Flash floods in Bangladesh – and the way forward for the anticipatory action community
Bangladesh has extensive experience in acting ahead of hazards and extreme events. But when severe floods hit in June 2022, no anticipatory action protocols were implemented. This article explains why – and considers what could be done differently in future to ensure that people are protected ahead of a hazard.
The flash floods
In June 2022, the north-east region of Bangladesh was hit by the worst floods in recent memory, with Sylhet and Sunamganj districts most severely affected. These flash floods were partly caused by record rainfall levels in the Indian states of Assam and Meghalaya, which lie upstream on rivers that flow into Bangladesh. These two states received 2,500mm of rain over three days, with Mawsynram – in Meghalaya, the wettest place in the world – receiving 972mm of rainfall in just 24 hours on 17 June. This was the third highest 24-hour amount of rainfall in June for 122 years (the highest was 1,563.3mm, recorded on 16 June 1995; the second highest was 973.8mm on 5 June 1956). At the same time as the upstream rainfall, there was heavy rain in Sylhet and Sunamganj, with the latter recording 185mm in 24 hours.
The ensuing floods affected 7.2 million people, with over 4.8 million displaced in nine districts of north-east Bangladesh. Among other impacts, this extreme event caused severe damages to crops, with 83,394 hectares damaged; to livestock, with losses worth 28.1 million US dollars; and to washing facilities and water sources (Need Assessment Working Group). By 20 June, the Directorate General of Health Services had recorded 2,492 cases of disease and injury, and according to the United Nations (UN) in Bangladesh, there were around 60,000 pregnant women in the affected area.
The flash floods in June were the third such event in the region during 2022, following earlier events in April and May. For each event, the Flood Forecasting and Warning Centre and the Bangladesh Meteorological Department worked together to forecast both rainfall and floods. A flood forecast with a two-day lead-time was issued for the floods in April, and with a five-day lead-time for May. However, while the rainfall in June was forecast three days beforehand, it turned into flash floods very rapidly due to the hilly topography and steep slopes in Bangladesh. This caught people in the region by surprise.
The humanitarian response
The government of Bangladesh, the Bangladesh Red Crescent Society, and several UN agencies and non-governmental organizations (NGOs), both national and international, started their responses immediately after the floods hit. A total of 82 humanitarian agencies co-developed a Joint Response Plan to consolidate their efforts into a coordinated appeal to manage the flood response in a more effective way. This plan targeted reaching 1,521,741 people and, as of 31 July 2022, these combined efforts had reached 450,498 people (30 per cent of the target).
Why were anticipatory actions not implemented?
Anticipatory action is an effective way to support vulnerable communities before a hazard strikes, and it has already been piloted for different hazards in Bangladesh. These include floods in the Jamuna River Basin, cyclones in the coastal belt, as well as landslides, river erosion and dengue in different parts of the country. Anticipatory approaches were not implemented ahead of the June flash floods in the north-east, however. To understand why, we first need to understand how anticipatory action works and what is needed to act early.
Bangladesh is ranked seventh in terms of countries affected by extreme weather events (1998-2017). While flash floods are a known risk, anticipatory action can only be implemented for hazards with existing forecasting systems, reliable data and predictions. Establishing a successful anticipatory action mechanism also requires huge efforts to design and test pilot projects beforehand, with the most crucial part being the development of a pre-agreed forecast threshold that gives sufficient lead time to act early before the impact. But flash floods are very difficult to predict with sufficient lead times – and without these, anticipatory action is not possible.
The anticipatory action community in Bangladesh has already developed models for various hazards. But the existing anticipatory action models – such as the one for the Jamuna River Basin – are designed specifically for riverine flooding and furthermore didn’t target the north-east of Bangladesh. There are further challenges with modelling flash floods in the north-east; for example, they are in part determined by rainfall in India and, due to the hilly topography and steep slopes in Bangladesh, can turn suddenly into flash floods.
The way forward
Considering Bangladesh is one of the countries most vulnerable to the impacts of the climate crisis, such extreme flash floods are expected to hit the north-east region more frequently in the future. This calls for forecasters, technical agencies and the wider anticipatory action community to work together to devise proactive measures. There is a need to further include anticipatory action in interventions for disaster risk reduction and climate change adaption, and to explore if and which early actions might be possible to reduce the impacts of flash floods.
While forecasting flash floods is still a challenge, technical agencies are working to develop and improve forecasts for such events in the region. This is, as noted, challenging due to factors such as complex topography, a lack of discharge data for stations in the region, and the limited accessibility of upstream/transboundary forecasts.
Alongside these efforts, developing community-based early warning systems could be a first step to providing some warning to communities, so that they can evacuate in time to save lives. This would be a good start, even if other anticipatory early actions will only be possible once forecasting improves.
Is this page useful?Yes No Report an issue on this page
Thank you. If you have 2 minutes, we would benefit from additional feedback (link opens in a new window).