Advantage of Multi-Band Solar-Induced Chlorophyll Fluorescence to Derive Canopy Photosynthesis in a Temperate Forest

Zhunqiao Liu,Xiaoliang Lu,Shuqing An,Mary Heskel,Hualei Yang,Jianwu Tang

DOI: https://doi.org/10.1016/j.agrformet.2019.107691

IF: 6.2

2019-01-01

Agricultural and Forest Meteorology

Abstract:When solar-induced chlorophyll fluorescence (SIF) is developed as a novel approach to quantify gross primary production (GPP), narrow SIF emission spectrum around the atmospheric oxygen absorption windows ((similar to) 761 nm and (similar to)687 nm) has been widely used to represent SIF and to derive GPP. As SIF is a continuous spectrum, deriving the full broadband SIF emission spectrum over 640-850 nm provides an opportunity to fully explore the potential of SIF in estimating GPP. Using high-frequency measurements of canopy carbon flux and SIF emissions at the atmospheric absorption bands, we reconstruct the full SIF spectrum from SIF signals at the absorption bands, and then analyze correlations between the GPP and the selected single SIF bands and their combinations using both linear regression (LR) and Gaussian processes regression (GPR). Our results indicate that (1) the red SIF bands (640-700 nm) shows low correlation with GPP due to the strong (re)absorption of red SIF emissions by leaf chlorophyll; (2) the individual bands in near-infrared area (at 720 nm, 740 nm, and 761 nm) can determine about 60% and 62% of the variance in GPP with hourly scale by LR and GPR, respectively, and the combination of those SIF bands provide increased predictive power, explaining 66% and 76% variance in GPP with hourly scale by LR and GPR, respectively; (3) the solar radiation saturation, fraction of direct solar radiation, air temperature and leaf area index may have negative impacts on the SIF-GPP correlations when they are beyond optimal thresholds; and (4) the temporal aggregation of SIF-GPP (daily scale) enhances the correlations as compared with the hourly scale: the daily combination of SIF bands (at 687 nm, 720 nm, and 761 nm) can account for 80% and 93% of variance in daily daytime GPP by LR and GPR, respectively, suggesting that the combination of these three bands of SIF on the daily scale is the best proxy for GPP. Our results provide a new approach to analyze the SIF-GPP correlations using the ground-based full broadband SIF emission, suggesting that multi-bands SIF has a stronger capacity in predicting plant GPP than traditionally used signal band SIF data.