SOCAT
Bakker et al. (2016)
The largest collection of surface ocean carbon observations
Open ocean, Coastal ocean, Surface
continuous
1957 - 2025
SOCAT
Bakker et al. (2016)
Description:
The Surface Ocean CO2 Atlas features surface fCO2 measurements from both the open ocean and the coastal ocean, predominantly sourced from research vessels, ships of opportunity, and autonomous platforms including fixed moorings and uncrewed surface vehicles (USVs) (Bakker et al., 2016). It represents the most extensive collection of observational ocean CO2 data for the global surface ocean. Since 2013, SOCAT has been updated annually. Dataset flags indicate the estimated uncertainty and completeness of metadata in SOCAT synthesis products. The SOCAT gridded product contains fCO2 values with an estimated uncertainty of less than 5 µatm.
Reference:
Bakker, D. C. E., Pfeil, B., Landa, C. S., Metzl, N., O'Brien, K. M., Olsen, A., Smith, K., Cosca, C., Harasawa, S., Jones, S. D., Nakaoka, S., Nojiri, Y., Schuster, U., Steinhoff, T., Sweeney, C., Takahashi, T., Tilbrook, B., Wada, C., Wanninkhof, R., Alin, S. R., Balestrini, C. F., Barbero, L., Bates, N. R., Bianchi, A. A., Bonou, F., Boutin, J., Bozec, Y., Burger, E. F., Cai, W.-J., Castle, R. D., Chen, L., Chierici, M., Currie, K., Evans, W., Featherstone, C., Feely, R. A., Fransson, A., Goyet, C., Greenwood, N., Gregor, L., Hankin, S., Hardman-Mountford, N. J., Harlay, J., Hauck, J., Hoppema, M., Humphreys, M. P., Hunt, C. W., Huss, B., Ibánhez, J. S. P., Johannessen, T., Keeling, R., Kitidis, V., Körtzinger, A., Kozyr, A., Krasakopoulou, E., Kuwata, A., Landschützer, P., Lauvset, S. K., Lefèvre, N., Lo Monaco, C., Manke, A., Mathis, J. T., Merlivat, L., Millero, F. J., Monteiro, P. M. S., Munro, D. R., Murata, A., Newberger, T., Omar, A. M., Ono, T., Paterson, K., Pearce, D., Pierrot, D., Robbins, L. L., Saito, S., Salisbury, J., Schlitzer, R., Schneider, B., Schweitzer, R., Sieger, R., Skjelvan, I., Sullivan, K. F., Sutherland, S. C., Sutton, A. J., Tadokoro, K., Telszewski, M., Tuma, M., van Heuven, S. M. A. C., Vandemark, D., Ward, B., Watson, A. J., and Xu, S. (2016). A multi-decade record of high-quality fCO2 data in version 3 of the Surface Ocean CO2 Atlas (SOCAT), Earth Syst. Sci. Data, 8, 383–413, https://doi.org/10.5194/essd-8-383-2016.
LDEO Surface pCO2 Database
Takahashi et a. (2009)
The LDEO database reports pCO2 exclusively from equilibrator-CO2 analyzer systems
Open ocean, Surface
continuous
LDEO Surface pCO2 Database
Takahashi et a. (2009)
Description:
Dr. Taro Takahashi [LDEO, Palisades, New York] started synthesizing global surface ocean CO2 data in 1997, compiling three decades of observations (~250,000 measurements) to create inaugural monthly global surface pCO2 maps (Takahashi et al., 1997; Takahashi et al., 2002). The most recent version (V2019) expanded this dataset to approximately 14.2 million surface water pCO2 measurements spanning from 1957–2019. Distinct from the SOCAT database, the LDEO database reports pCO2, instead of fCO2, exclusively from equilibrator-CO2 analyzer systems, with an average estimated uncertainty of ± 2.5 µatm. The database is also interpolated onto a global surface ocean 4° × 5° grid for a reference year 2000 (Takahashi et al., 2009) and 2010 (Fay et al., 2024).
Reference:
Takahashi, T., Sutherland, S. C., Wanninkhof, R., Sweeney, C., Feely, R. A., Chipman, D. W., Hales, B., Friederich, G., Chavez, F., Sabine, C., Watson, A., Bakker, D. C. E., Schuster, U., Metzl, N., Yoshikawa-Inoue, H., Ishii, M., Midorikawa, T., Nojiri, Y., Körtzinger, A., Steinhoff, T., and de Baar, H. J. W.: Climatological mean and decadal change in Surface Ocean pCO2, and net sea–air CO2 flux over the Global Oceans, Deep Sea Research Part II: Topical Studies in Oceanography, 56(8–10), 554–577, https://doi.org/10.1016/j.dsr2.2008.12.009, 2009.
GLODAPv2
Lauvset et al. (2024)
Adjustments are applied by comparing data in the deep ocean (>2000 m) using a crossover and inversion method as described by Johnson et al. (2001).
Open ocean, Water column
discrete
1960 - 2024
GLODAPv2
Lauvset et al. (2024)
Description:
The Global Ocean Data Analysis Project Version 2 (GLODAPv2) aggregates biogeochemical data collected from discrete bottle samples, offering extensive global coverage from the surface to depths (Key et al., 2015; Olsen et al., 2016; Lauvset et al., 2024). While GLODAP is primarily a product for basin-scale hydrographic data, it also includes coastal datasets and observations from a few time-series. The GLODAPv2 data product provides rigorously quality-controlled measurements for 14 essential oceanographic variables: temperature, salinity, dissolved oxygen (DO), nitrate, silicate, phosphate, DIC, TA, pH, chlorofluorocarbons (CFC-11, CFC-12, CFC-113), carbon tetrachloride (CCl4), and sulfur hexafluoride (SF6). These variables, excluding temperature, undergo both primary and secondary quality control procedures to detect outliers and adjust for significant measurement biases. GLODAPv2 was first published in 2016 and was updated annually through a living data process in Earth System Science Data from 2019 through “v2023,” which was published in 2024. For these updates, new data (including historical data not previously included in the data product) are quality controlled and adjusted to the 2016 version (Olsen et al., 2019; Olsen et al., 2020; Lauvset et al., 2021; Lauvset et al., 2022; Lauvset et al., 2024). Since the global repeat hydrography programs operate with decadal repetitions, the aim is to produce a completely new version of GLODAP, where all cruise datasets will be reevaluated, every decade. Release of the GLODAPv3 data product is planned for 2026, and is expected to evolve the secondary data quality control practices relative to those used in GLODAPv2. For more information on the secondary quality control process, refer to Tanhua et al. (2010) and Lauvset and Tanhua (2015). GLODAPv2 offers two kind of products: the collection of quality controlled data from discrete bottle samples taken at sampling location (Key et al., 2015; Olsen et al., 2016; Olsen et al., 2019; Olsen et al., 2020; Lauvset et al., 2021; Lauvset et al., 2022; Lauvset et al., 2024), and a gridded product, interpolated to a 1° × 1° grid and the 33 standard depth levels of World Ocean Atlas (WOA) (Lauvset et al., 2016).
Reference:
Lauvset, S. K., Lange, N., Tanhua, T., Bittig, H. C., Olsen, A., Kozyr, A., Álvarez, M., Azetsu-Scott, K., Brown, P. J., Carter, B. R., Cotrim da Cunha, L., Hoppema, M., Humphreys, M. P., Ishii, M., Jeansson, E., Murata, A., Müller, J. D., Pérez, F. F., Schirnick, C., Steinfeldt, R., Suzuki, T., Ulfsbo, A., Velo, A., Woosley, R. J., and Key, R. M.: The annual update GLODAPv2.2023: the global interior ocean biogeochemical data product, Earth Syst. Sci. Data, 16, 2047–2072, https://doi.org/10.5194/essd-16-2047-2024, 2024
Quality Edited Hydrographic Data
Swift et al. (2026)
Similar to GLODAPv2, with no adjustments
Open ocean, Water column
irregular
1967-2025
Quality Edited Hydrographic Data
Swift et al. (2026)
Description:
The collection contains ocean basin spanning vertical section data from the WOCE Hydrographic Program, CLIVAR Repeat Hydrography, GO-SHIP and other programs with similar intent and high quality standards. Many of the sections have been occupied multiple times. The data values for measured parameters are identical those available from the CLIVAR and Carbon Hydrographic Data Office (CCHDO), but the files have been vetted and groomed for improved readability and consistency, and assembled in sensible geographic order. Data are organized by ocean region, section name, and year. On occasion, a data file is replaced with a preferred version or new transects are added to the collection.
Reference:
Swift, James H. and Barna, A. (2026). Quality Edited Vertical Ocean Section and Gridded Hydrographic Data. UC San Diego Library Digital Collections. Dataset. https://doi.org/10.6075/J0K074P0
World Ocean Database
Mishonov et al. (2024)
Similar to GLODAPv2, with no adjustments
Open ocean, Water column
discrete
World Ocean Database
Mishonov et al. (2024)
Description:
In addition to the GLODAPv2 and JOA Suite, users can also access historical and recent original biogeochemical data collected from discrete bottle samples in a uniform format and units, along with their originator quality control (QC) flags, through the World Ocean Database (WOD) (Mishonov et al., 2024). Like the JOA Suite, these measured data remain unaltered. The WOD allows users to filter and subset data with specific variables, platforms, institutions, projects, regions, or time periods (Garcia et al., 2024). Users can visualize sampling locations on a “distribution plot” and access a cruise list for all selected data and variables. Users also have the option of exporting data in NetCDF or Comma-Separated Values (CSV) formats. Additionally, all data in the WOD are reproducible and traceable to their original data sources archived at NOAA’s National Centers for Environmental Information (NCEI).
Reference:
Mishonov, A. V., Boyer, T. P., Baranova, O. K., Bouchard, C. N., Cross, S. L., Garcia H. E., Locarnini, R. A., Paver, C. R., Wang, Z., Seidov, D., Grodsky, A. I., Beauchamp, J. G.: World Ocean Database 2023, C. Bouchard, Technical Ed., NOAA Atlas NESDIS 97, 206 pp., https://doi.org/10.25923/z885-h264, 2024.
SNAPO-CO2
Metzl et al (2024)
A compilation of cruises from multiple French initiatives
Open ocean, Coastal ocean, Surface, Water column
discrete
1993-2022
SNAPO-CO2
Metzl et al (2024)
Description:
Synthesis of 44,400 measurements of DIC and TA from a series of research cruises and ships of opportunity across various oceanic regions from 1993-2022, from several French research programs, to create a product called “Service National d’Analyse des Paramètres Océaniques du CO2 (SNAPO-CO2)”.
Reference:
Metzl, N., Fin, J., Lo Monaco, C., Mignon, C., Alliouane, S., Antoine, D., Bourdin, G., Boutin, J., Bozec, Y., Conan, P., Coppola, L., Diaz, F., Douville, E., Durrieu de Madron, X., Gattuso, J.-P., Gazeau, F., Golbol, M., Lansard, B., Lefèvre, D., Lefèvre, N., Lombard, F., Louanchi, F., Merlivat, L., Olivier, L., Petrenko, A., Petton, S., Pujo-Pay, M., Rabouille, C., Reverdin, G., Ridame, C., Tribollet, A., Vellucci, V., Wagener, T., and Wimart-Rousseau, C.: A synthesis of ocean total alkalinity and dissolved inorganic carbon measurements from 1993 to 2022: the SNAPO-CO2-v1 dataset, Earth Syst. Sci. Data, 16, 89–120, https://doi.org/10.5194/essd-16-89-2024, 2024.
CODAP-NA
Jiang et al. (2021)
A discrete bottle based data product similar to GLODAPv2, but for the North American coastal ocean
Coastal ocean, Water column
discrete
2003 - 2018
CODAP-NA
Jiang et al. (2021)
Description:
Jiang et al. (2021) synthesized two decades of discrete measurements of carbonate system variables, DO, and nutrient data from the North American continental shelves to generate the first version of Coastal Ocean Data Analysis Data Product in North America (CODAP-NA). The 2021 release encompasses 3,391 oceanographic profiles from 61 research cruises spanning the North American continental shelves from Alaska to Mexico in the west and from Canada to the Caribbean in the east. It includes 14 key variables, including temperature, salinity, DO, DIC, TA, pH, carbonate ion, fCO2, silicate, phosphate, nitrate.
