ALBERT

Description: Published empirical algorithms for oceanic total alkalinity (TA) and dissolved inorganic carbon (DIC) are used with monthly sea surface salinity (SSS) and temperature (SST) derived from satellite (SMOS, Aquarius, SST CCI) and interpolated in situ (CORA) measurements and climatological (WOA) ancillary data to produce monthly maps of TA and DIC at one degree spatial resolution. Earth system model TA and DIC (HADGEM2-ES) are also included. Results are compared with in situ (GLODAPv2) TA and DIC and results analysed in five regions (global, Greater Caribbean, Amazon plume, Amazon plume with in situ SSS 〈 35 and Bay of Bengal). Results are presented in three versions, denoted by 'X' in the lists below: using all available data (X = ''); excluding data with bathymetry 〈 500m (X = 'Depth500'); excluding data with both bathymetry 〈 500m and distance from nearest coast 〈 300 km (X = 'Depth500Dist300'). Datasets S1 to S5 are .csv lists of matchups in each region - date and location, in situ TA and DIC measurements and estimated uncertainties, all input datasets, estimates of TA and DIC from all outputs, and the best available output estimates of TA and DIC for each matchup. S1_GlobalAlgorithmMatchupsX.csv S2_GreaterCaribbeanAlgorithmMatchupsX.csv S3_AmazonPlumeAlgorithmMatchupsX.csv S4_AmazonPlumeLowSAlgorithmMatchupsX.csv S5_BayOfBengalAlgorithmMatchupsX.csv Datasets S6 to S10 are .csv statistical analyses of the performance of each combination of algorithm and input data - carbonate system variable, algorithm, input datasets used, (MAD, RMSD using all available data, output score, RMSD estimated from output score, output and in situ mean and standard deviation, correlation coefficient), all items in brackets presented both unweighted and weighted, number of matchups, number of potential matchups, matchup coverage, RMSD after subtraction of linear regression, percentage reduction in RMSD due to subtraction of linear regression and weighted score divided by number of matchups). S6_GlobalAlgorithmScoresX.csv S7_GreaterCaribbeanAlgorithmScoresX.csv S8_AmazonPlumeAlgorithmScoresX.csv S9_AmazonPlumeLowSAlgorithmScoresX.csv S10_BayOfBengalAlgorithmScoresX.csv Datasets S11 to S15 are zipped netCDF files containing error analyses of all outputs in each region, including the squared error of each output at each matchup, the weight of each squared error (1/squared uncertainty), weight * squared error, number of matchups available to each output, number of matchups available to each combination of two outputs, (score of each output in a given comparison of two outputs, overall output score and RMSD estimated from output score), all items in the last brackets presented both unweighted and weighted. S11_GlobalSquaredErrorsX.nc S12_GreaterCaribbeanSquaredErrorsX.nc S13_AmazonPlumeSquaredErrorsX.nc S14_AmazonPlumeLowSSquaredErrorsX.nc S15_BayOfBengalSquaredErrorsX.nc Datasets S16 to S20 are zipped netCDF files containing global maps of the mean and standard deviation of each of: in situ data; output data; output data - in situ data and number of matchups. Regional files show the same maps, but only including data within the region. S16_GlobalmapsX.nc S17_GreaterCaribbeanmapsX.nc S18_AmazonPlumemapsX.nc S19_AmazonPlumeLowSmapsX.nc S20_BayOfBengalmapsX.nc Datasets S21 and S22 are .csv files containing the effect on estimated RMSD of excluding various combinations of algorithms and/or inputs for TA and DIC in each region. For a given variable and region, the first line shows the algorithm, input data sources, estimated RMSD and bias of the output with lowest estimated RMSD. Subsequent lines show the effect of excluding combinations of algorithms and/or inputs, ordered first by the number of algorithms/inputs excluded (fewest first), then by effect on lowest estimated RMSD. So the first line(s) consist of the effects of excluding the best algorithm and each of the input sources to that algorithm, most important first. Each line consists of the item excluded, ratio of resulting estimated RMSD to original estimated RMSD, resulting bias and number of items excluded. Some exclusions are equivalent, for instance exclusion of WOA nitrate (the only nitrate source) is equivalent to excluding all algorithms using nitrate. Dataset S21 contains a comprehensive list of all possible exclusions, and so is rather hard to read and interpret. To mitigate this, Dataset S22 contains only those exclusion sets with effect greater than 1% and at least 0.1% greater than any subset of its exclusions. S21_importancesX.csv S22_importances2X.csv Dataset S23 is a .csv file containing like-for-like comparisons of RMSD between TA and DIC in each region. Bear in mind that the RMSD shown here is not the same as the estimated RMSD (RMSDe) shown elsewhere. S23_TA_DICcomparisonX.csv

