Purpose. The purpose of this study is to develop and implement a strongly coupled approach to the assimilation of available observational data in a coupled ocean-sea ice circulation model, and to test it for the Arctic region.

Methods and Results. In the coupled INMIO (ocean) and CICE 5.1 (ice) model at 0.25° resolution, the data were assimilated using the Compact Modeling Framework (CMF3.0) platform with the DAS (Data Assimilation Service) software based on the EnOI (Ensemble Optimal Interpolation) method. A strongly coupled assimilation approach was applied. This approach involves simultaneous adjustment of the fields of water temperature, salinity, sea level, and ice concentration using observational data (ARGO profiles, AVISO satellite altimetry, and OSI SAF ice concentration). A specialized interface was developed to redistribute the integrated ice concentration among the ice thickness categories. Numerical experiments with and without data assimilation were performed for 2020. It is shown that data assimilation through a strongly coupled regime reduces the average error in reproducing the ice area from 27 to 7% as compared to the NSIDC data. The standard error of ocean surface water temperature is reduced to 1.0 °C, and that of ice concentration in the edge area to 0.2. The model fields correspond better to the independent OSTIA data.

Conclusions. The developed approach to the strongly coupled assimilation of oceanic and ice data in the coupled ocean-ice model provides a significant increase in the accuracy of forecasting the condition both of the water and the ice field in the Arctic Ocean. The software can be adapted to other models.