Abstract. In this study, a new method to assimilate satellite radar altimetry derived freeboard instead of sea ice thickness is presented with the goal of improving the initial state of sea ice thickness predictions in the Arctic. In order to quantify the improvement in sea ice thickness gained by assimilating freeboard, we compare three different model runs. One reference run (refRun), one that assimilates only SIC (sicRun) and one that assimilates both SIC and FB (fbRun). It is shown that, estimates for both SIC and FB can be improved by assimilation, but only the fbRun improved the sea ice thickness estimates. The resulting sea ice thickness is evaluated by comparing it to Alfred Wegener Institute's (AWI) weekly CryoSat-2 sea ice thickness data product, which is based on the same FB observations as were assimilated in this study. It is shown that the sea ice thickness from the fbRun is closer to the traditional CryoSat-2 sea ice thickness than sea ice thickness from refRun or sicRun. Additionally, we compare independent sea ice draft measurements from the Beaufort Gyre Exploration Project to both fbRun sea ice thickness and observed CryoSat-2 sea ice thickness. This comparison shows that our new method provides equally good results as the AWI weekly CryoSat-2 product; in two of three locations even better results.

No abstract

No abstract