The contributions of urban morphology (UM) and their key thresholds to urban heat island (UHI) intensity during heatwave events lack clarity. In response to this problem, the impacts and thresholds of UM on summer daytime land surface temperature (LST) in megacity, Shanghai, during the heatwave event were quantified by high-resolution remote sensing and GIS data, including 57 2D/3D UM indices associated with building volumes, vegetation volumes and so on, using stepwise multiple linear regression (SMLR), and XGBoost-based SHAP interpretable methods. The results show that increasing the proportion of tree cover within urban functional zones (UFZs) no longer effectively mitigates UHI; instead, reducing the difference in building and vegetation volume is the key cooling factor related to vegetation. The 3D building-vegetation integrated morphology indices, we proposed, show a fine explanatory on LST, and predominantly govern LST variations in UFZs together with the 2D UM indices (29 %-44 %). Threshold effects are also observed in the impacts of UM during the heatwave event, and the proportion and cohesion of built-up generally has a low impact on LST, when their values are lower than 11.9 % and 98.9 % respectively. Based on these findings, we proposed that the reduction of the volume difference between buildings and vegetation depending on ecological construction of tree species with large canopy may serve as an effective approach to mitigate UHI in UFZs during heatwave events. © 2024 Elsevier Ltd