The ICOS Ocean Thematic Centre has published the results of the largest intercomparison study of its kind, comparing 27 instruments used for surface ocean pCO₂ measurements. The study provides a comprehensive comparison of the different measurement methods used by ICOS Ocean stations and other networks. Knowing the uncertainty of each method is crucial for estimating the global ocean CO₂ sink.
Surface ocean pCO₂ can be measured using a range of methods and instruments. Even within ICOS, stations apply several different approaches, reflecting the diversity of measurement platforms and the need to adapt to local conditions.
“The motivation was that we have various instruments in use in ICOS, and no standard system. Everyone must show that their instrument meets certain quality criteria, but still, it leaves the question whether this is the truth. We decided to make an intercomparison study to see how these instruments compare to each other,” explains Dr Tobias Steinhoff, lead author of the study who previously worked as a research scientist at the ICOS Ocean Thematic Centre.
The two-week experiment compared 27 instruments, representing 16 different models, in natural seawater at a tank facility at the Flanders Marine Institute (VLIZ). Eight of those instruments are used at ICOS Ocean stations. What began as an internal ICOS exercise grew into a wider international collaboration, involving both the scientific community and instrument manufacturers.
Direct air-water equilibration systems ranked highest in consistency
The instruments were tested side by side in a 5-m³ tank, with scientists carefully controlling water temperature and pCO₂ levels. High-quality reference gases from the ICOS Flask and Calibration Laboratory ensured thethat instruments that use reference gases for internal calibration could be checked and adjusted in the same way.
The instruments were divided into four categories: surface, underway air-water, underway membrane, and submersible. These reflect the different platforms on which they are deployed, from ships to buoys and autonomous platforms.
A major finding was that instruments using direct air–water equilibration delivered more consistent results and were less affected by external factors than those using membranes or photometric detection. This method proved especially reliable for ship-based measurements.
“The clarity of the results was a surprise. We expected to spend much more time working on the data before we could see any patterns. But once we grouped the instruments by type, it became obvious which measurement principles were the strongest”, Steinhoff describes.
Despite the results showing one measurement principle as the most reliable one, the authors are careful not prescribe a single “best” instrument. The choice depends on the platform, the environment, the required precision determined by the research question, and the expertise of the personnel handling the equipment and the data.
More observations are needed to address observation gaps
Globally, one of the main challenges is the scarcity of surface ocean CO₂ data. This lack of coverage leads to uncertainties, with models often underestimating the strength of the ocean carbon sink compared to observations.
Technology is advancing rapidly, offering more options than a decade ago. New instruments are increasingly compact and suitable for autonomous platforms, which are particularly important for filling observational gaps in remote regions such as the Southern Ocean.
Autonomous platforms and sensors, such as those developed in the ICOS-coordinated GEORGE project, can play a key role in accessing these remote areas.
“Several instruments designed for autonomous platforms took part, and they all performed very well. We couldn’t test the CaPASOS system, which is now being further developed in GEORGE, because it arrived damaged, but since a similar measurement principle proved successful, we can be confident that the work being done in GEORGE is on the right track”, Steinhoff says.
As the race to develop more autonomous technologies to address the observation gaps continues, the global community will need to find a balance between rapid technology development and quality data.
“We now have a range of tools with different levels of complexity, so the key is to combine them wisely. A backbone of high-quality systems can provide reference data, while simpler instruments can fill the gaps. This mix would give us both quality and coverage, and with technology developing so fast, we’ll likely need another intercomparison within five years”, Steinhoff says.
Results are being used to revise quality control systems
Between the study and the publishing of the results, some of the systems already been improved by the instrument manufacturers. One of the instrument manufacturers are continuing to develop their systems in cooperation with the ICOS community, most notably through testing environments offered by the GEORGE projects.
The results are also being implemented by The Surface Ocean CO₂ Atlas (SOCAT) to revise their quality control system.
“Until now, membrane-based systems - for example those used on sailing yachts - could receive a so-called ‘Flag C’, which meant their data could be included in the annual Global Carbon Budget. But to qualify for this, instruments must consistently perform within 5 microatmospheres, and our study showed that membrane systems could not reliably achieve that. As a result, SOCAT decided that these systems will no longer receive a Flag C. Their data will still be available, but it will no longer contribute to the global carbon budget. Of course, this was disappointing for the groups using them, but we cannot ignore the facts”, Steinhoff says.
“At the same time, it’s not a one-way decision. If there is evidence that performance improves, SOCAT will adapt. And on the positive side, several manufacturers really took the results on board. They’ve reflected on what can be improved and are already working to bring their instruments to a higher level. That’s exactly what we hoped to achieve.”
Read the full study: Steinhoff, T., Gkritzalis, T., Jones, S., Macovei, V. A., Neill, C., Schuster, U., ... & Wanninkhof, R. (2025). The ICOS OTC p CO2 instrument intercomparison. Limnology and Oceanography: Methods. https://aslopubs.onlinelibrary.wiley.com/doi/full/10.1002/lom3.10727