems exchange multiple greenhouse gases with the atmosphere, and forest soils are key regulators of both CO₂ and CH₄ fluxes. However, knowledge about how these fluxes vary across forest stand age classes and management systems remains limited. Improving our understanding of these dynamics is essential for reducing uncertainties in forest carbon balance and informing sustainable management strategies.

We present ongoing work based on manual chamber measurements of CO₂ and CH₄ fluxes across ten Norway spruce forest stands in northeastern Skåne, Sweden, managed with rotation forestry spanning a chronosequence from clear-cut to mature forest (0 to ~120 years). The majority of these stands are located around the ICOS station Hyltemossa. In addition, the study includes one mixed forest stand managed with selective cutting. Each stand contains untreated reference plots and root-exclusion treatments, enabling future partitioning of autotrophic and heterotrophic respiration. Preliminary results indicate differences among age classes, with younger stands exhibiting higher summer CO₂ fluxes compared to older stands, weaker CH₄ uptake fromt younger stands compared to older stands, which function as more consistent CH₄ sinks. 

This study is part of a larger research effort aimed at identifying the stand age at which optimum carbon uptake occurs and evaluating rotation forestry against alternative management practices, such as selective cutting. By contributing empirical observations from a managed forest landscape, this study aims to reduce uncertainties in greenhouse gas flux estimates, support improved parameterization of vegetation models, and enhance assessments of forest carbon balance relevant to climate mitigation and forest management planning.

222 Measuring forest carbon sequestration with eddy covariance and forest inventories: a comparison of wood growth, net primary production and forest carbon budgets in six contrasting forested sites

Poster

Jeanne Poughon¹*, Maxime Cailleret², Nicolas Delpierre³, Daniel Berveiller³, Christophe Chipeaux⁴, Pascal Courtois⁵, Matthias Cuntz⁵, Jean-Christophe Domec⁶, Joannes Guillemot⁷, Emilie Joetzjer⁵, Jean Kempf¹, Sébastien Lafont⁴, Guerric Le Maire⁷, Olivier Marloie⁸, Alexandre Morfin³, Yann Nouvellon⁷, Jean-Marc Ourcival¹, Guillaume Simioni⁸, Jean-Marc Limousin¹

¹CEFE, Univ Montpellier, CNRS, EPHE, IRD, Montpellier, France. ²RECOVER, Aix-Marseille Univ, INRAE, Aix-en-Provence, France. ³Ecologie Société Evolution, CNRS, AgroParisTech, Université Paris-Saclay, Paris, France. ⁴INRAE, Villenave d'Ornon, France. ⁵Silva, Université de Lorraine, AgroParisTech, INRAE, Nancy, France. ⁶Bordeaux Sciences Agro, UMR ISPA, INRAE, Gradignan, France. ⁷Eco&Sols, Univ Montpellier, CIRAD, INRAE, IRD, Institut Agro Montpellier, Montpellier, France. ⁸URFM, INRAE, Avignon, France

Session

Session 12: Comparing long-term eddy covariance measurements in terrestrial ecosystems with carbon stock variations: lessons and future challenges

Abstract text

Carbon sequestration by forest ecosystems is commonly quantified by two primary methods: biometric measurements of standing wood biomass and litter production, and eddy-covariance measurements of CO2 exchange between the forest and the atmosphere. Biometric approaches can easily be applied to a large number of forested sites, such as through national forest inventories, and have the advantage of giving direct insights into carbon partitioning to the perennial wood biomass production and the other short-lived organs. Eddy-covariance measurements provide continuous measurements at the ecosystem level encompassing both plant and soil exchanges, but may be prone to several methodological biases.

Using net ecosystem productivity (NEP) and photosynthetic gross primary productivity (GPP) obtained from eddy-covariance measurements and independent net primary productivity (NPP) measured over 6 to 20 years on six contrasting forest sites (5 French ICOS sites and one in Brazil), we compared the two methods to calculate forest carbon budgets and investigate carbon residence time in the ecosystems.

NPP of perennial organs varied twenty-fold among sites (from a minimum in a Mediterranean evergreen forest to a maximum in a tropical eucalyptus plantation), while NEP varied only six-fold. The proportion of short-lived organs produced over perennial organs was another major difference between sites, ranging from 17% to 235%, suggesting very different carbon turnover rates. While NPP values were generally correlated with eddy-covariance based estimates of carbon exchanges, large differences in carbon use efficiency, tree growth and organic matter decomposition can also be evidenced when comparing the two methods.

