Response of Solar-Induced Chlorophyll Fluorescence to Gross Primary Productivity, Evapotranspiration, and Transpiration Under Aridity Gradients of China

The intrinsic relationship between photosynthesis and solar-induced chlorophyll fluorescence (SIF) presents a valuable opportunity for estimating vegetation gross primary productivity (GPP), evapotranspiration (ET), and transpiration (Ec). However, characteristics of the relationship between SIF-GPP, SIF-ET, and SIF-Ec under different regional climatic conditions, such as in various climatic zones, land cover types, and aridity gradients, remain unclear. Thus, the GOSIF and PML_V2 (China) datasets including GPP, ET, and Ec were employed in the study. We first explored the correlations between SIF-GPP, SIF-ET, and SIF-Ec on annual and monthly time scales across different aridity index (AI). Then, the influence of climatic conditions on correlation coefficients (CCs) between SIF-GPP, SIF-ET, and SIF-Ec were investigated. Finally, the influences of environmental factors on their relationship were further examined. Results showed that SIF, GPP, ET, and Ec all increase as AI increases, with GPP exhibited the most similar variation trends to SIF in both spatial and temporal scales. For the annual time scale, the CC between SIF-GPP, SIF-Ec, and SIF-ET presented similar spatial distribution. Peak correlation values were all occurred in semiarid region (AI is around 0.4). The relationship between SIF-GPP in different climatic zones is generally linear and demonstrated a significant correlation. For monthly time scale, the positive and high CC between SIF-GPP, SIF-Ec, and SIF-ET exhibited obvious “south-north-south” swing trend with increasing AI, with the characteristics of SIF-Ec most consistent with that of SIF-GPP. The relationships between SIF-GPP and SIF-Ec are linear, but that between SIF-ET is polynomial nonlinear. We also argued that environmental factors should be considered when using SIF to estimate GPP, ET, and Ec. Results of this study will provide a new reference and possibility for estimating vegetation GPP, Ec, and ET on the monthly and annual scales from large regional scales.