Abstract

Seasonal delineation represents a critical interface between the natural environment and human activities. However, the conventional climatetemperature (C–T) method, which relies solely on thermal thresholds, has limited applicability in regions with complex climatic regimes. In this study, we develop and apply a composite seasonal index (CSI) integrating temperature and precipitation, using meteorological observations from 298 stations across mainland China during 1980–2020, with CSI calculation based on 278 stations that had valid paired temperature and precipitation records, to investigate spatial patterns of seasonal variability. The results show that incorporating precipitation improves the identification of regional heterogeneity in seasonal dynamics. In northeastern and northwestern China, spring rainfall advances spring onset, while autumn rainfall delays autumn termination, producing a CSI–defined spring duration 1–2 months longer than that derived from the C–T method and an autumn duration about one month longer. In some arid regions, concentrated precipitation prolongs summer by approximately 1–2 months. An independent comparison with land surface phenology metrics during 1982–2018 further shows that CSI–derived seasonal transition dates are broadly consistent with the spatial patterns of SOS, maturity, senescence, and EOS, especially in monsoonal and hydrothermally complex regions. Differences between the CSI and C–T methods are generally small (approximately ±1 month) where precipitation and temperature vary synchronously, but increase to approximately ±2 months where precipitation exerts stronger control. Overall, the CSI preserves the structure of the traditional C–T classification while accounting for hydrological influences, thereby enhancing seasonal delineation in climatically Eastern Monsoon Region and improving the ecological interpretability of hydrothermal seasonality assessment.

Keywords