Ainun Nishat, PhD, Professor Emeritus at BRAC University and a leading authority on water resource management and climate change in Bangladesh, speaks to Khairul Hassan Jahin of The Daily Star about the systemic management failures behind the recurring haor crisis.

The Daily Star (TDS): What makes the haor region structurally vulnerable to early flooding, and how has its agricultural system historically adapted to that water cycle?

Ainun Nishat (AN): Bangladesh’s north-eastern haor region comprises seven districts that geologically belong to a single category.

A defining factor is the massive rainfall it receives upstream. Across the border lies Cherrapunji, historically the wettest place on Earth. This geological formation creates a natural funnel: India’s high, hilly terrain channels water into Bangladesh’s low-lying plains, including Tanguar Haor and other districts.

Knee-deep in stagnant floodwater, a haor farmer races against time to cut and gather ripening paddy. Photo: Mintu Deshwara

Heavy rains in the Indian hills trigger rushing flash floods. While this year’s devastation evokes my deepest sympathy for the farmers, the severe impact fundamentally stems from management failures rather than nature alone.

Pre-monsoon rains typically begin in mid-April (Boishakh). This natural, revitalising phenomenon awakens biodiversity, prompting fish to spawn and various insects, animals, and birds to breed. It is an entirely normal cycle.

However, the inevitable late-April inundation leaves a narrow window for agriculture. To avoid damage, haor farmers must ideally complete their paddy harvest by mid-April, or Pohela Boishakh.

Over the last century, the haors were an open, wild landscape. Historically, regular late-April and May flooding meant farmers focused on a single, modest Boro rice crop.

My research shows that crops were destroyed three out of every four years. Despite this risk, the successful years yielded such an abundance that they sustained the farming communities of Sylhet, Netrokona, and Kishoreganj. This high-stakes gamble with nature has defined the region’s agricultural heritage for generations.

A haor is a backswamp. Between rivers descending from the hills lie depressions and wetlands, including numerous beels. During the monsoon months of Ashar and Shraban, the entire region is submerged under a vast body of water. During winter, only isolated water pockets remain; Tanguar Haor alone contains around fifty named pockets.

Crucially, the entire area is saucer-shaped. Around the rim are relatively flat lands, then rivers, with canals connecting to the beels. People build homes on slightly elevated areas along the riverbanks.

Thus, even when monsoon waters submerge the region, these densely populated settlements remain above water. Viewed from above, roofs practically touch one another, with people packed tightly onto these tiny, elevated patches of land.

As the water begins to recede, it drains from the entire beel region. Historically, water flowed through rivers, and once the rivers receded, the beels dried up. Because the beels are saucer-shaped, farmers first planted rice on the highest land (kanda), gradually moving lower as the deepest depressions dried and were planted last.

When river floodwaters returned, the lowest areas flooded first, followed by the middle areas, the upper beel parts, and finally the settlements. Consequently, the lowest areas—planted last—submerged first, leaving those crops with the least time to mature.

TDS: How did submersible dykes and embankment management become part of the current crisis?

AN: If we attempted to stop full monsoon flooding, we would need embankments 20 to 25 feet high. That would be enormously expensive. Around 1972–73, this problem was carefully analysed, and with Dutch expertise, we introduced submersible dykes. This was done experimentally. With permanent embankments, rivers would lose their natural connection with wetlands, and fish migration would also be disrupted.

With a submersible dyke—which remains underwater during the monsoon—wetlands still fill with water naturally. Fish stay in the beels during winter, so ecological connectivity remains intact.

The government wisely decided that embankments eight, ten, or twelve feet high would be sufficient to withstand the first shock of pre-monsoon floods—the sort we are seeing now. The Bangladesh Water Development Board planned to delay flooding until 15–30 May. This is where the problem began.

In years when the rice crop was good and the major flood arrived at the end of May, farmers harvested successfully. In years when rain came slightly earlier, embankments trapped the water, preventing drainage. In such circumstances, embankments should be cut open so water can escape whenever outside river levels are lower, because water levels naturally fluctuate.

Once harvesting is complete, if water arrives slightly later, villagers generally tolerate the situation because fish production becomes abundant. Since lower embankments do not prevent fish movement, fishing can become highly profitable.

Then came the fisheries leasing system. The beels were leased out to private individuals. Returning to the earlier point: by September or October, river water levels fall rapidly, but water trapped within submersible dykes cannot drain because the area remains enclosed by embankments. At that point, villagers cut the embankments themselves. This is a sensible practice. However, government authorities view it as unlawful tampering.

Once the embankment is cut, water drains away. The faster the water recedes, the sooner the land becomes suitable for cultivation. If villagers cut embankments and drain water by November, then December, January, February, and March provide four excellent months for cultivation.

After March, water will begin entering again. Therefore, embankments must then be repaired where they were cut.

Villagers often complain that the Bangladesh Water Development Board creates unnecessary complications because, under official procedures, by the time tenders are invited and processed, March has already arrived. How, then, can embankments be repaired in time?

