Ecological restoration of the Loess Plateau plays a pivotal role in mitigating land degradation and promoting regional sustainability. In this study, landscape pattern metrics were integrated into the MaxEnt model to evaluate the influence of landscape configuration on restoration planning. Nine representative species from three vegetation strata—herbs, shrubs, and trees—were selected based on ecological suitability. A comprehensive set of variables, including environmental, anthropogenic, and landscape metrics, was constructed for modeling. Results demonstrate that incorporating landscape metrics significantly enhanced the spatial explanatory power, providing a robust supplement to traditional ecological restoration assessments. Distinct responses to landscape structure were observed among vegetation types: herb species were more sensitive to patch aggregation and connectivity, shrubs preferred regular edges and larger patch size, while tree species favored extensive, low-fragmentation core habitats. Vertical structure optimization revealed that while large areas were suitable for single vegetation layers, composite vegetation configurations were more appropriate in certain central and southern subregions. These findings underscore the importance of landscape structure in guiding restoration strategies and serve as a basis for designing ecologically coherent and spatially targeted vegetation restoration plans on the Loess Plateau.

Against the backdrop of rapid urbanization and climate warming, the urban heat island (UHI) effect has severely affected ecological security and public health. Existing studies have often focused on single-city analyses or large-sample averages, with insufficient attention to the nonlinear driving mechanisms of UHI under different hydrothermal contexts. This study selects Wuhan and Xi’an as representative cities, constructing an explainable machine learning framework to interpret and compare UHI intensity across feature importance, nonlinear responses, factor interactions, and spatial differentiation. The results show that, in Wuhan, the top five factors contribute 62.4%, reflecting a composite dominance of ecology, spatial morphology, location, and human activities. In Xi’an, the top five factors contribute 72.0%, indicating a more concentrated dominant structure. Nonlinear responses reveal that key factors like NDVI have distinct effect thresholds and mechanisms in the two cities. Spatially, Wuhan displays a continuous gradient pattern characterized by center-promoting and peripheral-suppressing effects, whereas Xi’an presents a block-like mosaic structure composed of multiple juxtaposed districts. These differences suggest that UHI mitigation should move beyond a uniform control model and instead adopt climate-sensitive strategies that account for the dominant factor combinations, response thresholds, and spatial organization of each city. The proposed framework and findings provide scientific support for understanding UHI mechanisms under different hydrothermal contexts and offer targeted implications for thermal environment regulation and spatial planning in cities with similar climatic and environmental characteristics.

Arid regions are ecologically fragile and occupy a substantial portion of the global terrestrial surface. In these regions, ecosystem services (ESs) are strongly constrained by water availability and, more importantly, by the balance between water supply and demand. However, the nonlinear responses and threshold mechanisms linking water supply–demand balance to ES dynamics remain unclear. Taking the Kubuqi Desert in the “Great Bend” of the Yellow River as the study area, this study quantified the Comprehensive Ecosystem Service Index (CESI) and the Water Supply–Demand Ratio (WSDR) by integrating the InVEST model, RWEQ model, the RUSLE model, Water Balance Method, and so on. The dual-constraint line method and elasticity coefficient approach were integrated to identify the constraint effects and critical thresholds of WSDR on CESI. Ecological management zones were further delineated by integrating the inflection-point intervals of the dual-constraint lines with the threshold intervals identified by elasticity coefficients. The results showed that CESI remained relatively low, with a maximum value of approximately 0.5, suggesting that the overall ES level was still limited, but exhibited a continuous increasing trend. The regional water supply–demand pattern gradually shifted from deficit toward relative balance, although agricultural water use still accounted for about three-quarters of total consumption. CESI showed a nonlinear threshold response to WSDR: mild water deficit suppressed CESI growth, whereas moderate water surplus promoted CESI recovery by alleviating water constraints and improving ecosystem functioning. Thresholds identified by elasticity coefficients mainly occurred near critical transitions between water deficit and surplus. Based on ES supply and threshold sensitivity, nine ecological management zones were identified, with priority enhancement areas accounting for approximately 75%. These findings provide a threshold-based basis for ecological zoning and differentiated restoration in arid regions.