Modeling Skin-Layer Salinity: Focus on Seasonal Variability and Global Means

Abstract:

Due to near-surface salinity stratification, it is problematic to compare satellite-measured surface salinity within the first few centimeters (skin-layer) of the ocean with Argo-measured top-level salinity at 5 m, or with ocean models that do not resolve the skin layer. Although an instrument can be designed to measure the surface salinity, a global scale measurement is currently not available. A regional model can be configured to have a vertical grid in centimeters, but it would be computationally prohibited on a global scale due to time step constraints. Here, we propose an extended surface salinity layer (ESSL) within a global ocean circulation model to diagnose skin SSS without increasing the computational cost, while allowing comparable solutions with both satellite and Argo salinity at the respective depths within the corresponding uncertainty limits. Cross-comparisons with Aquarius and Argo data show that the gridded Aquarius surface salinity has a much stronger seasonal variability than the gridded Argo top-level salinity at 5 m or 10 m, particularly in regions of high precipitation variability, suggesting strong near-surface vertical salinity stratification. The near-surface stratification is well reproduced by the proposed ESSL model. In comparisons with data-assimilated HYCOM results, the ESSL provides much stronger seasonal variability of SSS, similar to the Aquarius observations. The ESS also provides a useful reference for the global mean SSS to constrain the global calibration constants in Aquarius SSS retrieval.