Im Rahmen des Projekts «Klimaanderung und Hydrologie in der Schweiz» (CCHydro) des Bundesamts fur Umwelt (BAFU) wurden die Auswirkungen des Klimawandels auf den Wasserhaushalt der Schweiz bis zum Jahr 2100 untersucht. Das Wasserdargebot wird sich bis dann nur wenig andern. Als Folge des Anstiegs der Schneefallgrenze parallel zur Zunahme der Lufttemperatur werden die in den Alpen gespeicherten Schnee- und Eismassen jedoch stark vermindert. Zusammen mit einer saisonalen Umverteilung des Niederschlags (trockener im Sommer, feuchter im Winter) wird dies eine jahreszeitliche Umverteilung der Abflusse hervorrufen. Hochwasser- und insbesondere Niedrigwasserereignisse werden wahrscheinlich vermehrt auftreten – vor allem in sensitiven Regionen wie dem Mittelland, dem Wallis oder dem Tessin.

Im Rahmen des Projekts «Klimaanderung und Hydrologie in der Schweiz» (CCHydro) des Bundesamts fur Umwelt (BAFU) wurden die Auswirkungen des Klimawandels auf den Wasserhaushalt der Schweiz bis zum Jahr 2100 untersucht. Das Wasserdargebot wird sich bis dann nur wenig andern. Als Folge des Anstiegs der Schneefallgrenze parallel zur Zunahme der Lufttemperatur werden die in den Alpen gespeicherten Schnee- und Eismassen jedoch stark vermindert. Zusammen mit einer saisonalen Umverteilung des Niederschlags (trockener im Sommer, feuchter im Winter) wird dies eine jahreszeitliche Umverteilung der Abflusse hervorrufen. Hochwasser- und insbesondere Niedrigwasserereignisse werden wahrscheinlich vermehrt auftreten – vor allem in sensitiven Regionen wie dem Mittelland, dem Wallis oder dem Tessin.

Hydrological models do not yet appropriately represent the processes in catchments leading to low-flow situations. Further, the model-related parameters are strongly based on the present-day climate, so that the applicability of such models for water-resources planning under future climate conditions remains limited. This paper investigates the applicability of baseflow separation for improved model calibration that is relying more strongly on low-flow processes within catchments. Three different baseflow separation procedures were applied in 59 catchments in Switzerland. The results show that baseflow is physically determined and can be used to improve the performance of conceptual hydrological models. Thus, uncertainties in low-flow projections can be reduced.

Base flow is a desirable entity to know, for water management in general and particularly for climate change impact studies. Base flow is most often defined as that part of total discharge which origins from delayed storages in a river catchment. During a prolonged period without rain, base flow is the sole contributor to discharge. Base flow therefore makes a river perennial. A high base flow contribution to total annual discharge makes a river more stable in respect of meteorological droughts. Annual base flow from a catchment cannot be determined exactly. Only total discharge can be measured with high accuracy. Therefore, base flow has to be estimated with appropriate methods. Calculating an entity which cannot be verified by measurements is easy. By defining the entity with a calculation procedure, the result is numerically always right. It is actually much more difficult to understand the results, i.e. how these outcomes should be interpreted. The present study investigates the application of three different base flow separation procedures for numerous (up to 40) meso-scale catchments in Switzerland. The methods Demuth (1993), Wittenberg (1999) and Institute of Hydrology (1980) are different approaches to determine base flow, based on daily runoff data. The method Demuth, and the separation of base flow according to Institute of Hydrology, are statistical methods. Demuth is based on the graphical approach of Kille (1970), and the procedure of the Institute of Hydrology is an empirical smoothing method. In contrast to this, the method by Wittenberg does not presume linearity between storage and outflow. Analyzing the results, among each other and in comparison with physiographic characteristics of the catchments under consideration, leads to a more detailed picture of the ongoing processes. At least the dominant control factors for base flow in the Swiss Midlands should be detectable. These are expected to be found first of all among geology and climate, which is generally accepted in the literature, secondly in land cover, and, especially for the Swiss Midlands, in aquifer area and aquifer volumes. In this contribution the results of the different methods are presented and conclusions as to control factors are drawn from the results. The data base for river flow analysis in the low flow range is ideal in Switzerland. There are long time series, a dense gauge network and a comprehensive knowledge about uncertainty of the runoff measurements during low flow. This allows, in addition to the obtained process understanding, a well-founded comparison between the methods applied, which is going to be presented as well.