Quantification of net ecosystem exchange sampling within two mature boreal aspen stands using airborne LiDAR and a flux footprint model: Scaling to MODIS

Exchanges of CO2 and H2O are often measured by eddy covariance within relatively homogeneous ecosystems, where the spatial variability of vegetation structural characteristics and ground surface topography are relatively non-varying. Therefore, scalars transported from source/sink areas are representative of site average characteristics, regardless of wind direction and atmospheric turbulence. Despite relatively high confidence in the efficacy of measured exchanges within homogeneous ecosystems (barring meteorological and technical problems, etc.), site representativeness of the larger area may be questionable. For example, ecosystems represented by eddy covariance are often more heterogeneous than those measured. On the other hand, deployment of eddy covariance within sites containing variable land cover types may be prone to biased site averages if sampling does not represent similar landscape characteristics within the region. By combining remote sensing data, estimates of source/sink areas, and net ecosystem exchanges (NEE), vegetation and topographical characteristics associated with the frequency of sampling may be quantified. Sample frequency may then be used to classify and scale fluxes to the larger region. The following study integrates a flux footprint model parameterisation with three-dimensional characteristics of the ground, understory, and canopy measured using airborne Light Detection and Ranging (LiDAR) and NEE within two contrasting mature boreal aspen/mixed ecosystems. The objectives of the study are to 1) quantify the frequency distribution of source/sink contributions of CO2 to eddy covariance; 2) classify canopy, understory and topographical characteristics of the footprint climatology to quantify biases in NEE; 3) determine the dominance of sites (e.g. their representativeness) within the larger basin; 4) Evaluate the effectiveness of eddy covariance placement for MODIS product validation.