Optimizing seasonally variable photosynthetic parameters based on joint carbon and water flux constraints

Abstract

Terrestrial biosphere models (TBMs) often adopt the Farquhar biochemical model coupled with the Ball-Berry stomatal conductance (\({g}_{s}\)) model to simulate ecosystem carbon and water fluxes. The parameters \(m\), representing the sensitivity of \({g}_{s}\) to the photosynthetic rate, and \({V}_{cmax}^{25}\), representing the leaf photosynthetic capacity, are two pivotal parameters but the two main sources of uncertainties in TBM simulations. The spatial and temporal variations of \(m\) in TBMs are still elusive, due to the lack of direct observations. It also remains unclear how accurate estimates of \(m\) and \({V}_{cmax}^{25}\) can improve the simulations of carbon and water fluxes. In this study, we used a Bayesian parameter optimization approach to infer seasonally varying \(m\) and \({V}_{cmax}^{25}\) from eddy covariance observations in a mixed forest stand at the Borden Forest Research Station located in southern Ontario, Canada, and used in-situ observations of \(m\) and \({V}_{cmax}^{25}\) for validation. Three strategies were tested for optimizing \(m\) and \({V}_{cmax}^{25}\), including the carbon, water, and carbon-water coupling scenarios. \(m\) and \({V}_{cmax}^{25}\) optimized from carbon-water coupling constraints shows best correlations with the measured \(m\) (R² = 0.70) and \({V}_{cmax}^{25}\) (R² = 0.70). By incorporating optimized \(m\) and \({V}_{cmax}^{25}\)with seasonal variations, we found considerable improvements in the estimated gross primary productivity (GPP) and evapotranspiration (ET) compared with constant \(m\) and \({V}_{cmax}^{25}\), with R² increasing from 0.78 to 0.85 for GPP, from 0.65 to 0.71 for ET and RMSE reducing from 2.579 g C m^− 2 d^− 1 to 2.038 g C m^− 2 d^− 1 for GPP, from 1.151 mm d^− 1 to 0.137 mm d^− 1 for ET. This study proposes an effective approach to retrieve \(m\) and \({V}_{cmax}^{25}\) for TBMs and demonstrates the efficacy of incorporating seasonally variable \(m\) and \({V}_{cmax}^{25}\) for reducing the uncertainties in GPP and ET simulations, which supports accurate quantifications of land-atmosphere exchanges.