基于TVDI的滇西北土壤水分反演研究

Soil Moisture Inversion in Northwestern Yunnan Based on TVDI

以香格里拉为研究区,以2002年、2005年、2009年、2013年以及2015年的Landsat系列影像和气象数据为数据源,利用单窗算法反演地表温度。以Ts-NDVI特征空间原理为基础,应用温度植被干旱指数(TVDI)反演香格里拉的土壤水分。通过GIS空间分析法完成对香格里拉2002—2015年的土壤水分动态反演。结果表明:2002年土壤水分含量最高,2005年土壤水分含量大幅度降低,2005—2015年呈持续上升的状态。宏观上比较典型的房屋建筑区及裸地处于严重干旱的状态,典型的耕地林地处呈现出中度干旱或轻度干旱的状态,植被覆盖度高的地区则受干旱影响较小,土壤较湿润。TVDI反演结果切合研究区的下垫面性质,与土地覆被类型密切相关。

Soil moisture is one important part in the surface water cycle and also one of the key factors of agricultur-al production. At the same time, it is an important indicator to guide drought monitoring. Taking Shangri - La as study area, based on a series of yearly Landsat images of 2002, 2005 , 2009 , 2013 , 2015 , and the meteorologi-cal data. The land surface temperature inversion was carried out by mono - window algorithm. The soil water of Shangri - la was calculated by the model of Temperature Vegetation Dryness Index on the basis of the principle of Ts - NDVI feature space. Shangri - La soil moisture dynamic inversion was completed by GIS spatial analysis meth-ods. The results showed that the content of soil moisture was the highest in 2002. Soil moisture content was greatly reduced in 2005 and increased from 2005 to 2015. From a macro perspective, typical housing construction area and bare land are in the state of severe drought. Cultivated land and forest land are moderate drought or drought. High vegetation coverage areas are less affected by the drought and the soil is wet. TVDI inversion results were in accord-ance with the land surface with high correlation with the land types.