Reference:
Jiang, L.-Q., Feely, R. A., Wanninkhof, R., Greeley, D., Barbero, L., Alin, S., Carter, B. R., Pierrot, D., Featherstone, C., Hooper, J., Melrose, C., Monacci, N., Sharp, J. D., Shellito, S., Xu, Y.-Y., Kozyr, A., Byrne, R. H., Cai, W.-J., Cross, J., Johnson, G. C., Hales, B., Langdon, C., Mathis, J., Salisbury, J., and Townsend, D. W.: Coastal Ocean Data Analysis Product in North America (CODAP-NA) – an internally consistent data product for discrete inorganic carbon, oxygen, and nutrients on the North American ocean margins, Earth Syst. Sci. Data, 13, 2777–2799, https://doi.org/10.5194/essd-13-2777-2021, 2021.
MOCHA - Multistressor Observations of Coastal Hypoxia and Acidification
Kennedy et al., 2023
A compilation of discrete and continuous carbonate system and dissolved oxygen observations from the U.S. West Coast, consistently formatted and quality controlled.
Open ocean, Coastal ocean, Surface, Water column, Regional
irregular, daily, monthly, yearly
1949-2022
MOCHA - Multistressor Observations of Coastal Hypoxia and Acidification
Kennedy et al., 2023
Description:
MOCHA encompasses temperature, salinity, DO, carbonate system variables (DIC, TA, pH, pCO 2 , fCO 2 ), nutrients, and chlorophyll measurements from the full water column along the U.S. west coast. The synthesis integrates observations from 71 different sources, including high-resolution autonomous sensors, synoptic oceanographic cruises, and shoreline samples, and extends to beyond the continental shelf. MOCHA includes observations from CODAP-NA, California Cooperative Oceanic Fisheries Investigations (CalCOFI), and other large-scale oceanographic cruises to facilitate linking nearshore, high-resolution observations to broader oceanographic conditions. As of 2025, MOCHA includes 15.9 million temperature readings, 5.0 million salinity measurements, 3.9 million DO records, and 2.3 million pH measurements, along with 8,368 DIC, 10,144 TA, and 505,000 pCO2/fCO2 measurements, with limited additional chlorophyll and nutrient observations. All data in the MOCHA synthesis product has been quality controlled to a “plausible and reasonable” standard, but researchers requiring high-precision coastal data may need to apply additional QC tests.
Reference:
Kennedy, E. G., Zulian, M., Hamilton, S. L., Hill, T. M., Delgado, M., Fish, C. R., Gaylord, B., Kroeker, K. J., Palmer, H. M., Ricart, A. M., Sanford, E., Spalding, A. K., Ward, M., Carrasco, G., Elliott, M., Grisby, G. V., Harris, E., Jahncke, J., Rocheleau, C. N., Westerink, S., and Wilmot, M. I.: A high-resolution synthesis dataset for multistressor analyses along the US West Coast, Earth Syst. Sci. Data, 16, 219–243, https://doi.org/10.5194/essd-16-219-2024, 2024.
ARIOS
Padin et al. (2020)
Data product of 17,653 samples from 3,343 stations measuring pH, alkalinity, and biogeochemical parameters off NW Iberian Peninsula (1976-2018). Data primarily from Ría de Vigo, spanning Bay of Biscay to Portuguese coast across 24 cruise projects.
Open ocean, Coastal ocean, Surface, Water column, Regional
irregular, discrete
1976-2018
ARIOS
Padin et al. (2020)
Description:
A data product of 17 653 discrete samples from 3343 oceanographic stations combining measurements of pH, alkalinity and other biogeochemical parameters off the northwestern Iberian Peninsula from June 1976 to September 2018 is presented in this study. The oceanography cruises funded by 24 projects were primarily carried out in the Ría de Vigo coastal inlet but also in an area ranging from the Bay of Biscay to the Portuguese coast. The robust seasonal cycles and long-term trends were only calculated along a longitudinal section, gathering data from the coastal and oceanic zone of the Iberian upwelling system. The pH in the surface waters of these separated regions, which were highly variable due to intense photosynthesis and the remineralization of organic matter, showed an interannual acidification ranging from −0.0012 to −0.0039 yr−1 that grew towards the coastline. This result is obtained despite the buffering capacity increasing in the coastal waters further inland as shown by the increase in alkalinity by 1.1±0.7 and 2.6±1.0 µmol kg−1 yr−1 in the inner and outer Ría de Vigo respectively, driven by interannual changes in the surface salinity of 0.0193±0.0056 and 0.0426±0.016 psu yr−1 respectively. The loss of the vertical salinity gradient in the long-term trend in the inner ria was consistent with other significant biogeochemical changes such as a lower oxygen concentration and fertilization of the surface waters. These findings seem to be related to a growing footprint of sediment remineralization of organic matter in the surface layer of a more homogeneous water column.
Reference:
Padin, X.A., Velo, A., Pérez, F.F., 2020. ARIOS: a database for ocean acidification assessment in the Iberian upwelling system (1976–2018). Earth Syst. Sci. Data 12, 2647–2663. https://doi.org/10.5194/essd-12-2647-2020
Marine Inorganic Carbonate Chemistry in the Northern Gulf of Alaska
Monacci et al. (2024)
A synthesis of twenty cruises from 2008 to 2017 on the Gulf of Alaska (GAK) Line
Coastal ocean, Water column
discrete
2008 - 2017
Marine Inorganic Carbonate Chemistry in the Northern Gulf of Alaska
Monacci et al. (2024)
Description:
Monacci et al. (2023) compiled a data product of discrete seawater samples collected each May and September over a 10-year period from 2008 to 2017 along the long-term hydrographic line in the Gulf of Alaska (GAK Line). Samples were collected from a sampling rosette on a profiling CTD. Data variables include profiled seawater temperature, salinity, and DO. Discrete sample variables include DO (i.e., Winkler titrations), macronutrients (nitrate, nitrite, phosphate, silicic acid), DIC, and TA. All carbonate system variables were analyzed at the Ocean Acidification Research Center (OARC) at the University of Alaska Fairbanks (UAF). The repeat hydrographic cruises were funded by the Alaska Ocean Observing System (AOOS), the Exxon Valdez Oil Spill Trustee Council (EVOS), Gulf Watch Alaska, and the North Pacific Research Board (NPRB) and were mostly conducted aboard the United States Fish and Wildlife Service (USFWS) R/V Tiĝlax̂.
Reference:
Monacci, N. M., Cross, J. N., Evans, W., Mathis, J. T., and Wang, H.: A decade of marine inorganic carbon chemistry observations in the northern Gulf of Alaska – insights into an environment in transition, Earth Syst. Sci. Data, 16, 647–665, https://doi.org/10.5194/essd-16-647-2024, 2024.
NCRMP Atlantic Carbonate Chemistry
Enochs et al. (2018)
Carbonate chemistry data from diverse Atlantic sites—hand sampling, autosamplers, and AOAT buoy CalVal—used to assess spatial and temporal variability in coral-reef seawater.
Coastal ocean
irregular, monthly, yearly
2012 - 2024
NCRMP Atlantic Carbonate Chemistry
Enochs et al. (2018)
Description:
This collection contains carbonate chemistry data collected at both random and fixed long-term sites in the Atlantic basin. These data are collected and analyzed to assess spatial and temporal variation in the seawater carbonate systems of coral reef ecosystems and include a variety of sampling methods. With the first method water samples are collected by hand at the surface, either from a boat or by SCUBA divers. The second method uses subsurface autosamplers to collect water samples. The final method is sampling of surface waters from specified locations around the Atlantic Ocean Acidification Test-Bed (AOAT) buoy. Water samples are gathered around the buoy on a monthly basis and the analyzed data are used to help calibrate buoy sensors. This is referred to as calibration validation (CalVal) data and is included within this collection as it still provides a useful look into the carbonate chemistry of the surrounding area. Samples within these datasets are either collected singularly or as part of a diurnal set.
Reference:
National Oceanic and Atmospheric Administration; NOAA Atlantic Oceanographic and Meteorological Laboratory; Cooperative Institute for Marine and Atmospheric Studies (2018). National Coral Reef Monitoring Program: Carbonate chemistry data collected in the Atlantic Ocean from 2012-2024. NOAA National Centers for Environmental Information. Dataset. https://doi.org/10.25921/vfz0-dg77.
Salish Cruise Data Package
Alin et al. (2025a)
A QCed compilation of cruise datasets from 2008 to 2024 in the southern Salish Sea and northern Washington coast.
Coastal ocean, Estuaries, Water column
seasonal, discrete, continuous
2008–2024
Salish Cruise Data Package
Alin et al. (2025a)
Description:
Alin et al. (2025a) compiled data from 61 individual cruise data sets that sampled marine waters of the southern Salish Sea and northern Washington coast (United States) from 2008 to 2024. Ongoing seasonal sampling occurred during April, July, and September for Puget Sound cruises has occurred since 2014 and most frequently during May and October for Sound-to-Sea cruises, which sample from Puget Sound through the Strait of Juan de Fuca to the northern Washington coast. The Salish cruise data package contains observations from a oceanographic water column profiles, with CTD sensor measurements of temperature, salinity, and oxygen; as well as discrete measurements of oxygen, nutrient, and inorganic carbonate content.
Reference:
Alin, S.R., J.A. Newton, R.A. Feely, B. Curry, D. Greeley, J. Herndon, and M. Warner (2024a). A decade-long cruise time-series (2008–2018) of physical and biogeochemical conditions in the southern Salish Sea, North America. Earth System Science Data, 16, 837–865, https://doi.org/10.5194/essd-16-837-2024
Salish Cruise Multi-stressor Data Product
Alin et al. (2025b)
A multi-stressor (ocean acidification, hypoxia, marine heatwaves) data product based on cruises from 2008 to 2024 in the southern Salish Sea and northern Washington coast (USA).
Coastal ocean, Estuaries, Water column
seasonal, discrete, continuous
2008–2024
Salish Cruise Multi-stressor Data Product
Alin et al. (2025b)
Description:
Alin et al. (2024a) compiled data from 35 individual cruise data sets that sampled marine waters of the southern Salish Sea and northern Washington coast (United States) from 2008 to 2018. Ongoing seasonal sampling occurred during April, July, and September for Puget Sound cruises has occurred since 2014 and most frequently during May and October for Sound-to-Sea cruises, which sample from Puget Sound through the Strait of Juan de Fuca to the northern Washington coast. The Salish cruise data package contains observations from a total of 715 oceanographic profiles, with > 7490 sensor measurements of temperature, salinity, and DO; ≥ 6070 measurements of discrete DO and nutrient (nitrate, phosphate, silicate, ammonium, nitrite) samples; and ≥ 4462 measurements of inorganic carbonate system variables (DIC and TA). A follow-on data product based on the Salish cruise data package, which only included the 3971 samples with complete information for temperature, salinity, DO, nutrients, DIC, and TA, is also available (Alin et al., 2023). This product additionally provides the most commonly used calculated carbonate system variables: pH (total scale), fCO2, pCO2, Ωarag, and Ωcalc.
Reference:
Alin, S.R., J.A. Newton, R.A. Feely, S. Siedlecki, and D. Greeley (2024b). Seasonality and response of ocean acidification and hypoxia to major environmental anomalies in the southern Salish Sea, North America (2014–2018). Biogeosciences, 21, 1639–1673, https://doi.org/10.5194/bg-21-1639-2024
Line P Marine Carbonate Chemistry Compilation
Franco et al., (2021)
A compilation of fifty-five Line P cruises containing discrete DIC and TA profiles at five stations in the Northeast Pacific Ocean. Sampled approximately three times per year from 1990 to 2019
Open ocean, Water column, Regional
irregular, sub-annual, discrete
1990 - 2019
Line P Marine Carbonate Chemistry Compilation
Franco et al., (2021)
Description:
This dataset contains marine carbonate system measurements collected during 55 Line P cruises from 1990 to 2019 in the subarctic Northeast Pacific. The dataset contains quality-controlled, discrete profiles of dissolved inorganic carbon (DIC), total alkalinity (TA), temperature, salinity, dissolved oxygen and nutrients. From a total of 27 hydrographic time-series stations, only the five major stations where DIC and TA are routinely sampled were included in this compilation. Cruises were conducted approximately three times per year, typically in February, May/June and August/September. Each vertical profile was individually inspected and contrasted with the whole pool of data relative to salinity, density, and oxygen to detect and flag poor quality data following the World Ocean Circulation Experiment (WOCE) quality control convention. The recommended cruise-specific adjustments from the Pacific Ocean Interior Carbon (PACIFICA) data synthesis were applied. The Line P carbonate chemistry timeseries is maintained by Fisheries and Oceans Canada and continues to the present day. Data are available and continuously updated in the Line P repository, which can be publicly accessed after generating an account at https://waterproperties.ca.
