Evaluation of a modified soil–plant–atmosphere model for CO2 flux and latent heat flux in open canopies

Moisture and carbon dioxide fluxes calculated by a modified soil–plant–atmosphere (mSPA) model were evaluated using the eddy correlation data collected above a pine forest located in a semi-arid region in central Oregon during the growing season of June and July in 2003. In June, soil water stress is relatively low with moderate vapor pressure deficit while in July both soil water stress and vapor pressure deficit are high. The comparison of the CO2 flux with observation has been made during daytime when the eddy flux is close to net ecosystem exchange. The model explained 88% of observed half hourly variance of latent heat flux and 86% of that of the CO2 flux. It was also found that the model simulates quite well the mean diurnal variations of latent and sensible heat flux and net ecosystem exchange in early June (Case 1) and that it reproduces the reduced carbon uptake due to the partial stomatal closure in the afternoon in late July (Case 2). However, the model overestimates the turbulent fluxes in the morning, in association with observed low energy balance closure due to the weak turbulent mixing in late July (Case 2). The simulated carbon uptake shows an earlier peak compared to the measured CO2 flux in Case 2 when weak wind prevails during the nights, probably due to the stored CO2 within the canopy which is assimilated by vegetation in the first few hours of the day. However, CO2 storage does not explain the measured low respiration rate throughout the night, suggesting the importance of other processes such as advection by drainage flows on the carbon budget during calm nights at this site.