Responsibility for repairing embankments should lie with the local villagers themselves. They are the ones benefiting from the system. They know exactly where canals are located and where embankments should be cut each year. The same places can be used repeatedly. The main challenge in rebuilding embankments is a shortage of soil.

When the embankment is cut, excavated soil is stored during October. Before March—during January or February—the gap can be closed again. In that case, any flooding that occurs afterwards may be caused solely by local rainfall, not by river water.

One must understand the region’s topography and river behaviour. The water level rises, then gradually falls. Afterwards, in Joishtho, it rises again to another level. That is the main flood period, when everything becomes submerged. Haor people do not particularly mind this seasonal inundation because it produces enormous quantities of fish.

Therefore, ownership of these fisheries should not be handed over to private parties. They ought to remain a common property resource. Whoever can catch fish should be allowed to do so.

TDS: How are climate change, forecasting gaps, and agricultural choices changing the nature of the crisis?

AN: The losses recently suffered in the haor region are nothing new. The same thing happened in 2017. After 2017, it happened several more times. We have repeatedly pointed out that climate change has intensified the situation.

Due to climate change, rainfall intensity, frequency, duration, and timing are changing. Timing, in particular, is shifting. Over the last few years, rainfall during Ashar and Shraban has not been normal. Instead, excessive rainfall occurred during Bhadro and Ashwin. Rainfall even began in March, during Falgun. Normally, Chaitra receives only scattered showers. But these early heavy rains occur in regions like Meghalaya and Arunachal Pradesh, where monsoon influence arrives much earlier.

We now need to rethink the entire system from the ground up. Government officials generally do not think this way; they need to be pushed to do so. The Bangladesh Meteorological Department should think about it first.

Presently, the weather office provides 12-hour and 24-hour forecasts. Recently, they began issuing indicative seven-day forecasts. If the United States can provide three-month forecasts, we should adopt that technology. Why can’t we provide reliable 10- or 15-day forecasts? They are giving indicative forecasts, but these are unreliable. So long-term forecasting is essential.

As mentioned earlier, farmers plant rice in January and February, aiming to harvest by late March or April. For that, short-duration rice varieties are needed. Such varieties already exist. Bangladesh Rice Research Institute developed them.

We also need to properly understand the relationship between rivers and beels. At one time, embankments must be cut open, and at another, sealed again.

The government’s solution is to construct sluice gates. But the design of these sluice gates is fundamentally flawed. In many cases, no one can properly operate the gates, and constructing enough sluices to adequately manage the system is financially impossible.So water management must be handled differently. I now place responsibility directly upon the Bangladesh Water Development Board.

The Bangladesh Water Development Board still operates exactly as it did in the 1970s. They begin repairing embankments in March, whereas I already said this work should be done in February. Instead, they carry out repairs during March and April. Then, when sudden floods arrive in late March, embankments cannot withstand the pressure and collapse. If the embankment breaks or repair work is damaged, blame is shifted onto nature. The explanation is that the contractor could not complete the work in time.

For that reason, during the 2008–09 caretaker government, the water resources adviser introduced a rule requiring all embankment work to be completed by 30 March. After that date, no further embankment work would be permitted.

The adviser in the previous interim government was herself an environmental expert. Since she understood water management, why was the embankment work not completed by the end of March this year?

TDS: Recently, cropping practices in the haor region have been changing. For example, maize cultivation is increasing. Does this carry any particular implications?

AN: This is precisely the harmful part. Maize is cultivated on higher land and even in some low-lying areas. Once water recedes, especially on medium- to high-lying land, which is not necessarily flood-free because even flood-free land may become submerged, farmers begin additional cultivation.

My suspicion—and I would happily be proven wrong—is that the Ministry of Agriculture advises farmers to plant Boro rice in February and grow maize or mustard during the extra month and a half or two beforehand. But by doing this, farmers delay harvesting rice. Instead of harvesting by 15 April, they push the harvest to 15 May or later. So, you are increasing vulnerability.

That is why I said earlier that the first responsibility lies with the Bangladesh Meteorological Department to provide warnings. The second responsibility lies with the Ministry of Agriculture, Bangladesh’s largest ministry. It has thousands of staff, including union-level personnel and several upazila-level gazetted officers.

Surely, they should properly advise farmers. They should say: If you cultivate maize, your rice cultivation becomes risky. If your land is relatively high, you may be able to manage it. But if your land is low-lying, after maize, your rice crop may or may not survive.

Farmers must receive this warning. These are forms of adaptation.

The National Adaptation Plan clearly states that farmers in these regions should cultivate early-ripening rice varieties. Previously, major floods arrived in May, but this year they came in April. In fact, they arrived about a month earlier, beginning in March. That is entirely normal under changing climatic conditions.

The entire haor region should not be treated as a single planning unit. Instead, it should be divided into multiple sub-planning units.

TDS: How are institutional failures, from the Haor Development Board to local embankment and sluice gate management, deepening the crisis?

AN: In our country, the word “development” itself is dangerous. If I had the authority, I would rename all development boards today. They should be called Management Boards. Once something is developed, it must be managed.

