Conceptual Runoff Models for Material Transport Estimations

Abstract: Catchment-scale material transport models are today important tools for water and environmental management. However, modelling of material transport is extremely complex due to the physical, chemical and biological processes involved. The complexity calls for spatial distribution and thereby a need for large amounts of input data that generally are not available. Lumped models based on conceptual rainfall-runoff models are therefore the only feasible solution in most applications, although doubts are raised regarding these models' suitability for material transport simulations. The objective of the presented work was therefore to evaluate the possibilities, limitations and uncertainties of applying conceptual rainfall-runoff models for material transport estimations. The thesis presents five applications in which the HBV-96 rainfall-runoff model was used as a base for estimating the riverine transport of three different substances, nitrogen, phosphorus and suspended sediment, in four different countries, Sweden, Estonia, Bolivia and Zimbabwe. The HBV-96 model was analysed in detail regarding model performance and parameter interdependence in different climates. The relation between internal hydrological model variables and nutrient transport was assessed and regression models based on internal variables were tested. A model for nitrogen source apportionment was developed from the existing nitrogen submodel and was analysed in comparison with a statistical method. A new submodel for suspended sediment transport in tropical and semi-arid regions was developed and tested. It was shown that riverine total nitrogen could be well simulated in the Nordic climate and riverine suspended sediment load could fairly well be estimated in tropical and semi-arid climates. The HBV-96 based models for material transport generally, but not always, estimated material transport loads better than simpler methods based on observed runoff alone. The main conclusion of the study is that the HBV-96 based models can be used to predict material transport on the catchment scale during stationary conditions. The major limitation for the use of HBV-96 based models for material transport is that they cannot be easily generalised to areas or periods outside the ones for which they have been calibrated. The main reasons are the large degree of parameter interdependence in both the rainfall-runoff and material transport models and the risk of compensating systematic errors in the input data by help of the model parameters.