Sensitivity test of Community Land Model version 3.5-Dynamic Global Vegetation Model over deciduous forest in Korea

In order to examine the effect of parameters of respiration and photosynthesis, and plant phenology on equilibrium carbon pool amount and consequently carbon exchange, we performed the sensitivity test of the parameterization using the Community Land surface Model 3.5-Dynamic Global Vegetation Model (CLM3.5-DGVM) for temperate deciduous forest in Gwangneung Korea. The model was run in an offline mode driven by meteorological data obtained from the deciduous forest site. The parameters of respiration and photosynthesis for the sensitivity test were carbon-to-nitrogen ratio of leaf (CNleaf) and the ratio of heterotropic respiration rate from litter decomposition to total litter decomposition (F). For the sensitivity test of plant phenology, we used a plant phenology module, which was modified to capture a growing period realistically. In each sensitivity test, we run the model for 400 years to get an equilibrium carbon pool amount in given conditions. The results of the sensitivity test show that 30% reduction of CNleaf mitigates slightly an overestimation of gross primary productivity (GPP) but intensified an underestimation of equilibrium soil carbon pool. Although F influences slowly decomposed soil carbon pool amount, it does not alter the heterotropic respiration significantly. Use of a modified phenology module significantly decreases equilibrium plant carbon pool, resulting in a remarkable reduction of GPP and ecosystem respiration (ER). Estimations of GPP, ER, and net ecosystem carbon exchange (NEE) are fairly comparable to the observations at this site. This confirms an importance of a proper representation of plant phenology in simulating of terrestrial carbon exchange.