Multifractal Prediction in Hydrology
Abstract
One of the main axes of the current hydrological research and engineering is the general forecast of extreme events and the development of new tools for their prediction, prevention and alert. Deterministic models based on various physical and/or statist ical approaches face concrete difficulties to capture the phenomena of extreme precipitation and discharges. The chain of 'precipitation-discharge-sedimentation' process remains even more unresolved issue. It is well known that one of the main difficulties for the description of hydro-meteorological extremes is the colossal variability of their intensities over a wide range of space-time scales. To contribute to the process of hydrological forecast improvement, our group uses the multifractal framework. It allows not only to explain the power-law fall-off of probability distributions for hydrological-meteorological extremes, but also to explore a connection of the observed variability with the physics, so to capture the phenomena within a full hierarchy of scales and intensities. First of all, we analyze space-time distributions of precipitation and discharges over different hydrological regions. A multifractal data analysis performed in the space-time domain produces - amongst other things - a physica lly-based tool for the clear distinction and multifractal description of flash-floods. We illustrate these methods on two recent flooding events in France: the Abbeville phreatic floods in 2001 and the flash floods in Gard in 2002. Furthermore, after multifractal analysis of several sediment time series, we obtain first results directed to the parameterization and prediction of the 'precipitation-discharge-sedimentation' process.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2003
- Bibcode:
- 2003AGUFM.H21B..04T
- Keywords:
-
- 1815 Erosion and sedimentation;
- 1821 Floods;
- 1860 Runoff and streamflow;
- 3200 MATHEMATICAL GEOPHYSICS (New field);
- 3250 Fractals and multifractals