A major problem of the earth-atmosphere system, relevant to empirical as well as modelling studies, is the degree to which land surface heterogeneity influences atmospheric processes. The significant scales at which the surface forces vertical fluxes and the relative strength of these surface sources and sinks are of particular interest. Our study uses data collected by the Canadian Twin Otter Aircraft during the BOREAS 1994 and 1996 Intensive Field Campaigns, satellite and land-surface data, to develop a spatial model for deriving flux estimates for CO2, water vapour and heat. The model sees the inter-relationships among the various fluxes and their spatial scales as a direct result of the prevailing mosaic of surface characteristics at varying scales of influence, as well as the dominant atmospheric conditions.
The spatial model is developed for the 16 km2 grids at both the northern and southern BOREAS sites. The relationships are derived from: (a) The correlations between the micro- and mesoscale flux events observed over the grid and the spatial scales of the surface characteristics. For example, while heat flux occurs predominantly at the local scale, fluxes of CO2 and water vapour have significantly larger mesoscale components. (b) The combinations of surface chracteristics such as stand age, stand type, productivity levels, LAI, NDVI, pure stands vs. mixed stands and the overstorey/understorey combinations. For example,under uniform, closed single level canopy, with average RH, one would expect a straightforward coupling between the surface temperature and heat and water vapour fluxes. However under conditions such as a relatively open overstorey of black spruce; which is resistant to water loss at high radiation levels and low RH, and consequently generates high levels of sensible heat flux, and an understorey of wet moss; generating a lower that expected surface temperature and increased water vapour flux, the observed relationships become more complex and may be scale-dependent. (c) Atmospheric conditions such as vapour pressure deficit, boundary layer height and level of solar radiation. The flux relationships and coherent structure characteristics generated by the model as a result of these varying conditions are then validated from other grid observations.