高密度城市三维建成环境与绿地空间格局协调关系及其非线性驱动机制研究OA

Rapid urbanization and intensified vertical densification have made the coordinated optimization of the three-dimensional built environment(3DBE)and green space spatial pattern(GSSP)a critical issue in high-density cities.In this context,the contradiction between urban development and ecological protection is no longer limited to two-dimensional land occupation,but increasingly manifests through the ways in which vertical development intensity,spatial enclosure,and morphological heterogeneity reshape ecological space,openness conditions,and landscape organization.Taking the central urban area of Chengdu as the study area,this paper aims to reveal the coordination status between 3DBE and GSSP,as well as its nonlinear driving mechanism,within a unified spatial framework.Chengdu provides a representative case because it combines high development intensity,a complex urban spatial form,and a relatively strong ecological foundation under the"Park City"strategy.Its central urban area is characterized by intensive population agglomeration,high-density construction,and increasing vertical differentiation,while ecological restoration and green space improvement have become important components of urban governance.At a unified grid scale,this study constructs a comprehensive evaluation framework integrating indicators of structure,form,and landscape configuration for both 3DBE and GSSP.Spatial statistical analysis,a coupling coordination degree model,and XGBoost-SHAP are employed to identify spatial differentiation,evaluate coordination status,and detect key drivers and threshold effects.The results indicate significant spatial differentiation and mismatch between 3DBE and GSSP in the central urban area of Chengdu.Green space is mainly distributed along ecological corridors and peripheral ecological areas,showing a directional and relatively continuous aggregation pattern.In contrast,high-intensity vertical development is concentrated along major urban development axes and core functional districts,with a clear tendency toward axial concentration and outward diffusion.This spatial mismatch constitutes an important basis for the uneven coordination status across the study area.Both systems exhibit strong spatial heterogeneity,but their organizational logics differ substantially.GSSP indicators are more strongly constrained by ecological background conditions and continuous green systems,whereas 3DBE is more strongly shaped by development intensity and functional agglomeration.In addition,3DBE indicators are generally highly associated with GSSP indicators,especially with landscape metrics such as fragmentation,patch size,and edge characteristics,suggesting that the influence of 3DBE on GSSP is mainly reflected in the restructuring of green space configuration rather than merely in changes in green quantity.In densely developed areas,reduced spatial openness may further constrain local ventilation and daylight access.In terms of coordination status,the study area is overall dominated by primary coordination,although local mismatch remains evident.High-value areas are mainly distributed in ecologically advantageous zones,including the Longquan Mountain area,the ring ecological belt,farmland patches,and the surroundings of major parks,where relatively complete ecological bases and continuous green space systems support a comparatively stable coordination status.By contrast,low-value areas are concentrated in high-intensity development zones,industrial sectors,and rapidly expanding new urban districts,where excessive development intensity,green space fragmentation,and insufficient spatial openness tend to overlap.The overall pattern can therefore be summarized as dispersed high values and contiguous low values,indicating pronounced intra-urban spatial disparities.Although the study area has generally moved beyond severe antagonism between development and ecology,its overall coordination status remains at a transitional stage from basic matching to stable synergy.The nonlinear driving mechanism analysis shows that coordination status is jointly shaped by a limited number of key structural factors.Building volumetric spatial density,the largest patch index,and floor area ratio are identified as the most important variables,while building height heterogeneity also plays an important role.Compared with linear regression,XGBoost provides stronger explanatory power,indicating that the formation of coordination status cannot be adequately captured by linear models alone.SHAP results further reveal significant nonlinear and threshold-dependent effects.In particular,building volumetric spatial density,floor area ratio,and the largest patch index all exhibit clear threshold characteristics:their contributions to coordination status are weak or even negative at low values,become positive after surpassing certain thresholds,and then show diminishing marginal effects in higher ranges.These findings suggest that the relationship between 3DBE and GSSP in high-density cities is not a simple trade-off,but is regulated by the structural matching among development intensity,spatial form,openness conditions,and green space configuration.Overall,this study incorporates 3DBE into a unified analytical framework for examining urban ecological spatial relationships,and conceptualizes its interaction with GSSP as the coordination status between two categories of spatial attributes within the same spatial unit.The findings suggest that ecological space optimization in high-density cities should pay greater attention to the structural matching and coordinated configuration among development intensity,spatial form,and GSSP.The results provide a quantitative reference for ecological space optimization,spatial zoning governance,and differentiated regulation in high-density urban areas.