Analyzing and modelling solute and sediment transport at different spatial and temporal scales

Abstract

Solute and sediment transport processes occur on a wide range of temporal and spatial scales. Differentiate between the observation and process scale which are defined by the extent, spacing and support of the samples and the modelling scale. Typical modelling scales are the local scale (1m), the hillslope scale (100 m), the catchment scale (10 km) and the regional scale (1000 km) and in time the event scale (1 day), the seasonal scale (1 yr) and the long-term scale (100 yrs). The modelling scale often differs from the process scale which necessitate an upscaling, downscaling or regionalization scheme. In this study process investigations at different scales are used for the temporal and spatial upscaling of a physically-based erosion model (OPUS) and its validation at different temporal and spatial scales.<br /> Numerous models for simulating erosion, transport and sedimentation are available. While the USLE model is able to simulate long-term erosion at single slopes without considering sedimentation, other approaches are calculating all processes. In this study a concept for continuously modelling of solute and sediment transport at the catchment scale for a period of 50 years is presented. The simulation is based on the application of the model system OPUS designed for long-term simulations of agricultural processes. As OPUS is a hillslope model the catchment is discretized into numerous slopes. These slopes are linked via a channel model (HEC6-model) which allows the simulation of erosion, transport and sedimentation in a catchment. After introducing some modifications and enhancements of the model it was possible to simulate interflow induced by macropores with OPUS. The simulation results are compared with measurements at different spatial and temporal scales of a meso-scale catchment used for drinking water supply. Three small sub-catchments of 22 to 29 ha are used to validate the model at the local scale. The modified and validated model is then applied to the Wahnbach catchment with an area of about 54 km². Simulations at the local and the catchment scales are presented and compared with measurements.<br /> To model the long-term behaviour of solute and sediment transport with a deterministic model a complete set of climate and landuse data is necessary. In this case study a method was developed to overcome the lack of data by using a weather generator. The long-term catchment simulations are validated by determining the mass of sediments trapped in the drinking water reservoir.<br /> In general, the obtained simulation results at different scales (from the sub-catchment scale to the catchment scale and from single events to long-term scales) are encouraging. It could be shown that process-based models can be applied at several scales in order to calculate the fluxes of matter. Furthermore, the results show that the considered flux types (water, solutes, sediment) are simulated with different success. The best results are obtained by simulating daily runoff at the sub-catchment scale and at the catchment scale. The solute concentration at the sub-catchment scale and the monthly nitrate discharge at the catchment scale is reproduced satisfactory, but the measures of model accuracy are less high. The sediment transport is simulated with the lowest accuracy.