A current challenge in boundary layer meteorology is to provide, either through modelling or measurements, estimates of turbulent fluxes that are representative of regions where the landscape is characterised by considerable surface heterogeneity. Simple models of the convective boundary layer (the CBL) offer a method of spatial integration, because of the averaging properties of the CBL. This paper explores the use of CBL methods to infer regionally averaged fluxes in a landscape that has at least three major sources of heterogeneity - irrigated and non-irrigated rural land-use and a large urban area (Sacramento region, California). Attention is directed to the differences between the Differential CBL (DCBL) and the Integrated CBL (ICBL) modelling schemes and their spatial scales of representativeness.
The first step is to evaluate the slab model of the CBL. In its differential form, the slab model uses measured time series of the surface heat and water vapour fluxes to predict time-varying mixed layer depths, temperatures and humidities. Of the four different CBL growth schemes tested, the Tennekes and Driedonks (1981) model is found to give the best performance. The effect on model performance of using different weightings for the calculation of regionally averaged surface heat and water vapour fluxes, used to force the slab model, is then analysed. This suggests that the source area for the CBL is larger than expected - extending beyond the urban area of Sacramento.
Finally, the measured time series of potential temperature are used to infer integral, regionally averaged heat fluxes with the ICBL method. These regionally averaged fluxes are compared with those measured at the local scale over the three land use types comprising the source area for the ICBL. The predicted integral heat fluxes are found to be closest to the average calculated by weighting the measured heat fluxes according to the location and dimensions of the ICBL source area. This confers some validity to the ICBL method of inferring regionally averaged land atmosphere exchanges.