P2.27
Major Bangladesh floods as coupled ocean-atmosphere phenomena
Peter J. Webster, Univ. of Colorado, Boulder, CO; and W. Han, K. Sahami, and R. Grossman
Bangladesh lies at the confluence of the Ganges and Brahmaputra rivers. Floods occur in the Bangladesh river delta region almost every summer. However, major floods, such as that which occurred in the summer of 1998 and which inundated nearly two-thirds of the country for nearly three months, occur only once or twice in a decade. The question is what causes these major floods and whether or not they can be forecast with sufficient lead time to be useful. We argue here that these major delta floods are a combination of atmospheric and ocean processes, both of which may be components of a prediction system.
A curious observation is that major Bangladesh floods are not always associated with excessive precipitation in the delta catchment areas. For example, the catchment rainfall during 1997 and 1998 was almost identical both being essentially normal monsoon rain years. Yet, major flooding only occurred during the summer of the latter year. However, a major difference between the two years, and in between other flood and non-flood years, was the sea-level height in the northern Bay of Bengal which was nearly 1/3 m higher in the flood year. We argue that the sea-level variability is a necessary condition for major delta flooding. It is noted that the slow oscillation of sea-level height is a function basin-wide coupled ocean-atmosphere processes associated with the Indian Ocean dipole. These low-frequency ocean modes are potentially predictable and their influence on the Bay of Bengal is well understood. Sea-level height forecasts, together with precipitation predictions in the catchment regions (obtained from either empirical SST-precipitation relationships or numerical ensemble predictions), form the basis of an experimental forecast scheme that has recently been developed and implemented.
Poster Session 2, Forecasting Climate Variability Posters
Tuesday, 16 January 2001, 2:15 PM-3:30 PM
Previous paper Next paper