Abstract:

Gansu Province， located in the arid and semi-arid transition zone of northwestern China， experiences significant spatiotemporal variations in drought that profoundly impact regional ecological security and agricultural sustainability. This study constructed a Temperature Vegetation Drought Index （TVDI） model based on Enhanced Vegetation Index （EVI） and Land Surface Temperature （LST） data from MODIS （2000-2022） to analyze the spatial patterns of drought intensity in Gansu Province over 23 years. The Sen trend analysis and Mann-Kendall （M-K） significance test were employed to examine drought evolution characteristics. The Hurst index combined with Sen trend analysis was used to predict future drought trends， and the Geodetector model was applied to explore the driving mechanisms of drought. The results indicate that： （1） The annual average TVDI value in Gansu Province from 2000 to 2022 was 0.57， indicating a mild drought level， with drought-affected areas accounting for 82.18% of the total area； （2） The annual TVDI decline rate was 0.0005， with significant drought alleviation observed in the eastern foothills of the Qilian Mountains， the Gannan Plateau， and the Longdong region； （3） The Hurst index combined with Sen trend analysis predicts that 61.4% of the study area will transition to a wetter state in the future， showing significant anti-persistence characteristics； （4） The Geodetector model reveals that elevation and temperature are the dominant factors driving drought， and their interaction significantly enhances explanatory power. These findings provide a scientific basis for drought risk early warning and ecological adaptive management in Gansu Province.

Key words: Gansu Province, TVDI, Hurst index, Geodetector