An Algorithm to Bias-Correct and Transform Arctic SMAP-Derived Skin Salinities into Bulk Surface Salinities

An algorithmic approach, based on satellite-derived sea-surface (“skin”) salinities (SSS), is proposed to correct for errors in SSS retrievals and convert these skin into comparable in-situ (“bulk”) the top-5 m of subpolar Arctic Oceans. In preparation routine assimilation operational ocean forecast models, Soil Moisture Active Passive (SMAP) satellite Level-2 observations are transformed using Argo float data from address mismatch between depth L-band measurements (∼1 cm) thickness top model layers (typically at least 1 m). Separate challenge availability most Oceans, products regions currently not suitable a myriad other reasons, including erroneous ancillary air-sea forcing/flux products. root-mean-square error (RMSE) SMAP product several salinity observational sets greater than 1.5 pss (Practical Salinity Scale), which can be larger their temporal variability. Thus, we train machine-learning algorithm (called Generalized Additive Model) an independent bulk-salinities, biases, quantify standard errors. The RMSE corrected bulk-salinities less open regions. Barring persistently problematic near coasts ice-pack edges, bulk-salinity better agreement with equivalent.