Flood Forecasting for Bangladesh with Satellite Data

Mohammad Khaled Akhtar

April 2006

Abstract

During recent years several large floods have caused disasters in Southeast Asia. The inundation and flooding in Bangladesh is a serious problem. Bangladesh is a low-lying country in the Meghna delta located at the confluence of three major rivers; the Ganges, the Brahmaputra and the Meghna. Monsoon river flooding is an annual event with river water levels rising and falling slowly during the monsoon season, though with some pulsation. Floods in Bangladesh can be divided into four categories: monsoon river flood, flash flood, local rainwater flood and storm surge flood. The monsoon river flood is an annual event, which is caused by over bank spills from major rivers and their tributaries and it causes the most extensive damage in Bangladesh. Flood forecasting in a deltaic region like Bangladesh is a difficult problem. Moreover hydrological modelling itself requires significant computer support and sophisticated models. On other hand, even if these factors are in place, a large amount of data is required in order to initialise the models. For example, rainfall data over the entire catchment regions of the Ganges, Brahmaputra and Meghna are required in addition to the river flow data, soil and crop information and the amount of water that is retained within each catchment by agriculture and water resources use. The present flood forecasting system only provides the short term (depending on location it varies from 1 to 3 day) forecasts of the monsoon river floods in Bangladesh.

The main barrier to improve flood forecasting and lead time in Bangladesh is the absence or unavailability of hydrological information of the crossboundary Ganges-Brahmaputra basins which together are around 10 times larger than the Bangladesh portion. Several efforts were made so far to get real time cross boundary information but achievement is insignificant. Considering the present situation, this study aims to develop a model which can forecast the discharge at the outlet of Ganges basin, with the help of freely available remotely sensed website data of NASA, USGS, FAO, USDA, ISRIC, etc. The ANN flood forecasting model that is developed for Ganges basin using discharge data of Bangladesh and freely available remotely sensed data, is capable of expanding the forecasting lead time of

Bangladesh by extra 3 days.

Keywords: Flood-forecasting, modelling, rainfall-runoff, ANN, satellite, remotesensing, satellite-derived precipitation, Bangladesh.

Back to the list of MSc abstracts