ÐÓ°ÉÔ­´´

How do you tame a river as wide as the Straits of Dover that moves its
course by hundreds of metres in a single year, eating up anything in its
path? The answer, say many hydrologists, is that you shouldn’t try. But
this year Western aid money will begin to pay for the first of a new generation
of high embankments designed to contain two of the largest and most violent
rivers in the world, the Ganges and the Brahmaputra, as they surge through
Bangladesh en route to the Indian Ocean.

Implementing the Bangladesh Flood Action Plan, drawn up two years ago
by the Bangladeshi government and the World Bank, is likely to cost around
$5 billion. The plan is a response to monsoon rains that flooded half the
country in 1987 and again in 1988. But some researchers cast serious doubt
on the project. The Eastern Waters Study, commissioned by the US government’s
Agency for International Development, castigates plans to channel the rivers
between high embankments as ‘engineering hubris likely to lead to massive
waste of scarce resources’. Barry Dalal-Clayton, field director of the International
Institute for Environment and Development in London, recently concluded
in a report on the plan that ‘it is difficult to be optimistic that man-made
interventions will be able to effectively constrain the powerful and highly
mobile rivers of Bangladesh’.

For Bangladesh, a country built on the delta formed by the rivers Ganges,
Brahmaputra and Meghna, floods have always been a fact of life. But they
are an increasing inconvenience for a nation attempting to modernise itself.
Flood control, the Bangladeshi government and Western banks agree, is a
prerequisite for investment in one of the world’s poorest countries. ‘If
you want a rural economy with workshops and industry, you need a reasonable
degree of safety for that capital investment,’ says Wybrand van Ellen, a
Dutch hydrologist and one of the foreign consultants overseeing the flood
plan.

But a trip on the rivers reveals the daunting reality behind this aspiration.
At Chandpur, a river port near the confluence of the Ganges and Brahmaputra,
the ferry station has continually had to be moved in recent years as half
the town has been ripped away by floods. In the space of a few days in 1988,
the river shifted its course by more than 500 metres, cutting a new channel
45 metres deep. Mosques, shops and part of the railway station were all
submerged and today the town, perched on a series of dykes that are surrounded
by water even in the dry season, is waiting to die. Could a new embankment
save it and prevent the same thing happening again? Peter Rogers of Harvard
University, author of the Eastern Waters Study, says that ‘no embankments
or river training works in the world can control these forces if they are
taken head-on’.

Waters beyond control

A few kilometres upstream from Chandpur, is the Meghna-Dhanagoda irrigation
project. Here $50 million has been spent over 20 years on ditches, banks
and pumping stations to keep out monsoon floods and to bring controlled
irrigation to 200 square kilometres of farmland. The aim is to ensure bumper
crops of high-yield varieties of rice that cannot tolerate flood waters.
The embankments tower above the fields, making the landscape look like a
tropical version of the Netherlands. But Europe has nothing to match the
rivers that surround the project on all sides. In 1988, the River Dhanagoda,
swollen by monsoon rains, washed away 200 metres of banks. It inundated
40 square kilometres of fields, plastering them with a thick layer of sand,
and carved out a new river channel 18 metres deep. Since then, new banks
have been built 3 kilometres from the river in places, in the hope that
they will then be safe from destruction.

But the rivers continue to encroach. Last summer, a year of comparative
calm, the river devoured more than 10 square kilometres in the Bera district
of Pabna, making 5000 families homeless-an event that merited just two paragraphs
in the Bangladeshi press. Huge beds of silt up to a kilometre long and 15
metres high have been recorded moving downstream en masse in the Brahmaputra
at rates of 600 metres a day during floods. One estimate is that the rivers
together bring down 2 billion tonnes of silt each year, transforming the
landscape.

Even in such an environment, the presumption persists that the floods
can be tamed. Just as environmentalists believe that halting deforestation
in the Himalayas will stop the great rivers from flooding in Bangladesh
(see ‘Floods in Bangladesh: who is to blame?’, this issue), so engineers
believe that they can quell those rivers.

During the 1960s, stretches of the right bank of the Brahmaputra were
protected with an embankment to stop flood waters from spilling across northwest
Bangladesh. The river took little notice of this impediment. Most of the
embankment has been rebuilt farther back in the past 20 years, but one section
still had to be rebuilt nine times. Under the first stages of the Flood
Action Plan, the right bank of the Brahmaputra is to be built higher and
a new embankment constructed on the left bank. As silt in the water settles
and the river bed rises, this construction work will make any breaches of
these banks even more dangerous and will probably create new threats downstream.

Nobody knows what these latest attempts to ‘train’ the tempestuous Brahmaputra
will do to the flow of the river downstream. The entire river shifted its
course dramatically at the turn of the century, when it largely abandoned
the old Brahmaputra and took a new, more westerly path. It is, says Rogers,
‘perhaps ready for a further major shift, which must not be provoked by
ill-judged works on the river’.