223 Detecting biased carbon balance from the day/night NEE ratio

Poster

Leonardo Montagnani*, Anna Candotti, Torben Callesen

Free University of Bolzano, Bolzano, Italy

Session

Session 12: Comparing long-term eddy covariance measurements in terrestrial ecosystems with carbon stock variations: lessons and future challenges

Abstract text

Eddy covariance (EC) sites serve a dual purpose: they provide irreplaceable information on the relationships between environmental drivers and biogenic fluxes and, at the same time, allow the estimation of ecosystem-scale carbon balances. Although this latter application was among the original motivations for establishing EC sites, its broader use remains limited, possibly because some sites have been shown to yield biased carbon balance estimates.

Based on a preliminary analysis of ecosystem sites in the Alpine region, we propose a simple site-level metric to identify potentially biased sites, particularly those affected by CO₂ advection (inflow or outflow). Specifically, we evaluate the ratio between daytime and nighttime NEE.

We found that most sites fall within a day/night NEE ratio range of −1 to −2, whereas others deviate substantially, with one site exhibiting an extreme value of −27. To further assess this discrepancy, we compared annual mean NEP derived from EC measurements with NEP estimates based on forest inventory and harvest data. This comparison confirmed that sites with anomalous day/night NEE ratios also show inconsistencies between inventory-based and EC-based NEP estimates.

We therefore recommend that sites exhibiting anomalous day/night NEE ratios be subjected to further scrutiny or, where appropriate, excluded from modelling efforts in which accurate CO₂ balance estimates are critical.

224 Energy balance partitioning over rice based agro-ecosystem through BREB approach: A study in Lower Gangetic Plains of India

Oral

Saon Banerjee*

Bidhan Chandra Krishi Viswavidyalaya, Mohanpur, India

Session

Session 11: Climate feedbacks from ecosystem management and natural disturbances

Abstract text

The Gangetic Plains of India is one of the most prominent rice growing areas in the world. To study the latent heat flux and its management option, a field experiment was conducted in a large continuous patch of rice-field in lower Gangetic Plains, India. Net radiation (R_n) data along with profile data (dry bulb and wet bulb temperature and wind speed) were collected at different critical crop growth stages. Simultaneously data on biometric parameters were measured. The energy balance components were partitioned in respect to net radiation to observe the variation of the different components over the crop growth stages. For monsoon-rice season, it is observed that the ratio between soil heat flux (G) and Rn varies from around 0.32 to 0.03 for early stages of crop growth and maturity respectively. The LE/R_n ratio is ranging between 0.7 and 1.12 and at the later growth stage, the ratio is always more than 1 depicting more amount of sensible heat advection.  At the same region, the LE/R_n ratio is always less than 1 for mustard, pulses and other crops with low water demand. The alternate wetting and drying methods also reduce the LE values considerably, thus can control the methane emission. The irrigation management was also explored to reduce the total water application in rice with a aim to narrow down the total LE fluxes in a season. 

225 Carbon uptake from year-round measurements at an open-ocean observatory in the productive northeast Atlantic

Poster

Susan Hartman*, Anita Flohr, Pablo Trucco Pignata, Andrew Gates, Richard Cornes

NOC, Southampton, United Kingdom

Session

Session 5: Advancing marine CO₂ observations through next-generation sensors, integration and platform innovation

Abstract text

The productive Northeast Atlantic is net sink for atmospheric CO2. Understanding changes in this sink, and in the associated acidification, are key research questions. The Porcupine Abyssal Plain sustained observatory (PAP-SO) is a labelled ICOS and OceanSITES site of year-round measurements situated in the deep Northeast Atlantic (49°N 16.5°W, 4850 m). PAP-SO is one of the longest running open-ocean multidisciplinary observatory in the oceans around Europe producing high-resolution datasets integrating environmental and ecologically relevant variables from the surface to the seabed.

Since 2002, a full-depth mooring has been in place with autonomous sensors measuring surface ocean biogeochemical data and delivering it in near real time. Essential Ocean Variables (EOVs) include pCO2 and pH, along with temperature, salinity, dissolved oxygen, chlorophyll-a fluorescence, nitrate. Since 2010, the collaboration between the UK’s Meteorological Office and Natural Environment Research Council delivers simultaneous, open-access atmospheric and surface ocean datasets. The PAP-SO time-series attracts process studies and novel technology deployment, which have been used to improve CO2 measurements and flux calculations in this region of the northeast Atlantic.

