While the role of green infrastructure in mitigating urban heat is acknowledged, the influence of its spatial configuration on thermal performance in arid climates has been less explored. This study quantitatively examines the relationship between the spatial structure of green infrastructure and land surface temperature in the historic desert city of Yazd. Using landscape ecology metrics and Sentinel-2 and Landsat 9 satellite imagery from summer 2024, green spaces were mapped and the Green Space Heat Mitigation Index was calculated. Spatial metrics at the patch, class, and landscape levels were extracted with FRAGSTATS software, and their association with cooling performance was analyzed via multivariate regression.
The results indicate that Yazd's green infrastructure has a highly fragmented and dispersed pattern, consisting of small, isolated patches and lacking large, contiguous green areas. Cooler patches exhibited a higher clumpiness index and lower edge density. Regression analysis revealed that approximately 68% of the variation in the heat mitigation index is explained by three spatial metrics: clumpiness, largest patch index, and edge density. This finding confirms that in arid environments, the spatial organization of green infrastructure has a greater influence on cooling efficiency than its total area does. Designing integrated, compact green spaces with simplified boundaries can optimize thermal performance and reduce the heat island effect. The presented methodology provides a generalizable framework for assessing structure–function relationships in urban landscapes, with practical implications for sustainable design in arid-region cities.