ÐÓ°ÉÔ­´´

IN THE AFTERMATH of Bangladesh’s worst-ever flooding in 1988,foreign
governments offered financial help and expert advice. They looked at how
technology might help 110 million Bangladeshis to live more safely in the
flood plains of South Asia’s most powerful and flood-prone rivers, the Ganges,
the Brahmaputra and the Meghna. The work was done against a backdrop of
uncertainty about the long-term implications of the greenhouse effect and
global warming, which could make Bangladesh even more vulnerable to flooding
in years to come.

France and the UnitedNations Development Programme (UNDP) quickly commissioned
flood studies in collaboration with the Bangladeshi government. The US and
Japan also asked their national specialists to report on the problem. Each
resulting report has a different emphasis. The most dramatic and costly
proposals come from the French engineers, who want to make the country virtually
impregnable. Their plan involvesthe construction of 4000 kilometres of high
embankments: it would take 20 years to complete and cost as much as $10
billion. Specialists from Japan and the US advise a different approach.
A report prepared for the US Agency for International Development (USAID),
the channel for government aid to the developing world, cautions against
relying on expensive engineering schemes. Peter Rogers, professor of Environmental
Engineering and City and Regional Planning at Harvard University, suggests
instead that techniques for forecasting floods should be greatly improved,
and that money should be spent studying and promoting the many ingenious
ways of living with floods that the people of the region have developed.
Japan’s International Cooperation Agency (JICA) was also wary of the heavy
engineering option. The UNDP report came halfway between the two extremes.
It recognised the need for embankments but pointed out that local social,
environmental and management issues would have to be addressed as well.

After months of work all agreed that the scale of the problem is too
vast to be tackled on the ‘project by project’ basis normally favoured by
national development aid agencies. Instead, they seem to be ready to gather
under the umbrella of the World Bank to cooperate in a five-year action
plan, which would be the first part of a longer-term flood control programme.
An international meeting in London later this month, hosted by Britain’s
Overseas Development Administration and chaired by the Bangladeshi government
and the World Bank, is expected to endorse the action plan. It will also
sanction Bangladesh’s proposal that the World Bank coordinates funds for
the programme.

The aim is to buy a little time for Bangladesh by protecting the country’s
agriculture and main population centres from the worst effects of recurrent
serious floods. This should help the fragile economy and encourage more
investment to grow the food needed for a population that has doubled since
1961 and is heading for 165 million by 2005.

Under the action plan, which runs from 1990 to 1995 (the period of Bangladesh’s
Fourth Five Year Plan), all the main rivers and areas that become flooded
will be thoroughly surveyed. Embankments along 260 kilometres of the right
(west) bank of the Brahmaputra river will be heightened and strengthened;
upgrading of a 90-kilometre stretch will begin immediately. Work will start
later on linking and upgrading embankments on the left bank; these were
built cheaply and not very well under a Food-For-Work Programme.

Four or five full-scale pilot projects, each costing up to$10 million,
will use mathematical models and satelliteobservation to assess the success
of techniques for controlling rivers. Some of these techniques were developed
for China’s Hwang He (Yellow) and Yangtze Kiang rivers before the start
of the Ming dynasty in 1368. Other pilot schemes will investigate how to
control flooding and drainage; how to protect river banks; and how to improve
freshwater fisheries. A series of studies will look in detail at local economic,
environmental and social problems before a costly programme of embankment
building starts.

Jim Dempster, a British water resources specialist with long experience
in Bangladesh, led the UNDP team. He says that it is neither possible nor
desirable to eliminate flooding completely. Instead, he suggests selective
use of engineering to protect life, vulnerable land, the main cities and
important parts of the country’s infrastructure such as hospitals, transport
links and power stations. Elsewhere, he advocates ‘controlled flooding’,
which would release a limited flow of water through regulators in the embankments.
Rice and jute, which need copious amounts of water, would benefit from this;
so would the country’s important inland fisheries. It would also help to
replenish the ground water for winter irrigation.

Dempster says that Bangladesh must be able to acquire more hydrological
data from India, which will help it to forecast floods better and to provide
more warning to people at risk. At present, Bangladesh receives no more
than 72 hours’ warning of floods in the Ganges and Brahmaputra rivers. Via
teleprinter, India tells Bangladesh when the water levels in the rivers
are within half a metre of the top of their banks and rainfall is greater
than 50 millimetres a day. Improved communication with India, preferably
including information from upstream rather than border gauging stations
on the Ganges and Brahmaputra, could give Bangladesh up to eight days’ warning
of floods.