226 Tracking urban and facility-scale methane emissions from Canadian infrastructure using atmospheric observations.

Oral

Felix Vogel*, Sebastien Ars, Lawson Gillespie, Shoma Yamanouchi, Kiran Ramlogan, Jordan Stuart, Meghan Flood, Cassandra Worthy

Environment and Climate Change Canada, Toronto, Canada

Session

Session 28: Combining data and models to support emissions estimation and policy at local to regional scales

Abstract text

To achieve its ambitious methane emission reduction goals, the Government of Canada is implementing mitigation policies and regulations in the energy and waste sector. The subnational greenhouse gas monitoring team of Environment and Climate Change Canada has developed a mobile Atmospheric Composition and Environmental Survey (ACES) platform, which has been used to survey over 75,000 km of roads across Canada in recent years. These surveys targeted natural gas distribution emissions in urban areas and major facilities of Canada’s downstream oil and gas infrastructure, including oil refineries, gas storage facilities as well as gas compressor and transmission stations. When combining these atmospheric observations with an improved open source Gaussian plume model we were able to estimate emission rates, which range from several kgCH₄ per hour for small sites in Ontario to over 10,000 kgCH₄ per hour for sites in Quebec. We furthermore deployed multiple observational techniques at one industrial site and during controlled-release experiments to assess their capabilities and limitations for methane emission monitoring.

In this presentation we will share lessons learned when designing and deploying our mobile ACES platform, how our Gaussian modelling framework was improved and how well different measurement techniques compare when tracking facility scale methane emissions. Lastly, we report emission rates quantified from our three-week methane survey across Eastern Canada through Ontario, Quebec, New Brunswick and Nova Scotia (~7000km).

227 Global Shocks and Disruptions to Scottish Peatlands – Exploring Carbon-Water Cycle Dynamics Using an Enhanced Modelling Approach

Poster

Luisa Orci Fernandez¹*, Mathew Williams¹, Roxane Andersen², Luke Smallman¹

¹University of Edinburgh, Edinburgh, United Kingdom. ²UHI North, West and Hebrides, Thurso, United Kingdom

Session

Session 15: Peatlands in the global climate system: fluxes, feedbacks, and human pressures

Abstract text

Peatlands occupy just 25% of Scotland land area but store >50% of its soil carbon. Peatlands are under constant threat by drainage, extraction for fuel, fire, and climate change. The Scottish Government has set ambitious net-zero goals, focusing on tree planting and peatland restoration. A robust understanding of the interactions between carbon and water cycles on peatland ecosystems is essential to support Scotland’s climate mitigation goals. 

Land ecosystem models calibrated with Eddy Covariance (EC) data can be used to conduct a more in-depth analysis between carbon, water cycle and climate feedbacks. In this study we use an ecosystem model (DALEC32) to provide uncertainty bounded analysis of the impacts of soil hydrology on peatland carbon cycling by combining it with a range of satellite-based and in-situ observations.

EC data from different peatland sites across Scotland and regional (EO) maps were used to calibrate and validate DALEC32 by implementing a model-data fusion framework (CARDAMOM). CARDAMOM employs a Bayesian approach within an adaptive-proposal-Markov-Chain Monte Carlo algorithm to retrieve model parameters and their uncertainties from site-specific (EC) and EO data.

Our CARDAMOM calibrated DALEC captures the overall trend of LAI (R²=0.67, RMSE=0.45 m²/m²) and was able to reproduce independent NEE observations (R²=0.73 and RMSE=0.48 gC/m²/day). Our analysis was also able to reproduce independent volumetric soil water content measurements at 10 cm depth (R² = 0.66, RMSE=0.14 m³/m³), and ET (R²=0.42, RMSE=0.95 mm/day). Here we present our efforts to implement water table depth estimation and explore its effect on peatland CO2 emissions.

228 Consistency across three independent observation-based ocean carbon sink estimates

Plenary

Neill Mackay¹*, Daniel Ford¹, Tobias Ehmen^2,1, Elena Koslova¹, Jamie Shutler¹, Andrew Watson¹

¹University of Exeter, Exeter, United Kingdom. ²University of Bergen, Bergen, Norway

Session

Session 1: How big is the open ocean carbon sink?

