Predicting canopy biophysical properties and sensitivity of plant carbon uptake to water limitations with a coupled eco-hydrological framework

Variations in soil water availability and atmospheric water demand impact seasonal canopy dynamics are often represented by the fraction of photosynthetically active radiation (FPAR) and leaf area index (LAI) under a coupled eco-hydrologic framework. Changes in FPAR and LAI lead to changes in evapotranspiration (ET) and gross primary productivity (GPP), coupling the water and carbon cycles. In this study, a predictive Dynamic Canopy Biophysical Properties (DCBP) model is adapted to predict daily FPAR and LAI forced by observed and modeled meteorological and root-zone soil moisture conditions, respectively. Vegetation green-up and die-off responses to temperature (T), vapor pressure deficit (VPD), soil water potential (ψ_soil), and photoperiod (Pht) are modeled through a modified form of the growing season index (GSI). The DCBP model parameterizations of seasonality (T and Pht) and intraseasonal water stress (VPD and ψ_soil) are calibrated separately for distinct plant functional types (PFTs) and soil types using a Bayesian estimator. To investigate the impact of dynamic phenology on modeled GPP and hydrologic processes, phenology predicted by the DCBP model is input to the Duke Coupled Hydrology Model with Prognostic Vegetation (DCHM-PV), and hydrologic conditions are input to the DCBP model to specify water availability constraints. The coupled model framework was evaluated against AmeriFlux tower data and remotely sensed FPAR and LAI products. Sensitivity analysis of the predicted daily FPAR, LAI, and GPP to the diurnal cycle of root zone water indicates that mid-day soil water availability is the primary control on seasonality across different PFTs and soil textures in the DCBP model. Further, calibrated parameters describing plant-water relations change significantly depending on whether the inference period used in the data assimilation includes persistent meteorological drought, thus effectively resulting in distinct plant water use strategies in the DCHM-PV. The dynamics of water stress recovery are examined by mapping seasonal phenology into the phase space of soil water stress and carbon uptake. In the Southeast U.S., simulated annual differences in GPP can be as high as 350 g C/m²/year with ET increases up to 125 mm/year during wet years. These values represent first order estimates of the dynamics of plant-water use feedbacks on the water and carbon budgets, and highlight the need to incorporate vegetation-specific phenology responses to water availability in order to accurately estimate the terrestrial carbon storage component of the global carbon budget at local and regional scales.