Spatiotemporal heterogeneity of agricultural dry-wet abrupt alternation and its driving factors in the southwest karst region of China

Under climate change, the intensification of agricultural dry-wet abrupt alternation (DWAA) events poses an increasingly severe challenge to agricultural production in ecologically fragile karst regions. However, the identification and subregional heterogeneity of agricultural DWAA events, as well as the spatiotemporally varying effects of their driving factors, remain insufficiently understood in the karst region of Southwest China, particularly from the perspective of crop water demand. To address these challenges, the standardized antecedent precipitation evapotranspiration index (SAPEI) combined with run theory was applied to identify agricultural DWAA events. Emerging Hot Spot Analysis (EHSA) was used to characterize their spatiotemporal evolution, while Rotated Empirical Orthogonal Function (REOF) analysis was employed to analyze the major spatial evolution patterns. In addition, the Geographical Detector Model (GDM) was used to identify the dominant driving factors, and Geographically and Temporally Weighted Regression (GTWR) was applied to quantify their spatiotemporally non-stationary effects across different subregions. The results show that: (1) the transition time of dry-to-wet (DTW) events was markedly shorter than that of wet-to-dry (WTD) events, and DTW events occurred approximately three times more frequently; (2) Subregion I was the only hotspot region, whereas Subregion IV was characterized by a dense cluster of cold spots for agricultural DWAA events; (3) three spatial evolution types of agricultural DWAA events were identified: east–west antiphase, central–periphery antiphase, and uniform–localized contrast; (4) the relationships between agricultural DWAA events and key driving factors exhibited spatiotemporal non-stationarity; and (5) the influence of precipitation temporal concentration strengthened in Subregions I, II, and III, whereas vegetation-related regulation became increasingly important in Subregions IV and V. These findings provide scientific support for region-specific agricultural water management and differentiated disaster prevention strategies.