B089-07
Towards improved simulations of carbon and water fluxes: evaluating competing stomatal optimisation hypotheses

Monday, 14 December 2020: 19:24
Virtual
Manon Sabot1, Martin Gerard De Kauwe1, Belinda Medlyn2, Andy Pitman1, Brendan Choat2, David Ellsworth2, Nicolas K. Martin-StPaul3 and Jean-Marc Limousin4, (1)ARC Centre of Excellence for Climate Extremes, University of New South Wales, Sydney, NSW, Australia, (2)Hawkesbury Institute for the Environment, Western Sydney University, Richmond, NSW, Australia, (3)Ecologie des Forêts Méditerranéennes, INRA Domaine St Paul, Avignon, France, (4)CEFE Centre d'Ecologie Fonctionnelle et Evolutive, Montpellier, France
Abstract:
Novel formulations of canopy-level gas-exchange regulation, that optimise stomatal conductance with respect to photosynthetic and hydraulic functions, have emerged as avenues to improve land-surface model (LSM) simulations of carbon and water fluxes. In the past five years, at least ten novel alternative formulations for coupling carbon-water fluxes through leaves have been proposed, but independent and systematic evaluations of how they impact predictions of these surface fluxes are lacking. This motivates an inter-model comparison to understand how different mechanistic assumptions across optimal models affect plant behaviour. Here, we embed eight competing optimality approaches within a single LSM framework and ask how, and under what conditions, their predicted behaviours differ from one another and from three empirical stomatal conductance models. Model behaviour was evaluated against leaf-level photosynthetic and hydraulic field/experimental data of species spanning a mean annual precipitation range of 360 - 1620 mm, and originating from Australia, France, and the US. We also conducted a series of idealised model experiments to isolate model-specific surface flux responses and sensitivities to (i) different rates of soil moisture dry-downs, (ii) high vapour pressure deficit, and (iii) elevated CO2. Finally, we assessed the suitability of the models for use at the global scale, depending on the number of model-specific parameters and on easiness of calibration.