Understanding the spatiotemporal dynamics of land surface temperature (LST) and its drivers is essential for mitigating urban heat and guiding climate-responsive planning. In this study, we integrate multitemporal remote sensing data derived from satellite sensors with multiresolution image segmentation and spatial regression models to investigate seasonal LST patterns and their influencing factors. Seasonal variability is assessed across the year. Analyses are conducted on spatial units consistent across time, and ordinary least squares (OLS) and geographically weighted regression (GWR) models are applied to identify both global and local driving mechanisms. In addition, Getis-Ord Gi* hotspot analysis is employed to delineate core high- and low-temperature zones for focused analysis. Results show pronounced seasonal and spatial heterogeneity. Forest cover and elevation consistently cool LST, whereas urban form effects vary by season and location. GWR reveals strong spatial non-stationarity, especially within thermal core zones. The findings highlight the importance of incorporating both spatial heterogeneity and seasonal dynamics in urban heat analysis. The results of this study provide a spatially explicit understanding of LST drivers and offer practical insights for climate-adaptive land use and urban design strategies.