Diagnostic Framework for Prediction of Floods

Thumbnail

Event details

Date 07.12.2009
Hour 16:15
Speaker Pr Witold F. Krajewski, director of Iowa Flood Center
Location
GR B3 30
Category Conferences - Seminars
The speaker outlines a diagnostic framework for modeling and prediction of floods. The framework takes advantage of the scaling of peak flows with drainage area. It has been known for quite some time that the annual statistics of flood peaks obey power laws with respect to areas. Less known is observational evidence that similar scaling relationships are valid at the event scale. Much of this evidence comes from data analysis of the recent June 2008 floods in Eastern Iowa. Data collected by the USGS from basins with area ranging over four orders of magnitude display strong power laws. This behavior of flood peaks can be explained by the role of rainfall variability in space and time, the drainage network topology, and the associated water transport. The speaker discusses examples that explore the role of rainfall variability and the associated rainfall estimation uncertainty in the context of flood forecasting. He also demonstrates the potential of the framework for the prediction of land use change effects such as urbanization on flood occurrence and characteristics of flooding. He closes with an outline of the implications of the presented framework for real time flood forecasting and for flood frequency prediction under changing climate.

Practical information

  • General public
  • Free

Contact

  • Alexis Berne

Tags

EESS

Share