基于改进FCLS算法的南极海冰密集度估算及算法比较

摘要

Sea ice has a heat isolation effect that could influence global climate change via modification of the exchange of heat between the ocean and the atmosphere. Sea ice concentration is one of the most important indicators in polar sea ice research. To achieve high spatial resolution in estimations of multitype sea ice concentrations, this study introduced a polarization gra-dient rate and a spectral gradient rate into a fully constrained least squares (FCLS) algorithm. Accuracy was validated using Antarctic Sea Ice Processes and Climate(ASPeCt) data, and the results were compared with NASA Team2(NT2) and Arctic Radiation and Turbulence Interaction Study Sea Ice(ASI) algorithms. The root mean square error(RMSE) and bias of the im-proved FCLS algorithm throughout the year were 13.8% and -0.7%, respectively. The accuracy of the improved algorithm was found better for multiyear ice than first-year ice.%海冰具有良好的热力隔绝效应, 它通过影响海洋和大气的热交换进而影响全球的气候变化.海冰密集度是极区海冰研究的重要指标之一.为实现高空间分辨率多类型海冰密集度的估算, 本文将亮温极化梯度率和光谱梯度率引入基于全约束最小二乘法(fully constrained least squares, FCLS)的海冰密集度估算方法,并利用南极海冰过程与气候计划(Antarctic Sea Ice Processes and Climate, ASPeCt)对改进方法的精度进行验证, 然后与NASA Team2(NT2)算法和ARTIST Sea Ice(ASI)算法获得的海冰密集度结果进行了对比分析.结果显示, 3种算法中本研究的方法精度最高, 全年均方差13.8%, 偏差为-0.7%; 改进的方法对多年冰的估算精度优于一年冰.