I. Iorgulescu

Uncertainty in hydrograph separations based on geochemical mixing models

A detailed uncertainty analysis of three-component mixing models based on the Haute–Mentue watershed (Switzerland) is presented. Two types of uncertainty are distinguished: the ‘model uncertainty’, which is affected by model assumptions, and the ‘statistical uncertainty’, which is due to temporal and spatial variability of chemical tracer concentrations of components. The statistical uncertainty is studied using a Monte Carlo procedure. The model uncertainty is investigated by the comparison of four different mixing models all based on the same tracers but considering for each component alternative hypotheses about their concentration and their spatio-temporal variability. This analysis indicates that despite the uncertainty, the flow sources, which generate the stream flow are clearly identified at the catchments scale by the application of the mixing model. However, the precision and the coherence of hydrograph separations can be improved by taking into account any available information about the temporal and spatial variability of component chemical concentrations.

Sensitivity analysis based on regional splits and regression trees (SARS-RT)

A global sensitivity analysis with regional properties is introduced. This method is demonstrated on two synthetic and one hydraulic example. It can be shown that an uncertainty analysis based on one-dimensional scatter plots and correlation analyses such as the Spearman Rank Correlation coefficient can lead to misinterpretations of any model results. The method which has been proposed in this paper is based on multiple regression trees (so called Random Forests). The splits at each node of the regression tree are sampled from a probability distribution. Several criteria are enforced at each level of splitting to ensure positive information gain and also to distinguish between behavioural and non-behavioural model representations. The latter distinction is applied in the generalized likelihood uncertainty estimation (GLUE) and regional sensitivity analysis (RSA) framework to analyse model results and is used here to derive regression tree (model) structures. Two methods of sensitivity analysis are used: in the first method the total information gain achieved by each parameter is evaluated. In the second method parameters and parameter sets are permuted and an error rate computed. This error rate is compared to values without permutation. This latter method allows the evaluation of the sensitivity of parameter combinations and thus gives an insight into the structure of the response surface. The examples demonstrate the capability of this methodology and stress the importance of the application of sensitivity analysis.

Validation of TOPMODEL on a small Swiss catchment

TOPMODEL, the topography-based variable contributing area model of Beven and Kirk by, was tested on two subcatchments of the Haute-Mentue (Switzerland) research basin. Simulations with field-estimated parameters gave poor results. Parameters were calibrated using response surfaces estimated with the Nash-Sutcliffe efficiency criteria. This approach provided useful insights into the structure of the model. Even if an acceptable numerical fit was reached (E = 0.84) and although it was possible to verify some of the underlying concepts of TOPMODEL for the Haute-Mentue basin, the model could not be validated fully with respect to field measurements and knowledge of the physical processes involved in catchment response.

Flow, mixing, and displacement in using a data‐based hydrochemical model to predict conservative tracer data

We extend the data-based hydrochemical model of Iorgulescu et al. (2005), able to simulate discharge and reactive chemical tracer concentrations (silica and calcium) in streamflow for subcatchments of the Haute-Mentue research basin (Switzerland), to the prediction of additional δ 18O values treated as a conservative tracer. The hydrochemical model is based on a parameterization of three runoff components (direct precipitation (DP), acid soil water (AS), and deep groundwater (GW)) in a chemical mixing model. Each component is modeled through an identical structure consisting of a nonlinear gain and a linear transfer function with two reservoirs (fast/slow) in parallel having a constant partition between them. We formulate a set of hypotheses concerning the isotope characterization of each component to provide additional information about how new rainfall inputs are processed in the hydrological response of the catchment. In particular, the AS component is modeled through a nested structure of hypotheses (models) of increasing complexity. It will be shown that hydrological processes in the hillslope associated with the DP, AS, and GW components are especially effective in filtering of higher-frequency fluctuations in precipitation isotopic ratios at the intraevent, interevent/seasonal, and annual/multiannual timescales. The highly nonlinear and nonstationary AS component represents predominantly “recent” water stored in the upper decimeters of the soil profile. Results also suggest that subsurface pathways are significant for the DP component. A local flow path mechanism is proposed for explaining the large fluxes of subsurface flows.

Nonparametric direct mapping of rainfall-runoff relationships: An alternative approach to data analysis and modeling?

We present a new approach for the analysis and modeling of catchment rainfall-runoff relationships that uses as predictor variables input history summary variables only. The latter are defined as linear combinations of inputs at a given number of previous time steps. This transforms the dynamic identification problem into a static one. As the identification algorithm we use regression trees, which act as a nonlinear nonparametric model. The original algorithm is adapted to account for serial correlation in variables. The new method is applied to two subcatchments of the U.S. Department of Agriculture Forest Service Andrews Experimental Forest Watershed (Oregon, United States). Simple and interpretable tree models explain more than 80% of the initial deviance of the observations in both calibration and validation. This suggests that the selected variables have a good predictive power and that further modeling attempts using them are warranted. The models show a distinct pattern of the selected explanatory variables. Applications of the method include data quality control, comparative analysis, assessment of hydrological change, and multicriterion evaluation of parametric hydrological models.

Management of forested landscapes in mountain areas: an ecosystem-based approach

Abstract The goal of this paper is to show how an ecosystem-based approach can contribute simultaneously to a multipurpose use of forest resources in mountain areas, and to the maintenance of the quality of the forests and forested landscapes producing the resources. An ecosystem-based approach in managing forest resources in mountain areas is considered as essential. Its principles, methods and instruments are introduced and illustrated with examples. The accent is put on the importance of the landscape (ecocomplex) level, the integration of ecological, economic and social considerations, and the use of multicriteria decision aid techniques.