Original scientific paper
PROBABILISTIC RUNOFF FORECASTING USING A COUPLED ATMOSPHERIC-HYDROLOGIC ENSEMBLE PREDICTION SYSTEM
M. Verbunt
; Institute for Atmospheric and Climate Science, Swiss Federal Institute of Technology (ETH), Winterthurerstrasse 190, 8057 Zurich, Switzerland
A. Walser
; MeteoSwiss, Krähbühlstrasse 58, 8044 Zurich, Switzerland
J. Gurtz
; Institute for Atmospheric and Climate Science, Swiss Federal Institute of Technology (ETH), Winterthurerstrasse 190, 8057 Zurich, Switzerland
A. Montani
; ARPA-SIM, Viale Silvani, 6, 40122 Bologna, Italy
C. Schär
; Institute for Atmospheric and Climate Science, Swiss Federal Institute of Technology (ETH), Winterthurerstrasse 190, 8057 Zurich, Switzerland
Abstract
A high-resolution atmospheric ensemble forecasting system, based on 51 runs of the Local Model (LM), has been used to make probabilistic runoff forecasts for a 5 day forecasting period in the alpine tributaries of the Rhine basin. The investigated cases are the spring 1999 flood, when a combination of snowmelt and heavy precipitation caused severe floods in Central Europe and the November 2002 flash flood in the Alpine Rhine area. This study focuses on the feasibility of ensemble prediction system (EPS) for runoff forecasting. Runoff predictions from the deterministic forecast are compared with those obtained from probabilistic atmospheric forecasts. For both cases, the deterministic simulations yield large forecast failures, while the hydrometeorologic EPS provides appropriate forecast guidance with proper uncertainty intervals. The use of clustering techniques showed that the clustering methodology does not reduce ensemble spread.
Keywords
COSMO-LEPS; runoff prediction; forecast uncertainties; hydrological ensemble prediction
Hrčak ID:
64507
URI
Publication date:
15.12.2005.
Visits: 959 *