Abstract
Karst regions are ecologically fragile and highly sensitive to both natural and anthropogenic disturbances, necessitating accurate and quantitative assessments of ecological quality and its driving forces. This study introduces a Karst-Specific Remote Sensing Ecological Index (KRSEI), a novel model tailored to the unique environmental conditions of karst landscapes. Using Guizhou Province in Southwest China as the study area, the KRSEI was applied to MODIS remote sensing data from 2002 to 2022, and further integrated with Theil-Sen-Mann-Kendall (Sen-MK) trend analysis and an optimized parameter geographical detector to explore spatiotemporal patterns and key driving mechanisms of ecological change. The findings reveal that: (1) KRSEI achieves higher information entropy and image contrast than traditional RSEI, effectively reflecting ecological heterogeneity; (2) ecological quality remained moderately stable (KRSEI ≈ 0.6) over two decades, with regions of lower ecological status exhibiting a "rise-decline-rebound" trajectory, and 58.25% of the area showing an improving trend; and (3) vegetation cover, precipitation, and population density were identified as dominant drivers, with most factor interactions showing synergistic (bivariate enhancement) effects. This research offers a robust methodological framework for ecological monitoring in karst regions and provides scientific support for targeted ecological protection and sustainable regional development.