The Ministry of Irrigation’s Flood Forecasting and Warning Centre in
Dhaka uses a wireless network to monitor water levels at 35 sites and rainfall
at another 34 gauging stations inside Bangladesh. Some of these stations
rely on local people telephoning the data to the wireless network, an unreliable
method, particularly during floods. The UNDP study says that data collection
should not depend on telephone or road links, or on public electricity supplies.
It recommends that readings at key rainfall and water level stations should
be automated as a first step towards the installation of a national telemetric
network. Information on how clouds move over Bangladesh and neighbouring
areas comes from a radar station at the national Meteorological Department.
A microwave link transmits the information to a video terminal at the flood
forecasting centre. The UNDP study says more radar stations are needed in
the hilly areas where the floods start.

In May this year the UNDP and the World Meteorological Organisation
started work on a project lasting 3 to 5 years to improve flood forecasting
and warning in Bangladesh. As a first phase, theyare applying a general
mathematical model that simulates flows and water levels in all the major
rivers, and uses this information to predict floods. More detailed regional
models are now being prepared with funds from Denmark’s international aid
agency, DANIDA. According to the Danes, the models will forecast the effects
of new embankments and other changes on floods and drainage within individual
upazilas, the smallest administrative units in Bangladesh. The upazilas,
each with a population of about a quarter of a million people, have the
difficult task of warning thousands of scattered villages. Experts working
in the country say how much they admire the resilience, ingenuity and cheerfulness
of Bangladeshis in the face of potential disaster. The flood policy seeks
to build on this by encouraging villagers to identify the best ways to protect
people and livestock.

Even if the large embankment projects are to be built, most of the country
will remain unprotected for the next 10 or 15 years. In these unprotected
areas the main improvements will come from ‘flood proofing’. This means
the development of irrigated agriculture in the dry season to help farmers
to avoid the risk of farming during the monsoon season; improved flood warning;
and better systems for helping people to cope with floods. According to
Sayeedul Huq, a Bangladeshi economist on the UNDP advisory team: ‘We have
to live with the flood and make it tolerable. We need to establish dry places
on high ground for people to escape to. Specially designed school buildings,
elevated marketplaces and roads, and high-ground shelters could be provided.
It is difficult for the government to evacuate people from flooded areas,
so local ‘self-help’ should be encouraged.’

As far as engineering is concerned, the country’s position – on a fragile
trident of land in the delta of three of the world’s most unstable rivers
– restricts what can be done. Even in a ‘normal’ year the average annual
flow of floodwater is so great that, if it were spread across the land,
it would cover the entire country to a depth of 7 metres. Diverting the
water into storage reservoirs is clearly impractical; there is just not
enough land in this crowded country for vast reservoirs. In theory, dredging
could improve the flow of water through the main rivers but it is almost
too expensive to think about: the rivers carry around 2400 million tonnes
of sediments per year. Dredged channels would become choked with siltagain
in one season. K K The favourite option is to build embankments along both
banks of the Brahmaputra and Ganges rivers to channel floodwater through
the country to the Bay of Bengal. This is the main theme of both the UNDP
and the French proposals. It would protect Bangladesh from the type of flood
that inundated the country in 1988: hydrologists predict that a flood this
great is likely once every 100 years. The embankments would rise to heights
of between 4.5 and 7.5 metres depending on the local topography, require
about 400 million cubic metres of fill, and take up over 20 000 hectares
of land. Roads could run along the tops of the larger embankments.

French engineers have produced two ambitious variations on this theme,
which would also provide protection from these major floods. They say that
Bangladesh could build high embankments within 1 kilometre of existing river
banks and protect them with breakwaters and tough synthetic mats ballasted
with concrete and anchored by galvanised steel cables. Alternatively, they
say, two sets of embankments could be built: lower, secondary embankments
along the banks of the rivers and higher ones between two and five kilometres
away. The secondary embankments would withstand the size of flood that hydrologists
predict occurs about every 10 years. They would be easy to repair and five
million people could live on the 300 000 hectares of irrigable land between
the two sets. However, as Rogers points out in his USAID study, ‘for (these
people) the flood danger will be increased rather than reduced’. For this
reason, and because of its high cost, the ‘distant embankments’ alternative
is unlikely to get beyond the drawing board.

