Aggregation of land surface parameters in the oasis–desert systems of north‐west China

Abstract

Aggregation rules for determining mesoscale representative land surface parameters from their patch-scale counterparts have been suggested from previous geld work over the last decade. A selected number of these rules are tested using field data collected in the framework of HEIFE (Heihe River Basin Field Experiment on Land Surface Processes, 1990-1993, China). Data from the stations of the Zhangye, Linze, Gobi and sand deserts have been used. Surface albedo and surface roughness for these four patches were aggregated in a relatively simple manner and the results match the physical expectations. The aggregation rules for surface resistance to evaporation, however, gave strongly diverging predictions (min. 277 to max. 5356 s m(-1)). Comparing with remote sensing date, which comprise a full natural variability of land surface types, revealed that area effective values for surface albedo, surface emissivity and surface roughness for the momentum obtained from the four patches were similar to the values obtained from the distributed remote sensing estimations and aggregation rules. The remote sensing algorithm SEBAL provides distributed estimates of vertical temperature and vapour pressure differences in association with the sensible and latent heat fluxes. Applying area average, flux matching schemes, area-representative values for aerodynamic and surface resistance were obtained. Because of the mathematical flux matching, the representative value for the areal surface roughness length for heat transport in HEIFE is three times larger than for momentum transport. It is demonstrated that, instead of four patches, the area-representative surface resistance and the surface roughness for heat transport need to be derived from a full spectrum of their component values, i.e. the full natural variability in the heterogeneous terrain should be understood. (C) 1998 John Wiley & Sons, Ltd.