Large-scale diurnal responses of solar-induced chlorophyll fluorescence (SIF) to varying heat and water stresses: Implications for monitoring SIF variations from satellite observations

Abstract

Satellite observations of solar-induced chlorophyll fluorescence (SIF) offer a promising approach for monitoring plant heat and water stresses across spatial scales. Most studies have focused on seasonal responses of satellite-observed SIF and its physiological component, fluorescence efficiency (${\Phi }_f$), to heat and water stresses. However, their diurnal responses remain poorly understood. Besides, polar-orbiting satellites typically use a daily correction factor to upscale fixed-time SIF observations into daily averages (${\mathrm{SIF}}_{\text{daily}}$). Given the diurnal SIF variations, the reliability of this approach under stress conditions is uncertain. In this study, we used the ratio of SIF to near-infrared radiance of vegetation (NIRvR) as a a linear approximation of ${\Phi }_f$. Then, we applied the eXtreme Gradient Boosting (XGBoost) algorithm to model hourly observations of summer (June–August) SIF and ${\Phi }_f$ from Orbiting Carbon Observatory-3 (OCO-3) data. Our modeling was constrained to mainland China from 2019 to 2022. We calculated anomalies in SIF and ${\Phi }_f$ at different times of the day and conducted linear regressions on them with daily air temperature or soil moisture anomalies. Our results showed that the responses of SIF and ${\Phi }_f$ to stresses varied significantly with time of day and vegetation type. On days experiencing high water and heat stresses, morning and afternoon SIF exhibited weaker declines than midday. In contrast, morning ${\Phi }_f$ generally exhibited stronger increases than midday, whereas afternoon ${\Phi }_f$ showed weaker increases. Such diurnal differences in SIF and ${\Phi }_f$ responses were more pronounced in forests than in grasslands and intensified with rising water and heat stress levels. Additionally, we found that morning polar-orbiting satellite SIF observations tended to overestimate ${\mathrm{SIF}}_{\text{daily}}$ changes, whereas midday observations tended to underestimate them. These biases also intensified with rising stress levels. Our findings emphasize the importance of diurnal satellite SIF observations in deepening our understanding of plant responses to water and heat stress, as well as in improving the monitoring of plant stresses.