Flood Forecasting & Warning Centre
Bangladesh Water Development Board (BWDB)

Water Level : 2.85
Danger Level : 9.50
Water Level : 8.32
Danger Level : 13.00
Water Level : 1.51
Danger Level : 6.25
Water Level : 5.38
Danger Level : 13.50
Water Level : 2.72
Danger Level : 11.25
Water Level : 12.60
Danger Level : 21.00
Water Level : 2.88
Danger Level : 8.65
Water Level : 3.99
Danger Level : 12.75
Water Level : 5.72
Danger Level : 14.25
Water Level : 12.74
Danger Level : 19.50
Water Level : 26.17
Danger Level : 30.00
Water Level : 22.82
Danger Level : 26.50
Water Level : 2.31
Danger Level : 5.80
Water Level : 5.38
Danger Level : 12.50
Water Level : 66.57
Danger Level : 70.75
Water Level : 6.41
Danger Level : 13.35

Increase the speed of the ticker
Set the ticker's scrolling direction to right
Severe Danger Warning Normal South Eastern Hill Ganges Meghna Brahmaputra Unit (mPWD)
River Based Map | Division Based Map | District Based Map | Google Map

Catchment Area:

A river's catchment (or basin) is the land area from which rainfall will ultimately contribute to the river discharge. The catchment area of the Ganges, Brahmaputra and Meghna Rivers are 907 X 103 km2, 583 X 103 km2 and 65 X 103 km2, respectively, of which only 8% lies in Bangladesh. More than 90% of the water that flows into the Bay of Bengal enters Bangladesh through its borders with India.

Danger level:

In Bangladesh danger level at a river location is the level above which it is likely that the flood may cause damages to nearby crops and homesteads. In a river having no embankment, danger level is about annual average flood level. In an embanked river, danger level is fixed slightly below design flood level of the embankment. The danger level at a given location needs continuous verification as e.g. embankments may be breached, but it is not done continuously by FFWC, whereby some danger levels may be not precise.

El Niño:

An El Niño/Southern Oscilation (ENSO) episode is a disruption of the ocean-atmosphere system in the tropical Pacific having important consequences for the weather and climate around the globe. Abnormally warm waters in the equatorial central and eastern Pacific characterize an ENSO episode. During ENSO episodes abnormally dry conditions over northern Australia, Indonesia and the Philippines are found, and the Indian monsoon rainfall tends to be less than normal. It is possible that ENSO episodes influence flood conditions in Bangladesh and research project are investigating the connection.


The floods in Bangladesh are divided into monsoon river flood, flash flood, local rainfall flood and storm surge flood. Monsoon river flood is an annual event forced mainly by intensive river inflow through Ganges, Brahmaputra and Meghna Rivers and rainfall over Bangladesh as causes the water level in the rivers to rise and fall slowly during the monsoon season. Flash flood occurs only in the northeastern Bangladesh in the period pre- to post-monsoon forced by intense rainfall in the Meghalaya Hills and in parts of eastern Bangladesh in the post-monsoon. Local rainfall flood is, as the name states, forced by local heavy rainfall over a location inside Bangladesh. Storm surge flood is a coastal phenomenon forced by cyclones hitting the Bangladeshi coastline.

Flood forecasting:

To predict water level conditions in Bangladesh FFWC collects measurements of water level and rainfall, satellite pictures and simulates the water level conditions by use of a numerical model of the Bangladeshi river network. Every day during most of the monsoon season the model simulates the water level conditions during the previous 7 days (hind-cast simulations) and during the coming 3 days (forecast simulation). More precisely the forecasting starts during early monsoon when one of measuring stations show a water level 60 cm below danger level. For obvious reasons no measurements exist in the forecasted period and simple relations estimate boundary conditions for the numerical model during this period. The uncertainties of the estimated boundary conditions propagate into the model domain. Results from the model simulations are used to provide flood forecasting and warning.

