基于三重配置分析TCA的全球日尺度土壤水分融合数据集（2011-2018）

本土壤水分数据集是一个包含8年（2011-2018）的全球时空连续一致的日尺度地表土壤水分数据集，空间格网分辨率为25km，时间分辨率为每天，数据单位为cm3/cm3。数据集采用基于三重配置分析 (TCA: Triple Collocation Analysis)的土壤水分线性融合算法，对SMOS，ASCAT，FY3B，CCI，SMAP五种土壤水分产品分两步进行了融合：第一步，融合2011~2018年SMOS，FY3B和ASCAT土壤水分数据产品；第二步，对第一步融合的2015~2018年间的结果与相应年份的CCI以及SMAP数据产品进行再融合，最终获得2011~2018年间融合的土壤水分产品。最终融合的土壤水分数据在全球空间覆盖比达80%以上。此外，利用全球7个地面观测网络的站点实测土壤水分数据对上述融合的土壤水分产品进行了评价分析，最小RMSE (Root Mean Square Error) 为0.036 cm3/cm3。This soil moisture dataset is an 8-year (2011–2018) globally spatially and temporally consistent daily-scale surface soil moisture dataset, with a spatial grid resolution of 25 km, daily temporal resolution, and a data unit of cm³/cm³. The dataset employs a linear soil moisture fusion algorithm based on Triple Collocation Analysis (TCA) to fuse five soil moisture products, including SMOS, ASCAT, FY3B, CCI, and SMAP, in two stages: first, fuse the SMOS, FY3B and ASCAT soil moisture products from 2011 to 2018; second, re-fuse the 2015–2018 results from the first stage with the CCI and SMAP products of the corresponding years, ultimately obtaining the fused soil moisture product spanning 2011–2018. The final fused soil moisture dataset has a global spatial coverage of over 80%. Additionally, the fused soil moisture product was evaluated and analyzed using in-situ soil moisture measurements from seven global ground observation networks, with a minimum Root Mean Square Error (RMSE) of 0.036 cm³/cm³.