Flächenhafte Modellierung der Evapotranspiration mit TRAIN
L. Menzel (August 1999)
The present publication gives an overview of the evapotranspiration model TRAIN. This model includes information from comprehensive field studies of evapotranspiration and associated sub-processes and has been designed to simulate the spatial pattern of actual evapotranspiration. The overall structure and goals of the model are presented in an example application, the modelling of areal evapotranspiration for the territory of Switzerland (ca. 41.000 km²). For the individual squares of a grid covering Switzerland in 1 x 1 km resolution daily evapotranspiration rates are calculated over a 20 year period.
As a preliminary study, Switzerland is subdivided into nine climate zones. For each of these zones the spatial interpolation of the meteorological data is carried out by a combination of altitude dependent regression and inverse distance weighting interpolation (Schulla, 1997). As a result, daily interpolated data of precipitation, temperature, relative humidity, wind speed and sunshine duration are available. Another model input are gridded data of topography (digital elevation model, slope, aspect), soil properties and land-use, including bare rock and glaciated areas.
TRAIN is based on a modular structure. This facilitates the individual simulations of processes at the soil-vegetation-atmosphere interface which are relevant for evapotranspiration. At the beginning of a model run radiation is calculated with meteorological data and slope geometry as principal inputs (Oke, 1987). Next, radiation is modified due to topographic shade effects. Another module simulates snow accumulation, snow melt and variations of snow albedo. Interception and interception evaporation are modelled according to Menzel (1997), where the non-uniform distribution of intercepted precipitation inside a canopy is considered. A more simplified version is applied for bare surfaces. The modelling of transpiration is based on the Penman-Monteith equation (Monteith, 1965), which is combined by a submodel to deliver canopy resistances (Menzel, 1996). Canopy resistances are modified by meteorological conditions, phenology and soil moisture, which is again an output of the soil moisture module. Finally, the various evaporation components are combined with actual evapotranspiration.
A thorough discussion of the model results is included in this report. An isoline and a pixel map of mean actual evapotranspiration are presented for the investigated area. The temporal and spatial characteristics of evapotranspiration in Switzerland are described by example profiles across the Alps and an elevation dependent representation of evapotranspiration. For different types of land cover and forested areas mean daily values of evapotranspiration are presented.
