基于决策树模型的武汉市地表比辐射率反演研究

摘要

地表比辐射率值是反演地表温度中的关键参数，它通过高效的遥感影像分类方法来获取。利用NDVI指数，M NDWI指数及NDBI指数，探讨基于决策树模型的比辐射率估算方法，并通过与传统方法的对比研究，认为针对武汉市的特点，采用决策树的比辐射率估算方法精度高于传统模型，较适合研究区的地表组成特征。%T he land surface emissivity is a key parameter for retrieving land surface tem-perature ,w hich gains by higher remote sensing classification algorithm .In this paper ,a new estimation method of land surface emissivity is presented based on MNDWI (Modified Nor-malized Difference Water Index )、NDBI(Normalized Difference Built-up Index ) .According to the attributes of Wuhan city ,the accuracy of the results show that the estimating result of LSE with Decsion Tree is better than other models ,which is propitious to the study area .