Stefan Uhlenbrook

Prediction uncertainty of conceptual rainfall–runoff models caused by problems in identifying model parameters and structure

Abstract The uncertainties arising from the problem of identifying a representative model structure and model parameters in a conceptual rainfall-runoff model were investigated. A conceptual model, the HBV model, was applied to the mountainous Brugga basin (39.9 km”) in the Black Forest, southwestern Germany. In a first step, a Monte Carlo procedure with randomly generated parameter sets was used for calibration. For a ten-year calibration period, different parameter sets resulted in an equally good correspondence between observed and simulated runoff. A few parameters were well defined (i.e. best parameter values were within small ranges), but for most parameters good simulations were found with values varying over wide ranges. In a second step, model variants with different numbers of elevation and landuse zones and various runoff generation conceptualizations were tested. In some cases, representation of more spatial variability gave better simulations in terms of discharge. However, good results could...

Hydrograph separations in a mesoscale mountainous basin at event and seasonal timescales

[1]xa0The spatial and temporal (event and seasonal timescale) variability of major runoff components in the mountainous Brugga basin (Black Forest, Germany) were examined. The mesoscale (40 km2) study basin represented an extraordinary challenge as comparable studies have been undertaken mainly in smaller headwater basins. Discharge data, tracer concentrations of 18O, 3H, CFCs, and dissolved silica, and major anions and cations were analyzed during single events and over a period of 3 years. Three main runoff components were defined: event water with a residence time of several hours to a few days contributed up to 50% during flood peaks, quantified by a classical hydrograph separation technique using 18O. However, this component is of minor importance for longer periods, comprising ∼11.1% of total runoff as estimated for the period August 1995 to April 1998. The other two flow components originated from shallow and deep groundwater. Source areas for these are the upper drift and debris cover for the shallow groundwater and the deeper drift, weathering zone and hard rock aquifer for the deep groundwater. Mean residence times ranged from 28 to 36 months on the basis of 18O data for the shallow groundwater and from 6 to 9 years on the basis of 3H and CFC data for the deep groundwater. The importance of the upper drift and debris cover of the slopes for runoff generation at the test site was clearly demonstrated at the seasonal timescale, showing a contribution of 69.4% based on a mixing model with a monthly time step. The deep groundwater contribution was 19.5%. With this information a conceptual model of runoff generation for the study site was constructed.

Modeling spatial patterns of saturated areas: An evaluation of different terrain indices

[1]xa0A key component to understanding and predicting water fluxes and water quality in river basins is the spatial distribution of water-saturated areas. There is limited knowledge on spatial patterns of saturated areas, their relation to landscape characteristics and processes, and the ability of hydrological models to represent the observed spatial patterns, particularly at the large scales most relevant for water resources management. In this study, saturated areas were mapped in two mesoscale (18 and 40 km2), humid temperate basins. Geobotanical and pedological criteria were used to achieve a consistent time-integrated delineation of saturated areas. Using commonly available spatial data on landscape characteristics, various terrain indices were evaluated for their ability to predict the observed patterns. Quantitative performance criteria describing the agreement of modeled and observed spatial patterns included cell-by-cell and cell-neighborhood approaches. Upslope contributing area was the most important single factor explaining the observed pattern. An improved pattern was obtained for the topographic wetness index (TOPMODEL index). However, the performance was markedly sensitive to the algorithms used for calculation of upslope contributing area and slope gradient. Other factors such as soil or climate were of less value for improving the predictions. The optimum spatial agreement of observed and modeled saturated areas was about 50% for a combined soil-climate-topographic index. Geological features (bedrock fractures) partly explained the residual pattern. Using an independent test catchment, it was shown that the index approach can be transferred to basins with similar physiographic characteristics for estimating the general pattern of saturated areas.

On the value of experimental data to reduce the prediction uncertainty of a process-oriented catchment model

Predicting hydrological response to rainfall or snowmelt including an estimation of prediction uncertainty is a major challenge in current hydrological research. The process-based catchment model TACD (tracer aided catchment model, distributed) was applied to the mountainous Brugga basin (40 km2), located in the Black Forest Mountains, southwest Germany. The Monte Carlo-based generalized likelihood uncertainty estimation (GLUE) framework was used to analyse the uncertainty of discharge predictions. The model input parameter sets were generated using the Latin Hypercube sampling method, which is an efficient way to sample the parameter space representatively. It was shown that the number of investigated parameters should exceed the number of varied parameters by at least a factor of 10. Even if the process basis and suitability of the model could be proven, relatively large uncertainty ranges of the discharge predictions still occurred during the simulation of floods. Prediction uncertainty varied both temporally and spatially. Incorporating additional data, i.e. sub-basin runoff and observed tracer concentrations, reduced the prediction uncertainty. However, the potential restriction of the uncertainty clearly depends on the goodness of the simulation of the additional data set. Knowledge of the uncertainty of model predictions and of the potential for experimental data to reduce it are crucial to sustainable environmental management, and should be considered more thoroughly during the planning of future field studies.

An empirical approach for delineating spatial units with the same dominating runoff generation processes

Abstract A method is presented that delineates spatial units with the same dominating runoff generation processes at meso-scale, mountainous basins. It is applied to two basins (40.12 and 17.81 km 2 ) in the Black Forest Mountains, south-western Germany. The characteristics of the soil and drift cover and the topography served as main criteria to identify runoff generation units. The method is based on previous experimental investigations including tracer studies. The comparison of the two neighbouring basins showed the plausibility of the method. One catchment has a larger runoff response than the other, which is documented in specific flood runoff values and the spatial delineation. Finally, the potential of such a spatial delineation for distributed hydrological models and further practical issues is discussed.

Runoff Generation Processes on Hillslopes and Their Susceptibility to Global Change

Global change will influence hillslope hydrological processes for a variety of reasons. On the one hand, climate change might alter the hydrological input, i.e. precipitation and snow melt, which might cause an increase or decrease in the intensity of specific hillslope processes. For instance, overland flow might be amplified by increased rain intensities (Horton 1933) or by reduced infiltration due to surface crusts (Yair 1990) or increased hydrophobicity (Doerr et al. 2002), triggered by longer and more pronounced drought periods. However, overland flow could also be significantly influenced by antecedent moisture conditions of the substrate that were either altered due to wetter climate and reduced evapotranspiration at a site or due to different snow and snow melt regimes, changing the hydrological input for a specific precipitation event. On the other hand, global change in the form of land use changes will play a key role in defining the dominant runoff generation processes on hillslopes (cf. summary given in DVWK 1999).