摘要
本项工作在黑河流域中上游区域内,利用地下4cm深度的地面实测土壤水分数据验证了2012年7月至2014年12月期间AMSR2的两种算法产品——日本宇航局标准算法土壤水分产品(JAXA产品)和阿姆斯特丹自由大学联合美国宇航局开发的陆表参数反演模型算法土壤水分产品(LPRM产品)。验证结果显示:与地面实测数据相比,所有验证像元上两种土壤水分产品的均方根误差RMSE(Root Mean Square Error)普遍超过了0.1m^3/m^3。JAXA产品动态变化范围较小,升轨产品的总体精度略高于降轨,相比地面实测数据均存在明显的低估,在冻季与实测数据比较接近。LPRM产品动态范围较大,降轨产品在冻季不可用,在未冻季升轨产品精度高于降轨,相比地面实测数据有高估的倾向。同时,还进一步讨论并分析了两种算法对土壤温度和植被的不同处理方式对土壤水产品精度的可能影响,指出了算法可能的改进方向。
A comprehensive evaluation of two kinds of the AMSR2 soil moisture products was provided during the period 2012.7.1～2014.12.31 on the upper and middle reaches of Heihe River basin compared with-4cm ground measurements：one is retrieved by the Japan Aerospace Exploration Agency（JAXA）algorithm（hereafter cited as JAXA product）,the other one is the land parameter retrieval model algorithm developed by the Free University of Amsterdam Joint the National Aeronautics and Space Administration（NASA）（hereafter cited as LPRM product）.The result revealed that the root mean square errors（RMSEs）of both two soil moisture products exceeds 0.1 m^3/m^3 compared with ground measurements.JAXA soil moisture product has a restricted range and quite underestimates ground measurements.Values of JAXA product is close to insitu in frozen seasons.Performance of JAXA soil moisture product ascending data is better than descending data while the difference between them is small.LPRM product has a large dynamic range and overestimates ground measurements.The descending LPRM product is not available in frozen seasons.As for unfrozen seasons,performance of the ascending LPRM product is better than the descending data with closer estimations but worse correlation.Analysis and discussion were then given based on their different approaches towards soil temperature and vegetation factor.Reasonable results to soil moisture products and feasible improving methods on the algorithms were also discussed.
出处
《遥感技术与应用》
CSCD
北大核心
2017年第2期324-337,共14页
Remote Sensing Technology and Application
基金
国家重点基础研究发展计划资助项目(2013CB733406)
国家自然科学基金项目(41531174)资助
