Non-uniform climatic responses of land surface phenology derived from optical-, fluorescence-, and microwave-based satellite observations

Abstract

Satellite remote sensing has greatly advanced the study of land surface phenology, providing crucial insights into large-scale vegetation dynamics in the context of climate change. Multi-sensor satellite-derived vegetation proxies, such as the normalized difference vegetation index (NDVI) in optical mode, solar-induced chlorophyll fluorescence (SIF) in fluorescence mode, and vegetation optical depth (VOD) in microwave mode, are effective indicators of vegetation dynamics in greenness, biomass, and productivity. However, a comprehensive understanding of phenological patterns derived from these proxies and their climatic responses remains limited. In this study, spanning 1988–2021, we analyzed spatio-temporal patterns and climatic responses of spring green-up dates (GUDs) and autumn dates of foliar senescence (DFSs) across the three vegetation proxies. We found broadly consistent trends of advancing GUDs (−2.4 to −1.1 days decade⁻¹) across proxies, while DFSs exhibited divergent temporal patterns: VOD-derived DFSs showed significant delays (+1.1 days decade⁻¹), whereas NDVI- and SIF-derived DFSs displayed no clear trends. Rising temperatures were the primary driver of earlier GUDs, while precipitation and insolation had limited influence. In autumn, warming delayed VOD-based DFSs but had minimal effects on NDVI- and SIF-based estimates. Spatial attribution analyses indicated that both biotic and abiotic factors contributed to the spatial variability in climatic responses, with species richness and tree density modulating temperature effects. These findings highlight the importance of proxy selection in phenological studies and underscore the need for integrated investigations into the biophysical mechanisms driving non-uniform climatic responses across vegetation proxies.