基于地形校正TVDI的地下水位埋深反演

Estimation of groundwater level from TVDI after topographic correction

在干旱、半干旱地区,地下水是陆生植被生存的重要水源,而传统的地下水位监测方法费时费力,及时获取大尺度高精度的地下水位埋深显得十分重要。在温度植被干旱指数(TVDI)的基础上,分别提取三期Landsat 8遥感数据的归一化植被指数(NDVI)、修正土壤调节植被指数(MSAVI)、比值植被指数(RVI)、差值植被指数(DVI)和增强型植被指数(EVI)和地表温度(Ts),并引入DEM数据对Ts进行地形校正,减少地形起伏对能量二次分配的影响,并选择最佳TVDI反演地下水位埋深状况,结果表明:(1)在传统TVDI的基础上引入DEM进行地形校正,分析校正前后TVDI与地下水位埋深的决定系数,R^2从0.4381提高到0.5053,这说明地形校正能够有效地提高地下水位埋深的反演精度;(2)通过三期影像对比分析了五种不同的TVDI值分别与其对应栅格点的地表土壤湿度和地下水位埋深的决定系数,总体上都是Ts-MSAVI较好,最高R^2分别为0.5547、0.5202;(3)在实测土壤含水量缺失的情况下,可以根据反映土壤湿度高低的因子(TVDI)间接地反演地下水埋深分布状况。

In arid and semiarid areas,groundwater is the important water source for terrestrial vegetation survival,but traditional method of monitoring underground water is waste time and energy. Acquiring large-scale and high-precision groundwater level in real or near-real time is critical. Based on the temperature vegetation drought index（ TVDI）,three phase of Landsat 8 extract RVI（ Ratio Vegetation Index）,NDVI（ Normalized Difference Vegetation Index）,EVI（ Enhanced Vegetation Index）,MSAVI（ Modified Soil Adjusted Vegetation Index）,DVI（ Difference Vegetation Index） and Ts（ Land Surface Temperatures）,the DEM data were used to correct the Ts data for the reduction of the topographic influences by topographic relief,and the best groundwater depth inversion model was selected. The results indicated that：（ 1） DEM was used to regulate Ts based on the traditional TVDI,the coefficient of determination between TVDI before and after topographic correction and groundwater depth was improved from 0.4381 to 0.5053,which showed that the terrain correction could effectively improve the inversion accuracy of groundwater depth;（ 2） Compared the coefficient of determination between five different TVDI and soil moisture and groundwater depth by three phase of the image,in the whole,Ts-MSAVI was best,and the highbest R^2 was 0.5547 and 0.5202 respectively;（ 3） In case the soil moisture data were lacking,the groundwater depth distribution could be estimated by factors reflecting the relative soil water content.