Reference:
Franco, A. C., Ianson, D., Ross, T., Hamme, R. C., Monahan, A. H., Christian, J. R., Davelaar, M., Johnson, W. K., Miller, L. A., Robert, M., and Tortell, P. D. 2021. Anthropogenic and climatic contributions to observed carbon system trends in the northeast Pacific. Global Biogeochemical Cycles, 35(7). https://doi.org/10.1029/2020GB006829
Anthropogenic Carbon in the Arctic Ocean
Terhaar et al. (2020)
Observation-based estimates of anthropogenic carbon in the Arctic Ocean
Open ocean
climatology
2005
Anthropogenic Carbon in the Arctic Ocean
Terhaar et al. (2020)
Description:
This dataset includes anthropogenic carbon estimates in the Arctic Ocean based on measurements of transient tracers, such as CFC-12 and SF6 (Terhaar et al., 2020; Tanhua et al., 2009). Using the transient time distribution (TTD) method, anthropogenic carbon estimates were estimated at measurement locations across all basins of the Arctic Ocean between 1983 and 2005. In addition to these estimates, adjusted estimates of anthropogenic carbon at these locations are provided to account for differences in the saturation of transient tracers and anthropogenic carbon in Arctic Ocean surface waters that caused anthropogenic carbon estimates to be biased low (Terhaar et al., 2020). It is recommended to use the adjusted estimates. This dataset can be accessed at https://doi.org/10.17882/103920 (Terhaar et al., 2024).
Reference:
Terhaar, J., Tanhua, T., Stöven, T., Orr, J. C., and Bopp, L.: Evaluation of data‐based estimates of anthropogenic carbon in the Arctic Ocean, Journal of Geophysical Research: Oceans, 125(6), https://doi.org/10.1029/2020jc016124, 2020.
Bermuda Atlantic Time-series Study (BATS)
Bates and Johnson (2023)
Forty years of ocean acidification observations (1983-2023) in the Sargasso Sea at the Bermuda Atlantic Time-series Study (BATS) site.
Open ocean
monthly
1988-2025
Bermuda Atlantic Time-series Study (BATS)
Bates and Johnson (2023)
Description:
Ocean physical and biogeochemical conditions are rapidly changing over time. Forty years of observations from 1983 to 2023 collected at the Bermuda Atlantic Time-series Study (BATS) site near Bermuda in the North Atlantic Ocean shows continuing trends of surface warming, increase in salinity, loss of dissolved oxygen (DO), increase in carbon dioxide (CO2), and ocean acidification (OA) effects. Over this period, the ocean has warmed by about +1°C, increased in salinity by +0.136, and lost DO by 12.5 µmol kg−1 or ~6%. Since the 1980s, ocean dissolved inorganic carbon (DIC), total alkalinity (TA), a tracer of anthropogenic CO2 (CTrOCA), and fugacities/partial pressures of CO2 (i.e., fCO2 and pCO2) have continued to increase substantially, with no evidence of a reduction in the rates of change over time. Contemporaneously, ocean pH has decreased by ~0.1 pH units [with ocean acidity (i.e., H+) increasing by >30%], and the saturation states of calcium carbonate minerals (Ωcalcite and Ωaragonite) have decreased. These OA indicators show that the chemical conditions for calcification have become less favorable over the past 40 years. Updating of data and trends at the BATS site show how ocean chemistry of the 2020s is now outside the range observed in the 1980s, and how essential these data are for predicting the response of ocean chemistry and marine ecosystems to future shifting earth and ocean conditions.
Reference:
Bates, N., Johnson, R. J., Lomas, M. W., Smith, D., Lethaby, P. J., Bakker, R., Davey, E., Derbyshire, L., Enright, M., Garley, R., Hayden, M. G., Lomas, D., May, R., Medley, C., Stuart, E., Chambers, E. (2025) Discrete bottle samples collected at the Bermuda Atlantic Time-series Study (BATS) site in the Sargasso Sea from October 1988 through December 2024. Biological and Chemical Oceanography Data Management Office (BCO-DMO). (Version 8) Version Date 2025-06-27 [if applicable, indicate subset used]. doi:10.26008/1912/bco-dmo.3782.8 [access date];Bates, N., and Johnson, R.J., 2023. Forty years of ocean acidification observations (1983-2023) in the Sargasso Sea at the Bermuda Atlantic Time-series Study (BATS) site. Frontiers in Marine Science, 10, https://doi.org/10.3389/fmars.2023.1289931.
HOT
Dore et al. (2009)
35+ years of inorganic carbon dynamics in the open waters of the central North Pacific
Open ocean, Water column
monthly
1988-2025
HOT
Dore et al. (2009)
Description:
The Hawaii Ocean Time-series (HOT) CO2 measurement program documents 35+ years of inorganic carbon dynamics in the open waters of the central North Pacific. Since October 1988, full ocean depth profiles of DIC and TA have been analyzed, and direct measurements of pH have been made over most of this longest-running Pacific Ocean time-series study. The program is based on shipboard observations and experiments conducted on ~10 expeditions per annum to Station ALOHA (22.75°N, 158°W).
Reference:
Dore, J. E., Lukas, R., Sadler, D. W., Church, M. J., and Karl, D. M.: Physical and biogeochemical modulation of ocean acidification in the central North Pacific, Proc. Natl. Acad. Sci., 106, 12235-12240, https://doi.org/10.1073/pnas.0906044106, 2009.
Point B Time-series
Kapsenberg et al. (2017)
Carbonate chemistry at a coastal site of the Bay of Villefranche, France
Coastal ocean
discrete
Point B Time-series
Kapsenberg et al. (2017)
Description:
The Point B Time-series documents the carbonate chemistry at a coastal site of the Bay of Villefranche (43.686200N 7.314800E) in Villefranche-sur-mer, France, northwestern Mediterranean Sea. Since January 2007, seawater is sampled weekly at 1 and 50 m, and analyzed for DIC and TA. Salinity and temperature are extracted from CTD profiles. Variables of the carbonate system such as pH (total scale) are calculated using the R package seacarb.
Reference:
Kapsenberg, L., Alliouane, S., Gazeau, F., Mousseau, L., and Gattuso, J.-P.: Coastal ocean acidification and increasing total alkalinity in the northwestern Mediterranean Sea, Ocean Sci., 13, 411–426, https://doi.org/10.5194/os-13-411-2017, 2017.
Ny-Ålesund Time-series
Gattuso et al. (2023)
Carbonate chemistry at a coastal site of Kongsfjorden, Spitsbergen
Coastal ocean
continuous, discrete
Ny-Ålesund Time-series
Gattuso et al. (2023)
Description:
The Ny-Ålesund Time-series documents the carbonate chemistry at a coastal site of Kongsfjorden, Spitsbergen (78.930660N 11.920030E) during the period 2015-2021. It is the first high-frequency (1 hour), multi-year (6 years) dataset of salinity, temperature, pCO2, pH, as well as calculated DIC and TA in the High-Arctic Ocean (Gattuso et al., 2023).
Reference:
Gattuso, J.-P., Alliouane, S., and Fischer, P.: High-frequency, year-round time series of the carbonate chemistry in a high-Arctic fjord (Svalbard), Earth Syst. Sci. Data, 15, 2809–2825, https://doi.org/10.5194/essd-15-2809-2023, 2023.
pCO2 and pH Time-series from 40 Surface Buoys
Sutton et al. (2019)
Based on 40 moored surface pCO2 time-series, with 17 of them containing pH
Open ocean, Coastal ocean, Surface
continuous
pCO2 and pH Time-series from 40 Surface Buoys
Sutton et al. (2019)
Description:
Sutton et al. (2019) established a living dataset comprising 40 individual autonomous moored surface ocean pCO2 time-series established between 2004 and 2013, 17 of them also include autonomous pH measurements. These time-series characterize a wide range of surface ocean carbonate system conditions, across a variety of environments, including 17 oceanic and 13 coastal locations, as well as 10 coral reefs.
Reference:
Sutton, A. J., Feely, R. A., Maenner-Jones, S., Musielwicz, S., Osborne, J., Dietrich, C., Monacci, N., Cross, J., Bott, R., Kozyr, A., Andersson, A. J., Bates, N. R., Cai, W.-J., Cronin, M. F., De Carlo, E. H., Hales, B., Howden, S. D., Lee, C. M., Manzello, D. P., McPhaden, M. J., Meléndez, M., Mickett, J. B., Newton, J. A., Noakes, S. E., Noh, J. H., Olafsdottir, S. R., Salisbury, J. E., Send, U., Trull, T. W., Vandemark, D. C., and Weller, R. A.: Autonomous seawater pCO2 and pH time series from 40 surface buoys and the emergence of anthropogenic trends, Earth Syst. Sci. Data, 11, 421–439, https://doi.org/10.5194/essd-11-421-2019, 2019.
Takahashi ΔfCO2 & Flux Climatology
Fay et al. (2024)
Does not use proxy variables for extrapolation. Only produced as monthly climatology.
Open ocean
monthly
1° x 1°
2010
Takahashi ΔfCO2 & Flux Climatology
Fay et al. (2024)
Description:
Following on previous climatologies published by the late Taro Takahashi in 1997 and 2009, Fay et al. (2024) created a legacy climatology using his methodology and the updated SOCAT database of observations. This product provides 12 months of delta fCO2 values and corresponding fluxes for a base year of 2010 at 4° × 5° resolution subsequently subgridded
to 1° × 1° resolution and near-global coverage. This climatology represents the mean of ocean conditions over the last four decades and is distinctive relative to many other mechanistic machine learning approaches in that it interpolates in time and space using only the available fCO2 data and a surface water advection scheme rather than using proxy variables for gap filling. It uses the median of observations to determine a reference year of 2010 and fluxes are provided using air-sea partial pressure differences and inputs from the SeaFlux product (Fay et al., 2021).
Reference:
Fay, A. R., Munro, D. R., McKinley, G. A., Pierrot, D., Sutherland, S. C., Sweeney, C., and Wanninkhof, R. 2024. Updated climatological mean ΔfCO2 and net sea–air CO2 flux over the global open ocean regions, Earth Syst. Sci. Data, 16, 2123–2139, https://doi.org/10.5194/essd-16-2123-2024.
MPI-ULB-SOM-FFN
Landschützer et al. (2020)
Monthly gridded pCO2 without adjusting for a specific reference year, high-resolution coastal ocean coverage.
Open ocean, Coastal ocean, Surface
monthly, climatology
0.25° x 0.25°
MPI-ULB-SOM-FFN
Landschützer et al. (2020)
Description:
Landschützer et al. (2020a) created a uniform pCO2 climatology combining open and coastal oceans. It is a monthly gridded global surface ocean pCO2 data product without adjusting for a specific reference year. Developed on a higher-resolution 0.25° × 0.25° global surface-ocean grid, this product is the result of combining two neural network-based pCO2 products: the open ocean product described below (i.e., Landschützer et al., 2016) and the coastal product created by Laruelle et al. (2017). Consequently, it represents coastal zones better. Data collected between 1998 and 2015 from the SOCAT database (Version 5) were used to create this data product.
Reference:
Landschützer, P., Laruelle, G. G., Roobaert, A., and Regnier, P.: A uniform pCO2 climatology combining open and coastal oceans, Earth Syst. Sci. Data, 12, 2537–2553, https://doi.org/10.5194/essd-12-2537-2020, 2020.
VLIZ-SOM-FFN
Landschützer et al. (2016)
Monthly gridded pCO2 from 1982 through near present
Open ocean
monthly, climatology
1° x 1°
1982 - 2025
VLIZ-SOM-FFN
Landschützer et al. (2016)
Description:
Landschützer et al. (2016) employed the Self-Organizing-Map Feed-Forward Network (SOM-FFN) neural network method (Landschützer et al., 2013) to map sea surface pCO2 from SOCAT (see No. 1 above) (Bakker et al., 2014) to generate monthly pCO2 fields on a 1° × 1° global surface ocean grid, covering the period from 1982 to near present. It is based on the gridded pCO2 measurements from SOCAT and is updated regularly. The creation of the pCO2 fields involve a two-step neural network approach, which has been extensively detailed and validated in previous works by Landschützer et al. (2013, 2014, 2016). In the initial step, the global ocean is clustered into biogeochemical provinces, and subsequently, the non-linear relationship between CO2 driver variables and gridded data from SOCAT (Bakker et al., 2016) is reconstructed. Air-sea CO2 fluxes are also computed based on the air-sea pCO2 difference, utilizing a bulk gas transfer formulation as described by Landschützer et al. (2013, 2014, 2016).
