AUTHOR=Wang Sinan , Xu Kairan , Ma Xiaoming , Hu Wei , Zhao Jianwei , Wang Fuqiang , Su Donghui , Zhang Zhuo , Wu Yingjie , Li Mingyang 

TITLE=Characteristics of spatial and temporal changes of drought in the Inner Mongolia section of the Yellow River Basin based on geodetector

JOURNAL=Frontiers in Climate

VOLUME=Volume 7 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/climate/articles/10.3389/fclim.2025.1656424

DOI=10.3389/fclim.2025.1656424

ISSN=2624-9553

ABSTRACT=The Inner Mongolia section of the Yellow River Basin is a critical ecological barrier and core production base (energy, agriculture, animal husbandry) in northern China, where water resource security and ecological sustainability directly affect regional and national development. Drought is a major constraint on the basin’s ecosystem and production activities, but its long-term spatial-temporal patterns and driving mechanisms remain insufficiently understood. Here, we used MOD16 evapotranspiration data (2001-2024) to construct the Crop Water Stress Index (CWSI), aiming to clarify drought evolution in the region, and applied the Geodetector model to identify key drivers of CWSI spatial heterogeneity and their interactions. Our results showed that: (1) Potential evapotranspiration (PET) and CWSI had stable interannual fluctuations, while actual evapotranspiration (ET) exhibited significant interannual variability; the average CWSI was 0.85, indicating long-term severe drought in the study area. (2) Spatially, most regions were in severe drought, but >76% of the area showed a significant downward CWSI trend (drought alleviation). (3) All land use types had decreasing CWSI; water bodies had the highest CWSI and cropland the lowest—opposite to the ET ranking. (4) NDVI and air temperature were the primary drivers of CWSI spatial variability (average q-values >0.5). The strongest interactive effects on CWSI heterogeneity were between precipitation and temperature (q: 0.72-0.95) and between precipitation and DEM (q: 0.78-0.93), with both interactive q-values >0.81. These findings reveal that drought in the basin is regulated more by ET than PET, and is driven by the synergy of meteorological factors (precipitation, temperature) and topographic factors (DEM). They provide a scientific basis for drought prediction, water resource management, and ecological protection in the Yellow River Basin, and offer a reference for similar arid/semi-arid regions globally.