Attributing the divergent changes of drought from humid to dry regions across China

Different drought indices can significantly influence our understanding of drought dynamics, particularly as evaporative demand intensifies with ongoing warming. This study investigated long-term drought variations across China using the Standardized Precipitation Index (SPI) and the Standardized Precipitation-Evaporation Index (SPEI). Additionally, we attributed divergent drought dynamics to climate and vegetation variations by applying Multivariate Nonlinear Regression models and Random Forest models. From 1982 to 2022, SPI exhibited a significant wetting trend (0.05/decade, p < 0.01), while SPEI transitioned from wetting to a drying trend in 1993, followed by a significant drying trend (−0.19/decade, p < 0.01), particularly in arid regions. The two indices were highly correlated in humid regions (r = 0.89–0.94) but showed weaker correlations in arid regions (r = 0.57–0.81). The influence of climate variables on SPEI increased progressively from humid (87.6 %) to hyper-arid (95.6 %) regions. Precipitation and relative humidity were the most influential factors in humid and non-humid regions, respectively. While the combined effects of precipitation and relative humidity on SPEI were dominant in most regions, relative humidity and minimum temperature played a more significant role in arid and hyper-arid regions. Stronger drought-vegetation coupling was observed in humid (32.8 %) and hyper-arid (32.5 %) regions. Notably, SPEI was sensitive to evaporation and transpiration in densely vegetated humid regions, while in sparsely vegetated hyper-arid regions, transpiration was the dominant factor. These findings underscore the importance of selecting appropriate drought indices based on regional climatology to enhance drought monitoring and support agricultural and water management strategies.