Reference:
Landschützer, P., Gruber, N., and Bakker, D. C. E.: Decadal variations and trends of the Global Ocean Carbon Sink, Global Biogeochemical Cycles, 30(10), 1396–1417. https://doi.org/10.1002/2015gb005359, 2016.
JMA-MLR
Iida et al. (2021)
Temporal trends of DIC, TA, pCO2, air-sea CO2 flux, pH, and Ωarag.
Open ocean, Surface
monthly
1° x 1°
1990-2023
JMA-MLR
Iida et al. (2021)
Description:
Iida et al. (2021) developed a monthly data product for inorganic carbonate variables on a 1° × 1° global surface ocean grid for the period 1993?2018. Variables include DIC, TA, pCO2, sea-air CO2 flux, pH, and Ωarag. They leveraged data products such as SOCAT.v2019 (Bakker et al., 2016) and GLODAPv2.2019 (Olsen et al., 2019), as well as satellite-based variables, including sea-surface dynamic height (SSDH), mixed layer depth (MLD), and chlorophyll-a. The product is updated annually using the latest SOCAT and GLODAPv2 data.
Reference:
Iida, Y., Takatani, Y., Kojima, A., and Ishii, M. 2021. Global trends of ocean CO2 sink and ocean acidification: An obsevation-based reconstruction of surface ocean inorganic carbon variables, J. Oceanogr. 77, 323-358, https://doi.org/10.1007/s10872-020-00571-5
OceanSODA-ETHZv1
Gregor and Gruber (2021)
Temporal trends of DIC, TA, pCO2, pH, Ωarag, and Ωcalc
Open ocean, Surface
monthly
1° x 1°
1982 - 2022
OceanSODA-ETHZv1
Gregor and Gruber (2021)
Description:
A monthly gridded global surface ocean data product for multiple ocean carbonate system variables, including DIC, TA, pCO2, pH (total scale), Ωarag, and Ωcalc (Gregor and Gruber, 2020; Gregor and Gruber, 2021; Gregor and Gruber, 2023; Ma et al., 2023). This dataset is structured on a 1°×1° global surface ocean grid with monthly resolution from 1982–2022, facilitating research on OA over seasonal to decadal scales. The OceanSODA-ETHZ data product was created by extrapolating in time and space the surface ocean observations of fCO2 from SOCATv2022 (Bakker et al., 2016) and TA from GLODAPv2.2022 using the newly developed Geospatial Random Cluster Ensemble Regression (GRaCER) method (Gregor, 2021). TA and pCO2 were then used to calculate the remaining variables of the marine carbonate system with the PyCO2SYS software (Humphreys et al., 2022). Phosphate and silicate from WOA 2018 product were used (Boyer et al., 2018; Garcia et al., 2018a).
Reference:
Gregor, L.; Gruber, N. OceanSODA-ETHZ: A Global Gridded Data Set of the Surface Ocean Carbonate System for Seasonal to Decadal Studies of Ocean Acidification. Earth System Science Data 2021, 13 (2), 777–808. https://doi.org/10.5194/essd-13-777-2021.
OceanSODA-ETHZv2
Gregor et al. (2024)
Highlighting fine-scale and short-term variability of the ocean carbon sink
Open ocean, Surface
8-day
0.25° x 0.25°
1982 - 2023
OceanSODA-ETHZv2
Gregor et al. (2024)
Description:
A surface fCO₂ product with a 0.25° × 0.25° spatial resolution and an 8-day temporal resolution, providing estimates starting from 1982 (Gregor et al., 2024a; Gregor et al., 2024b). The high-resolution outputs are suitable for investigating the shorter- and finer-scale dynamics of surface fCO2. Despite sharing a name with its predecessor, OceanSODA-ETHZv2 does not provide TA estimates and employs a different methodology, as described in the following steps: 1) The atmospheric trend of CO2 is removed by subtracting marine boundary layer CO2 concentrations from SOCAT fCO2 producing a new target ∆*CO2 to reduce the biases at the start and end of the time-series. 2) An 8-day seasonal climatology of ∆*CO2 is estimated using Gradient Boosted Decision Trees (GBDT), which is later used as a predictor. 3) The non-seasonal thermal component is removed from ∆*CO2, resulting in a new target, ∆*CO2nonT. 4) The new target is estimated using a feed-forward neural network, with the GBDT as one of the forcing variables. 5) Steps 4 through to 1 are inverted to arrive at fCO2. 6) Air-sea CO2 fluxes are computed using ERA5 winds.
Reference:
Gregor, L.; Shutler, J.; Gruber, N. High-Resolution Variability of the Ocean Carbon Sink. Global Biogeochemical Cycles 2024, 38 (8), e2024GB008127. https://doi.org/10.1029/2024GB008127.
LDEO-HPD fCO2, with Extended Temporal Coverage
Gloege et al. (2022), Bennington et al. (2022)
LDEO-HPD uses model-data misfit climatology to extend estimate back in time to 1959
Open ocean
monthly
1° x 1°
1959-2023
LDEO-HPD fCO2, with Extended Temporal Coverage
Gloege et al. (2022), Bennington et al. (2022)
Description:
LDEO-HPD with Extended Temporal Coverage builds on the work of Gloege et al. (2022), the LDEO-HPD product, extending the timeseries back in time to predict fCO2 for all available model years. Bennington et al. (2022) find that the largest component of the GOBM corrections is climatological. The smaller corrections at other timescales suggest either that these are well captured by the GOBMs or the data are insufficient. The dominance of climatological corrections supports the extension of the LDEO-HPD fCO2 product backwards in time. A climatology of model-observation misfits for the best-observed period (2000–present) is applied to the GOBMs for 1959–1981, while an inter-annually varying correction is used for 1982 onward. (Bennington et al., 2022a). This results in reconstructed monthly surface ocean fCO2 and air–sea CO2 fluxes on a 1° × 1° grid covering the open ocean, beginning in 1959.
Reference:
Gloege, L., Yan, M., Zheng, T. and McKinley, G. A.: Improved quantification of ocean carbon uptake by using machine learning to merge global models and pCO2 data, Journal of Advances in Modeling Earth Systems, 14(2), https://doi.org/10.1029/2021MS002620, 2022; Bennington, V., Gloege, L., and McKinley, G. A.: Variability in the global ocean carbon sink from 1959 to 2020 by correcting models with observations, Geophysical Research Letters, 49(14), https://doi.org/10.1029/2022GL098632, 2022
LDEO fCO2 - Residual Method
Bennington et al. (2022)
Removes the temperature component before Machine Learning
Open ocean
monthly
1° x 1°
1982-2023
LDEO fCO2 - Residual Method
Bennington et al. (2022)
Description:
LDEO fCO2 - Residual Method estimates full-coverage fCO2 by training a machine learning algorithm on sparse in situ fCO2 data and associated physical and biogeochemical observations. While these associated variables have well-known relationships to fCO2 , it is often unclear how they mechanistically drive fCO2 around the world. The LDEO fCO2 -Residual method takes the basic approach and enhances connections between physical understanding and reconstructed fCO2 . The novel approach used here includes applying pre-processing to the fCO2 data to remove the direct effect of temperature – a relationship well-documented in literature and lab experiments. This enhances the biogeochemical/physical component of fCO2 in the target variable (now fCO2 -Residual) and reduces the complexity that the machine learning must disentangle. The resulting algorithm has physically understandable connections between input data and the output biogeochemical/physical component of fCO2. This results in reconstructed monthly surface ocean fCO2 and air–sea CO2 fluxes on a 1° × 1° grid covering the open ocean, beginning in 1982 and extended to the most recent year of available data.
Reference:
Bennington, V., Galjanic, T., & McKinley, G. A. (2022). Explicit physical knowledge in machine learning for ocean carbon flux reconstruction: The pCO2-Residual method. Journal of Advances in Modeling Earth Systems, 14, e2021MS002960. https://doi. org/10.1029/2021MS002960
CMEMS-LSCEv1
Chau et al. (2022)
seamless reconstruction from coastal to open ocean
Open ocean, Coastal ocean, Surface
monthly, climatology
1° x 1°
1985 - 2019
CMEMS-LSCEv1
Chau et al. (2022)
Description:
Monthly surface ocean pCO2 and air–sea CO2 fluxes on a 1° × 1° grid in both the open ocean and coastal seas from 1985–2019 were reconstructed by Chau et al., (2022). CMEMS-LSCE is short for Copernicus Marine Environment Monitoring Service - Laboratoire des Sciences du Climat et de l’Environnement. This product is generated from an ensemble-based reconstruction of pCO2 maps trained with gridded data from SOCATv2020 (Bakker et al., 2016). Sea-surface pCO2 values (converted from the original fCO2 values in SOCATv2020) were regressed against a set of predictors with non-linear functions, i.e., feed-forward neural network (FFNN) models. The predictors include: sea-surface height (SSH), SST, SSS, MLD, chlorophyll a (Chl-a), atmospheric CO2 mole fraction (xCO2), and geographical coordinates (longitudes and latitudes).
Reference:
Chau, T.-T.-T., Gehlen, M., and Chevallier, F.: A seamless ensemble-based reconstruction of surface ocean pCO2 and air–sea CO2 fluxes over the global coastal and open oceans. Biogeosciences, 19(4), 1087–1109, https://doi.org/10.5194/bg-19-1087-2022, 2022.
CMEMS-LSCEv2
Chau et al. (2024)
Yearly extension of time-series & monthly reconstruction at low latency
Open ocean, Coastal ocean, Surface
monthly, climatology
0.25° x 0.25°
1985 - 2025
CMEMS-LSCEv2
Chau et al. (2024)
Description:
CMEMS-LSCEv2 corresponds to the latest version of the CMEMS-LSCE FFNN. It uses the same ensemble-based reconstruction method for pCO2 maps as CMEMS-LSCEv1. Improvements include downscaling the spatial resolution to 0.25° × 0.25° and reproducing additional surface ocean carbonate system variables on a global grid from 1985 onwards (Chau et al., 2024a). The additional surface ocean carbonate system variables are: pCO2, DIC, TA, pH, Ωarag, and Ωcalc. Surface ocean pCO2 is reconstructed based on an ensemble of neural network models mapping gridded observation-based data provided by SOCATv2022 (Bakker et al., 2016). Surface ocean TA is estimated with a multiple linear regression approach (Carter et al., 2016, 2017). The remaining carbonate variables are calculated from pCO2 and TA using a MATLAB version of CO2SYS (Lewis and Wallace, 1998; Van Heuven et al., 2011). The CMEMS-LSCE product is updated yearly for surface ocean pCO2, air-sea fluxes, and the carbonate system variables. Updates are phased with release of the SOCAT database. For surface ocean pCO2 and air-sea fluxes the temporal coverage is extended to the present date with a latency of 1 month (Chau et al., 2024b).
Reference:
Chau, T.-T.-T., Gehlen, M., Metzl, N., and Chevallier, F.: CMEMS-LSCE: a global, 0.25°, monthly reconstruction of the surface ocean carbonate system, Earth Syst. Sci. Data, 16, 121–160, https://doi.org/10.5194/essd-16-121-2024, 2024.
UoEX-Watson
Watson et al (2020)
Air-sea fluxes of CO2 with adjusted skin temperature effect
Open ocean
monthly
1° x 1°
1985-2019
UoEX-Watson
Watson et al (2020)
Description:
This is a gridded estimate of the atmosphere-ocean flux of CO2. It takes into account near-surface temperature and salinity deviations due to the surface "skin effect", and small deviations in temperature caused by sampling below the actual surface of the ocean, which most estimates ignore. The result is a substantially larger estimate of net flux into the ocean than most other products. Apart from the surface correction, the SOM-FFN neural net interpolation method developed by Landschutzer et al is followed.
