Compared to the Arctic, there are few in-situ sea-ice thickness (SIT) datasets for the Southern Ocean (SO), and an estimation method using data from the Synthetic Interferometric Radar Altimeter (SIRAL) on board Cryosat-2 (CS-2) has not yet been developed. In this study, as a precursor for the SIT estimation method for the SO, we improved the estimation method proposed by the Alfred Wegener Institute (AWI) and applied it to the SO. To validate the ESAL2 datasets in the Arctic, we used the freeboard and SIT dataset of the CS-2 SIT product from the AWI and the National Snow and Ice Data Center (NSIDC), the aircraft mission IceBridge dataset, and the ice mass balance (IMB) buoy dataset. The root mean square errors (RMSEs) were 0.47 and 0.97m for ESAL2 and buoy or IceBridge, respectively. These RMSEs differed from those of the AWI and NSIDC products by 0.1m or less. Accordingly, ESAL2 may be used to estimate SIT with the same precision as that obtained from other CS-2 products. Applying this method to the SO and comparing it with in-situ observational data yielded a mean square error of 1.00 m, which is larger than the mean sea ice thickness of 2.06 m. However, there was a positive correlation
coefficient of 0.70. Because the estimated SIT depends on sea ice classification, such as fast-ice margin and first ice area, it is necessary to develop a more accurate method for classification of sea ice to improve estimation accuracy in the future.
View full abstract