Bangladesh: Monsoon Floods 2004 - Post-flood needs assessment summary report

Originally published


Executive Summary
The Floods

The 2004 monsoon floods commenced around 8 July. This followed early flooding in the northwest districts of Bangladesh in April, that had destroyed much of the main annual rice crop in that region just before it was harvested. They rose more swiftly than usual in the same area, and in the northern districts either side of the Brahmaputra/Jamuna River. Water persisted in these regions for some 3 to 4 weeks whilst gradually draining southwards, severely flooding most of Central Bangladesh and the Greater Dhaka area, especially districts adjacent to the two confluence points of the three great rivers.

The high water level and widest extent of the flood - about 35,000 km2 - was reached on 24 July. The water had receded in most places by mid-August, although it persisted and stagnated in areas behind protective embankments without adequate drainage. In total 39 out of 64 districts and 36 million people (25% of the total population) were affected.

In mid-September, a localised depression caused continuous torrential rain and high winds over a six-day period, bringing renewed flooding to many parts of Central Bangladesh, but also flooding areas never normal flooded by the rivers, including Dhaka and other urban areas and some of the most productive agricultural land.

The worst impact of this rain-fed flooding were in Comilla and Feni to the east, where the Meghna River breached embankments already weakened by the earlier floods, and in the southwest border districts of Jessore, Satkhira, and Magura and parts of three others. Here the impact was a combination of the rainfall, water released from West Bengal, and poor drainage because of the region's low elevation and inappropriate coastal protection.

Although the monsoon normally ends by mid-October, there is still the risk of heavy rainfall upstream in the catchment areas of the major rivers, and the attendant risk of water being released without warning as occurred in 2000. The cyclone season then prevails until the end of the year, posing a significant threat to coastal communities.

The Assessment

The Disaster and Emergency Response (DER) LCG Sub-Group conducted a quick assessment at the height of the floods, 20-27 Jul 04. By late August the situation had changed as the floodwater receded and more information was required about relief and recovery needs, so the Sub-Group decided to conduct a more thorough assessment. Its aim was to obtain accurate independently verified information on the short, medium and longer-term needs of the poor specifically caused by the floods, to enable planning, budgeting and implementation of relief and rehabilitation programmes in each sector.

Eight multi-agency teams comprising volunteers from 17 different agencies were tasked to assess the 27 worst-affected districts across the country. The assessment methodology placed equal emphasis on information directly from the flood victims themselves, and from officials at district and upazila level. Some 88 focus group discussions were conducted with communities, as well as interviews with hundreds of officials. It took place 6-15 Sep 04, with a further 10 days to consolidate information and prepare reports.

This information was then entered onto 79 separate upazila report forms, and combined with district-level information to produce 27 district summaries, now available on the LCG website <>. Based on these, the findings were summarised across the following sectors: Health and Nutrition; Water and Sanitation; Agriculture; Food; Shelter and Housing; Education; Economic Recovery and Infrastructure.

Taking into account the limitations of time, lack of training, and unfamiliarity with an untried questionnaire format, the output of the teams was good, and will provide a basis for future development in this area. Teams pointed out that time for the assessment was very tight given the scale of the task and that interview respondents often could not provide the kind of data they were seeking. This suggests that local officials may be collecting data of limited value for determining appropriate responses to disasters.

Main Conclusions and Recommendations

In the health, nutrition, water and sanitation sectors it is hard to distinguish acute flood-related needs from the chronic 'normal' situation. Similarly, riverbank erosion is an ongoing problem in many districts, creating a growing number of completely destitute families probably worse off than the victims of the recent flooding. Nevertheless, in general this assessment corroborated the findings of the quick assessment in July, with the overall level of needs increasing due to the second episode of flooding in September.

People are highly resilient and self-reliant, and they have a number of coping strategies such as advance selling of labour or migration, although these options are not open when so much of the country is similarly affected. Whilst the immediate national response was fair given the circumstances and resource constraints, the overall response was not commensurate with the needs in any sector. For instance the 10 kg/family/month rice ration is meagre, and in practice results in people receiving only about 40 gr./person/day. Resource constraints prevented even the Agricultural Rehabilitation Programme from meeting more than a small proportion of the full needs with a comprehensive package.

The greatest area of concern is that the flood will push large numbers of the poorest families deeper into poverty. They will need targeted assistance if they are to avoid this, including food relief, support to agriculture employment opportunities and micro-credit. In addition, many will not be able to rebuild their homes without external assistance.

Urgent response actions are still required in some areas in terms of bulk food relief and supplementary feeding for vulnerable groups, support to housing rehabilitation and sanitation, and emergency healthcare interventions. Beyond this districts will need assistance with the significant costs to repair or reconstruct public infrastructure such as road networks, health centres and school buildings. Employment generation, whether through Food / Cash For Work or support to cottage industries, and widespread provision of affordable credit will be essential for recovery from the floods. In all cases the new infrastructure should be able to resist future major flooding.

Bangladesh clearly needs to improve disaster response and preparedness at local level, with provision of immediate rescue resources, emergency funding mechanisms, and better information management and contingency planning. In the medium to longer term, more emphasis is needed on mitigating and managing future flood disasters rather than attempting to prevent them completely. A large part of this is advocacy for a safer future through appropriate development that reduces rather than increases vulnerability.