Reference:
Watson, A. J., Schuster, U., Shutler, J. D., Holding, T., Ashton, I. G., Landschützer, P., Woolf, D., and Goddijn-Murphy, L.: Revised estimates of ocean-atmosphere CO2 flux are consistent with ocean carbon inventory, Nature Communications, 11(1), 1–6, https://doi.org/10.1038/s41467-020-18203-3, 2020.
NIES-ML3
Zeng et al. (2022)
Three machine learning methods were used to simulate surface ocean CO2 trends and construct monthly CO2 maps.
Open ocean, Surface
monthly
1° x 1°
1982 - 2024
NIES-ML3
Zeng et al. (2022)
Description:
A feed forward neural network (NN), a random forest (RF) and a gradient boost machine (GB) were used to simulate the time variant trends of the surface ocean CO2 and construct monthly global surface ocean CO2 maps. The trends of ocean CO2 were estimated from the annual increase rate of atmospheric CO2 first and then corrected by using the NN, RF and GB with a leave-one-year-out validation method. Including different machine learning methods has the advantages of preventing over fitting by inter-comparison and reducing hot-spot by using the ensemble of their results.
Reference:
Zeng, J., Iida, Y., Matsunaga, T., and Shirai, T. 2022. Surface ocean CO2 concentration and air-sea flux estimate by machine learning with modelled variable trends. Front. Mar. Sci. 9, 989233. https://doi.org/10.3389/fmars.2022.989233
CSIR-ML6
Gregor et al. (2019)
Various ML methods produce different results when data is sparse, but all still achieving roughly the same uncertainty
Open ocean, Surface
monthly
1° x 1°
1982 - 2020
CSIR-ML6
Gregor et al. (2019)
Description:
Provides monthly 1° × 1° estimates of surface pCO2 (Gregor et al., 2019a). The approach uses the conceptual two-step approach of clustering and performing regressions for each cluster as Landschützer et al. (2016). CSIR-ML6 investigates the efficacy of various machine learning (ML) methods in estimating surface pCO2, namely, feed-forward neural networks (FFNN), extremely randomized trees (ERT), gradient boosting machines (GBM), and support vector regression (SVR). It is found that the ensemble of all but the ERT method resulted in the best estimate, highlighting the fact that various ML methods do not produce the same outcome, particularly when data is sparse. Further, the variance between ensemble members can inform us about regions where uncertainty may be large due to methodological differences. Despite this, all methods achieve roughly the same uncertainty – a barrier, or wall beyond which the community has yet to overcome.
Reference:
Gregor, L., Lebehot, A. D., Kok, S., and Scheel Monteiro, P. M.: A comparative assessment of the uncertainties of global surface ocean CO2 estimates using a machine-learning ensemble (CSIR-ML6 version 2019a) – have we hit the wall?, Geosci. Model Dev., 12, 5113–5136, https://doi.org/10.5194/gmd-12-5113-2019, 2019.
Stepwise-FFNN
Zhong et al. (2022)
ML-based selection of predictors considering regional differences of pCO2 drivers
Open ocean, Surface
monthly
1° x 1°
1992 - 2024
Stepwise-FFNN
Zhong et al. (2022)
Description:
A monthly global 1° × 1° surface ocean pCO2 product from January 1992 to December 2024, constructed by combining the stepwise regression algorithm and a feed-forward neural network (FFNN) to select predictors of pCO2 based on the mean absolute error in each of the 11 biogeochemical provinces defined by the self-organizing map (SOM) method.
Reference:
Zhong, G., Li, X., Song, J., Qu, B., Wang, F., Wang, Y., ... & Duan, L. 2022. Reconstruction of global surface ocean pCO2 using region-specific predictors based on a stepwise FFNN regression algorithm, Biogeosciences, 19, 845-859, https://doi.org/10.5194/bg-19-845-2022.
AOML-ET
Wanninkhof et al. (2015)
Monthly global sea-air CO2 flux maps in modern era
Open ocean, Surface
monthly, climatology
1° x 1°
1998 - 2023
AOML-ET
Wanninkhof et al. (2015)
Description:
Wanninkhof et al. (2025) developed a monthly global ocean data product of seawater pCO2 and sea-air CO2 fluxes, referred to as AOML-ET, using an extremely randomized trees (ET) machine learning technique. These maps are created on 1° × 1° spatial grids, providing global surface ocean coverages from 1998 to 2023. AOML-ET incorporates several predictor variables, including time, location, SST, SSS, MLD, and chlorophyll-a. The model was trained using the v2020 and v2023 releases of the SOCAT data product (No. 1). Sea-air CO2 fluxes were calculated using the air-sea CO2 partial pressure difference (∆pCO2) and a bulk gas transfer formulation incorporating windspeed.
Reference:
Wanninkhof, R., Triñanes, J., Pierrot, D., Munro, D. R., Sweeney, C., and Fay, A. R.: Trends in sea-air CO2 fluxes and sensitivities to atmospheric forcing using an extremely randomized trees machine learning approach, Global Biogeochemical Cycles, 39, https://doi.org/10.1029/2024GB008315, 2025.
ULB-SOM-FFN-Coastalv2.1
Roobaert et al. (2024)
Global temporal trends of coastal pCO2 and air-sea CO2 fluxes based on SOCATv2022 with data collected from 1982–2020
Coastal ocean
continuous, monthly
0.25° x 0.25°
1982-2020
ULB-SOM-FFN-Coastalv2.1
Roobaert et al. (2024)
Description:
Roobaert et al. (2024) present high-resolution (0.25° × 0.25° grid) monthly maps showing the distribution of sea surface pCO2 across the global coastal oceans, spanning from 1982 to 2020. This product (ULB-SOM-FFN-coastalv2.1) builds upon the work by Laruelle et al. (2017), incorporating a two-step methodology that utilizes Self Organizing Maps (SOM) and Feed Forward Networks (FFN). This updated product now captures temporal variability, enabling the assessment of interannual variability and long-term trends in coastal air-sea CO2 exchange, unlike the product by Laruelle et al. (2017), which only offers a climatology for a short period (1998-2015). The enhancements include additional environmental predictors and an expanded dataset for training and validation, featuring approximately 18 million direct coastal observations from the SOCAT database, specifically the SOCATv2022 release (Bakker et al., 2016). The product is available at OCADS: https://www.ncei.noaa.gov/data/oceans/ncei/ocads/metadata/0279118.html (Roobaert et al., 2023).
Reference:
Roobaert, A., Regnier, P., Landschützer, P., and Laruelle, G. G.: A novel sea surface pCO2-product for the global coastal ocean resolving trends over 1982–2020, Earth System Science Data, 16(1), 421–441, https://doi.org/10.5194/essd-16-421-2024, 2024.
RFR-LME
Sharp et al. (2024)
Temporal trends of OA indicators and estimated uncertainties across 11 U.S. Large Marine Ecosystems (LMEs), with monthly coverage from 1998–2024.
Coastal ocean, Surface, Regional
monthly
0.25° x 0.25°
1998 - 2024
RFR-LME
Sharp et al. (2024)
Description:
A data product delineating the temporal trends of ocean acidification indicators mapped on a 0.25°× 0.25° spatial grid, across eleven U.S. Large Marine Ecosystems (LMEs), with monthly coverage from 1998–2024. The method combines Gaussian Mixture Models to categorize the data into environmentally similar subregions, Random Forest Regressions (RFRs) for the spatial and temporal extrapolation of observational fCO2 data, and regressions to estimate total alkalinity (Carter et al. 2021) to provide a second carbonate system constraint for the computation of multiple ocean acidification indicators.
Reference:
Sharp, J. D., Jiang, L.-Q., Carter, B. R., Lavin, P. D., Yoo, H., and Cross, S. L., 2024. A mapped dataset of surface ocean acidification indicators in large marine ecosystems of the United States, Sci. Data, 11, 715, https://doi.org/10.1038/s41597-024-03530-7.
Gridded Surface OA Indicators in the Northern Caribbean Sea
Wanninkhof et al. (2020)
A 17-year record of fCO2, TA, pH, Ωarag, and air-sea CO2 flux in the Caribbean Sea
Coastal ocean
monthly, climatology
1° x 1°
2002 - 2019
Gridded Surface OA Indicators in the Northern Caribbean Sea
Wanninkhof et al. (2020)
Description:
This dataset contains a high-quality dataset of derived products from over a million observations of surface water partial pressure/fugacity of carbon dioxide (pCO2w/fCO2w), for the Caribbean Sea, Gulf of Mexico/Gulf of America and North-West Atlantic Ocean covering the timespan from 2002-01-01 to 2019-12-30. The derived quantities include TA, acidity (pH), Ωarag and air-sea CO2 flux (Wanninkhof et al., 2020).
Reference:
Wanninkhof, R., Pierrot, D., Sullivan, K., Barbero, L., and Triñanes, J.: A 17-year dataset of surface water fugacity of CO2 along with calculated pH, aragonite saturation state and air–sea CO2 fluxes in the northern Caribbean Sea, Earth Syst. Sci. Data, 12, 1489–1509, https://doi.org/10.5194/essd-12-1489-2020, 2020.
INCOIS-ReML
Joshi et al. (2024)
This data product integrates publicly available open-ocean observations with data from the Indian EEZ region in the Bay of Bengal to provide surface pCO2 and air-sea CO2 flux estimates.
Regional
monthly
0.083° x 0.083°
2015
INCOIS-ReML
Joshi et al. (2024)
Description:
The Indian National Centre for Ocean Information Services-Regional Machine Learning model (INCOIS-ReML) pCO2 data product offers machine learning based monthly climatological sea surface pCO2 and the corresponding air-sea CO2 flux for the Bay of Bengal (Joshi et al., 2024). This high-resolution (0.083° × 0.083°) monthly climatological pCO2 data product is available from the INCOIS Portal: https://las.incois.gov.in, and from OCADS: https://www.ncei.noaa.gov/data/oceans/ncei/ocads/metadata/0307627.html (Joshi et al., 2025a).
Reference:
Joshi, A., Ghoshal, P.K., Chakraborty, K. et al. Sea-surface pCO2 maps for the Bay of Bengal based on advanced machine learning algorithms. Sci Data 11, 384 (2024). https://doi.org/10.1038/s41597-024-03236-w
INCOIS_TA
Joshi et al. (2025)
This data product integrates publicly available open-ocean observations with data collected during Indian scientific expeditions and from the Indian Exclusive Economic Zone to provide surface TA estimates for the North Indian Ocean.
Regional
monthly
0.083° x 0.083°
1993-2020
INCOIS_TA
Joshi et al. (2025)
Description:
The Indian National Centre for Ocean Information Services-Total Alkalinity (INCOIS_TA) data product offers a machine learning based monthly interannual surface TA from 1993-2020 for the North Indian Ocean (Joshi et al., 2025). This data product integrates publicly available open-ocean observations with data collected during Indian scientific expeditions and from the Indian Exclusive Economic Zone. This high-resolution (0.083° × 0.083°) long-term monthly TA data product is available from the INCOIS Portal: https://las.incois.gov.in, and from OCADS: https://www.ncei.noaa.gov/data/oceans/ncei/ocads/metadata/0307789.html (Joshi et al., 2025).
Reference:
Joshi, A. P., Ghoshal, P. K., Chakraborty, K., Roy, R., Jayaram, C., Sridevi, B., & Sarma, V. V. S. S. (2025). Long-term changes of surface total alkalinity and its driving mechanisms in the north Indian Ocean. Global Biogeochemical Cycles, 39, e2024GB008344. https://doi.org/10.1029/2024GB008344
GLODAPv2 Climatology
Lauvset et al. (2016)
Ocean interior climatology for multiple variables from surface to the bottom of the ocean (referenced to year 2002)
Open ocean, Water column
climatology
1° x 1°
2002
GLODAPv2 Climatology
Lauvset et al. (2016)
Description:
Lauvset et al. (2016) generated a comprehensive set of global interior ocean climatologies, mapping key biogeochemical variables on a 1° × 1° grid for 33 depth levels from surface to 5500 m. These climatologies cover temperature, salinity, DO, nitrate, phosphate, silicate, DIC, TA, pH, Ωarag, and Ωcalc). This data product was created based on the quality-controlled and internally consistent GLODAPv2.2016 (Olsen et al., 2016) using the data-interpolating variational analysis (DIVA) method (Barth et al., 2014). The conceivably confounding temporal trends in DIC, pH, Ωarag and Ωcalc due to anthropogenic influence were removed prior to mapping by normalizing their values to a reference year of 2002 using first-order calculations of anthropogenic carbon accumulation rates. For all variables, all data from the full 1972–2013 period were used, including data that did not receive full secondary quality control. This data product is not updated each year along with the main GLODAPv2 data product.
