UBC Theses and Dissertations
Numerical simulation of canopy flow and carbon dioxide flux at the west coast flux station Sun, Haizhen
Using an eddy covariance-approach to estimate net-ecosystem-exchange (NEE) of carbon dioxide (CO₂) requires that advection is negligible. This approach assumes flat and homogenous terrain. Sloping and heterogeneous topography may contribute to errors in estimating carbon sequestration or loss of an ecosystem. The complex topography at the West Coast Flux station on Vancouver Island raises uncertainties on the estimation of NEE. This research employs a high-resolution atmospheric model to study the effects of a deep forest canopy on atmospheric boundary-layer flow, and to evaluate the roles of mesoscale flow and the advection effects on NEE estimates. The model is refined to include tree-drag, radiation effects of canopy on surface energy budget, CO₂ budget and soil conduction. Over an idealized two-dimensional mountain of similar horizontal and vertical scales as Vancouver Island, the numerical simulations capture the first-order effects of diurnal heating/cooling on the sloping terrain during fair-weather. The volume exchange of heat, momentum and CO₂ above and within the canopy appears to be strongly affected by the local flows resulting from diurnal thermal forcing, land/sea breezes and convective thermals. Synoptic-scale forcing is neglected. The simulated CO₂ concentration shows significant variation and gradients near the surface, corresponding to the diurnal change of stability in the atmospheric boundary layer. The time evolution of CO₂ spatial field exhibits the transport processes at night by drainage flows. The CO₂ budget analysis further quantifies the contributions of horizontal and vertical advection. The estimated NEE by only using the sum of the storage change and vertical flux results in large fluctuation during daytime and underestimation of nocturnal respiration by about 40%. When we reduce the effects of strong convection by subtracting a half-hour average from instantaneous variables, as is done in eddy-covariance approach, the daytime fluctuations of estimated NEE are strongly decreased and the estimated nocturnal respiration is within 20% of the prescribed source. Back-streakline analysis shows that the advection source areas of CO₂ reaching the flux towers during fall are mainly from the northwest and southeast along Georgia Strait with a larger advective source area during daytime and a smaller one at night.
Item Citations and Data