干旱区生态环境知识资源中心

Potential ET (PET) and partitioning of evaporation and transpiration are important information for hydrologic, ecologic, forest, and agricultural studies. Most PET models were developed in flat areas for agricultural purposes, with potential evaporation (PE) and potential transpiration (PT) lumped together. To quantify the evaporative demand for sloped surfaces with a wide range of vegetation coverage, a topography-and vegetation-based surface energy partitioning algorithm for PE and PT estimates (TVET) is developed. In this paper, vegetation-based part of the TVET model is presented. TVET employs a hybrid of layer and patch approaches in partitioning energy and routing vapor and sensible heat. It first uses a layer approach to partition available energy for the canopy and the soil components. The available energy of each component is then partitioned into potential latent heat and sensible heat, using a patch approach. Hybrid of these two approaches results in simple model formulae, while coupling the two components in terms of energy partitioning and aerodynamic resistances for heat and vapor transfer. TVET is different from a layer-approach model in that it distinguishes the difference in evaporation from inter-canopy soil and from under-canopy soil, and limits convective transfer contribution to transpiration only for vegetation-cover fraction. TVET is different from a patch-approach model in that it allows evaporation occurring from under-canopy soil, and that vegetation effect on both evaporation and transpiration is well considered. These features make TVET sensitive to vegetation effect on surface energy partitioning. The model is demonstrated and tested with Penman-Monteith and Shuttleworth-Wallace models, and with observations, at four sites covering mountain, basin floor, and riparian environments. The results indicate that TVET can be used to estimate PE and PT partitioning for a wide range of surfaces with different fractional vegetation cover. Good estimates of riparian surface evapotranspiration at the Rio Grande in the central New Mexico suggest its capacity to estimate ET in similar environments. (C) 2009 Elsevier B.V. All rights reserved.