Reference:
Lauvset, S. K., Key, R. M., Olsen, A., van Heuven, S., Velo, A., Lin, X., Schirnick, C., Kozyr, A., Tanhua, T., Hoppema, M., Jutterström, S., Steinfeldt, R., Jeansson, E., Ishii, M., Perez, F. F., Suzuki, T., and Watelet, S.: A new global interior ocean mapped climatology: The 1° × 1° GLODAP version 2, Earth System Science Data, 8(2), 325–340, https://doi.org/10.5194/essd-8-325-2016, 2016.
Aragonite Saturation Climatology
Jiang et al. (2015)
Ocean interior climatology for (Ωarag) from surface to 4000 m (referenced to year 2000)
Open ocean, Water column
climatology
1° x 1°
2000
Aragonite Saturation Climatology
Jiang et al. (2015)
Description:
Jiang et al. (2015a) developed an interior ocean Ωarag climatology (referenced to 2000), on a 1° × 1° grid at 9 standardized depth levels from the surface down to 4000m. This was accomplished by integrating data from the first version of GLODAP (Key et al., 2004), CARINA (Key et al., 2010), and PACIFICA (Suzuki et al., 2013), along with additional recent cruise datasets up to 2012.
Reference:
Jiang, L.-Q., Feely, R. A., Carter, B. R., Greeley, D. J., Gledhill, D. K., and Arzayus, K. M.: Climatological distribution of aragonite saturation state in the global oceans, Global Biogeochemical Cycles, 29(10), 1656–1673, https://doi.org/10.1002/2015GB005198, 2015.
MOBO-DIC (Version 2020)
Keppler et al., 2020
Seasonal variability of DIC in the interior ocean from surface to 2000 m
Open ocean
monthly, climatology
1° x 1°
2010
MOBO-DIC (Version 2020)
Keppler et al., 2020
Description:
Keppler et al. (2020) produced a global interior ocean DIC monthly climatology (average climatological values for January through December) on a 1° × 1° grid at 33 standardized depth levels from the surface to 2000 m.
Reference:
Keppler, L., Landschützer, P., Gruber, N., Lauvset, S. K., and Stemmler, I.: Seasonal carbon dynamics in the near-global ocean, Global Biogeochemical Cycles, 34(12), e2020GB006571, https://doi.org/10.1029/2020GB006571, 2020.
MOBO-DIC (Version 2023)
Keppler et al. (2023)
Temporal trends and interannual variability of DIC in the interior ocean from surface to 1500 m
Open ocean
monthly
1° x 1°
2004-2020
MOBO-DIC (Version 2023)
Keppler et al. (2023)
Description:
Keppler et al. (2023) extended the temporal resolution of MOBO-DIC to resolve monthly fields from January 2004 to December 2019, as opposed to the average climatological values in Keppler et al. (2020).
Reference:
Keppler, L., Landschützer, P., Lauvset, S. K., and Gruber, N.: Recent trends and variability in the oceanic storage of dissolved inorganic carbon, Global Biogeochemical Cycles, 37(5), https://doi.org/10.1029/2022gb007677, 2023.
Monthly Interior Ocean TA Climatology
Broullon et al., (2019)
Ocean interior climatology for TA from surface to bottom
Open ocean, Water column
monthly, climatology
1° x 1°
Monthly Interior Ocean TA Climatology
Broullon et al., (2019)
Description:
Global climatologies of the seawater CO2 chemistry variables are necessary to assess the marine carbon cycle in depth. The climatologies should adequately capture seasonal variability to properly address ocean acidification and similar issues related to the carbon cycle. Total alkalinity (AT) is one variable of the seawater CO2 chemistry system involved in ocean acidification and frequently measured. We used the Global Ocean Data Analysis Project version 2.2019 (GLODAPv2) to extract relationships among the drivers of the AT variability and AT concentration using a neural network (NNGv2) to generate a monthly climatology. The GLODAPv2 quality-controlled dataset used was modeled by the NNGv2 with a root-mean-squared error (RMSE) of 5.3 µmol kg−1. Validation tests with independent datasets revealed the good generalization of the network. Data from five ocean time-series stations showed an acceptable RMSE range of 3–6.2 µmol kg−1. Successful modeling of the monthly AT variability in the time series suggests that the NNGv2 is a good candidate to generate a monthly climatology. The climatological fields of AT were obtained passing through the NNGv2 the World Ocean Atlas 2013 (WOA13) monthly climatologies of temperature, salinity, and oxygen and the computed climatologies of nutrients from the previous ones with a neural network. The spatiotemporal resolution is set by WOA13: 1∘ × 1∘ in the horizontal, 102 depth levels (0–5500 m) in the vertical and monthly (0–1500 m) to annual (1550–5500 m) temporal resolution. The product is distributed through the data repository of the Spanish National Research Council (CSIC; https://doi.org/10.20350/digitalCSIC/8644, Broullón et al., 2019).
Reference:
Broullón, D., Pérez, F. F., Velo, A., Hoppema, M., Olsen, A., Takahashi, T., Key, R. M., Tanhua, T., González-Dávila, M., Jeansson, E., Kozyr, A., and van Heuven, S. M. A. C.: A global monthly climatology of total alkalinity: a neural network approach, Earth Syst. Sci. Data, 11, 1109–1127, https://doi.org/10.5194/essd-11-1109-2019, 2019.
Monthly Interior Ocean DIC Climatology
Broullón et al., (2020)
Ocean interior climatology for DIC from surface to bottom (referenced to year 1995)
Open ocean, Water column
monthly
1° x 1°
1995
Monthly Interior Ocean DIC Climatology
Broullón et al., (2020)
Description:
Anthropogenic emissions of CO2 to the atmosphere have modified the carbon cycle for more than 2 centuries. As the ocean stores most of the carbon on our planet, there is an important task in unraveling the natural and anthropogenic processes that drive the carbon cycle at different spatial and temporal scales. We contribute to this by designing a global monthly climatology of total dissolved inorganic carbon (TCO2), which offers a robust basis in carbon cycle modeling but also for other studies related to this cycle. A feedforward neural network (dubbed NNGv2LDEO) was configured to extract from the Global Ocean Data Analysis Project version 2.2019 (GLODAPv2.2019) and the Lamont–Doherty Earth Observatory (LDEO) datasets the relations between TCO2 and a set of variables related to the former's variability. The global root mean square error (RMSE) of mapping TCO2 is relatively low for the two datasets (GLODAPv2.2019: 7.2 µmol kg−1; LDEO: 11.4 µmol kg−1) and also for independent data, suggesting that the network does not overfit possible errors in data. The ability of NNGv2LDEO to capture the monthly variability of TCO2 was testified through the good reproduction of the seasonal cycle in 10 time series stations spread over different regions of the ocean (RMSE: 3.6 to 13.2 µmol kg−1). The climatology was obtained by passing through NNGv2LDEO the monthly climatological fields of temperature, salinity, and oxygen from the World Ocean Atlas 2013 and phosphate, nitrate, and silicate computed from a neural network fed with the previous fields. The resolution is in the horizontal, 102 depth levels (0–5500 m), and monthly (0–1500 m) to annual (1550–5500 m) temporal resolution, and it is centered around the year 1995. The uncertainty of the climatology is low when compared with climatological values derived from measured TCO2 in the largest time series stations. Furthermore, a computed climatology of partial pressure of CO2 (pCO2) from a previous climatology of total alkalinity and the present one of TCO2 supports the robustness of this product through the good correlation with a widely used pCO2 climatology (Landschützer et al., 2017). Our TCO2 climatology is distributed through the data repository of the Spanish National Research Council (CSIC; https://doi.org/10.20350/digitalCSIC/10551, Broullón et al., 2020)
Reference:
Broullón, D., Pérez, F. F., Velo, A., Hoppema, M., Olsen, A., Takahashi, T., Key, R. M., Tanhua, T., Santana-Casiano, J. M., and Kozyr, A.: A global monthly climatology of oceanic total dissolved inorganic carbon: A Neural Network approach, Earth Syst. Sci. Data, 12(3), 1725–1743, https://doi.org/10.5194/essd-12-1725-2020, 2020
Acidification Metrics in the Ocean Interior
Fassbender et al. (2023)
Metrics of acidification in the ocean interior (to 2000 m) and the component of those changes caused by carbonate system nonlinearities
Open ocean, Water column
climatology
1° x 1°
1850 - 2002
Acidification Metrics in the Ocean Interior
Fassbender et al. (2023)
Description:
Fassbender et al. (2023) generated estimates of global interior ocean changes to pH, [H+], Ωarag, pCO2, and the Revelle sensitivity factor driven by the accumulation of anthropogenic carbon (Cant) from the preindustrial period to the year 2002, and quantified the component of those changes caused by nonlinearities in the carbonate system. For each OA metric, the dataset includes year 2002 values and quasi-preindustrial values, which were estimated by subtracting Cant from the year 2002 carbonate chemistry information and recomputing each OA metric without considering any warming, circulation, or biological changes that may have occurred since the preindustrial era.
Reference:
Fassbender, A. J., Carter, B. R., Sharp, J. D., Huang, Y., Arroyo, M. C., and Frenzel, H.: Amplified subsurface signals of ocean acidification, Global Biogeochemical Cycles, 37, e2023GB007843, https://doi.org/10.1029/2023GB007843, 2023.
Ocean Interior Acidification over the Industrial Era
Müller et al. (2024)
Temporal trends in the progression of acidification in the interior ocean are resolved
Open ocean, Water column
decadal
1° x 1°
1800 - 2014
Ocean Interior Acidification over the Industrial Era
Müller et al. (2024)
Description:
Building on the total anthropogenic carbon estimates for 1994 from Sabine et al. (2004) and the decadal changes between 1994 and 2014 reconstructed by Müller et al. (2023a), Müller and Gruber (2024a) quantified ocean interior acidification over the industrial era. To convert the increasing anthropogenic carbon concentrations into acidification estimates, their approach relied on time-invariant climatologies of ocean interior DIC, TA, temperature, salinity, and other relevant variables to determine the background state of the marine carbonate system. Hence, their estimates resolve exclusively the acidification driven by the anthropogenic carbon accumulation. In contrast to direct observations of acidification variables, such as those collected at time-series stations, this approach does not account for changes in the natural carbon cycle or the displacement of water masses.
Reference:
Jens D. Müller, Nicolas Gruber, Progression of ocean interior acidification over the industrial era. Sci. Adv.10, eado3103(2024). DOI:10.1126/sciadv.ado3103
Anthropogenic CO2 from 1994 to 2007
Gruber et al. (2019)
The oceanic sink for anthropogenic CO2 over the period 1994 to 2007
Open ocean, Water column
climatology
1° x 1°
1994 - 2007
Anthropogenic CO2 from 1994 to 2007
Gruber et al. (2019)
Description:
Gruber et al. (2019a) estimated the decadal time-scale changes in the oceanic content of anthropogenic CO2 (∆Cant) between 1994 to 2007. The results were derived from the GLODAPv2.2016 product (Olsen et al., 2016), utilizing the eMLR(C*) methodology pioneered by Clement and Gruber (2018). The product is combined with the estimated amount of Cant for 1994 derived by Sabine et al. (2004) from GLODAPv1 to infer Cant for 2007.
Reference:
Gruber, N., Clement, D., Carter, B. R., Feely, R. A., van Heuven, S., Hoppema, M., Ishii, M., Key, R. M., Kozyr, A., Lauvset, S. K., Lo Monaco, C., Mathis, J. T., Murata, A., Olsen, A., Perez, F. F., Sabine, C. L., Tanhua, T., and Wanninkhof, R.: The oceanic sink for anthropogenic CO2 from 1994 to 2007, Science, 363(6432), 1193–1199, https://doi.org/10.1126/science.aau5153, 2019.
