Dual-polarized ratio algorithm for retrieving Arctic sea ice concentration from passive microwave brightness temperature.

We present a new algorithm for retrieving sea ice concentration from the AMSR-E data, the dual-polarized ratio (DPR) algorithm. The DPR algorithm is developed using vertically and horizontally polarized brightness temperatures at the same channel of 36.5 GHz. It depends on the ratio of dual-polarized emissivity, α, which is determined empirically at about 0.92 by remotely sensed brightness temperature in winter and used for the other seasons as well. The ice concentration retrieved by the DPR is compared with those by the NT2 and ABA algorithms. Since the main difference among these algorithms takes place in marginal ice zones, 17 marginal ice zones are chosen. The retrieved ice concentrations in these zones are examined by the ice concentration obtained by the MODIS data. The mean error, root-mean-square error and mean absolute error of the DPR algorithm are relatively better than those from the other two algorithms. The results of this study illustrate that the DPR algorithm is a more accurate algorithm for retrieving sea ice concentration from the AMSR-E brightness temperature, and can be used for operational purposes. [ABSTRACT FROM AUTHOR]