Assimilation of sea surface salinities from SMOS in an Arctic coupled ocean and sea ice reanalysis

In the Arctic, the sea surface salinity (SSS) plays a key role in processes related to water mixing and sea ice. However, the lack of salinityobservations causes large uncertainties in Arctic Ocean forecasts and reanalysis. Recently the Soil Moisture and Ocean Salinity (SMOS) satellitemission was used by the Barcelona Expert Centre to develop an Arctic SSS product. In this study, we evaluate the impact of assimilating this data ina coupled ocean-ice data assimilation system. Using the deterministic ensemble Kalman filter from July to December 2016, two assimilation runsrespectively assimilated two successive versions of the SMOS SSS product on top of a pre-existing reanalysis run. The runs were validated againstindependent in situ salinity profiles in the Arctic. The results show that the biases and the root-mean-squared differences (RMSD) of SSS arereduced by 10 % to 50 % depending on the area and highlight the importance of assimilating satellite salinity data. The time series offreshwater content (FWC) further shows that its seasonal cycle can be adjusted by assimilation of the SSS products, which is encouraging of theassimilation of SSS in a long-time reanalysis to better reproduce the Arctic water cycle.

Automated summary

This study evaluates the impact of assimilating Soil Moisture and Ocean Salinity (SMOS) satellite data in an Arctic ocean-ice data assimilation system. The results show that assimilating satellite salinity data can reduce biases and root-mean-squared differences in sea surface salinity (SSS) in the Arctic, highlighting the importance of incorporating satellite data in Arctic Ocean forecasts.