Abstract text

The ocean provides one of only two observational constraints on the global carbon budget, the other being the atmosphere. Quantifying the ocean carbon sink is therefore critical to accurate annual assessments of the budget that inform policy. In recent years, a discrepancy has emerged between estimates of the sink based on observational data products and those using global ocean biogeochemical models (GOBMs), with the former showing a larger sink and stronger increasing trend. This discrepancy adds uncertainty to the budget, compromising its ability to guide policy recommendations. We present a comparison of three completely independent observation-based estimates of the ocean carbon sink: one based on revised surface ocean fCO₂ data, one on atmospheric O₂/N₂ ratios, and one on a reconstruction of ocean interior carbon combined with an ocean inverse method. The three independent methods agree closely with one another over the period analysed (1996-2017), and by propagating uncertainties in each method we constrain the annual mean ocean carbon sink to within ± 0.35 Pg C yr^-1. Our combined estimate confirms the larger and more quickly increasing ocean carbon sink suggested by fCO₂ data product ensemble, and we show that the ensemble mean of the GOBM-based estimates is unlikely to be plausible. Combining our new ocean sink estimate with published data, we reconstruct a global carbon budget, inferring the land sink as a residual. Our results suggest the land sink stagnated over the period from 2003-2017, implying that efforts to enhance it through policy interventions have been largely ineffective.

229 Strong CH₄ emissions offset CO₂ sink in the largest lake wetland in the North China Plain

Poster

Kai Wang¹*, Yujie Ning¹, Ran Tian², Yuting Zhang¹, Xunhua Zheng¹

¹Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China. ²Xiong’an New Area Meteorological Service, Xiong'an, China

Session

Session 4: Carbon cycling in the land ocean aquatic continuum

Abstract text

Lake wetlands are critical components of the global carbon cycle, functioning as dynamic zones of CO₂ uptake and CH₄ emission. However, their complex heterogeneity and lack of multi-site observations create large uncertainties in assessing net radiative forcing. This study established a representative multi-site eddy covariance network in Baiyangdian (BYD) Lake, the largest freshwater wetland in the North China Plain.

Four towers measured CO₂ and CH₄ fluxes over open water, reed marsh, and lotus-dominated areas. Results from 2024-2025 reveal contrasting dynamics. The reed site was a strong CO₂ sink ( -304.2 g C m⁻² yr⁻¹), while open water was a consistent CO₂ source (167.2 g C m⁻² yr⁻¹). The lotus site showed a seasonal pattern with a near‑neutral annual budget (89.1 g C m⁻² yr⁻¹). Critically, both water and reed sites were persistent CH₄ sources (29.8 and 41.6 g C m⁻² yr⁻¹). Reed CH₄ emissions were 38.5 % higher than open water, attributed to plant‑mediated transport and root substrate supply. Temperature was the dominant control, with air temperature more strongly correlated at the reed site due to plant transport.

By integrating the multi-site EC data with remote sensing and machine learning, we found that when converting CH₄ to CO₂‑equivalents, the net greenhouse gas budget of the BYD Lake shifts from a CO₂ sink to a net source. This underscores the necessity of including CH₄ in wetland assessments and demonstrates the value of observational networks for constraining complex biogeochemical processes in lake wetlands.

230 Improving observational capability of carbon fluxes in the river–sea system of the Northern Adriatic.

Poster

Carolina Cantoni¹*, Mengjie Zhao^2,3, Antonio G. Antico¹, Vittorio E. Brando², Debora Bellafiore⁴, Stefano Cozzi¹

¹CNR-ISMAR, Istituto di Scienze Marine, Consiglio Nazionale delle Ricerche, Trieste, Italy. ²CNR-ISMAR, Istituto di Scienze Marine, Consiglio Nazionale delle Ricerche, Rome, Italy. ³Department of Civil, Building and Environmental Engineering, Sapienza University of Rome, Rome, Italy. ⁴CNR-ISMAR, Istituto di Scienze Marine, Consiglio Nazionale delle Ricerche, Venezia, Italy

Session

Session 4: Carbon cycling in the land ocean aquatic continuum

Abstract text

The Northern Adriatic is a dynamic land–ocean interface influenced by agricultural lowlands and karstic catchments, where rivers supply both organic carbon and total alkalinity to coastal waters. The magnitude and composition of these loads vary substantially across seasons and discharge regimes, yet their variability and fate in the coastal zone remain poorly constrained.