Decadal Trends in Anthropogenic CO2 from 1994 to 2014
Müller et al. (2023)
Temporal trends in the accumulation of anthropogenic CO2 in the interior ocean are resolved
Open ocean, Water column
decadal
1° x 1°
1994 - 2014
Decadal Trends in Anthropogenic CO2 from 1994 to 2014
Müller et al. (2023)
Description:
Müller et al. (2023a) extended the analysis by Gruber et al. (2019a) to reconstruct decadal trends in the oceanic storage of ∆Cant in the global ocean interior from mid-year 1994 to mid-year 2004, and further to mid-year 2014. They applied the extended multiple linear regression (eMLR) method (Clement and Gruber, 2018) to ship-borne observations of DIC and other biogeochemical variables from GLODAPv2.2021 (Lauvset et al., 2021).
Reference:
Müller, J. D., Gruber, N., Carter, B., Feely, R., Ishii, M., Lange, N., et al. (2023). Decadal trends in the oceanic storage of anthropogenic carbon from 1994 to 2014. AGU Advances, 4, e2023AV000875. https://doi.org/10.1029/2023AV000875
Tracer-based Rapid Anthropogenic Carbon Estimation (TRACE) gridded Canth
Carter et al. (2025)
New methods for rapidly estimating anthropogenic carbon from temperature, salinity, and date are presented. The methods are then used with SSPs to estimate anthropogenic carbon in the ocean globally from 1750 to 2500.
Open ocean, Surface, Water column
irregular
1° x 1°
1750-2500
Tracer-based Rapid Anthropogenic Carbon Estimation (TRACE) gridded Canth
Carter et al. (2025)
Description:
Here, we make modifications to the transit time distribution approach for Canth estimation that render the method more accessible. We also release software (BRCScienceProducts, 2025) called “Tracer-based Rapid Anthropogenic Carbon Estimation version 1” (TRACEv1) that allows users – with one line of code – to obtain Canth and water mass age estimates throughout the global open ocean from user-supplied values of geographic location, pressure, salinity, temperature, and the estimate year. We use this code to generate a data product of global gridded open-ocean Canth distributions (TRACEv1_GGCanth; Carter, 2025) that ranges from the preindustrial era through 2500 under a range of Shared Socioeconomic Pathways (SSPs, or atmospheric CO2 concentration pathways). We estimated the skill of these estimates by reconstructing Canth in models with known distributions of Canth and transient tracers and by conducting perturbation tests. In the model-based reconstruction test, TRACEv1 reproduces the global ocean Canth inventory to within ±10 % in 1980 and 2014. We discuss implications and limitations of the projected Canth distributions and highlight ways that the estimation strategy might be improved. One finding is that the ocean will continue to increase its net Canth inventory at least through 2500 due to deep-ocean ventilation, even with the SSP in which intense mitigation successfully decreases atmospheric Canth by ∼60 % in 2500 relative to the 2024 concentration. A notable limitation of this and similar projections made with TRACEv1 is that ongoing and potential future warming and changing oceanic circulation patterns with climate change are not captured by the method. The data products generated by this research are available as MATLAB code (https://doi.org/10.5281/zenodo.15692788, BRCScienceProducts, 2025) and a spatially and temporally gridded data product (https://doi.org/10.5281/zenodo.15692788, BRCScienceProducts, 2025).
Reference:
Carter, B. R., Schwinger, J., Sonnerup, R., Fassbender, A. J., Sharp, J. D., Dias, L. M., & Sandborn, D. E. (2025). Tracer-based Rapid Anthropogenic Carbon Estimation (TRACE). Earth System Science Data, 17(6), 3073-3088.
Preformed properties for marine organic matter and carbonate mineral cycling quantification
Carter et al. (2021)
We have updated global preformed biogeochemical property estimates and provided the estimates for analysis and model validation
Open ocean, Water column
climatology
1° x 1°
Preformed properties for marine organic matter and carbonate mineral cycling quantification
Carter et al. (2021)
Description:
We estimate preformed ocean phosphate, nitrate, oxygen, silicate, and alkalinity by combining a reconstruction of ventilation pathways in the ocean interior with estimates of submixed layer properties. These new preformed property estimates are intended to aid biogeochemical cycling studies and validation of modeled preformed property distributions and are available online. Analyses of net property accumulations (observed minus preformed properties) indicate net remineralization ratios in the ocean interior of [1 P]: [14.1 ± 0.6 N]: [−141 ± 12 O2]: [95 ± 25 Si]: [89 ± 9 TA]. These ratios imply that the interior ocean stores 1,300 (±230) PgC through organic matter remineralization and 540 (±60) PgC through carbonate mineral dissolution and that apparent oxygen utilization can overestimate the interior ocean oxygen consumption by ~25%. Further, only 4 (±1%) and 46 (±5%) of the total alkalinity accumulated from carbonate mineral dissolution are found in seawater that is supersaturated with respect to the aragonite and calcite mineral forms of calcium carbonate, respectively. These small excess alkalinity inventories are due to smaller volumes of the supersaturated water masses and shorter ventilation timescales, as carbonate mineral dissolution rates appear nearly independent of depth and saturation state.
Reference:
Carter, B. R., Feely, R. A., Lauvset, S. K., Olsen, A., DeVries, T., & Sonnerup, R. (2021). Preformed properties for marine organic matter and carbonate mineral cycling quantification. Global Biogeochemical Cycles, 35(1), e2020GB006623.
Monthly Interior Ocean pH Climatology
Zhong et al. (2025)
Temporal variability of pH in the interior ocean from surface to 2000 m over 3 decades
Open ocean, Water column
monthly
1° x 1°
1992 - 2020
Monthly Interior Ocean pH Climatology
Zhong et al. (2025)
Description:
A monthly 1° × 1° gridded global seawater pH (total scale) climatology from 1992 to 2020 at in situ temperature, derived using a machine learning algorithm trained on pH observations from GLODAPv2. The product spans from 1992 to 2020 and covers depths from the surface to 2000 m across 41 vertical levels.
Reference:
Zhong, G., Li, X., Song, J., Qu, B., Wang, F., Wang, Y., Zhang, B., Cheng, L., Ma, J., Yuan, H., Duan, L., Li, N., Wang, Q., Xing, J., and Dai, J. 2025. A global monthly 3D field of seawater pH over 3 decades: a machine learning approach, Earth Syst. Sci. Data, 17, 719–740, https://doi.org/10.5194/essd-17-719-2025.
CODAP-NA Climatology
Jiang et al. (2024)
The first discrete bottle based climatology in the North American ocean margins
Coastal ocean, Water column
climatology
1° x 1°
2010
CODAP-NA Climatology
Jiang et al. (2024)
Description:
Jiang et al. (2024) developed a coastal OA indicators climatology on a 1°×1° grid, covering North American ocean margins from surface to 500 m at 14 standardized depth levels. This product includes 10 key oceanographic variables: fCO2, pH, [H+]total, free hydrogen ion content ([H+]free), carbonate ion content ([CO32-]), Ωarag, Ωcalc, DIC, TA, and Revelle Factor (RF), as well as temperature and salinity. The climatology was produced with the WOA gridding technologies of the NOAA National Centers for Environmental Information (NCEI), based on the recently released Coastal Ocean Data Analysis Product in North America (CODAP-NA) (Jiang et al., 2021), along with GLODAPv2.2022 (Lauvset et al., 2022). The relevant variables were adjusted to the year of 2010 before the gridding.
Reference:
Jiang, L.-Q., Boyer, T. P., Paver, C. R., Yoo, H., Reagan, J. R., Alin, S. R., Barbero, L., Carter, B. R., Feely, R. A., and Wanninkhof, R.: Climatological distribution of ocean acidification variables along the North American ocean margins, Earth System Science Data, 16(7), 3383–3390, https://doi.org/10.5194/essd-16-3383-2024, 2024.
SeaFlux
Fay and Gregor et al. (2021)
Careful consideration of flux calculation provides a resource and code to the community for independent flux calculations
Open ocean, Surface
monthly
1° x 1°
1990 - 2022
SeaFlux
Fay and Gregor et al. (2021)
Description:
Harmonization of air–sea CO2 fluxes from surface pCO2 data products using a standardized approach (Gregor and Fay, 2021). This resource provides an ensemble of six pCO2 products with air-sea CO2 fluxes computed consistently. The six included products are: CMEMS-LSCEv1, CSIR-ML6, JENA-MLS, JMA-MLR, MPI-SOMFFN, and NIES-FNN. First, missing areas of pCO2 estimates (mostly high-latitude and marginal seas) are filled using a linear-regression approach, thus addressing differences in spatial coverage between the mapping products. Further, also accounts for methodological inconsistencies in flux calculations. Fluxes are calculated using three wind products (CCMPv2, ERA5, and JRA55) along with the application of a scaled gas exchange coefficient for each of the wind products. Through these steps, SeaFlux presents a product ensemble of interpolated global surface ocean pCO2 and air–sea carbon flux estimates for the years 1990–2019
Reference:
Fay, A. R., Gregor, L., Landschützer, P., McKinley, G. A., Gruber, N., Gehlen, M., Iida, Y., Laruelle, G. G., Rödenbeck, C., Roobaert, A., and Zeng, J.: SeaFlux: harmonization of air–sea CO2 fluxes from surface pCO2 data products using a standardized approach, Earth Syst. Sci. Data, 13, 4693–4710, https://doi.org/10.5194/essd-13-4693-2021, 2021
Decadal Trends in the Ocean Carbon Sink
DeVries et al. (2019)
Climate variability drove weakened ocean CO2 uptake in the 1990s, and strengthened CO2 uptake in the 2000s
Open ocean, Surface
yearly
1959 - 2016
Decadal Trends in the Ocean Carbon Sink
DeVries et al. (2019)
Description:
The DeVries et al. (2019) analysis examines decadal trends in global and regional air-sea CO2 fluxes from a variety of ocean biogeochemical models that contributed to the GCB (see No. 60). Three sets of model simulations were performed. Simulation A uses variable climate forcing (e.g., variable wind stress, heat and freshwater fluxes) and observed atmospheric CO2 forcing, Simulation B uses constant (repeated) climate forcing and observed atmospheric CO2, and simulation C uses both constant climate forcing and constant atmospheric CO2 concentrations. With these simulations, the authors partitioned decadal trends in ocean CO2 uptake into those driven by climate variability and those driven by atmospheric CO2. They found that climate variability drove a weakening trend of the ocean carbon sink during the 1990s, and a strengthening trend during the first decade of the 2000s. The magnitude of these trends agreed with those of an OCIM that was trained to replicate tracer data from the 1990s and 2000s (DeVries et al., 2017), indicating that the decadal trends may be driven by variability in ocean circulation.
Reference:
DeVries, T., Le Quéré, C., Andrews, O., Berthet, S., Hauck, J., Ilyina, T., Landschützer, P., Lenton, A., Lima, I. D., Nowicki, M., Schwinger, J., and Séférian, R.: Decadal trends in the ocean carbon sink, Proceedings of the National Academy of Sciences, 116(24), 11646–11651, https://doi.org/10.1073/pnas.1900371116, 2019.
ECCO-Darwin
Carroll et al. (2020)
Model-data synthesis product based on the Estimating the Circulation and Climate of the Ocean (ECCO) ocean state estimate. Fully-closed, physically-consistent 3-D biogeochemical budgets.
Open ocean, Coastal ocean, Surface, Water column
daily, monthly
0.33° x 0.33°
1992 - 2018
ECCO-Darwin
Carroll et al. (2020)
Description:
Carroll et al. (2022) used the Estimating the Circulation and Climate of the Ocean-Darwin (ECCO-Darwin) global-ocean biogeochemistry state estimate to generate a data-constrained DIC budget and investigate how spatiotemporal variability in advection and mixing, air-sea CO2 flux, and the biological pump have modulated the ocean sink for 1995–2018. ECCO-Darwin assimilates ocean circulation and physical tracers, including temperature, salinity, and sea ice, derived from the Estimating the Circulation and Climate of the Ocean (ECCO) LLC270 global-ocean and sea-ice data synthesis (Zhang et al., 2018). Additionally, it assimilates biogeochemical observations encompassing the cycling of carbon, nitrogen, phosphorus (PO 4 ), iron (Fe), silica (SiO 2 ), DO, and TA. This inclusive approach enhances the model’s fidelity by aligning it with a diverse array of observations. All ECCO-Darwin model output is available on the ECCO Data Portal: https://data.nas.nasa.gov/ecco/. The model code and platform-independent instructions for running ECCO-Darwin simulations can be found at: https://github.com/MITgcm-contrib/ecco_darwin.
