Heat requirement, not warming, governs the urban-rural disparity in spring phenology advancement.

Urbanization has profoundly altered local climates and vegetation structures, resulting in pronounced differences in spring phenology between urban and rural areas. However, the mechanistic drivers of this urban-rural divergence, specifically regarding the rates of phenological advancement, remain poorly understood. Using satellite-derived phenology observations across 354 cities in China from 2001 to 2020, we show that the advancement of the spring green-up date (GUD) was significantly more rapid in urban areas than in their rural surroundings under continuous warming. This urban-rural divergence was observed in 47.7% of the studied cities, where urban advancement trends outpaced rural ones by an average of 0.37 ± 0.24 days/year, particularly in temperate warm-summer regions. By integrating chilling accumulation and thermal requirement models with interpretable machine learning, we identify accumulated heat demand as the primary driver of this divergence. Partial correlation analysis reveals that thermal requirements dominate GUD shifts, explaining 74% of the variance in urban areas and surpassing the influence of chilling accumulation (51%). Moreover, we find that vegetation structure modulates these shifts, as increased species richness amplifies urban GUD trends while tree density buffers rural shifts. Notably, current empirical models largely failed to reproduce both the spatial patterns and the magnitude of the urban-rural divergence in GUD trends. Our results redefine the hierarchy of controls on spring phenology in human-dominated landscapes, with critical implications for predicting ecosystem responses under accelerating urbanization.