We want to develop roads. Fine, we build first-class roads. But within three years, they are filled with enormous potholes because nobody manages them properly. Roads are not even shaped correctly to drain rainwater.

All this must be environmentally compatible and properly managed. Why should the Haor Development Board not manage embankments in cooperation with local communities? Why should local people not be actively involved?

We lack long-term weather forecasts. We lack institutional infrastructure. And there is confusion over operational responsibility.

Ten or fifteen years ago, during the caretaker government, responsibility was assigned to the Upazila Nirbahi Officer (UNO), who delegated it to local Union Parishad chairpersons.

But Union Parishad chairpersons lack the necessary technical knowledge. Technical expertise exists within the Bangladesh Water Development Board. The Board became resentful because it had previously controlled everything, but now responsibility had shifted to Union Parishad chairpersons.

Frankly, I cannot say with confidence that most Union Parishad chairpersons or members are honest. Consequently, there are leakages and corruption within the system.

The designs of these sluice gates are completely wrong. The gates cannot be opened properly. Their sill and bottom levels are incorrect. For water to drain effectively, the bottom level must be low enough. But in many cases, the bottom level is too high, preventing proper drainage.

Therefore, sluice gates connected to these submersible dykes under the Bangladesh Water Development Board must be reviewed and redesigned immediately. Urgently. We need to identify which functions and which do not. Then comes the question: who is responsible for managing them?

At present, no one is effectively responsible. Once, a post called Khalashi was assigned specifically to operate each sluice gate. Then, a secretary, through a single decision, reduced staffing levels from 24,000 to 7,000 employees. Many lower-level posts—guards, attendants, sluice operators—were abolished. Previously, an executive engineer could order a sluice gate to be opened or closed. Now there is no staff to carry out those instructions.

Later, the idea emerged that local village committees would manage the sluices. But those committees do not function effectively either.

TDS: Beyond flood control, what should a serious haor policy address for farmers and communities?

AN: Ten, fifteen, or fifty years ago, the major challenge was harvesting the rice itself. Vast paddy expanses stretched across enormous fields. Harvesting by hand took immense time—often 1 to 1.5 months to cut and bring the crop home.

Fortunately, over the last two or three years, combine harvesters have enabled much faster harvesting. But another problem remains: where will the rice be dried afterwards?

Drying requires elevated platforms. Therefore, I propose building drying yards alongside residential settlements. Presently, where roads exist, people dry rice on the roads themselves. Proper rice-drying infrastructure must be provided.

Wet rice can still be harvested quickly after soaking this year. But if it remains fully submerged for too long, seedlings begin sprouting from the grain, and the rice turns black.

So, we must think about rice drying. We must also consider fisheries, as the haor region produces enormous quantities of fish. I have seen huge fish-drying racks opposite Sunamganj town. There are large numbers of small fish, such as puti and similar species. Therefore, if we take the haor region seriously, we must consider many interconnected issues.

I have worked extensively in the haor region, especially Tanguar Haor. One thing I noticed was the complete lack of vegetables and nutritional diversity. Many things need reconsideration: marketing, fishing, ownership systems, flooding, and its timing.

Have you heard the term “afal”? When water suddenly produces large waves during strong winds, the waves strike settlements. The wind may stop, but the waves continue moving due to a resonant effect. So those villages also need protection from these wave impacts.

Since we are specifically discussing the haor region, not flooding in general, let me mention another issue: dry-season water scarcity. During winter, there is often no drinking water.

Imagine one village on an elevated riverbank, and another on another raised riverbank. Between them lies a sloping area leading down to flat land and eventually the beel. Water remains only in the beel itself. Consequently, villagers may walk three kilometres to collect drinking water.

Yet because this is considered a wetland, the government often does not install tube wells. The assumption is that water exists everywhere. That is misleading. For five or six months, water is abundant, yet for three or four months, people suffer severe shortages.

There is also almost no sanitation infrastructure.

Regarding human suffering, in our country, the government primarily focuses on relief distribution. They announced that three crore people will be fed for three months. But simply deciding who exactly should receive assistance may itself take three months.

There is another issue in the haor region: social structure. There are three groups of people. First is the local population. Second are the “abadi” settlers—migrants who, over the last forty or fifty years, came from places like Noakhali, Faridpur, and Barishal to cultivate land.

Sylhetis themselves generally are not inclined towards heavy agricultural labour. The same applies to people from Netrokona. Across the border in Meghalaya and Assam, local people also tend not to do such labour-intensive work. So settlers—the abadi people—carry out much of the cultivation. These abadi settlers are often looked down upon as lower-class.

The third category is “jirati”—people whose land and property have eroded, leaving them with nothing. They live in makeshift huts wherever they can. During the monsoon, they often stay in towns, and after floods, during planting and harvesting seasons, they return to sell their labour. We must also understand these social dynamics. Who exactly are we trying to help?

Send your articles for Slow Reads to [email protected]. Check out our submission guidelines for details.