The modulation of synoptic weather patterns and human activities on the diurnal cycle of the summertime canopy urban heat island in the Yangtze River Delta Urban Agglomeration, China

Abstract. Synoptic weather patterns (SWPs) and human activities are significant driving factors of the canopy urban heat island effect (CUHI), and the CUHI phenomenon exhibits a pronounced diurnal cycle. However, to date, there has been a significant knowledge gap in understanding how the combination of SWPs and human activities modulates the diurnal cycle of CUHI. This study systematically analyzed the diurnal patterns of CUHI intensity (CUHII) in the Yangtze River Delta Urban Agglomeration (YRDUA) by integrating multiple source datasets and utilizing objective classification methods. Among all SWPs, type 2 (dominated by subtropical high pressure) exhibited the strongest CUHII, while type 4 (affected by southwestern moisture and southward cold air) had the lowest. In terms of human activities, landscape percentage (PLAND), largest patch index (LPI), and anthropogenic heat flux (AHF) showed an increasing trend, with higher values in the east and lower values in the west, coinciding with the spatiotemporal patterns of CUHII. These findings collectively confirm the pivotal roles of SWPs and human activities in CUHI phenomena. More importantly, this paper quantifies the contributions of SWPs and human activities to the diurnal cycle of CUHI using a random forest (RF) model. Shapley additive explanation (SHAP) revealed that SWPs had a more pronounced influence on daytime CUHII, whereas human activities dominated nighttime CUHII. During the daytime, the partial dependence plot (PDP) of relative humidity (RH) decreased more pronouncedly, explaining why type 4 contributed more significantly to CUHII during this period. When the wind speed (WS) exceeded a threshold, PDP rapidly increased, which might account for the higher contribution of type 5 to CUHII during this period. During the nighttime, there appeared to be a threshold for PLAND, beyond which its impact on improving CUHII became markedly more prominent. These results indicated that there was a clear diurnal asymmetry in the modulation of CUHI by SWPs and human activities. In conclusion, this study not only provided scientific insight into the complex driving mechanisms of the CUHI diurnal cycle in YRDUA, but also offered a theoretical foundation for evaluating urban overheating issues and developing effective mitigation strategies.