Abstract
Sustainable urban development requires a comprehensive understanding of land surface temperature (LST) and the factors that influence it, especially in rapidly urbanizing areas. This study introduces an approach that integrates Multi-Criteria Evaluation (MCE) with Geographically Weighted Regression (GWR) to enhance green space (GS) planning in Bojnourd City, Iran. By merging these techniques, the research identifies critical areas where high suitability for green space development coincides with LST hotspots, offering a robust tool for urban planners. LST mapping was conducted using the radiative transfer equation (RTE) method with Landsat 8 data. GS suitability was evaluated through an MCE framework incorporating Weighted Linear Combination (WLC), fuzzy set analysis, the Analytic Hierarchy Process (AHP), and Zonal Land Suitability (ZLS) methods. The GWR model demonstrated superior predictive accuracy over the Ordinary Least Squares (OLS) model, evidenced by a higher R² and lower AIC. This combined analysis identified 255 hectares within the city and 4,393.9 hectares in the expansion zones as prime locations for urban green spaces (UGS) and peri-urban agriculture and forestry (P-UGS) development. Overall, the study's approach demonstrates a valuable framework for advancing green space management and enhancing climate change adaptation strategies.