Optimal inverse estimation of ecosystem parameters from observations of carbon and energy fluxes

Canopy structural and leaf photosynthesis parameterizations such as maximum carboxylation capacity (V-cmax), slope of the Ball-Berry stomatal conductance model (BBslope) and leaf area index (LAI) are crucial for modeling plant physiological processes and canopy radiative transfer. These parameters are large sources of uncertainty in predictions of carbon and water fluxes. In this study, we develop an optimal moving window nonlinear Bayesian inversion framework to use the Soil Canopy Observation Photochemistry and Energy fluxes (SCOPE) model for constraining V-cmax, BBslope and LAI with observations of coupled carbon and energy fluxes and spectral reflectance from satellites. We adapted SCOPE to follow the biochemical implementation of the Community Land Model and applied the inversion framework for parameter retrievals of plant species that have both the C-3 and C-4 photosynthetic pathways across three ecosystems. We present comparative analysis of parameter retrievals using observations of (i) gross primary productivity (GPP) and latent energy (LE) fluxes and (ii) improvement in results when using flux observations along with reflectance. Our results demonstrate the applicability of the approach in terms of capturing the seasonal variability and posterior error reduction (40 %-90 %) of key ecosystem parameters. The optimized parameters capture the diurnal and seasonal variability in the GPP and LE fluxes well when compared to flux tower observations (0.95 > R-2 > 0.79). This study thus demonstrates the feasibility of parameter inversions using SCOPE, which can be easily adapted to incorporate additional data sources such as spectrally resolved reflectance and fluorescence and thermal emissions.