基于大气温度和水汽含量改进HJ-1B地表温度反演

为了提高地面气象站稀少地区地表温度遥感反演的精度,本文基于多源遥感数据的优势,首先利用MODIS影像获取研究区像元尺度上平均大气水汽含量;然后利用同时相的HJ-1B影像估算区域地表比辐射率,再采用温度-植被指数法获取近地表大气温度;最后将以上3个参数输入单窗体算法,改进其地表温度反演的精度。研究结果表明,改进单窗体算法反演地表温度与地面实测温度的偏差小于1 K,为地面气象站点稀少的植被覆盖区域提供了一种可行的精确遥感反演地表温度方法。

Improvements in Land Surface Temperature Retrieval from HJ-1 B Data Based on Atmospheric Water Vapor Content and Temperature

To improve the accuracy of land surface temperature retrieval based on remote sensing in the regions with rare distribution of in-situ meteorological stations.In this paper,we present an improved methodology to retrieve land surface temperature from HJ-1 B data using atmospheric water vapor content and atmospheric temperature which could be obtained by remote sensing images.First the atmospheric water vapor content on pixel scale was calculated from the MODIS data.Then the emissivity of various land covers and land uses were defined by HJ-1B data.What's more,the temperature-vegetation index method was applied to mapping area-wide instantaneous near surface air temperature.Finally,the parameters of mean atmospheric water vapor content,near surface air temperature and land surface emissivity were inputted to the mono-window algorithm in order to improved land surface temperature retrieval precision.The results showed that compared with in situ land surface temperature,this improved mono-window algorithm made significantly better retrieval of LST in estimation than did mono-window algorithm,it can be widespread used to retrieve LST in the vegetation covered regions where with few in-situ meteorological stations.