Modeling the impact of drought on canopy carbon and water fluxesfor a subtropical evergreen coniferous plantation in southern Chinathrough parameter optimization using an ensemble Kalman filter

摘要

Soil and atmospheric water deficits have signif-icant influences on CO_2and energy exchanges between theatmosphere and terrestrial ecosystems. Model parameteriza-tion significantly affects the ability of a model to simulatecarbon, water, and energy fluxes. In this study, an ensem-ble Kalman filter (EnKF) and observations of gross primaryproductivity (GPP) and latent heat (LE) fluxes were used tooptimize model parameters significantly affecting the calcu-lation of these fluxes for a subtropical coniferous plantationin southeastern China. The optimized parameters include themaximum carboxylation rate ( Vcmax), the slope in the mod-ified Ball-Berry model (M) and the coefficient determiningthe sensitivity of stomatal conductance to atmospheric watervapor deficit (Do). Optimized Vcmax and M showed largervariations than Do. Seasonal variations of Vcmax and Mweremore pronounced than the variations between the two years.Vcmax and M were associated with soil water content (SWC).During dry periods, SWC at the 20 cm depth explained 61%and 64% of variations of Vcmax and M, respectively. EnKFparameter optimization improved the simulations of GPP, LEand SH, mainly during dry periods. After parameter opti-mization using EnKF, the variations of GPP, LE and SH ex-plained by the model increased by 1% to 4% at half-hourlysteps and by 3% to 5% at daily time steps. Further efforts areneeded to differentiate the real causes of parameter variationsand improve the ability of models to describe the change ofstomatal conductance with net photosynthesis rate and thesensitivity of photosynthesis capacity to soil water stress un-der different environmental conditions.