Improving the Capability of the SCOPE Model for Simulating Solar-Induced Fluorescence and Gross Primary Production Using Data from OCO-2 and Flux Towers

Abstract

Solar-induced chlorophyll fluorescence (SIF) measured from space has shed light on the diagnosis of gross primary production (GPP) and has emerged as a promising way to quantify plant photosynthesis. The SCOPE model can explicitly simulate SIF and GPP, while the uncertainty in key model parameters can lead to significant uncertainty in simulations. Previous work has constrained uncertain parameters in the SCOPE model using coarse-resolution SIF observations from satellites, while few studies have used finer resolution SIF measured from the Orbiting Carbon Observatory-2 (OCO-2) to improve the model. Here, we identified the sensitive parameters to SIF and GPP estimation, and improved the performance of SCOPE in simulating SIF and GPP for temperate forests by constraining the physiological parameters relating to SIF and GPP by combining satellite-based SIF measurements (e.g., OCO-2) with flux tower GPP data. Our study showed that SIF had weak capability in constraining maximum carboxylation capacity (Vcmax), while GPP could constrain this parameter well. The OCO-2 SIF data constrained fluorescence quantum efficiency (fqe) well and improved the performance of SCOPE in SIF simulation. However, the use of the OCO-2 SIF alone cannot significantly improve the GPP simulation. The use of both satellite SIF and flux tower GPP data as constraints improved the performance of the model for simulating SIF and GPP simultaneously. This analysis is useful for improving the capability of the SCOPE model, understanding the relationships between GPP and SIF, and improving the estimation of both SIIF and GPP by incorporating satellite SIF products and flux tower data.

The Relationship of Gross Primary Productivity with NDVI Rather than Solar-Induced Chlorophyll Fluorescence Is Weakened under the Stress of Drought

“…Some researchers have found that GOSAT SIF and GPP were strongly correlated [26,28]. Other studies utilizing SIF have discovered a consistently significant relationship of GPP with SIF in terrestrial ecosystems across different temporal scales [29,30]. SIF has also demonstrated good performance in estimating GPP under drought conditions.…”

Improving E3SM Land Model Photosynthesis Parameterization via Satellite SIF, Machine Learning, and Surrogate Modeling

“…The optimized parameters greatly improved the ELM's prediction accuracy and efficiency of major carbon fluxes. Others have also used surrogate models in simulating SIF and GPP (e.g., Verrelst et al., 2017; Wang & Xiao, 2021).…”

“…B. Fisher et al., 2014; Houborg et al., 2013; R. Ma et al., 2017). In contrast to reflectance‐based satellite GPP products, recent emergence of solar‐induced chlorophyll fluorescence (SIF) from satellite passive measurements provides direct information relating to the photosynthesis process (Chen, Mao, Ricciuto, Xiao, et al., 2021; Li et al., 2018; Zhang et al., 2018), and thus offers a new empirical tool for benchmarking GPP simulations (Damm et al., 2010; Guanter et al., 2014; Wang & Xiao, 2021). In particular, many studies have demonstrated an empirical linear relationship between SIF and GPP (e.g., Li et al., 2018), which is also easy to interpret.…”

Assessment of Six Machine Learning Methods for Predicting Gross Primary Productivity in Grassland

“…Among them, BGCM estimates GPP by simulating vegetation physiological processes [23,24]. BGCM requires a lot of knowledge about the vegetation growth process, and the establishment of the model is accompanied by a lot of mechanism simplification [25,26].…”