Spatiotemporal Non‐Stationarity and Driving Mechanisms of Ecological Risk in the Three Gorges Reservoir Area: Integrating Land Use Patterns and Terrain Positioning Into Multi‐Scale Geospatial Modeling

ABSTRACT To address the gaps in existing research regarding multiscale coupling mechanisms of ecological risk and integrated analysis across vertical–horizontal dimensions in mountainous reservoir catchments, this study focuses on the Three Gorges Reservoir Area (TGRA) in China and aims to investigate the spatiotemporal patterns of ecological risk and their underlying drivers through an integrated framework that combines land use analysis with risk assessment. The study demonstrates theoretical novelty by being the first to apply spatial autocorrelation, Geodetector, and Geographically Weighted Regression (GTWR) models to comprehensively evaluate ecological risk across both horizontal and vertical dimensions in reservoir ecosystems—an approach rarely employed in prior studies of such systems. Based on Landsat imagery, population density, nighttime light data, and climate variables (2000–2020), the study maps risk distribution and identifies key drivers. Results reveal a southwest‐northeast expanding risk pattern, with high‐risk zones concentrated in low‐altitude urbanized areas. Between 2000 and 2020, construction land expanded by 1769.13 km 2 , primarily at the expense of cultivated and ecological land, reflecting intense urbanization pressure. Ecological land remained dominant but declined gradually, while other land types—mainly construction land—increased from 2.18% to 5.89% of the total area. Human activities, especially land development, are the dominant drivers, showing significant spatial heterogeneity. Ecological risk exhibited complex temporal dynamics: high‐risk areas decreased by 2.79% in area proportion, yet remained concentrated in low‐elevation zones, where risk levels continued to rise. Low‐risk areas expanded by 3.50% from 2010 to 2020, particularly in mountainous regions, due to ecological restoration efforts such as the Grain for Green Program. The centroid of ecological risk shifted slightly northward, with migration distance accelerating after 2015, indicating increasing spatial uncertainty. The study proposes tiered land management strategies: buffer zone establishment in low‐risk areas, ecological restoration in mid‐risk zones, and systemic governance with early warning systems in high‐risk areas. Offering a scientific basis for ecological risk governance, this research provides actionable insights for sustainable management of large‐scale reservoir ecosystems globally.