The relationship between long-term reconstructed solar-induced chlorophyll fluorescence and tower-derived transpiration across diverse ecosystems in China

Satellite-based solar-induced chlorophyll fluorescence (SIF) provides novel capabilities for global vegetation monitoring, with most previous studies focusing on its relationship with gross primary productivity (GPP). Although photosynthesis and transpiration (T) are closely linked through stomatal regulation, the covariation between SIF and T remains insufficiently understood. Here, we conduct a multi-timescale assessment of the SIF-T statistical relationship across six vegetation types using three long-term reconstructed SIF datasets, including contiguous SIF (CSIF), reconstructed TROPOMI SIF (RTSIF), and China high spatial resolution SIF (HCSIF), together with tower-derived transpiration observations from 46 eddy covariance sites across China. Our findings showed positive relationships between SIF and tower-derived T across the six vegetation types based on the pooled observations at the 8-day timescale, with R 2 values ranging from 0.52 to 0.78 for CSIF-T, 0.49 to 0.76 for RTSIF-T, and 0.36 to 0.70 for HCSIF-T. Temporal aggregation from 8-day to monthly resolution generally increased the explained variance of the SIF-T relationship across most ecosystems. The slopes of the linear regression (SLR) of CSIF-T and RTSIF-T were generally comparable across vegetation types, with no significant differences in most pairwise comparisons. The strength of SIF-T linear relationship in C3 crops was not superior to that in C4 crops, but the regression slopes were generally higher than those in C4 crops. Environmental conditions further affected the SIF-T relationship. Vapor pressure deficit showed a potential nonlinear influence, with weaker SIF-T relationships observed under extremely humid or dry atmospheric conditions in grasslands. The SIF-T relationship also varied along soil water content gradients across vegetation types, with stronger relationships observed in grasslands under moist soil conditions. Based on reconstructed SIF, this study clarifies the SIF-T relationship in various ecosystems, offering a critical reference to improve regional and global vegetation transpiration models.