Deriving photosystem-level red chlorophyll fluorescence emission by combining leaf chlorophyll content and canopy far-red solar-induced fluorescence: Possibilities and challenges

Abstract

Solar-induced chlorophyll fluorescence (SIF) emitted from photosystem I (PSI) and photosystem II (PSII) is characterized by two peaks centered in the red and far-red spectral regions. SIF provides a unique remotely sensible signal to track plant photosynthetic dynamics. Compared with far-red SIF, red SIF (RSIF) is more strongly linked to PSII and thus with plant photosynthetic activity, but is subject to stronger reabsorption within leaves and canopies. This hinders the understanding and use of canopy RSIF observations (RSIFobs), which is only a small fraction of the total RSIF emitted by the photosystems (RSIFtotal). Deriving RSIFtotal from RSIFobs is still challenging due to retrieval uncertainty, limited availability of RSIFobs and spectral overlap with chlorophyll absorption. To address the challenges associated with deriving RSIFtotal, we propose an exploratory method framework that combines canopy far-red SIF observations (FRSIFobs) and leaf chlorophyll content (LCC) to derive RSIFtotal. We first downscale FRSIFobs from canopy to leaf, and then leverage LCC information to estimate RSIF at the leaf level. Finally, we incorporate LCC information in the subsequent downscaling of RSIF from leaf to photosystem. To evaluate our approach, we use ground-based observation data in three crop types (rice, wheat, and maize) and SCOPE model simulations. Our results demonstrate that the seasonal patterns of RSIFtotal show a close agreement with the seasonal patterns of gross primary production (GPP) and absorbed photosynthetic active radiation (APAR). More importantly, RSIFtotal slightly outperforms FRSIFobs in estimating GPP for the three crop types. Our study has also revealed a strong linear relationship between the escape probability of RSIFtotal (fesc_R) and the RSIFobs/FRSIFobs ratio affected by LCC. The simplicity and robustness of our approach, along with its potential application in satellite remote sensing, will contribute to the improvement of large-scale GPP estimation and photosynthetic phenology detection. Moreover, our investigation of fesc_R will contribute to a better understanding the physiological and non-physiological dynamics of RSIFobs.