Every year during the mon-soon season, the Himalayan region appears in the headlines because of large scale flooding in the plains of the Ganga and Brahmaputra. In general, this is also the time to go through the annual ritual of accusing the peasants of Nepal and the nearby mountain regions for sending down the floods in ever-higher volumes. They are to blame, it is said, because it is the deforestation in the Himalaya which leads to devastating inundation, particularly in Bangladesh. The truth is, it has never been clear to what extent the floods are a natural phenomenon occurring through history, and to what extent human activities either forest-cutting upstream or building of embankments downstream have played a role in increasing the inundation in modern times. It is not even clear that the floods are increasing in frequency and intensity over the decades, as is claimed.
However, this lack of scientific confirmation has not deterred politicians, engineers and journalists from engaging in passionate condemnation of upstream inhabitants for the inundation, particularly in years when the floods are high. Blame for the Bangladesh floods became a geopolitically sensitive matter because the Ganga and Brahmaputra are both international rivers. Even more interesting was the fact that plans for flood management and even flood control, involving vast sums of foreign aid, were activated on the basis of incomplete knowledge.
The hypothesis regarding the domino-effect of human activities in the Himalaya on the ecological processes in the lowlands can be summarised by the following, superficially convincing, sequence: population growth in the mountains Õincreasing demand for fuelwood, fodder and timberÕuncontrolled forest removal in more and more marginal areasÕintensified erosion and higher peak flows in the riversÕsevere flooding and siltation on the densely populated and cultivated plains of the Ganga and Brahmaputra.
Such a supposedly scientific chain of events has served as an expedient tool for both the plains politician and his counterpart in the hills. For the former, it has been useful in times of flood-related crises to pin the blame on the peasantry of a remote region. His hill counterpart, meanwhile, was amenable to accepting the blame because bad science was presented to him as fait accompli, and also because the aid agencies funded reforestation programmes in the bargain.
The Himalayan Degradation Theory, as it came to be known, was proposed in the early 1970s based on anecdotal information, and was immediately picked up by the media, bureaucracy and the aid community for its plausibility. However, as empirical research was carried out through the 1980s, holes began to appear in the fabric of the hypothesis. It turned out that explanations for floods were not really that simple, and that the inhabitants of the Himalaya had less to do with the ecological processes in the lowlands than what had been too easily assumed.
Unfortunately, such was the inherent attraction of the Himalayan Degradation Theory that the increasing accumulation of scientific evidence against it failed to gain publicity. This remained true even after, in the late 1980s and early 1990s, two publications came out which did a creditable job of presenting the case against the Theory; they were The Himalayan Dilemma, edited by Jack D. Ives and Bruno Messerli, and The Citizen´s Report of the Indian Environment, edited by Anil Agarwal and Sunita Narain. The two reports confirmed that it was not possible to prove on the basis of available data that the Himalaya contributed to floods in Bangladesh, nor that there has been any increasing trend in flood frequency or flood volume.
What the Data Says
Data collected and analysed between 1992 and 1996 by a Bangla-Swiss team now provides scientific evidence to further disprove the Himalayan Degradation Theory. It also provides new suggestions as to the cause of Bangladesh floods, and indicates that the perception of the Bangladesh peasantry regarding floods displays considerably less panic than the engineers, politicians and even academics in Dhaka. The detailed scientific results of the study, of which this writer is principal author, is to be presented in a book next year, The Floods in Bangladesh: Processes and Impacts, published by the United Nations University Press.
The study, which was a joint effort of the University of Berne, Department of Geography, the Swiss Agency for Development and Cooperation, and the United Nations University shows, firstly, that floods in Bangladesh and India are largely independent of human activities in the upper catchment areas. Neither the frequency nor the volume of flooding has increased in Bangladesh over the last 120 years. Precipitation and runoff in the Himalaya do not seem to be important causes of the floods in Bangladesh. The rainfall patterns in the Meghalaya Hills, however, seem to be decisive. Comparing the three major river systems, the Meghna catchment is of primary importance, followed by the Brahmaputra catchment. The Ganga catchment plays only a minor role in Bangladesh floods.
