Estimating and Tracking the Exchange of Carbon and Water between the Vegetation and the Atmosphere using a Model-Data approach

Plant water-use efficiency, i.e., the carbon gained through photosynthesis per unit of water lost through transpiration (WUE hereafter) is a key tracer of the plant physiological controls on the exchange of carbon and water between terrestrial ecosystems and the atmosphere. Changes in atmospheric CO₂ concentrations and climate variations both affect the stomatal regulation of leaf gas exchange, and thus ecosystem photosynthesis and transpiration rates, thereby modulating WUE. Yet, current vegetation models continue to differ greatly in their representations of terrestrial carbon uptake and water loss, leading to major uncertainties in the predictions of present and future CO₂ uptake, vegetation cover and WUE. Long-term values of WUE reconstructed from different types of observations representing processes from the leaf to the ecosystem scales can be used for evaluating and improving vegetation models. Here, we will examine and evaluate long-term simulations of WUE changes (and associated variables) from vegetation models of varying complexity against in-situ observations that are rarely studied together (i.e., eddy-covariance flux measurements, stable carbon isotopes measured in leaves and in tree rings). In particular, we will compare historical simulations from CMIP6 models and from a simple model based on optimality principles with available observations at the site level. Our aim is to highlight the impact of using either of the vegetation models that incorporate different assumptions about stomatal conductance, photosynthesis and scaling on the simulated WUE, and to estimate the relative contributions from the environmental drivers to changes in WUE across scales.