摘要：Global climatologies of the seawater CO2 chemistry variablesare necessary to assess the marine carbon cycle in depth. The climatologiesshould adequately capture seasonal variability to properly address oceanacidification and similar issues related to the carbon cycle. Totalalkalinity (AT) is one variable of the seawater CO2 chemistrysystem involved in ocean acidification and frequently measured. We used theGlobal Ocean Data Analysis Project version 2.2019 (GLODAPv2) to extractrelationships among the drivers of the AT variability and ATconcentration using a neural network (NNGv2) to generate a monthlyclimatology. The GLODAPv2 quality-controlled dataset used was modeled by theNNGv2 with a root-mean-squared error (RMSE) of 5.3µmolkg−1.Validation tests with independent datasets revealed the good generalizationof the network. Data from five ocean time-series stations showed anacceptable RMSE range of 3–6.2µmolkg−1. Successful modeling ofthe monthly AT variability in the time series suggests that the NNGv2is a good candidate to generate a monthly climatology. The climatologicalfields of AT were obtained passing through the NNGv2 the World OceanAtlas 2013 (WOA13) monthly climatologies of temperature, salinity, and oxygenand the computed climatologies of nutrients from the previous ones with aneural network. The spatiotemporal resolution is set by WOA13:1∘×1∘ in the horizontal, 102 depth levels(0–5500m) in the vertical and monthly (0–1500m) to annual (1550–5500m)temporal resolution. The product is distributed through the data repositoryof the Spanish National Research Council (CSIC;https://doi.org/10.20350/digitalCSIC/8644, Broullón et al., 2019).
