Performance and relationship of four different agricultural drought indices for drought monitoring in China's mainland using remote sensing data

Increasing frequency and intensity of extreme drought events have harmed the environment, ecosystem, and agricultural productivity. However, the characteristics of agricultural drought in China have not been well understood. The remote sensing (RS) based gridded monthly precipitation, soil moisture, land surface temperature (LST), and normalized difference vegetation index (NDVI) datasets over 1982–2018 were utilized to derive standardized precipitation index (SPI), standardized soil moisture index (SSI), multivariate standardized drought index (MSDI), and vegetation health index (VHI). The variation patterns and trends of SPI, SSI, and MSDI at the 1-, 3-, and 6-month timescales against monthly VHI anomaly were compared to identify the best agricultural drought index in China. The drought variations in the four sub-regions (northwest, north, Qinghai-Tibet area, and south area) were also investigated. The results showed that: (1) Temporal patterns of VHI anomaly were similar to relative soil moisture and slightly different from precipitation. The spatial patterns of MSDI matched with VHI the best than SPI and SSI. (2) All three indices showed positive correlations with VHI at the three timescales. The highest correlation coefficients (r) between MSDI and VHI ranged from 0.25 to 0.67, 0.22 to 0.78, 0.23 to 0.69, and 0.19 to 0.74 in northwest China, north China, Qinghai-Tibet Plateau, and south China, respectively. (3) The connections between monthly VHI and the three drought indices were weaker at the 1-month timescale (0.16 < r < 0.25) than the 3-month (0.39 < r < 0.78) and 6-month (0.26 < r < 0.68) timescales. (4) The VHI significantly increased in most of China except north China. Overall, MSDI showed better performance for monitoring agricultural drought in China's mainland.