Abstract
Against the backdrop of global warming, urban heat island (UHI) are driven by both physical land use change and dynamic functional intensity, yet their independent and interactive effects remain poorly understood. This study deconstructs these dual drivers in Beijing. First, we employed the Logistic-Harmonic model and random forest classifier to conduct fine classification of land use changes, and constructed the Urban Function Intensity Index (IDI) based on nighttime light and NDVI time series, analyzing their individual and combined effects on land surface temperature change (ΔLST). Results reveal distinct spatial patterns: the core area is dominated by intensification, while the suburbs are the primary locus of expansion. The thermal response to IDI is spatially and seasonally heterogeneous, with the suburban areas being most sensitive and winter responses being strongest. In suburban areas, high IDI was +0.27°C higher than low IDI in winter, whereas low IDI was -0.24°C higher than high IDI in summer. Critically, the interaction between IDI and LCintensity reveals unique synergistic warming pathways, such as a ‘low-IDI + high-intensification’ hot spot in the core area driving ΔLST to 0.74°C, and unexpected warming of 0.64°C in suburban areas experiencing ‘medium-low IDI + de-urbanization’ degradation. Collectively, these findings reveal an evolutionary trajectory of thermal environmental responses across the city’s development gradient, providing a new perspective on the "UHI life cycle" in megacities. This study offers a scientific basis for developing targeted, zone-specific thermal regulation strategies.