TY - JOUR
T1 - Ensemble flood forecasting in Africa
T2 - A feasibility study in the Juba-Shabelle river basin
AU - Thiemig, Vera
AU - Pappenberger, Florian
AU - Thielen, Jutta
AU - Gadain, Hussein
AU - de Roo, Ad
AU - Bodis, Katalin
AU - Del Medico, Mauro
AU - Muthusi, Flavian
PY - 2010/4/1
Y1 - 2010/4/1
N2 - The European flood alert system (EFAS) achieves early flood warnings for large to mediumsize river basins with lead times of 10 days. This is based on probabilistic weather forecasts, the exceedance of alert thresholds and persistence. The methodologies have been tested for different events and time scales in mid-latitude basins in Europe. In this article, the transferability of the EFAS-methodologies to equatorial African basins is assessed with the analysis of the Juba-Shabelle river basin as an example using a variety of different meteorological data sources. In this context, ERA-40 and CHARM have been used for the calculation of climatologies; re-forecasts of the current operational European Centre for Medium-Range Weather Forecasts model provided hindcasts of historic flood events. The results show that flood events have been detected successfully in more than 85% of all cases, with a high accuracy in terms of timing and magnitude.
AB - The European flood alert system (EFAS) achieves early flood warnings for large to mediumsize river basins with lead times of 10 days. This is based on probabilistic weather forecasts, the exceedance of alert thresholds and persistence. The methodologies have been tested for different events and time scales in mid-latitude basins in Europe. In this article, the transferability of the EFAS-methodologies to equatorial African basins is assessed with the analysis of the Juba-Shabelle river basin as an example using a variety of different meteorological data sources. In this context, ERA-40 and CHARM have been used for the calculation of climatologies; re-forecasts of the current operational European Centre for Medium-Range Weather Forecasts model provided hindcasts of historic flood events. The results show that flood events have been detected successfully in more than 85% of all cases, with a high accuracy in terms of timing and magnitude.
KW - Africa
KW - Ensemble prediction system
KW - European flood alert system
KW - Flood warning
KW - Hydrological ensemble prediction systems
KW - Juba-Shabelle river basin
KW - LISFLOOD
KW - Probabilistic flood forecasting
UR - http://www.scopus.com/inward/record.url?scp=77954439472&partnerID=8YFLogxK
U2 - 10.1002/asl.266
DO - 10.1002/asl.266
M3 - Article (Academic Journal)
AN - SCOPUS:77954439472
SN - 1530-261X
VL - 11
SP - 123
EP - 131
JO - Atmospheric Science Letters
JF - Atmospheric Science Letters
IS - 2
ER -