Ensemble prediction of floods – catchment non-linearity and forecast probabilities

Quantifying the uncertainty of flood forecasts by ensemble methods is becoming increasingly important for operational purposes. The aim of this paper is to ex- amine how the ensemble distribution of precipitation fore- casts propagates in the catchment system, and to interpret the flood forecast probabilities relative to the forecast errors. We use the 622 km 2 Kamp catchment in Austria as an exam- ple where a comprehensive data set, including a 500 yr and a 1000 yr flood, is available. A spatially-distributed contin- uous rainfall-runoff model is used along with ensemble and deterministic precipitation forecasts that combine rain gauge data, radar data and the forecast fields of the ALADIN and ECMWF numerical weather prediction models. The analy- ses indicate that, for long lead times, the variability of the precipitation ensemble is amplified as it propagates through the catchment system as a result of non-linear catchment re- sponse. In contrast, for lead times shorter than the catch- ment lag time (e.g. 12 h and less), the variability of the pre- cipitation ensemble is decreased as the forecasts are mainly controlled by observed upstream runoff and observed pre- cipitation. Assuming that all ensemble members are equally likely, the statistical analyses for five flood events at the Kamp showed that the ensemble spread of the flood forecasts is always narrower than the distribution of the forecast errors. This is because the ensemble forecasts focus on the uncer- tainty in forecast precipitation as the dominant source of un- certainty, and other sources of uncertainty are not accounted for. However, a number of analyses, including Relative Op- erating Characteristic diagrams, indicate that the ensemble spread is a useful indicator to assess potential forecast errors for lead times larger than 12 h.

[1]  Günter Blöschl,et al.  Das Katastrophenhochwasser vom 7. August 2002 am Kamp — Eine erste Einschätzung , 2002 .

[2]  Andrew F. Loughe,et al.  The Relationship between Ensemble Spread and Ensemble Mean Skill , 1998 .

[3]  A. Hollingsworth,et al.  Probabilistic Predictions of Precipitation Using the ECMWF Ensemble Prediction System , 1999 .

[4]  B. Golding.  Nimrod: a system for generating automated very short range forecasts , 1998 .

[5]  Michael Steinheimer,et al.  Advances in Geosciences Improved nowcasting of precipitation based on convective analysis fields , 2008 .

[6]  R. Buizza.  WEATHER PREDICTION | Ensemble Prediction , 2003 .

[7]  R. Buizza,et al.  Using weather ensemble predictions in electricity demand forecasting , 2003 .

[8]  T. McMahon,et al.  Precipitation elasticity of streamflow in catchments across the world , 2006 .

[9]  Erwin Zehe,et al.  On hydrological predictability , 2005 .

[10]  G. Evensen.  Sequential data assimilation with a nonlinear quasi‐geostrophic model using Monte Carlo methods to forecast error statistics , 1994 .

[11]  G. Blöschl,et al.  Regionale Wasserbilanzkomponenten für Österreich auf Tagesbasis , 2005 .

[12]  Giorgio Boni,et al.  A hydrometeorological approach for probabilistic flood forecast , 2005 .

[13]  Roman Krzysztofowicz,et al.  Integrator of uncertainties for probabilistic river stage forecasting: precipitation-dependent model , 2001 .

[14]  Nicholas E. Graham,et al.  Conditional Probabilities, Relative Operating Characteristics, and Relative Operating Levels , 1999 .