Capabilities and limitations of detailed hillslope hydrological modelling

Hillslope hydrological modelling is considered to be of great importance fo r the understanding and quantification of hydrological processes in hilly o r mountainous landscapes. In recent years a few comprehensive hydrological models have been developed at the hillslope scale which have resulted in an advanced representation of hillslope hydrological processes (including the ir interactions), and in some operational applications, such as in runoff a nd erosion studies at the field scale or lateral flow simulation in environ mental and geotechnical engineering. An overview of the objectives of hills lope hydrological modelling is given, followed by a brief introduction of a n exemplary comprehensive hillslope model, which stimulates a series of hyd rological processes such as interception, evapotranspiration, infiltration into the soil matrix and into macropores, lateral and vertical subsurface s oil water flow both in the matrix and preferential flow paths, surface runo ff and channel discharge. Several examples of this model are presented and discussed in order to determine the model's capabilities and limitations. F inally, conclusions about the limitations of detailed hillslope modelling a re drawn and an outlook on the future prospects of hydrological models on t he hillslope scale is given. The model presented performed reasonable calculations of Hortonian surface runoff and subsequent erosion processes, given detailed information of init ial soil water content and soil hydraulic conditions. The vertical and late ral soil moisture dynamics were also represented quite well. However, the g iven examples of model applications show that quite detailed climatic and s oil data are required to obtain satisfactory results. The limitations of de tailed hillslope hydrological modelling arise from different points: diffic ulties in the representations of certain processes (e.g. surface crusting, unsaturated-saturated soil moisture flow, macropore flow), problems of smal l-scale variability, a general scarcity of detailed soil data, incomplete p rocess parametrization and problems with the interdependent linkage of seve ral hillslopes and channel-hillslope interactions. Copyright (C) 1999 John Wiley & Sons, Ltd.