BANGLADESH - Monsoon Floods 2004

DER Post-Flood Needs Assessment Report, Sep 04


Development of the Flood Emergency

1. The 2004 monsoon arrived early and Bangladesh has experienced heavy rainfall since late June. Persistent rainfall within its borders but, much more significantly, in the vast 1.72 million km2 catchment basins across India, Nepal, Bhutan and China, caused first the Surma River which feeds the Meghna, and then the Brahmaputra which becomes the Jamuna in Bangladesh, to register rises above their danger levels from around 8 Jul 04. The river rises caused widespread flooding in the flat Northeast 'haor' area, which had already lost 75% of its main annual crop (the 'boro' variety of rice) through unexpected flooding from 12 Apr 04.

2. The wave of flooding across the haor floodplain that commenced on 8 Jul 04 started falling in mid-July, but then rose again due to upstream rainfall in Assam and Meghalay States of India, reaching a peak around 24 Jul 04. Water flow down the Brahmaputra/Jamuna to the west had increased a few days later than in the haor but reached its peak in the northern districts around the same date.

3. From mid-July water started to drain slowly from the northeast and the northern districts, and to rise in the central districts around Dhaka, at the confluence of the Ganges/Padma and Jamuna, and their confluence with the Meghna. Districts along these rivers experienced severe flooding until the second week of August. The floodwaters overtopped many embankments and raised roads, inundating and stagnating in areas not well provided with drainage. This included Dhaka City, where waters persisted in many residential and slum areas until the third week of the month.

4. According to Government of Bangladesh (GoB) figures 39 out of 64 districts, 265 out of 507 sub-districts (upazila), 35,000 km2, and 36 million people (25% of the total population) were affected. The official death toll, however, was under 800, a mark of the effectiveness of disaster response measures and emergency health interventions.

5. The flooding caused huge dislocation to normal life: homes, courtyards, kitchen outhouses, latrines, livestock sheds, haystacks, vegetable gardens, and paddy fields were inundated, damaged and eroded. Many people had to move into makeshift shelters on higher drier land, into a school building or with relatives; others chose to live in their flooded homes to safeguard their property; whilst still others migrated to the cities. Road and rail transport, power supplies and telecommunications were disrupted in many parts of the country, and the receding water in August revealed extensive damage to agriculture, fisheries, housing, educational institutions, to roads, bridges and culverts, and to embankments and flood-protection systems.

The Current Flood Situation

6. From 10 to 16 Sep, a localised low-pressure depression centred over Bangladesh dropped up to three times the normal rainfall, including 341 mm of rain in Dhaka in one 24 hour period (13/14 Sep 04), the heaviest rainfall recorded in 50 years. This intense rainfall caused another round of flooding in Dhaka and central and southwest districts, killing another 19 people and unknown number of fishermen - possibly hundreds - in the Bay of Bengal. According to GoB more than a million people were again isolated by the floodwaters, whilst slum-dwellers in Dhaka were forced to flee their homes and set up makeshift shelters in the streets for a second time.

7. This flooding was of different character to that in July and August, affecting districts away from the main rivers in the southwest that had previously been spared, and in some places where cultivation had previously been hampered by a lack of water. The rain also caused sudden widespread inundation and water-logging behind the protective embankments and brought the Meghna into spate; with the Gumti and Kakri tributaries bursting their banks and flooding Comilla and Feni Districts. The south-western border districts of Jessore, Satkhira, Magura and others have also now been inundated, causing major losses to newly transplanted paddy fields and vegetable beds, as well as to fish-farming.

Projected Evolution of the Flood Emergency

8. The monsoon normally ends by mid-October with the onset of the cyclone season, and more heavy rainfall can be expected across the great South Asia river basins until then. The most immediate threat is from a sudden onrush of more water from India down the Ganges and also in lesser rivers flowing into the southwest border districts of Jessore, Satkhira and Magura. They have already experienced flooding due to the heavy rainfall and the rise of their smaller rivers, and are now seriously threatened by vast amounts of water that may yet be suddenly released - perhaps due to spontaneous local action - to take the pressure off dams and embankments in West Bengal. This occurred without warning in 2000 and is still a real concern.



9. The Disaster and Emergency Response (DER) Sub-Group of the Bangladesh Local Consultative Group (LCG) conducted a quick assessment of 30 districts divided into 6 zones, from 20 to 27 Jul 04. The draft report was produced on 28 Jul and was deemed valuable because it was timely, broad in coverage, and represented a DER consensus position on the severity of the flood impact.

10. The possibility of further assessment was raised at the DER Group meeting on 11 Aug, justified by the fact that the first DER assessment was quick, did not reveal enough about needs, and much of its data was taken straight from the district authorities without any independent verification. Furthermore the overall needs were believed to have changed and possibly grown as the water receded and revealed the full impact of the floods, and more accurate assessment could help to target the ongoing relief effort as well as identifying sectoral or geographical gaps.

11. The DER Group decided at its meeting on 18 Aug that further assessment was required and assigned a working group to plan it. The DER multi-agency multi-sectoral assessment working group met on 23, 26 and 29 Aug 04 to determine the objectives, methodology and timeframe for the assessment.


12. The objectives of the DER Post-Flood Needs Assessment were to:

(1) Identify the short, medium and longer-term needs of the poor that were specifically caused by these floods, rather than the damage or losses.

(2) Obtain detailed information on the needs in each sector, to enable agencies to plan, budget, and implement their relief / rehabilitation programmes.

(3) Report accurate, independently verified information to give a clear picture of new needs created by these floods, on top of pre-existing chronic needs.

(pdf* format - 1.41 MB)