Whatever embankment scheme is chosen, there are still many problems
to contend with. The immense power of the water, which scours river beds
and erodes embankments, makes ‘river-training’ extremely difficult. Studies
show that flood waters have dug 53 metres into the downstream bed of the
Meghna river, 32 metres into the beds of the Brahmaputra and the Ganges,
and 45 metres into the bed of the Padma river, which forms at the confluence
of the Brahmaputra andthe Ganges. To make matters worse, there is not enoughrock
within Bangladesh to reinforce exposed embankments;finding substitutes or
importing material adds to the construction costs. K K Another drawback
of building embankments is that fishing is likely to be disrupted. Bangladeshis
eat a lot of fish, which supplies almost 80 per cent of their animal protein.
The fish flourish in rivers, canals and flood plain depressions, or beels.
As the monsoon floods rise, newly hatched fish and shrimps, born in rivers
and estuaries, move out into the flood plain to feed and grow. When the
floods recede, they return to rivers and permanent beels to breed. Embankments
would prevent this easy migration unless, as French engineers suggest, fish
‘routes’ and ‘ladders’ are built into them.

According to Steve Jones, the European Commission’s adviser on the World
Bank team, both types of embankment scheme would have a huge social impact
on rural Bangladeshis. If the embankments envisaged under the French plan
are built and 20 000 hectares are requisitioned, the lives of more than
180 000 people will be affected. He stresses that special provision must
be made to teach these people new skills and to help them to find employment.
An urgent study, aimed at finding out how people who lost their land to
earlier embankment schemes have coped since, will start early next year
as part of the action plan.

Besides the need for rural embankments, special bunds are planned to
protect Dhaka; these would be capable of withstanding a flood that might
occur once every 1000 years. Bunds would protect other main towns and cities
from flooding that hydrologists expect to occur once every 500 years.

Protecting the cities is an urgent problem because buildings sit packed
close together on the river banks. The government estimates that floodwater
is eroding the foundations of at least 85 towns or commercial centres; it
has not published their names. The French study highlights the problems
in Sirajganj on the Brahmaputra river, where protection works were badly
damaged during last year’s floods, and in Chandpur on the lower Meghna river.
Chandpur lies on the fringe of the coastal zone where cyclones and storm
surges periodically take a toll on life and often sweep over newly formed
islets, or chars. Poverty and land shortage in Bangladesh are so intense
that, despite government regulations, poor farmers rush to settle on the
chars as soon as they appear. In 1970, 300 000 people died when a cyclone
swept across the Bay of Bengal, and in 1985 tens of thousands were killed.
The government has already built 3600 kilometres of embankments in the delta
and Dutch engineers have proposed two dams for the region. The European
Community is financing further detailed studies.

The sediment that stifles Chandpur and creates unstable chars in the
delta is Himalayan soil. Monsoon rains and melted snow wash the sediment
from mountainsides; rivers carry it in suspension. Over the past 10 000
years or so, the sediment has settled at the confluence of the rivers to
form the flood-plain delta that is now Bangladesh. Some experts believe
that forest felling on the Himalayan slopes is accelerating this centuries-old
process. Others say that there is no evidence for this, and that Himalayan
slopes are naturally unstable. The danger of earthquakes in the Himalayas
and the certainty of rapid siltation of reservoirs limit the opportunities
for building high dams in India and Nepal to mitigate flooding in Bangladesh.

Environmental change within Bangladesh is being studied by the Ministry
of Irrigation’s Space Research and Remote Sensing Organisation. This agency
operates a ground station that receives data from Japanese, American Landsat
and French SPOT satellites. USAID has provided most of the equipment and
has trained 15 Bangladeshi staff to interpret the images from space. The
remote sensing organisation uses these facilities to monitor changes in
the courses of the rivers and in the size of flood areas in Bangladesh.
The UNDP suggests they could also be used to investigate how much forest
has disappeared from the mountain slopes in the past 10 to 15 years, and
whether the torrents that feed the rivers have grown wider.

But even if links were established between increasing erosion in the
Himalayas and more frequent and severe floods in Bangladesh, the area to
be covered is so vast that it would take many decades of reforestation and
soil conservation to contain the problem, let alone reverse trends. This
reinforces Bangladesh’s pleas for something concrete to be done now to protect
the country from the floods that plague it. The portfolio of studies and
embankment projects in the action plan to be discussed at the London conference
this month needs to be considered in that light.

