Data assimilation system to improve sea ice predictions in the Arctic Ocean using an ice-ocean coupled model

Abstract

With the recent changes in the Arctic climate, transportation in the Arctic Ocean has increased. It is important to accurately predict the ice conditions to navigate safely in the Arctic Ocean. In this study we have improved the ice-ocean coupled Ice-POM model by implementing a data assimilation system. This system incorporates sea ice observation data such as sea ice concentration, sea ice thickness and sea ice velocity to improve the numerical predictions. Ocean part of the model is based on the Princeton Ocean Model while the Ice model considers the discrete characteristics of the ice along the ice edge. The model domain consists of Arctic Ocean, Greenland-Iceland-Norwegian (GIN) seas and the Northern Atlantic Ocean. The model grid is with 25kms resolutions. Several data assimilation techniques have been tested in the study. We have used a nudging method that incorporates background and observation errors and an ensemble assimilation approach that uses different atmospheric forcing conditions. The observation errors and background errors were taken in to account. Observation errors were varied according to the location and the season. Assimilation has improved the ocean conditions significantly. This is evident from the changes in the ocean salinity compared to the model predictions. We have also compared the results from different assimilation techniques. Ensemble Kalman filter method outperforms the nudging method however it comes with a greater computational cost. The assimilation time interval was also varied. While daily assimilation window produced the most accurate results, the weekly and monthly assimilation window also produce results with reasonable accuracy about 5 months after assimilation is begun.