In this region, observations are conducted within the complementary frameworks of DANUBIUS-ERIC and ICOS-ERIC. To strengthen monitoring of the river–sea continuum, a new DANUBIUS riverine station was installed in spring 2025 at the mouth of the Isonzo River. High-frequency measurements are complemented by discrete sampling for dissolved organic carbon, dissolved inorganic carbon, and total alkalinity. These data provide a land-based counterpart to offshore carbonate system observations at the PALOMA ICOS station.

We present the first results of this integrated observational approach developed within the LandSeaLot Project and explore the potential of satellite remote sensing products to improve flux estimates and spatial representativeness. By integrating organic and inorganic carbon dynamics, this study reduces uncertainties in carbon partitioning along the land–ocean aquatic continuum and advances understanding of its role in the coastal carbon cycle under ongoing climatic and anthropogenic pressures.

231 Land Use Changes in the Savanna Belt of Sub-Saharan Africa: Meeting Local Community Needs and Creating Benefits When Planning REDD+ Projects

Poster

Abdelsalam Elfahal*

Independent Professional, Guelph, Canada

Session

Session 7: CO2 and CH4 cycle dynamics in Africa: observations, processes, and climate implications

Abstract text

In Sub-Saharan Africa, carbon loss is primarily driven by changes in land use. Across Africa, much of the land use change is due to extensive transformation of natural ecosystems into areas managed by humans, largely because of swift deforestation for agriculture and charcoal production.   The value chain varies depending on the scale of production.  Extensive production leads to deforestation and negatively affects rural communities' economies and livelihoods.  The REDD+ model suggests reserving an additional area of land, separate from protected areas, to be sustainably managed for local community benefits.  The case study utilizing data from Sudan demonstrated significant findings.  Financial analysis showed that sustainably managed fuel wood land yields similar revenue to land reserved for carbon credits.

232 Monitoring near-real-time climate impacts on biospheric CO₂ fluxes over Europe with a modified Vegetation Photosynthesis Respiration Model (VPRM)

Oral

Otto Briner^1,2*, Hassan Bazzi³, Philippe Ciais¹, Diego Santaren¹

¹Laboratoire des Sciences du Climat et de l’Environnement, UMR 1572 CEA-CNRS-UVSQ, Université Paris-Saclay, 91191, Gif-sur-Yvette, France. ²Atos France, Technical Services, 80 Quai Voltaire, 95870, Bezons, France. ³UMR TETIS, AgroParisTech, University of Montpellier, INRAE, CIRAD, CNRS, 34093, Montpellier, France

Session

Session 22: Remote sensing and flux observations to link disturbance, ecosystem states, and carbon-water fluxes

Abstract text

The rise of the atmospheric concentration of carbon dioxide (CO₂) has been buffered in part by the sustained absorption of CO₂ into the terrestrial biosphere. The impacts of climate change on ecosystem carbon sequestration are uncertain, so continuous monitoring of biospheric CO₂ fluxes is crucial to understand and detect carbon loss from ecosystems. Here, we present 0.1° × 0.1° gridded hourly ecosystem CO₂ fluxes upscaled at low latency and aimed toward capturing climate impacts on carbon cycling in the heterogeneous landscapes of temperate Europe. We integrate temporally dense 30 m Harmonized Landsat Sentinel-2 indices into a modified Vegetation Photosynthesis Respiration Model (VPRM) with restructured functions for soil moisture stress, respiration, and temperature dependence. We invert and validate model parameters in a Bayesian framework against eddy covariance observations of net ecosystem exchange (NEE) and gross primary production (GPP) at 43 ICOS sites. The resulting upscaled CO₂ fluxes over temperate Europe are examined over past periods of extreme climate, including 2021’s late spring frost and the summer heatwaves in 2022 and 2023. We place particular emphasis on extreme climate events in recent years for which impacts on ecosystem carbon cycling are more uncertain. This multiyear time series of gridded fluxes provides representation of sub-grid cell landscape heterogeneity at the European regional scale over the preceding decade and can be updated within hours of satellite data availability, enabling a detailed and rapid assessment of the impact of climate extremes on terrestrial carbon cycling.

