SeaFlux v2023: harmonised sea-air CO2 fluxes from surface pCO2 data products using a standardised approach

version 2023.02: fixed a bug where sol was 1e3 too large for the units (mol m3 uatm). Now corrected to the right order of magnitude. Mean for sol = 4.547e-5 version 2023.01: released alongside the GCBv2023. Note that I have not provided NCEPv2 since this was dropped during the analysis of the paper. Further, the salinity for the solubility is the same as that used in OceanSODA-ETHZ (see the documentation of that product). Temperature is still NOAA's OISST. I do not provide pco2atm but rather fco2atm since the GCB now requests fco2 instead of pco2. Note that the alpha (coefficients of gas transfer) are still the same as before. version 2021.04.03: includes globally integrated fluxes including all wind and pco2 products. version 2021.04.02: removed some files that remained from v2021.02 that I didn't delete in v2021.04 version 2021.04 now extends the variables to calculate fluxes from 1982-2020. The comparison period for fluxes is now limited to 1990-2019 (30 years). The area contains coastal fraction coverage. A missing strip along the longitude 179.5°E is filled in. Negative values of pCO2 are limited to 50 µatm (primarily affects JENA in Hudson Bay). version 2021.02 calibrates kw with 14-C bomb estimated global average kw (16.5 ± 3.2 cm/hr) where the average is calculated without ice weighting (v2021.01 included ice weighting). Further, the date range of the data has been increased from 1982 to 2020 where possible (not possible for the scaling factor and pCO2 product).