Monitoring monthly soil moisture conditions in China with temperature vegetation dryness indexes based on an enhanced vegetation index and normalized difference vegetation index

Abstract

Soil moisture is a key land surface parameter that reflects conditions of drought in agricultural areas, such that variations in soil moisture significantly impact agricultural production. At a large spatial scale, soil moisture can be indicated using the temperature vegetation drought index (TVDI), which is based on soil moisture monitoring. Such metrics allow us to know whether or not the soil in a region is experiencing drought. The TVDI can be estimated by using either the normalized difference vegetation index (NDVI) or the enhanced vegetation index (EVI) combined with land surface temperature (LST) data. Our objectives were to study the applicability of the NDVI-based TVDI (TVDINDVI) and EVI-based TVDI (TVDIEVI) in the monitoring of soil moisture at a large time-space scale. The TVDINDVI and TVDIEVI were calculated by constructing a feature space based on the NDVI (or EVI) and LST to obtain the corresponding dry/wet-edge equation. By establishing a correlation relationship between the TVDIs and soil moisture at soil depths of 0–10, 0–40, 0–100, and 0–200 cm, the applicability of the TVDIEVI and TVDINDVI was investigated for seven different climatic regions in mainland China. The relationships between soil moisture at the 0–10-cm depth and the NDVI (or EVI) in cold and warm seasons were obtained. Linear functions of soil moisture and the TVDINDVI (TVDIEVI) were constructed. From the results, both the TVDINDVI and TVDIEVI were more suitable for monitoring soil moisture at the 0–10-cm depth than at the other depths. The regions that had the best correlations between soil moisture and the TVDINDVI (TVDIEVI) were in northwestern China during the cold season and southeastern China during the warm season (from April to September), meaning that the TVDI was more applicable in these region/season combinations. The TVDINDVI showed better performance in reflecting soil moisture conditions than the TVDIEVI, especially in the warm season. The applicability of the TVDIEVI for soil moisture at the 0–10-cm depth was better than that of the TVDINDVI in northwestern China. However, in the other regions, the TVDINDVI performed better. The performance of TVDINDVI (TVDIEVI) was better than TVDIEVI (TVDINDVI) under different circumstances, which was related to the goodness of fit for wet/dry-edge equations, natural variability of NDVI/EVI/LST, depths, changes in seasons (vegetation cover), and regions. This research compared the applications of the TVDI in China comprehensively.