Goodness of results:

The outcome of the work at FFWC is to at large extent water levels. The water levels can be divided into measured water levels and simulated water levels. The simulated water levels can be further divided into results from hind-cast simulation and results from forecast simulation. The water levels are sometimes transformed before being presented. Water levels are e.g. interpolated to produce inundations maps and other maps. Measured, hind-casted, forecasted and transformed water levels are all determined with a given precision, but also influenced by errors. Measured water levels as collected in Bangladesh have a precision of approximately plus-minus 10 cm. The precision of a measured water level is the sum of the precision of the leveling of the station and the precision of the reading of the instrument (often a wooden stick). Sometimes the gauge reader e.g. misread the instrument or writes a wrong number and it is then considered an error, and the water level can be several meters wrong. Many of the erroneously measured water levels are found while quality assuring the water levels at FFWC. Water levels from hind-cast simulations are also not precise. The Super Model is forced and adjusted by not precise measurements and the model is not prefect. Thus the precision of the hind-casted water levels is some integration of measurements precision and model precision. Water levels from forecast simulations are not precise for the reasons given at hind-cast simulation and further because the Super Model boundary conditions needs to be extrapolated in time. The boundary conditions include water level boundaries at the Indian border and at the Bay of Bengal and rainfall over Bangladesh. The uncertainty of the extrapolated water level boundaries propagates as flood waves through the rivers into Bangladesh. Especially the upstream water level boundaries at the Indian border limits the possible forecast time. Today it is not possible to forecast the flood level in Bangladesh more than 3 days ahead. If the boundaries could be moved further upstream, i.e. inside India, it would be possible to give longer forecast. When water levels are transformed, it may be measured, hind-casted or forecasted water levels, the outcome is not precise as the used water levels are not precise and because the transformation most often also demands use of other not precise data. To produce e.g. a flood inundation map for Bangladesh a digital elevation map (DEM) of Bangladesh is required. Thus the precision of flood maps is some integration of used water level precision, DEM precision and interpolation precision.

Green house effect:

The green house effect is a change in the global climate due to mans activities. The IPCC "Business-as-Usual emissions scenario" predicted water level rise forced by green house effect is about 20 cm in global mean sea level by 2030, and 65 cm by the end of next century. There will be significant regional variations. Also the climate of Bangladesh may get warmer and wetter. The monsoon rainfall is estimated to increase 10-15% by the year 2030. Little is known on changes in cyclone frequency and intensity. Presently FFWC does not investigate the consequences of the green house effect.


Bangladesh is characterized by a tropical climate, i.e. 'cool', dry winter from November to January with predominantly northeasterly monsoon winds; hot, humid summer from April into October; and wet monsoon from June to September with predominantly southwesterly monsoon winds. Flooding in Bangladesh occurs during the wet monsoon period.

Super Model:

FFWC simulates the water level conditions in Bangladesh by use of a numerical model of the Bangladeshi river network. The model is named the "Super Model" and is based on measurements of the topography of Bangladesh and the general numerical hydrodynamic model system named MIKE 11 developed at DHI Water & Environment. To run the Super Model information on water level and rainfall is needed to force the model: water level boundaries are e.g. found upstream in Ganges and Brahmaputra and downstream in Meghna River. The Super Model has been run at the FFWC since 1998.

Flood Condition:

FFWC disseminates flood warnings during most of the monsoon season. The warning is related to the measured and forecasted water levels and the danger levels: a) Normal condition: water level at any particular location(gauge point) is more than 50 cm below the danger level at that point; b) Warning: water level is below but within the 50 cm of the danger level c) Flood: Water level is above danger level and 100 cm above danger level; and d) Severe Flood: water level is more than 100 cm above danger level. Warnings are disseminated through a daily flood bulletin, e-mail, FFWC home page, IVR, newspapers, radio and television.


Water level:

BWDB and other government departments refer water levels to the Public Works Datum (PWD). PWD is a horizontal datum believed originally to have zero at a determined Mean Sea Level (MSL) at Calcutta. PWD is located approx. 1.5 ft below the MSL established in India under the British Rule and brought to Bangladesh during the Great Trigonometric Survey.