Description: This submission includes the reference data required to perform a complete verification of the FluxEngine v3.0 install. All data are in netCDF-3 format. Note that this dataset is greater than 100 MB in size. FluxEngine is an open source software toolkit for calculating in situ, regional or global gas fluxes between the atmosphere and ocean. It can be used with model, in situ or satellite Earth observation data. A full description of the toolkit is provided in Shutler et al. (2016) and the FluxEngine software can be freely downloaded from GitHub: https://github.com/oceanflux-ghg/FluxEngine Input data are for the year 2010 with the exception of the Takahashi climatology, which is for the year 2000. The following input data are included: * input_data/air_pressure - atmospheric pressure data from the European Centre for Medium-Range Weather Forecasts (ECMWF, www.ecmwf.int) * input_data/ice - fraction ice coverage from the Ocean and Sea Ice Satellite Application Facility (OSI SAF, www.osi-saf.org) * input_data/rain_gpcp - total precipitation data from the National Oceanic and Atmospheric Administration (NOAA) Global Precipitation Climatology Project (v2.2) * input_data/sig_wv_ht - significant wave height data from the GlobWave project (www.globwave.org) * input_data/sigma0 - radar backscatter from the GlobWave project (www.globwave.org) * input_data/windu10 - wind speed at 10 m above sea level from the GlobWave project (www.globwave.org) * input_data/SMOS - ocean salinity data from Centre Abal de Traitement des Données Soil Moisture and Ocean Salinity (CATDS SMOS v01, www.catds.fr) * input_data/SOCATv4 - The Surface Ocean CO₂ Atlas (SOCAT) (Bakker et al., 2016) version 4 reanalysed to a CO₂ climatology using (Goddijn-Murphy et al., 2015) * input_data/SST - skin sea surface temperature (SST) data from the ATSR Reprocessing for Climate (ARC) project. * input_data/sstfnd_Reynolds - sub skin sea surface temperature (Optimally Interpolated SST, OISST project: Reynolds et al., 2007, Banzon et al., 2016) * input_data/takahashi09 - Takahashi CO₂ climatology dataset (Takahashi et al., 2009) Where appropriate input data have been re-gridded from their original resolution into a monthly 1 × 1 degree resolution. The verification process involves comparing a newly generated output (e.g. generated by a local install of FluxEngine) to a reference dataset for each verification scenario. These reference datasets are included in the 'reference_data' directory, and use the following naming convention: * socatv4, takahashi09_pco2 - uses partial pressure of CO₂ (pCO₂) data from SOCATv4 or Takahashi2009 climatologies, respectively * sst, no_gradients - does/does not use sea surface temperature gradients, respectively * salinity - applies a correction for skin layer salinity * N00 - uses (Nightingale et al., 2000) parameterisation for calculating gas transfer velocity * K0, K1, K2, K3 - uses the generic formulation for calculating gas transfer velocity (using only the 0th, 1st, 2nd and 3rd order components, respectively) * takahashi09_all_inputs - The Takahashi et al. (2009) dataset is used for all inputs in order to reproduce the results in Shutler et al. (2016). * FEv1, FEv2, FEv3 - Reference data was generated using FluxEngine version 1, 2 or 3, respectively Configuration files for each validation set are included in the 'configs' directory. These files are well commented and fully specify the input data to be used, data pre-processing, gas transfer velocity parameterisation and the structure of the gas flux calculation to be performed. Acknowledgements: The Surface Ocean CO₂ Atlas (SOCAT) is an international effort, endorsed by the International Ocean Carbon Coordination Project (IOCCP), the Surface Ocean Lower Atmosphere Study (SOLAS) and the Integrated Marine Biosphere Research (IMBeR) program, to deliver a uniformly quality-controlled surface ocean CO₂ database. The many researchers and funding agencies responsible for the collection of data and quality control are thanked for their contributions to SOCAT.