The flood plan proposes completing the upstream banks before trying
to protect the flood plain downstream. According to Ainun Nishat, professor
of civil engineering at the Bangladesh University of Engineering and Technology
and a member of the flood plan’s panel of independent experts: ‘The idea
is to embank the Brahmaputra, then the Ganges and then look for effects
of both on the combined reaches downstream.’ But, says Dalal-Clayton, there
are no baseline hydrological data to measure changes against, and the flood
plan includes no provision for collecting such data.

Opponents of the flood plan expect the worst. As the Eastern Waters
Study puts it: ‘Embankments do not reduce flood water, but merely move it;
excess water that is confined in the stream bed higher on the river will
increase the volume and velocity, and perhaps the depth, of the flow that
has to be managed in the districts closer to the sea.’ This depicts a dim
future for towns such as Chandpur and irrigation projects such as Meghna-Dhanagoda,
where flooding will probably be worse under the flood plan. And it suggests
that along the coast, where in 1970 river flood waters met a cyclone from
the Indian Ocean and killed up to half a million people, the risks of catastrophe
will also increase.

Van Ellen says the risks of devastation can be reduced by building ‘compartments’-self-contained
areas of land protected by embankments, rather like giant polders-behind
the main banks. ‘If there are high discharges of water coming downstream,
then the authorities can decide to flood a particular compartment in order
to protect downstream areas,’ he says. ‘The Chinese do this, though of course
Bangladesh is not such an organised society.’ Van Ellen agrees that it will
be hard to justify flooding some areas to protect others. And in a society
where public policy decisions are frequently driven by patronage and corruption,
it is hard to believe that, when the day of reckoning comes and land has
to be sacrificed, the technocrats will be in charge of the sluice gates.
The strategy is also at variance with another stated objective of the flood
plan, which is to delegate control of the operation of compartments to local
people.

The big mistake may have been to assume that it is desirable to banish
floods from Bangladesh. Engineers in charge of the flood plan say that the
floods bring nothing but harm. But, says Rogers: ‘Without the annual flood,
Bangladesh would find itself in very serious economic and environmental
straits.’ In a typical year, 20 per cent of the country is covered by water
during the monsoon season. Yet far from bringing disaster, in most years
the flood waters renew the soil. During the 1980s, crop yields rose fastest
on areas most prone to flooding.

Famine was predicted after the floods in 1987 and 1988; in fact, record
crops followed, largely because the floods left behind moisture, fertile
silt and algae. There was, of course, suffering too. Around 2000 people
lost their lives, hundreds of thousands were made homeless and property
was destroyed. But building embankments on a grand scale is not the cure.

There were other misconceptions. Government ministers and most journalists
blamed the floods on water pouring into the country from Nepal and India.
Yet the extent of the flooding was caused largely by the ponding of local
rainfall. Certainly, one reason for the water staying on the land was that
the rivers were high. But another, especially in the weeks following the
height of the flood, was that embankments prevented the water draining off
fields and back into rivers. There were many reports of farmers breaking
embankments and dykes in order to let their water out. Parts of the Khulna
district on the Ganges were still under water at the end of 1990, says Nishat,
more than two years after being inundated. Though a member of the flood
plan’s panel of independent experts, Nishat has mixed feelings about the
scheme.

Embankments of all sorts cover Bangladesh, preventing flooding on modern
farms and urban areas, and carrying roads and railways. All of them interrupt
the natural drainage of the delta region. Rarely are openings for flood
waters built and maintained. ‘Floods from local rainfall have been made
much worse by road networks,’ says Nishat. ‘The Water Board is supposed
to certify that none of these projects upsets drainage. But in practice
local pressure means they let them through regardless.’

Many modern irrigation projects in Bangladesh rely not on river water,
but on wells to tap the country’s copious quantities of underground water.
The growing use of wells is one of the great successes of modern Bangladesh,
says van Ellen. They are allowing a desperately poor country to escape famine
at a time when its population has doubled within a generation.

Yet the flood plan could end up undermining that success by threatening
the underground water stores that are currently replenished each year by
flood waters. The flood plan in theory provides for controlled flooding
of land to preserve ground waters, but it is unclear how this would be organised.

Lowered water tables will have far-reaching effects in a country where
most wildlife depends on a wet environment. Conservationists have railed
at the prospect of losing the Chalan Beel, an ecologically rich wetland
in the west of the country. ‘Wetland resources are also of great economic
importance to local people and to Bangladesh as a whole, and over 5 million
people are dependent on fishing for their livelihoods,’ says Dalal-Clayton.
Fishing supplies 80 per cent of the nation’s animal protein.

Most fish and shrimps hatch in rivers and estuaries but migrate to the
flood plains, where they feed and grow in rivulets and wet depressions,
known as beels, before returning to the main rivers. In many parts of the
country, says Nishat, land drainage and embankments have already ensured
that ‘natural stocking of beels has declined as connections between beels
and rivers are cut off’. The construction of an embankment at Chandpur reduced
local fisheries’ production by more than a third within two years. The flood
plan will prevent more fish making the trip and, as it lowers water tables,
will destroy the wetland habitats that the fish grow in.

