Optimizing urban patterns is increasingly recognized as an effective technological strategy to mitigate the urban heat island (UHI) effect. Taking Xiamen City as a case study, this research extracts and quantifies city spatial characteristics from multiple data sources. Key factors influencing the urban thermal environment were integrated into three primary urban pattern elements: Mountain, City, and Sea. The spatial autocorrelation and heterogeneous impacts of these urban pattern elements on the thermal environment were analyzed using Moran’s I and Geographically Weighted Regression (GWR) modeling, followed by impact-based zoning using K-means clustering algorithms. The results revealed a significant positive correlation between Mountain and City elements and the thermal environment, whereas Sea elements exhibited a notable cooling effect. Furthermore, each factor demonstrated significant spatial heterogeneity. Based on local GWR regression coefficients and spatial variations in factor intensity and directionality, Xiamen was partitioned into four distinct regulatory zones: City-dominated zones, Sea-dominated zones, Mountain–Sea co-dominated zones, and Comprehensive transitional zones influenced by Mountain–City–Sea interactions. Customized, spatially targeted regulatory strategies were subsequently proposed for each zone. This study provides an innovative methodological framework for targeted, region-specific policy interventions to alleviate urban thermal stress under climate change, thereby contributing to the optimization of future urban planning and promoting sustainable and adaptive development.