Description: The Surface Ocean CO₂ Atlas (SOCAT) version 2019 dataset (Bakker et al., 2016) is a quality-controlled dataset containing 25.7 million surface ocean gaseous CO₂ measurements collated from thousands of individual submissions. These gaseous CO₂ measurements are typically collected at many different depths (of the order of several metres below the surface) using many different systems, and the sampling depth varies dependent upon the sampling platform and/or setup. Different platforms (e.g. ships of opportunity, research vessels) and systems will collect water samples at different depths, and the sampling depth can even vary dependent upon sea state. Therefore, the collated SOCAT dataset contains high quality data, but these data are all valid for different and inconsistent depths. Therefore the SOCAT provided individual gaseous CO₂ measurements and gridded data are sub-optimal for calculating global or regional atmosphere-ocean gas exchange (and the resultant net CO₂ sinks) and sub-optimal for verifying gas fluxes from (or assimilation into) numerical models. Accurate calculations of CO₂ flux between the atmosphere and oceans require CO₂ concentrations at the top and bottom of the mass boundary layer, the ~100 μm deep layer that forms the interface between the ocean and the atmosphere (Woolf et al., 2016). Ignoring vertical temperature gradients across this very small layer can result in significant biases in the concentration differences and the resulting gas fluxes (e.g. ~5 to 29% underestimate in global net CO₂ sink values, Woolf et al., 2016). It is currently impossible to measure the CO₂ concentrations either side of this very thin layer, but it is possible to calculate the concentrations either side of this layer using the SOCAT data, satellite observations and knowledge of the carbonate system. Therefore to enable the SOCAT data to be optimal for an accurate atmosphere-ocean gas flux calculation, a reanalysis methodology was developed to enable the calculation of the fugacity of CO₂ (fCO₂) for the bottom of the mass boundary layer (termed sub-skin value). The theoretical basis and justification for this is described in detail within Woolf et al., (2016) and the re-analysis methodology is described in detail in (Goddijn-Murphy et al., 2015). The re-analysis calculation exploits paired in situ temperature and fCO₂ measurements in the SOCAT dataset, and uses an Earth observation dataset to provide a depth-consistent (sub-skin) temperature field to which all fugacity data are reanalysed. The outputs provide paired fCO₂ (and partial pressure of CO₂) and temperature data that correspond to a consistent sub-skin layer temperature. These can then be used to accurately calculate concentration differences and atmosphere-ocean CO₂ gas fluxes. This data submission contains a reanalysis of the fugacity of CO₂ (fCO₂) from the SOCAT version 2019 dataset to a consistent sub-skin temperature field. The reanalysis was performed using a tool that is distributed within the FluxEngine open source software toolkit (https://github.com/oceanflux-ghg/FluxEngine) (Shutler et al., 2016; Holding et al., in-review). All data processing and driver scripts are available from the FluxEngine ancillary tools repository https://github.com/oceanflux-ghg/FluxEngineAncillaryTools. The NOAA Optimum Interpolation Sea Surface Temperature (OISST) dataset (Reynolds et al., 2007) was used to provide the climate quality and depth consistent temperature data. The original OISST data were first resampled to provide monthly mean values on a 1º by 1º degree grid. These data were then used as the temperature input for the reanalysis. The resulting reanalysed data are provided as a tab-separated value file (individual data points) and as netCDF-5 file (gridded monthly means). These are the same file formats as provided by SOCAT and analogous to the SOCAT single data point and gridded data. Each row in the tab-separated value file corresponds to a row in the original SOCAT version 2019 dataset. The original SOCAT version 2019 data are included in full, with four additional columns containing the reanalysed data: * T_reynolds - The temperature (in degrees C) taken from the consistent OISST temperature field for the corresponding time and location. * fCO2_reanalysed - The fugacity of CO₂ (in μatm) reanalysed to the consistent surface temperature indicated by T_reynolds. * pCO2_SST - The partial pressure of CO₂ (in μatm) corresponding to the in situ (measured) temperature. * pCO2_reanalysed - The partial pressure of CO₂ (in μatm) reanalysed to the consistent surface temperature indicated by T_reynolds. The netCDF gridded version of the reanalysed dataset contains monthly mean data, binned into a 1º by 1º grid and uses the same units, missing value indicators and time and space resolution as the original SOCAT gridded product to maximise compatibility. The gridding is performed using the SOCAT gridding methodology (Sabine et al. 2013). The implementation of the gridding has been verified by performing the gridding on the original (non-reanalysed) SOCAT data and all results were identical to 8 decimal places. The result of gridding the original SOCAT data are included within these netCDF data, along with additional variables containing the equivalent results for the reanalysed SOCAT data. Statistical sample mean, minimum, maximum, standard deviation and count data for each grid cell are included, with unweighted and cruise-weighted versions (following the convention used by SOCAT). Full meta data are included within the file. Notes: 1. Due to the temporal range of the OISST dataset the reanalysed values are only available from 1981 onwards. Pre-1981 rows contain "NaN" (not-a-number) in the reanalysis columns. 2. The download for this submission is provided as a single .zip file (1.1 GB, uncompressed: 10.7GB) containing two files: SOCATv2019_reanalysed_subskin.tsv (containing every data point, ungridded) and SOCATv2019_reanalysed_subskin.nc (the gridded monthly mean data). How to cite these data: Please cite this PANGAEA submission, the theory (Woolf et al., 2016), the reanalysis methodology (Goddijn-Murphy et al., 2015), the FluxEngine toolbox which was used to perform the reanalysis (Shutler et al., 2016, Holding et al. in review) and the original SOCAT dataset (Bakker et al., 2016) and/or gridded equivalent (Sabine et al., 2013). Acknowledgements: The Surface Ocean CO₂ Atlas (SOCAT) is an international effort, endorsed by the International Ocean Carbon Coordination Project (IOCCP), the Surface Ocean Lower Atmosphere Study (SOLAS) and the Integrated Marine Biosphere Research (IMBeR) program, to deliver a uniformly quality-controlled surface ocean CO₂ database. The many researchers and funding agencies responsible for the collection of data and quality control are thanked for their contributions to SOCAT. These data were provided by two Integrated Carbon Observing System (ICOS) European Union (EU) readiness projects, Ringo (grant no. 730944) and BONUS Integral (grant no. 03FO773A).

Description: The data set consists of interpolated fields of global surface ocean partial pressure of carbon dioxide (pCO2sw) and the flux of CO2 between ocean and atmosphere, on monthly time and 1-degree latitude and longitude, between January 1992 and December 2018. The pCO2sw is interpolated to this grid from the data set of Holding et al: https://doi.org/10.1594/PANGAEA.905316, which is in turn derived from the the SOCATv2019 observational data (https://www.socat.info/) where the observations have been corrected to the surface subskin temperature using satellite-derived surface temperature products. The interpolation uses the neural net technique of Landschützer et al. ( Biogeosciences 10, 7793-7815, doi:10.5194/bg-10-7793-2013 2013). The ocean-atmosphere flux is then calculated, using the gas transfer equation with the gas transfer velocity parameterized as a function of wind speed and atmospheric mixing ratio of CO2, with a further correction for the cool (and salty) surface ocean skin. These corrections for near-surface temperature deviations increase the net negative (e.g. into the ocean) flux. Full details are given in the corresponding article in Nature Communications (doi:10.1038/s41467-020-18203-3) and its accompanying extended data.

