In the context of accelerating global warming, which profoundly affects ecosystems and human societies, it is crucial to understand the complex mechanisms through which climate, urban form, and vegetation greening influence land surface temperature (LST). This study investigates the distribution of LST across five urban agglomerations within the Yangtze River Economic Belt (YREB) of China, analyzing both seasonal (summer-winter) and diurnal variations. A range of modeling techniques, including linear regression, Random Forest (RF), and Shapley's Additive interpretation (SHAP), are employed to reveal the complex relationships between LST and its driving factors. The results indicate significant seasonal and diurnal differences in LST trends and their determinants among the urban agglomerations. Notably, during summer daytime, LST exhibits a pronounced increase along the urban-rural gradient in most agglomerations (with an average gradient of 0.046 ± 0.0017), where the Yangtze River Delta urban agglomeration (YRDUA) experiences the most substantial rise, characterized by a gradient of 0.059, with the Enhanced Vegetation Index (EVI) playing a critical role during this period. Conversely, in winter, Temperature (TMP) emerges as the primary determinant of both daytime and nighttime LST, while Precipitation (PRE) significantly affects nighttime LST, particularly in the Chengdu-Chongqing urban Agglomeration (CCUA) and YRDUA, with importance scores of 0.59 and 0.69, respectively. In comparison, urban form indicators demonstrate a less direct impact on LST within these urban agglomerations. These findings offer valuable guidance for future climate mitigation strategies, scientific urban planning, and natural resources management in the YREB.