* * *

A land made for natural disasters

BANGLADESH is the delta of one of the wettest catchment areas on Earth.
The three main rivers and their tributaries drain more than 1.5 million
square kilometres. The combined discharge is two-and-a-half times that of
the Mississippi. Only the Amazon is mightier. At Cherrapunji on the Meghalaya
plateau of northeast India, which overlooks northern Bangladesh, more than
11 metres of rain falls in an average year. In a wet year, rainfall exceeds
25 metres. Most of this water finds its way into the rivers and finally
flows into Bangladesh. With less than 10 per cent of the catchment area
within its borders, Bangladesh has no control over the vast flows that pour
into its territory.

Heavy rainfall and floods in the monsoon season and frequent drought
in the spring make farming in Bangladesh an extremely hazardous business.
Yet farmers are adept at making the most of their remarkably difficult environment
and many produce two or three crops a year.

Rice is the most popular crop and, over the centuries, farmers have
developed many thousands of rice varieties to suitlocal conditions. Short-stemmed,
drought tolerant strains are grown on land that is not expected to flood;
long-stemmed ‘deep-water’ rice, which can grow up to 5 metres long, is planted
in the flood plains at the beginning of the monsoon season. But prolonged
drought in the spring can destroy short-stemmed rice and the quick-growing,
deep-water rice may be drowned if water levels rise earlier, higher or more
quickly than ‘normal’.

In 1987 and 1988 the floods were exceptional. All previous records were
broken in 1987 when 40 per cent of the country was flooded. The next year,
62 per cent of the land was inundated in two calamitous weeks. The floods
in 1988 killed at least 2000 people, closed Dhaka airport for five days
and brought the country to a standstill for a month. A stretch of the embankment
on the right bank of the Brahmaputra river collapsed, releasing water that
swamped 100 000 hectares of crops. When the water receded it left a swathe
of sand over land that was fertile before.

Experts say that exceptional rainfall over northwest Bangladesh between
July and September, the monsoon season, was the main cause of the 1987 disaster.
The highest flow ever recorded in the Ganges aggravated the situation, blocking
drainage from flooded land. It roared past a gauging station close to the
Indian border at Hardinge Bridge (the only bridge over the Ganges in Bangladesh)
at a rate of 76 000 cubic metres per second, half as much again as its normal
peak flow.

In 1988, the problem started outside Bangladesh. Concentrated heavy
rainfall on the Meghalaya plateau caused an exceptionally high flow in the
Meghna river in early June. Further storms over north and northeast Bangladesh,
India, Nepal and Bhutan brought the Brahmaputra and one of its tributaries,
the Tista, down in spate. Both rivers quickly reached record high levels
and water spread rapidly from them across adjacent flood plains.

On 30 August, the Brahmaputra flow at Bahadurabad in northern Bangladesh
reached 98 600 cubic metres per second, breaking another record. Three days
later, water in the Ganges sped under Hardinge Bridge at 72 300 cubic metres
per second, not much less than the record flow of1987. In early September,
water spilledacross the flood plain between the two greatrivers and Dhaka
was seriously floodedfor the first time.

In both years the long-standing problemsof the country – poor maintenance,
lack ofplanning controls and bureaucratic inefficiency – made matters worse.
Water easily breached embankments that were riddled with ratholes and poorly
maintained: unplanned or badly designed roads, built without culverts to
drain water, were soon closed. In some cases farmers cut through dykes as
they fought to remove water from their land. Sprawling, uncontrolled development
on the outskirts of the major cities compounded the difficulties.

As international donors rushed to offer emergency assistance and help
with studies for long-term solutions, the country’s Ministry of Planning
and the United Nations Development Programme assembled their own team of
local and foreign advisers.

The team was asked to come up with a scheme to mitigate the effects
of future floods on lives, infrastructure, industry and agriculture. There
was no time to negotiate water management agreements for the Ganges/Brahmaputra
catchment with upstream neighbours India, Nepal, Bhutan and China. The need
for urgent action meant the scheme had only to be politically acceptable
in Bangladesh and to persuade potential donors that it would work.

Annette Bingham is a freelance journalist with a special interest in
Asian water affairs.