A touchstone for the worth of aid projects is that they should help
the poorest people. But the flood plan appears destined to help the better
off in the cities of Bangladesh at the expense of the rural majority. Urban
areas need flood protection most, rural areas least. And in the countryside,
it is the richer landowners who want to invest in modern methods of irrigated
farming, using high inputs of chemicals that need protection from floods.
Many poor farmers and fishing communities, by contrast, will find their
crops and catches dwindling if the annual flood is lost. ‘Some would argue,’
says Dalal-Clayton, that the flood plan amounts to ‘a political response
to the clamour for flood protection from wealthier, influential, urban-based
groups . . . A great many people who presently live along the main rivers
and on islands in the river channels will be exposed to increased risks
from flooding’.

Nobody has bothered to count how many people live in these areas, though
Hugh Brammer, a British soil researcher and acknowledged authority on Bangladesh,
estimates that more than 6 million people could face additional risks. The
hand-ful of social impact studies proposed under the plan have yet to begin.
What is certain is that the numbers will be swelled by people made homeless
as their land is acquired to accommodate the embankments and diversion canals
and to provide a source of the huge amounts of earth that will be needed
in construction.

Bangladeshi ministers routinely claim that the floods that their country
regularly suffers have recently become fiercer and more extensive. But this
is far from clear. There is a good case for saying that Bangladesh should
do nothing about its river flooding problem and should concentrate instead
on building protection against cyclones along the heavily populated mouth
of the delta.

Rogers dismisses the embankments proposed in the flood plan as ‘extremely
unlikely to be a justifiable investment’ for Bangladesh. He proposes instead
a switch from ‘hard’ engineering options to ‘soft’ engineering-adapting
to rather than defying the natural forces unleashed on the country each
summer. ‘People in this region have adopted many ingenious ways of living
through the floods. Resources should be allocated to helping them do this
more effectively,’ he says. The need is for refuge areas that people can
move to when the flood waters rise, for emergency food and medical services,
for improved flood forecasting and warning systems. Some of this work has
been included in the flood plan, but appears to have a low priority with
the Bangladeshi authorities. The Central Flood Plan Coordination Organisation,
which is responsible for implementing the flood plan, is composed entirely
of engineers. The organisation has delegated the ‘soft’ options to other
government agencies, such as the Water Development Board, which was denied
control of the overall scheme at the insistence of Western aid agencies.
The agencies regard the board as a slow, cumbersome and inefficient bureaucracy.

The board has responsibility for flood warnings. Currently, warnings
of high waters in the Ganges and Brahmaputra upstream of Bangladesh reach
officials by a tortuous route via New Delhi and the President’s office in
Dhaka. ‘Delays of up to 12 hours in receiving data are normal,’ says the
board’s director of surface hydrology, Alam Mrah. Since his hydrological
models allow predictions of major floods only 24 hours ahead, this is a
lot of time to lose. But as Nishat points out, ‘though there is a flood
forecasting system of sorts, we have no method of flood warning for the
population, even now.’

Learning from the past

Amid the rush to build embankments, little work is being done on creating
effective flood warnings. And schemes to improve flood preparedness suffer
similarly. According to the coordinating committee’s chief engineer, Nurul
Huq, no progress has been made because of the protracted dispute over whether
to concentrate on creating refuges on dry land or to supply villages with
emergency boats.

Bangladesh is, by tradition, a land that relies on boats, fish and floods
to get by. These remain the cornerstone of life in the wet and stupendously
green countryside. Now, in their urge to modernise, the urban elites want
to do away with the old. They want to build roads and high embankments,
to tame their mighty rivers. Van Ellen believes that in the future ‘Bangladesh
will look a little like my country, with a shorter more protected coastline
and better water management inside the barriers. That’s what you see all
over the world. The only question is when it will happen.’ He may be right,
but the casualties of such a strategey will be high, especially among the
poorest people in a poor land. The betting must be that the rivers will
remain in control of this delta nation.

The final irony is that Bangladesh once had a perfectly adequate flood
defence system. When the British colonists arrived in 1757, they reported
seeing thousands of kilometres of embankments, mostly built in the Mogul
period of two centuries before. The banks were relatively modest affairs
designed to prevent major damage, rather than routine flooding. The system
was operated by feudal landlords and paid for by stiff local taxes, which
the British abolished.

These banks, entirely forgotten by modern hydrological engineers, have
proved a useful resource for others. All along the rivers downstream of
Dhaka, clay embankments are being dismembered and carried in small baskets
on the heads of labourers to brickworks that line the river for many miles,
coughing fluoride-laden fumes across the pitted, quagmire landscape. As
urban Bangladesh pleads for foreign aid to build embankments to protect
Dhaka and other cities from floods, the cities themselves are consuming
existing embankments in order to grow.

Fred Pearce is an environment journalist. He is currently writing a
book on the politics of water.