<p>The influence of urban landscape patterns on land surface temperature and urban heat island intensity is well recognized, yet the distinct roles of landscape composition and configuration across thermal zones remain insufficiently quantified. This relationship was quantitatively investigated in Gorgan City, Iran as a humid subtropical city to identify landscape characteristics contributing to cooling or warming effects. From summer 2022 Landsat 8 imagery, a land surface temperature map was generated using the improved Mono-Window Algorithm and classified into six thermal zones (Jenks natural breaks). A nine-class landuse/land cover map was produced via supervised/unsupervised classification. After calculating landscape metrics per zone, analysis of the relationship between LST and landscape patterns was performed. Forests dominated the coolest zone (Zone 1: 37.11% cover, Aggregation Index = 97.48), demonstrating the strongest cooling effect. Built-up areas (Zone 5: Aggregation Index = 83.66) and croplands (Zone 6: Aggregation Index = 85.57, Area_Mean = 6.05) were both associated with higher land surface temperature. Highest landscape heterogeneity (Landscape Shape Index) was found in Zones 4 and 5. Croplands showed dual aggregation peaks (Zone 2: Aggregation Index = 85.16; Zone 6: Aggregation Index = 85.57), suggesting thermal effects vary by crop type. It was concluded that dense, continuous forests are effective natural coolers, while fragmented built-up areas and open fallow croplands intensify surface temperatures. Strategic planning of peri-urban croplands with year-round vegetation and the expansion of compact green spaces is essential for mitigating urban heat islands and enhancing thermal comfort.</p>