Description: The Surface Ocean CO2 Atlas (SOCAT) version 2021 (v2021) dataset (Bakker et al., 2016, Bakker et al., 2021) is a quality-controlled dataset containing 30.6 million surface ocean gaseous CO2 measurements collated from thousands of individual submissions. These gaseous CO2 measurements are typically collected at many different depths (of the order of several metres below the surface) using many different systems, and the sampling depth varies dependent upon the sampling platform and/or setup. Different platforms (e.g. ships of opportunity, research vessels) and systems will collect water samples at different depths, and the sampling depth can even vary dependent upon sea state. Therefore, the collated SOCAT dataset contains high quality data, but these data are all valid for different and inconsistent depths. This means that the SOCAT provided individual gaseous CO2 measurements and gridded data are sub-optimal for calculating global or regional atmosphere-ocean gas exchange (and the resultant net CO2 sinks) and sub-optimal for verifying gas fluxes from (or assimilation into) numerical models. Accurate calculations of CO2 flux between the atmosphere and oceans require CO2 concentrations at the top and bottom of the mass boundary layer, the ~100 μm deep layer that forms the interface between the ocean and the atmosphere (Woolf et al., 2016). Ignoring vertical temperature gradients across this very small layer can result in significant biases in the concentration differences and the resulting gas fluxes (e.g. ~5 to 29% underestimate in global net CO2 sink values, Woolf et al., 2016). It is currently impossible to measure the CO2 concentrations either side of this very thin layer, but it is possible to calculate the concentrations either side of this layer using the SOCAT data, satellite observations and knowledge of the carbonate system. Therefore to enable the SOCAT data to be optimal for an accurate atmosphere-ocean gas flux calculation, a reanalysis methodology was developed to enable the calculation of the fugacity of CO2 (fCO2) for the bottom of the mass boundary layer (termed sub-skin value). The theoretical basis and justification for this is described in detail within Woolf et al., (2016) and the re-analysis methodology is described in detail in (Goddijn-Murphy et al., 2015). The re-analysis calculation exploits paired in situ temperature and fCO2 measurements in the SOCAT dataset, and uses an Earth observation dataset to provide a depth-consistent (sub-skin) temperature field to which all fugacity data are reanalysed. The outputs provide paired fCO2 (and partial pressure of CO2) and temperature data that correspond to a consistent sub-skin layer temperature. These can then be used to accurately calculate concentration differences and atmosphere-ocean CO2 gas fluxes. This data submission contains a reanalysis of the fugacity of CO2 (fCO2) from the SOCAT version 2021 dataset to a consistent sub-skin temperature field. The reanalysis was performed using a tool that is distributed within the FluxEngine V4.0.1 open source software toolkit (https://github.com/oceanflux-ghg/FluxEngine) (Shutler et al., 2016; Holding et al., 2019). All data processing and driver scripts are available from the FluxEngine Ancillary Tools (FEAT) repository https://github.com/oceanflux-ghg/FluxEngineAncillaryTools. The National Oceanic and Atmospheric Administration (NOAA) Optimum Interpolation Sea Surface Temperature (OISST) dataset (Reynolds et al., 2007) were used to provide the climate quality and depth consistent temperature data. The original ¼ degree OISST weekly data (v2.1) were first resampled to provide monthly mean values on a 1º by 1º degree grid (using the Python tools provided in the FEAT repository). These monthly 1º by 1º data were then used as the temperature input for the reanalysis. The resulting reanalysed data are provided as a tab-separated value file (individual data points) and as netCDF-5 file (gridded monthly means). These are the same file formats as provided by SOCAT and analogous to the SOCAT single data point and gridded data. Each row in the tab-separated value file corresponds to a row in the original SOCAT version 2021 dataset. The original SOCAT version 2021 data are included in full, with four additional columns containing the reanalysed data: * T_reynolds - The temperature (in degrees C) taken from the consistent OISST temperature field for the corresponding time and location. * fCO2_reanalysed - The fugacity of CO2 (in μatm) reanalysed to the consistent surface temperature indicated by T_reynolds. * pCO2_SST - The partial pressure of CO2 (in μatm) corresponding to the in situ (measured) temperature. * pCO2_reanalysed - The partial pressure of CO2 (in μatm) reanalysed to the consistent surface temperature indicated by T_reynolds. The netCDF gridded version of the reanalysed dataset contains monthly mean data, binned into a 1º by 1º grid and uses the same units, missing value indicators and time and space resolution as the original SOCAT gridded product to maximise compatibility. The gridding is performed using the SOCAT gridding methodology (Sabine et al. 2013). The implementation of the gridding has been verified by performing the gridding on the original (non-reanalysed) SOCAT data and all results were identical to 8 decimal places. The result of gridding the original SOCAT data are included within these netCDF data, along with additional variables containing the equivalent results for the reanalysed SOCAT data. Statistical sample mean, minimum, maximum, standard deviation and count data for each grid cell are included, with unweighted and cruise-weighted versions (following the convention used by SOCAT). Full meta data are included within the file.