Catchments as non-linear filters: understanding catchment similarities for the regionalisation of rainfall-runoff transformation processes using wavelets
Zusammenfassung der Projektergebnisse
This postdoc research project entitled "Catchments as non-linear filters: understanding catchment similarities for the regionalisation of rainfall-runoff transformation processes using wavelets" had as main objective to develop a new time series analysis method to quantify the hydrological filtering characteristics of landscapes. The main assumption was hereby that if we understand how landscapes filter rainfall and transform into discharge and if we can quantify these filtering characteristics, we can later-on relate these filtering characteristics to observable landscape characteristics. We can then make predictions for similar landscapes for which we do not have detailed long-term observations of the hydrological response (i.e. of the river discharge). This research project had thus as an overall objective to contribute to the prediction in so-called ungauged basins, i.e. catchments with no or very few observed hydrological data. Hydrological prediction in these catchments is particular challenging and of crucial importance for water management around the world. The used time series analysts method was the so-called wavelet spectral analysis, a method to analyze the spectra of time series as a function of time (i.e. a frequency-time resolved analysis). During the work on this research project, it became clear that, contrary to what was suggested in the available literature, the method was not ripe for the proposed application. The completed review of the state-of-the-art in this field is likely to become an important reference paper in the field of wavelet spectral analysis in geosciences. Accordingly, we first had to develop a method to test the significance of wavelet spectral analysis for hydrological applications. The work on this method is ongoing but produced as a side-result a new method for model calibration and diagnosis in the wavelet spectral domain, which has to be further developed and tested but which opens promising new perspectives for hydrological model calibration. The above results could not have been achieved without to the very fruitful interdisciplinary collaboration with colleagues from different institutes of the University of Potsdam and from several other European and American Universities.
Projektbezogene Publikationen (Auswahl)
- 2007. What drives high flow events in the Swiss Alps? Recent developments in wavelet spectral analysis and their application to hydrology. Advances in Water Resources
Schaefli, B., Maraun, D. and Holschneider, M.