Reference:
Carroll, D., Menemenlis, D., Adkins, J. F., Bowman, K. W., Brix, H., & Dutkiewicz, S., et al. (2020). The ECCO-Darwin data-assimilative global ocean biogeochemistry model: Estimates of seasonal to multidecadal surface ocean pCO2 and air-sea CO2 flux. Journal of Advances in Modeling Earth Systems, 12, e2019MS001888. https://doi.org/10.1029/2019MS001888
Surface pH and Revelle Factor
Jiang et al. (2019)
A model-observation fusion product for pH, acidity and Revelle Factor, leveraging GFDL-ESM2M and SOCATv6
Open ocean, Surface
decadal, climatology
1° x 1°
1770 - 2100
Surface pH and Revelle Factor
Jiang et al. (2019)
Description:
Jiang et al. (2019a) produced a high-resolution (1°×1°) data product delineating regionally varying view of global surface ocean pH, acidity, and Revelle Factor (RF) from 1770 to 2100 by amalgamating recent observational seawater CO2 data from the SOCAT database (Version 6, 1991–2018, ~23 million observations) (Bakker et al., 2016), and temporal trends at individual locations of the global surface ocean from an Earth System Model, i.e., GFDL-ESM2M (Dunne et al., 2013). The calculations were conducted under historical atmospheric CO2 levels (pre-2005) and four Representative Concentrations Pathways (post-2005) corresponding to the Intergovernmental Panel on Climate Change (IPCC)’s 5th Assessment Report, specifically RCP2.6, RCP4.5, RCP6.0, and RCP8.5.
Reference:
Jiang, L.-Q., Carter, B. R., Feely, R. A., Lauvset, S., and Olsen, A.: Surface ocean pH and buffer capacity: Past, present and future, Scientific Reports, 9(1), 18624, https://doi.org/10.1038/s41598-019-55039-4, 2019.
Surface OA Indicators
Jiang et al. (2023)
A model-observation fusion product for all major OA indicators, leveraging a consortium of 14 Earth System Models and 3 observational data products
Open ocean, Surface
decadal, climatology
1° x 1°
1750 - 2100
Surface OA Indicators
Jiang et al. (2023)
Description:
Jiang et al. (2023) developed a comprehensive model-data fusion product that delineates the trajectory of 10 OA indicators: fCO2, pH, [H+]total, [H+]free [CO32-], Ωarag, Ωcalc, DIC, TA, and RF, as well as temperature and salinity at all locations of the global surface ocean from 1750 to 2100. This product marks a significant breakthrough in OA forecasting by refining temporal trends with data from 14 Earth System Models (ESMs) within CMIP6, and by applying bias and drift corrections from three updated observational ocean carbonate system data products: SOCAT (Version 2022) (Bakker et al., 2016), GLODAPv2.2022 (Lauvset et al., 2022), and CODAP-NA (Jiang et al., 2021). This dataset offers 10-year averages on a 1° × 1° global surface ocean grid, capturing trends from preindustrial times (1750), through historical conditions (1850–2010), and projects future conditions to 2100 across five Shared Socioeconomic Pathways: SSP1-1.9, SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5.
Reference:
Jiang, L., Dunne, J., Carter, B. R., Tjiputra, J. F., Terhaar, J., Sharp, J. D., Olsen, A., Alin, S., Bakker, D. C., Feely, R. A., Gattuso, J., Hogan, P., Ilyina, T., Lange, N., Lauvset, S. K., Lewis, E. R., Lovato, T., Palmieri, J., Santana‐Falcón, Y., Schwinger, J., Séférian, R., Strand, G., Swart, N., Tanhua, T., Tsujino, H., Wanninkhof, R., Watanabe, M., Yamamoto, A., and Ziehn, T.: Global surface ocean acidification indicators from 1750 to 2100, Journal of Advances in Modeling Earth Systems, 15(3), https://doi.org/10.1029/2022ms003563, 2023.
Simulated and Constrained Southern Ocean Carbon Sink
Terhaar et al. (2021)
A constrained estimate of the ocean carbon sink based on the simulated carbon sink from CMIP5 and CMIP6 models and constrained with observations of the ocean physics and carbonate chemistry
Open ocean, Regional
yearly
1850-2100
Simulated and Constrained Southern Ocean Carbon Sink
Terhaar et al. (2021)
Description:
These two datasets include spatially-integrated and annually averaged values for the ocean carbon sink from 1850 to 2100 for different scenarios over the 21st century for the global ocean (Terhaar et al., 2022a, 2022b) and the Southern Ocean (Terhaar et al., 2021b, 2021c). All results are based on CMIP5 and CMIP6 models. For the global ocean carbon sink, values are available for SSP1-2.6, SSP2-4.5, and SSP5-8.5. For the Southern Ocean, values are also available for SSP1-2.6, SSP2-4.5, and SSP5-8.5 and additionally also for RCP2.6, RCP4.5, and RCP8.5. In addition, to the raw simulated values, constrained estimates of the annually averaged ocean carbon sink estimates are available. These constrained estimates adjusted the simulated carbon sink estimates for biases on the ocean’s circulation and surface carbonate chemistry (see Terhaar et al., 2021b, 2022a for details). It is recommended to use the constrained estimates. The datasets are available at https://doi.org/10.17882/103934 and https://doi.org/10.17882/103938.
Reference:
Terhaar, Jens, Frölicher, T. L., & Joos, F.: Southern Ocean anthropogenic carbon sink constrained by sea surface salinity, Science Advances, 7(18), https://doi.org/10.1126/sciadv.abd5964, 2021.
Simulated and Constrained Global Ocean Carbon Sink
Terhaar et al., (2022)
A constrained estimate of the ocean carbon sink based on the simulated carbon sink from CMIP5 and CMIP6 models and constrained with observations of the ocean physics and carbonate chemistry
Open ocean
yearly
1850-2100
Simulated and Constrained Global Ocean Carbon Sink
Terhaar et al., (2022)
Description:
These two datasets include spatially-integrated and annually averaged values for the ocean carbon sink from 1850 to 2100 for different scenarios over the 21st century for the global ocean (Terhaar et al., 2022a, 2022b) and the Southern Ocean (Terhaar et al., 2021b, 2021c). All results are based on CMIP5 and CMIP6 models. For the global ocean carbon sink, values are available for SSP1-2.6, SSP2-4.5, and SSP5-8.5. For the Southern Ocean, values are also available for SSP1-2.6, SSP2-4.5, and SSP5-8.5 and additionally also for RCP2.6, RCP4.5, and RCP8.5. In addition, to the raw simulated values, constrained estimates of the annually averaged ocean carbon sink estimates are available. These constrained estimates adjusted the simulated carbon sink estimates for biases on the ocean’s circulation and surface carbonate chemistry (see Terhaar et al., 2021b, 2022a for details). It is recommended to use the constrained estimates. The datasets are available at https://doi.org/10.17882/103934 and https://doi.org/10.17882/103938.
Reference:
Terhaar, J., Frölicher, T. L., and Joos, F.: Observation-constrained estimates of the Global Ocean Carbon Sink from earth system models, Biogeosciences, 19(18), 4431–4457, https://doi.org/10.5194/bg-19-4431-2022, 2022.
Composite model-based estimate of the ocean carbon sink from 1959 to 2022
Terhaar (2025)
A model-based estimate of the ocean carbon sink combining the respective strengths of hindcast simulations and simulations by coupled earth system models
Open ocean
yearly
1959-2022
Composite model-based estimate of the ocean carbon sink from 1959 to 2022
Terhaar (2025)
Description:
This data product, developed by Terhaar (2025), presents an estimate of the global ocean carbon sink by combining forced hindcast simulations and simulations made by coupled earth system models. Hindcast models manage to adequately simulate the short-term variability of the ocean, but struggle to simulate the long-term climate change trend (Huguenin et al., 2022; Takano et al., 2023; Hollitzer et al., 2024). Earth system models cannot simulate the observed short-term variability by definition, but accurately simulate long-term trends (Takano et al., 2023; Hollitzer et al., 2024). The composite model-based estimate combines the simulated short-term variability from hindcast simulations and the long-term trend from earth system models. The output is supplied with the associated study (https://bg.copernicus.org/articles/22/1631/2025/) (Terhaar, 2025).
Reference:
Terhaar, J.: Composite model-based estimate of the ocean carbon sink from 1959 to 2022, Biogeosciences, 22, 1631–1649, https://doi.org/10.5194/bg-22-1631-2025, 2025.
pCIBR_Clim and pCIBR_Int
Ghoshal et al. (2025)
This product uses a hybrid approach, applying machine learning to correct 1/12° ocean–ecosystem model outputs with SOCAT (1984–2019) and SAS (1991–2019) observations for the Indian Ocean.
Regional
monthly
0.083° x 0.083°
1980-2019
pCIBR_Clim and pCIBR_Int
Ghoshal et al. (2025)
Description:
A machine learning (ML) model is employed to correct biases in surface pCO2 simulations generated by the INCOIS-BIO-ROMS model (pCO2model) over the period 1980–2019. The ML model is trained using the differences between observed (pCO2obs) and modeled pCO2 to estimate the spatio-temporal deviations (pCO2obs − pCO2model). These interannually and climatologically varying deviations are then added back to the original model output, resulting in two improved data products: pCIBR_Int and pCIBR_Clim (Ghoshal et al., 2025). Evaluation against independent datasets, including moored observations (BOBOA), the gridded SOCAT product, and other ML-based pCO2 products (such as CMEMS-LSCEv2 and OceanSODA), demonstrates a significant improvement of approximately 40% ± 3.31% in RMSE compared to the original model. This high-resolution (0.083° × 0.083°), long-term monthly pCO2 data product is available from the INCOIS Portal (https://las.incois.gov.in) and from OCADS: https://www.ncei.noaa.gov/data/oceans/ncei/ocads/metadata/0307788.html (Ghoshal et al., 2025).
Reference:
Ghoshal, P.K., Joshi, A. & Chakraborty, K. An improved long-term high-resolution surface pCO2 data product for the Indian Ocean using machine learning. Sci Data 12, 577 (2025). https://doi.org/10.1038/s41597-025-04914-z
INCOIS‐BIO‐ROMS Surface pCO2 and pH
Chakraborty et al. (2024)
INCOIS-BIO-ROMS, developed following the RECCAP-2 Ocean Modeling Protocol, integrates model simulations with field and reconstructed data to enhance understanding of Ocean Acidification in the Indian Ocean.
Regional
monthly
0.083° x 0.083°
1980-2019
INCOIS‐BIO‐ROMS Surface pCO2 and pH
Chakraborty et al. (2024)
Description:
INCOIS‐BIO‐ROMS Simulated Surface pCO2 and pH for the Indian Ocean: This data product presents a comprehensive assessment of OA trends across the Indian Ocean and its sub-regions from 1980 to 2019, leveraging outputs from a regional, high-resolution coupled ocean-ecosystem model (INCOIS-BIO-ROMS), an offline biogeochemical (BGC) model, and two machine learning-based products (Chakraborty et al., 2024). INCOIS-BIO-ROMS, configured at 1/12° resolution for the Indian Ocean, was developed in accordance with the “RECCAP-2: Ocean Modeling Protocol” for regional oceans. The INCOIS‐BIO‐ROMS simulated surface pCO2 and pH data product is available from the INCOIS Portal (https://las.incois.gov.in) and from OCADS: https://www.ncei.noaa.gov/data/oceans/ncei/ocads/metadata/0307663.html (Chakraborty et al., 2025).
Reference:
Chakraborty, K., Joshi, A. P., Ghoshal, P. K., Baduru, B., Valsala, V., Sarma, V. V. S. S., et al. (2024). Indian Ocean acidification and its driving mechanisms over the last four decades (1980–2019). Global Biogeochemical Cycles, 38, e2024GB008139. https://doi.org/10.1029/2024GB008139
OCIM2-48L
DeVries (2022b)
Data-constrained estimate of anthropogenic CO2 accumulation in the ocean from inverting physical ocean circulation tracers
Open ocean
yearly
2° x 2°
1780-2020
OCIM2-48L
DeVries (2022b)
Description:
Data-constrained estimate of anthropogenic CO2 accumulation in the ocean from inverting physical ocean circulation tracers
Reference:
DeVries, T. (2022). Atmospheric CO2 and sea surface temperature variability cannot explain recent decadal variability of the ocean CO2 sink. Geophysical Research Letters, 49(7), e2021GL096018.