Automated catena-based discretization of landscapes for the derivation of hydrological modelling units

In hydrological and soil erosion modelling at large spatial scales, semi-distributed approaches may use representative hillslope profiles to reproduce landscape variability. Until now, the process of delineating landscape units as homogeneous parts of the landscape with regard to their terrain, vegetation, and soil properties required expert knowledge and familiarity with the study area. In addition, the delineation procedure was often highly time-consuming and included a high degree of subjectivity. This paper presents a novel, semi-automated approach for the delineation of landscape units, the derivation of representative toposequences, and their partitioning into terrain components. It incorporates an algorithm to retrieve representative catenas and their attributes for elementary hillslope areas based on elevation and other key spatial data frequently required as environmental model input, e.g. vegetation and soil data. An example application for the sera catchment in Spain illustrates that with the presented approach, upscaling of hillslope properties becomes feasible for environmental modelling at large scales while ensuring reproducible results.