闪电河流域SMAP L4被动微波土壤水分降尺度研究  

Downscaling of SMAP L4 Passive Microwave Soil Moisture in Shandian River Basin

作　　者：文静 杨娜 WEN Jing;YANG Na(School of Surveying and Land Information Engineering,Henan Polytechnic University,Jiaozuo,Henan 454000,China;School of Cyber Science and Engineering,Ningbo University of Technology,Ningbo,Zhejiang 315211,China)

机构地区：[1]河南理工大学测绘与国土信息工程学院,河南焦作454000 [2]宁波工程学院网络空间安全学院,浙江宁波315211

出　　处：《遥感信息》2024年第6期114-121,共8页Remote Sensing Information

基　　金：国家自然科学基金重点项目(U21A20108);国家自然科学基金(42271124)。

摘　　要：作为全球水循环中的关键变量,土壤水分在农业生产、灾害监测、气候研究等诸多领域具有重要的支撑作用。为优化星载被动微波遥感土壤水分数据的空间尺度,使其更好地满足区域性的研究与应用,针对闪电河流域,文章在“地形-植被-温度”传统因子随机森林模型的基础上,进一步引入地表反照率、蒸散发、覆被类型3个参量,构建改进因子模型(地表反照率、蒸散发)和加权融合模型(覆被类型),将9 km的SMAP L4土壤水分数据降至1 km尺度。结果表明:改进因子模型的建模精度优于(R=0.778)传统因子模型(R=0.766);加权融合模型能够实现传统和改进因子模型的集成,降尺度结果与地面实测数据具有良好的相关性(R=0.502),能够保持SMAP L4数据所反映的土壤水分空间特征。As a key parameter of the global water cycle,soil moisture plays an important supporting role in agricultural production,disaster monitoring,climate research and many other fields.In order to optimize the spatial scale of space-borne passive microwave remote sensing soil moisture data(SMAP L4)for regional research and application,this paper constructs an improved factor model(surface albedo(ALBEDO),evapotranspiration(ET))and a weighted fusion model(land cover(LC))for the Shandiang river basin.Building on the traditional random forest model(RF)based on“terrain-vegetation-temperature,”three additional parameters ALBEDO,ET,and LC are introduced.These models are then used to downscale the 9 km SMAP L4 soil moisture data to a 1 km resolution.The results show that:the modeling accuracy of the improved factor model(R=0.778)is better than that of the traditional factor model(R=0.766);the weighted fusion model is able to realize the integration of the traditional and the improved factor models,and the downscaling results have a good correlation with the in-situ data(R=0.502)and are able to maintain the spatial characteristics of soil moisture reflected by the SMAP L4 data.

关 键 词：土壤水分 随机森林 SMAP 降尺度 闪电河流域 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]