Correcting Clear‐Sky Bias in Gross Primary Production Modeling From Satellite Solar‐Induced Chlorophyll Fluorescence Data

Abstract

Satellite solar-induced chlorophyll fluorescence (SIF) has been demonstrated the potential to monitor photosynthesis, quantified as gross primary production (GPP), for large areas. However, satellite SIF retrievals are only reliable for clear-sky conditions, creating overestimation in estimating GPP for all-sky conditions, called clear-sky bias. Clouds can reduce the total radiation and increase the diffuse radiation, which can have counteracting effects on photosynthesis since plants are more efficient in using diffuse radiation. To mitigate this uncertainty in estimating global GPP from satellite SIF, we propose a clear-sky bias correction approach that considers both the increase in light use efficiency (LUE) and the decrease in photosynthetically active radiation (PAR) under cloudy conditions. An enhancement factor (α) of LUE under cloudy conditions is first parameterized from FLUXNET data set, which is found to converge to 1.21 among all vegetation types. Then, a correction factor of clear-sky bias is developed to account for the variations of LUE under cloudy conditions, which is used to estimate global terrestrial GPP in 2019 with spaceborne SIF retrievals from TROPOspheric Monitoring Instrument (TROPOMI). We estimate global annual terrestrial GPP of 135.47 ± 3.36 and 142.03 ± 3.53 PgC/year, respectively, with and without correcting clear-sky bias. This approach is also able to identify regions with large uncertainties in GPP estimation from satellite SIF data due to the effects of clear-sky bias, such as eastern America, western Europe, and southern China. Our results highlight the importance of considering clear-sky bias in estimating terrestrial photosynthesis from satellite SIF data. Plain Language Summary Photosynthesis is an important process by which plants remove the atmosphere of carbon dioxide (CO 2) and fuel the life on Earth. Satellite solar-induced chlorophyll fluorescence (SIF) shows great potential in monitoring terrestrial photosynthesis. However, satellite SIF retrievals are only reliable for clear-sky days, creating biases in estimating photosynthesis for all-sky condition. Cloud can reduce the total radiation and increase the diffused radiation, which can have counteracting effects on photosynthesis since plants are more efficient in using diffused radiation. Under slightly cloudy conditions, the decrease and increase from these two effects are approximately equal. However, under extremely cloudy conditions, the decreased radiation outweighs the efficiency enhancement, leading to an overall negative effect on photosynthesis. In this study, we propose a practical approach to mitigate this uncertainty in estimating global photosynthesis from SIF. Using the TROPOspheric Monitoring Instrument (TROPOMI) SIF during the year of 2019, we estimate a global photosynthesis for all sky conditions to be 135.47 ± 3.36 PgC/year. We also recognize three regions including eastern America, western Europe, and southern China, where clear-sky bias is clearly corrected with our approach. Mitigating clear-sky...

Representation of Leaf‐to‐Canopy Radiative Transfer Processes Improves Simulation of Far‐Red Solar‐Induced Chlorophyll Fluorescence in the Community Land Model Version 5

“…The conversion factor may also vary spatially and temporally due to variations of canopy structure and leaf biochemical and biophysical properties. The illumination conditions for CLM simulations and satellite observations are also not identical, as all satellite SIF products applied cloud filtering while no filter was applied to CLM simulations based on illumination conditions (Zhang et al., 2020 ). This can contribute to discrepancies between simulated and observed SIF and lead to clear‐sky bias when linking satellite SIF observation with GPP.…”

“…Filtering simulations based on cloud condition may improve the comparison between simulations and observations and the assimilation of observed SIF. And, the clear‐sky bias can be mitigated by applying corrections to SIF simulations (Hu et al., 2021 ; Zhang et al., 2020 ). The spatial resolutions are also different between the satellite SIF products and CLM simulations, and may affect the simulation‐observation comparisons.…”

Progress of studies on satellite-based terrestrial vegetation production models in China

“…Zhang et al (2021) explained that the hotspot effect in remote sensing observed SIF is important for GPP estimation. Zhang et al (2020) used tower-based data to correct all clear-sky biases for GPP estimation by SIF. Overall, there are some specific conditions in which SIF cannot track GPP changes, so further investigation of the SIF-GPP pattern is vital for global-scale GPP estimation.…”

Remote Sensing of Ecosystem Water Use Efficiency: A Review of Direct and Indirect Estimation Methods

“…Global SIF products have also been retrieved with several time and space resolutions from spaceborne spectrometers, such as the Orbiting Carbon Observatory2 (OCO-2), the TROPOspheric Monitoring Instrument (TROPOMI), the Global Ozone Monitoring Experiment 2 (GOME2), and the Greenhouse Gases Observing Satellite (GOSAT) [78][79][80]. A previous study has developed an algorithm of GPP estimation using SIF products, which have global coverage and can be used in all weathers [81]. Studies have also shown the empirical and mechanistic relationship between T and SIF to improve the T estimation [82][83][84].…”