High resolution data set of thin sea ice thickness in the Arctic Ocean

New and improved estimates of thin sea-ice thickness in the Arctic Ocean are derived from AMSR-E satellite for the period 2002-2011, at a spatial resolution of 6.25 km and using a newly developed fast-ice mask. The algorithm is based on comparisons between the polarization ratio (PR) of AMSR-E brightness temperatures from the 89 and 36 GHz channels (PR89 and PR36) and the thermal ice thickness. The thermal ice thickness is estimated from a heat budget calculation using the ice surface temperature from clear-sky Moderate-Resolution Imaging Spectroradiometer (MODIS) infrared data. Whereas coastal polynyas have been the main target of previous algorithms, this algorithm is also applicable for marginal ice zones. AMSR-E has twice the spatial resolution of Special Sensor Microwave/Imager (SSM/I) data and can therefore resolve polynyas at a smaller scale. Although the spatial resolution of the 89 GHz data (6.25 km) is twice that of the 36 GHz data (12.5 km), the 89 GHz data can be contaminated by atmospheric water vapor. We propose an exclusion method of data affected by water vapour to resolve this issue.本数据集针对2002-2011年的北冰洋薄海冰厚度提供了全新改进的估算结果，数据源自先进微波扫描辐射计-地球观测系统（AMSR-E）卫星观测，空间分辨率为6.25 km，并采用了新研发的固定冰掩膜。该算法基于89 GHz与36 GHz频段下AMSR-E亮温的极化比（PR89与PR36）与热学海冰厚度之间的对比关系构建。热学海冰厚度通过热量收支计算得到，其输入数据来自晴空条件下中分辨率成像光谱仪（MODIS）的红外冰面温度数据。过往算法主要以沿岸冰间湖为核心研究目标，而本算法同样适用于边缘冰区。AMSR-E的空间分辨率是专用传感器微波成像仪（SSM/I）数据的两倍，因此能够以更小尺度分辨冰间湖。尽管89 GHz频段数据的空间分辨率（6.25 km）为36 GHz频段数据（12.5 km）的两倍，但89 GHz频段数据易受大气水汽污染。针对这一问题，我们提出了一种受水汽影响数据的剔除方法。