Saturation effect of background temperature and aridity on vegetation phenology sensitivity to urban warming

Abstract. In this study, urban warming effects on vegetation phenology were assessed for 293 cities in China. The variations in urban warming effects were expected to be attributed to the baseline land surface temperature (LST) and the aridity index (AI) of each locale. LST and AI related phenology and their temperature sensitivity (R_t-SOS and R_t-EOS) was quantified. We observed an urban-rural phenological disparity of 12.06 days for Start of Season (ΔSOS) and 9.86 days for End of Season (ΔEOS) among the studied cities. Spatially, cities in high latitude regions and coastal areas exhibited pronounced negative ΔSOS shifts and positive ΔEOS shifts, positively correlating with R_t-SOS and R_t-EOS, respectively. Employing a continent-wide preseason temperature (T), we observed a logistic decrease for SOS and an increase for EOS, illustrating the "saturated effect" of warming on plant phenology-patterns echoed within urban settings. First-order derivatives of those logistic curves identified a highest phenological sensitivity at T = 4 °C and T = 6 °C, as well as the warming benefit range of −3.5 °C–10 °C and 2 °C–14 °C for SOS and EOS respectively. Substituting T with LST, weaker ΔSOS and ΔEOS would be presented in warmer regions only when LST exceeded 12.5 °C and 4 °C for spring and autumn, respectively. Except for LST, AI exhibited a positive correlation with ΔSOS and ΔR_t-SOS, but a negative one with ΔEOS and ΔR_t-EOS. Collectively, LST and AI explained 75.05 % and 76.21 % of the phenological variance across the continent for ΔSOS and ΔEOS, respectively. These findings lay the groundwork for predicting vegetation changes under global warming at large scales.