Retrieving Soil Moisture by TVDI Based on Different Vegetation Index: A Case Study of Shanxi Province

Combined with ground weather observation data from March 2005 to October, the MODIS vegetation index products was used to analysis the sensitivity of Ratio Vegetation Index (RVI), Normalized Difference Vegetation Index (NDVI) and the Enhanced Vegetation Index (EVI) under different coverage of surface. Together with the MODIS land surface temperature products to form the feature space of Ts/RVI, Ts/NDVI and Ts/EVI, from which obtaining the Temperature-Vegetation Dryness Index (TVDI). Based on TVDI, the 0-20 cm relative soil moisture is acquired. Results show that the average soil moisture retrieved from Ts/EVI is closer to in situ data. Average retrieved soil moisture relative to average verification soil moisture has a deviation of 2%-12%. However, the Ts/RVI has a better accuracy of retrieving soil moisture under vegetation coverage of higher than 0.6, while Ts/EVI performs better for inversing soil moisture under low vegetation coverage of lower than 0.4.