Robust statistical analysis of multivariate hydrological time series and their application in regionalization of design floods (B06)

• Subject Area: Statistics and Econometrics
• Term: from 2013 to 2017
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 68236791

The project extends the analysis of extremes of important flood characteristics such as peak and volume. Key challenges are heteroscedastic uncertainties in hydrological time series and insufficient information about the tails of the distribution. Given these restrictions the project develops models and robust estimators for extreme quantiles and for copulas of different flood characteristics. In view of several severe storms in 2016, it also considers heavy rain events.

• DFG Programme: Collaborative Research Centres
• Subproject of: SFB 823: Statistical Modelling of Nonlinear Dynamic Processes
• Applicant Institution: Technische Universität Dortmund
• Co-Applicant Institution: Ruhr-Universität Bochum
• Project Heads: Professor Dr. Roland Fried; Professor Dr. Andreas H. Schumann