OceanSODA-UNEXE: Gridded surface ocean carbonate system datasets in the Amazon and Congo River outflows

Within the European Space Agency funded Oceanographic datasets for acidification (OceanSODA) project, the University of Exeter (UNEXE) produced the OceanSODA-UNEXE dataset (v1.0) which is an optimal dataset of the surface ocean carbonate system in the Amazon and Congo River outflows. All four main carbonate system variables, total alkalinity (TA), dissolved inorganic carbon (DIC), the partial pressure of carbon dioxide (pCO2) and pH are provided on monthly 1° × 1° grids along with additional carbonate system parameters. The uncertainties within these data have been assessed using independent in situ database (Land et.al 2022). A paper detailing the methodology used to optimally construct and then evaluate this dataset is currently being written. Each netCDF4 dataset file contains 10 or more years of data; the full carbonate system is provided for 2010-2020 in the Amazon outflow (defined as 2°S and 24°N and between 70°W and 31°W) datasets and the full carbonate system is provided for the period 2002-2016 in the Congo outflow (defined as 10°S and 4°N and between 2°W and 16°E). Variables are stored on a 180° by 360° latitude grid with a time dimension (defined as the months from January 1957 to December 2021). Following the methodology of Land et al. (2019), TA and DIC were derived using empirical algorithms from the published literature that use combinations of inputs that include sea surface temperature (SST), sea surface salinity (SSS) datasets and nutrients (silicate (SiO4-), nitrate (NO3-), phosphate (PO4-) or dissolved oxygen (DO). TA and the inputs used to derive it (e.g. SST and SSS) are within the netCDF files prefixed with _TA, whereas DIC and the inputs used to derive it (SST and SSS) are within the netCDF files prefixed with _DIC. The full carbonate system equations (calculating for surface waters) were run twice with PyCO2SYS V1.7 (Humphreys et al., 2022), using the same TA, DIC, SiO4- and PO4- along with the SST and SST datasets from the respective DIC or TA netCDF files. The variables computed with PyCO2SYS are the carbonate ion (CO3-2), the bicarbonate ion (HCO3-), hydrogen ions (H+) ,pH on the total scale, pH on the free scale, pH on the seawater scale, the partial pressure of carbon dioxide (pCO2), the fugacity of carbon dioxide (fCO2),the saturation state of calcite and the saturation state of aragonite. A full list of variables and references for all input data can be found in Table 1. All variable fields have an associated uncertainty field; this uncertainty has the same abbreviated variable name along with the suffix uncertainty (e.g. TA_uncertainty). SST, SSS and nutrient input data uncertainties come from their respective dataset accuracy assessments and dataset references (Table 1). TA and DIC uncertainty is the combined standard uncertainty from the algorithm and input data evaluation determined using the methods of Land et al. (2019) which are consistent with the uncertainty methods of (JCGM, 2008). Uncertainties for the remaining variables were determined by propagating the TA, DIC, SST and SSS uncertainties through PyCO2SYS using a forward finite difference approach (Humphreys et al., 2022).