基于多源遥感数据的TVDI方法在荒漠草原旱情监测的应用

为探讨近年来广泛使用的低空间分辨率的MODIS数据以及高空间分辨率的Landast 8数据对同一地区的旱情状况,选择内蒙古自治区干旱频发的乌审旗荒漠草原为研究区,借助分裂窗算法反演地表温度(Ts),获取归一化植被指数(NDVI),建立温度植被干旱指数(TVDI)的干旱监测模型,分别反演MODIS-TVDI和Landast8-TVDI,并与同期野外实测的不同深度土壤含水量进行回归分析。结果发现,基于MODIS和Landast8 2种遥感数据计算得到的TVDI与各层的土壤水分线性相关显著,两者都能表征地表的干旱分布,且Landast8-TVDI与各层土壤含水量的相关性大于MODIS-TVDI与各层土壤含水量的相关性,其中0~10 cm表层土壤含水量的相关性要好于0~20 cm、0~30 cm的相关性。因此Landast8-TVDI能够更好地反映乌审旗荒漠草原的土壤水分状况,更适宜于旱情监测。

Application of the multi-source remote sensing data-based TVDI method in monitoring desert grassland drought

An investigation of the drought condition was conducted in Wushen County Desert Grassland, a drought prone region of the Inner Mongolia Autonomous Regions using the MODIS data with low spatial resolution and the Landast 8 data with high spatial resolution. With a drought monitoring model that was built on temperature vegetation drought index (TVDI) in the split window algorithm for retrieving land surface temperature (T_s) and the normalized difference vegetation index (NDVI), the inversion of MODIS-TVDI and Landast8-TVDI were conducted and a regression analysis was made with the measured soil moisture in different depths. The results showed that the linear correlation was significant between soil the moisture in different layers and the TVDI calculated from two remote sensing data of MODIS and Landast8. Both of them could characterize the drought distribution in the land surface. Besides, the correlation between Landast8-TVDI and the soil moisture of each layer was greater than that between MODIS-TVDI and the soil moisture. Of which, the correlation in the surface layer of 0-10 cm was better than that in the layer of 0-20 and 0-30 cm. Therefore, Landast8-TVDI could better reflect the status of soil moisture in the desert grassland and is more suitable for drought monitoring.