Solar-induced fluorescence-based phenology of subtropical forests in China and its response to climate factors

Vegetation phenology refers to the cyclical changes of plant development in different seasons of the year, and is the result of the interaction between environmental factors and biological internal regulatory mechanisms. Current large-scale phenology studies mainly use vegetation indices for remote sensing retrieval, while the greenness information of subtropical forests is inconspicuous, causing regular vegetation indices to retrieve their phenology ineffectively. SIF (Solar-induced chlorophyll fluorescence) is strongly coupled with GPP (Gross Primary Productivity), and has great potential in vegetation phenology research. In this study, we used m-SG (modified Savitzky-Golay) filter to reconstruct OCO-2 (Orbiting Carbon Observatory 2) GOSIF time-series, and used D-L (Double-Logistic) curve fitting to retrieve subtropical forests phenology in China from 2001 to 2018. Then we used phenology retrieved from GPP products and flux GPP as well as phenology observation data to validate the GOSIF phenology. Finally, we analyzed the spatiotemporal distribution pattern of subtropical forests phenology and the influence of climate factors on it. The results showed that: (1) GOSIF can retrieve subtropical forests phenology, and the R2 of SOS (Start of Growth Season), EOS (End of Growth Season) and LOS (Length of Growth Season) based on GOSIF and GPP products reached 0.50, 0.68, 0.59, respectively, while the R2 of SOS, EOS, and LOS based on GOSIF and flux GPP as well as phenology observation data were 0.20, 0.43, 0.56, respectively. (2) The spatial distribution of subtropical forests phenology showed that SOS gradually delayed, EOS gradually advanced and LOS gradually shortened from southeast to northwest. Moreover, different types of forests showed different trends of advanced SOS, delayed EOS and prolonged LOS over time. (3) The annual average minimum temperature was the main climate factor affecting the subtropical forests phenology while the impact of annual total precipitation also cannot be ignored. For every 1 °C increase in the annual average minimum temperature, the overall SOS advanced by 2.94 days, EOS delayed by 2.14 days and LOS prolonged by 4.91 days.