233 Linking Satellite Observations and ICOS Flux Towers: Integrating PRISMA and Sentinel-2 for Ecosystem Productivity Monitoring

Poster

Carlos Camino¹*, Raphael Zuurbier¹, José Luis Pancorbo²

¹Laboratory of Geo-Information Science and Remote Sensing, Wageningen University, Wageningen, Netherlands. ²2. Consiglio Nazionale delle Ricerche – Istituto per la BioEconomia, Fiorenze, Italy

Session

Session 22: Remote sensing and flux observations to link disturbance, ecosystem states, and carbon-water fluxes

Abstract text

Accurate monitoring of vegetation productivity is essential for understanding ecosystem functioning and the terrestrial carbon cycle. Gross Primary Productivity (GPP) represents the main pathway through which carbon enters terrestrial ecosystems and is therefore a key indicator of ecosystem health and climate–biosphere interactions. While ICOS flux towers provide high-quality in-situ measurements of carbon fluxes, their spatial representativeness is limited to local tower footprints. Earth Observation (EO) data offer an opportunity to upscale these measurements and monitor vegetation productivity across larger spatial scales. This study evaluates the potential of integrating multispectral and hyperspectral satellite observations with ICOS flux tower measurements to estimate GPP in Mediterranean forest ecosystems. Sentinel-2 multispectral imagery and PRISMA hyperspectral data are analysed for two ICOS forest sites in Italy: San Rossore and Castelporziano. Satellite reflectance observations are combined with in-situ GPP estimates. Two machine learning modelling strategies are evaluated. First, empirical models based on vegetation indices (e.g., NDVI, EVI and red-edge indices) are used to estimate GPP. Second, a hybrid modelling framework combining radiative transfer model simulations (PROSAIL) with machine learning is implemented to retrieve functional plant traits such as leaf chlorophyll content and leaf area index, which are then used as predictors for GPP estimation. The study further assesses cross-sensor consistency through spectral harmonization and time-series analysis to evaluate the contribution of hyperspectral information to GPP prediction. The results provide insights into the integration of EO data with ICOS observations and highlight the potential of hyperspectral missions to improve ecosystem productivity monitoring.

234 Carbon Flux Monitoring: Comparing Emerging Sensor Networks with Eddy Covariance in a Cropland Ecosystem

Poster

Matthew Mullin¹, Andy Suyker², Matthew Saunders¹*

¹Trinity College Dublin, Dublin, Ireland. ²University of Nebraska-Lincoln, Lincoln, USA

Session

Session 31: Flux measurements for immediate societal benefits

Abstract text

This study presents an analysis of conventional eddy covariance carbon dioxide flux measurements compared to emerging sensors to determine cost-efficient, yet accurate alternatives for estimating ecosystem carbon budgets. A suite of LI-COR sensors is deployed at an agricultural soybean-corn crop site in Mead, Nebraska, to evaluate ecosystem carbon dioxide flux and associated environmental parameters. Four measurement positions are established throughout the reference site: a central location co-located with a traditional eddy covariance tower, surrounded by three discrete points to assess spatial variability.  Carbon nodes (LI-720) are deployed at each point of interest, providing carbon flux, water flux, sensible heat flux, and wind vectors. Each carbon node is accompanied by two water nodes (LI-710), which monitor evapotranspiration and atmospheric conditions. At fortnightly intervals, sensors are rotated to facilitate cross sensor comparison. A co-located carbon node is situated at the central position to provide sensor inter-comparison of the same signal.

 This analysis investigates emerging sensor comparability and spatial coherence relative to traditional carbon emission measurements across an agricultural site. To achieve this, historical crop yield data will be integrated with sensor data to assess the impact of land-use heterogeneity.  Ancillary environmental data, such as remote/near-Earth sensing, soil moisture content and high frequency foot-printing packages are integrated to provide spatial variability insights on inter-site carbon dioxide flux. Results demonstrate the viability of distributed low-cost sensors for carbon budget monitoring, robustness of benchmarking against eddy covariance flux towers, and capture heterogenous signal in an “homogenous” cropland ecosystem.

235 The R/V Gaia Blu: a new FerryBox pCO₂ line for the Mediterranean Sea

Poster

Carolina Cantoni¹*, Florian V. Kokozka², Tobias Steinhoff³, Marcello Felsani⁴, Melf Paulsen³, Malek Belgacem^5,6, Davide Vernazzani⁷, Giovanna Inserra⁷, Emanuele Organelli², Katrin Schroeder⁸