HR: 11:50h
AN: H12B-07 INVITED [Abstracts]
TI: Forecasting Flooding in the Brahmaputra and Ganges Delta of Bangladesh on Short (1-10 days), Medium (20-30 days) and Seasonal Time Scales (1-6 months)
AU: * Webster, P J
EM: pjw@eas.gatech.edu
AF: School of Earth and Atmospheric Sciences, Georgia Institute of Technology, 311 Ferst
Drive, Atlanta, GA 30332, United States
AU: Hoyos, C D
EM: choyos@eas.gatech.edu
AF: School of Earth and Atmospheric Sciences, Georgia Institute of Technology, 311 Ferst
Drive, Atlanta, GA 30332, United States
AU: Hopson, T M
EM: hopson@ucar.edu
AF: National Center of Atmospheric Research, 1850 Table Mesa Dr, Boulder, CO 80305,
United States
AU: Chang, H
EM: hrc@eas.gatech.edu
AF: School of Earth and Atmospheric Sciences, Georgia Institute of Technology, 311 Ferst
Drive, Atlanta, GA 30332, United States
AU: Jian, J
EM: jun.jian@eas.gatech.edu
AF: School of Earth and Atmospheric Sciences, Georgia Institute of Technology, 311 Ferst
Drive, Atlanta, GA 30332, United States
AB:
Following the devastating flood years of 1998 during which 60% of Bangladesh was under water for a period of 3
months, the Climate Forecast Applications in Bangladesh (CFAB) project was formed with funding by USAID and
NSF which eventually resulted in a joint project with the European Centre for Medium Range Weather Forecasting
(ECMWF), the Asian Disaster Preparedness Centre (ADPC) and the Bangladesh Flood Forecasting and Warning
Centre. The project was organized and developed through the Georgia Institute of Technology.
The aim of CFAB was to develop innovative methods of extending the warning of flooding in Bangladesh noting
that there was a unique problem: India provided no upstream discharge data to Bangladesh so that before CFAB
the maximum lead time of a forecast was that given by measuring river discharge at the India-Bangladesh border:
no lead-time at the border and 2 days in the southern parts of the country. Given that the Brahmaputra and
Ganges catchment areas had to be regarded as essentially unguaged, it was clear that innovative techniques
had to be developed. On of the basic criterion was that the system should provide probabilistic forecasts in order
for the Bangladeshis to assess risk.
A three-tier system was developed to allow strategic and tactical decisions to be made for agricultural purposes
and disaster mitigation: seasonal (1-6 months: strategic), medium range (20-30 days: strategic/tactical) and
short range (1-10 days: tactical). The system that has been developed brings together for the first time operational
meteorological forecasts (ensemble forecasts from ECMWF), with satellite and discharge data and a suite of
hydrological models. In addition, with ADPC and FFWC we have developed an in-country forecast dispersion
system that allows a rapid dissemination.
The system has proven to be rather successful, especially in the short range. The flooding events of 2004 were
forecast with all forecasting tiers at the respective lead time. In particular, the short-term forecasts picked 10 days
ahead of time the double flooding peak. In 2007, the system forecast the commencement and retreat of the July-
August floods allowing for the first time for the Bangladesh Disaster Management Committee to act proactively
rather than reactively. As a result, many thousands of villagers were evacuated out of harms way. The forecasting
system will be discussed in some detail together with examples of forecasts made during the last 5 years. Most
importantly, we see the method we have developed as a template for flood forecasting in the developing world
where modern technology from the United States and Europe interfaces, interacts and supports local
infrastructure.
DE: 1816 Estimation and forecasting
DE: 1821 Floods
DE: 1847 Modeling
SC: Hydrology [H]
MN: 2007 Fall Meeting