Before providing the details of the study, it should be stated that the water resources of the Ganga-Brahmaputra-Meghna basin are shared by China, India, Nepal, Bhutan and Bangladesh. The availability and sharing of water is therefore a geopolitically very sensitive issue, and so even when there is reliable hydrological data available it tends to be treated as classified information. Climatological information is less restricted, but the “station density” is very low, especially for the Himalaya.
All in all, this makes it difficult to pursue scientific studies of the region´s water-related problems and to investigate the framework of highland-lowland interactions, which can therefore only be based on case studies rather than systematic analyses. In that context, and as far as the Himalayan catchment area is concerned, the present study has restricted itself to the region which may be called “Ganga Himalaya”. No conclusions can be formulated for the “Arunachal Himalaya” in the east, which feeds the Brahmaputra before it enters Bangladesh via Assam, due to the near-total lack of access to data for this area.
Cloudburst in Nepal
An opportunity presented itself in the monsoon of 1993 to study the correlation between heavy rainfall in Nepal and the flood mark in Bangladesh. An extended cloudburst on 19-20 July 1993 in eastern and central Nepal had catastrophic effects on the local population and infrastructure. Several districts were hit by floods and landslides, and many people died or became homeless. There was widespread destruction of crops, and flash floods destroyed the main highway into Kathmandu, cutting off the capital for days. Due to very high sedimentation, the life-span of the Kulekhani Reservoir was reduced from 50 years to about 25 years.
Data shows that there were two flood periods in Bangladesh in 1993, one from 18-25 June and the other from 10-26 July. Both events were the result of heavy rainfall in Bangladesh and, significantly, were concentrated on the Meghna and Brahmaputra systems. Western Bangladesh, that is, the Ganga system, was almost completely unaffected by flooding. What significant floods there were in Bangladesh that year, therefore, were concentrated on a different river system than the one into which the Nepali deluge of July emptied. In other words, the 1993 flood in Nepal had no connection to the floods in Bangladsh.
In order to further support this finding, and to estimate the effect of the Nepal flood on the hydrology of the Ganga in Bangladesh, the water level graph of the Ganga at Hardinge Bridge was analysed. Hardinge Bridge is located approximately 600 km downstream of the flood-affected areas of Nepal in 1993. Assuming an average flow velocity of 1.5 m/sec, the flood wave from Nepal should have reached Hardinge Bridge five to six days later, around 25 July.
It is indeed the case that a short-term, very moderate fluctuation of the Ganga´s flow was recorded from 24 July to 30 July, with a peak on 27 July. This situation may be interpreted as a result of the Nepal floods, but can just as easily be attributed to the intense local rainfall from 20 July to 24 July, recorded at Pabna in western Bangladesh.
In any case, it is clear that the flood flow of the Ganga´s Nepali tributaries had levelled off over the course of their passage from the hills and through the plains, and had only a very small effect, if any, on the discharge of the Ganga as it entered Bangladesh. The short-term peak of the Ganga on 24-30 July was not significant in terms of constituting a flood in Bangladesh.
It is thus clear that even an extraordinary flood event of rare dimension in the Himalayan foothills of Nepal has almost no impact on hydrological conditions in Bangladesh. The effect of the 1993 cloudburst in Nepal having had almost negligible consequence on the Ganga discharge at Hardinge Bridge in Bangladesh, we must conclude that the Ganga Himalaya does not contribute significantly to the floods in Bangladesh.
Rain in Meghalaya
Through the testimony of millions of schoolbooks, the small town of Cherrapunji in the southern slopes of the Meghalaya Hills has attained mythic status as “the place where it rains the most in the world”. It is interesting, therefore, that no one has thought to study a possible link between heavy rains in this region of Meghalaya and downstream flooding in Bangladesh. For, the precipitation in this region of Meghalaya can only be described as massive, and far above the average rainfall anywhere else further up the catchment areas of the Ganga or Brahmaputra.
The average monsoon rainfall (May to September) in Cherrapunji amounts to 9527 mm. In some years, it can be much higher; there was 19,728 mm of deluge during the same period in 1974. It helps in the understanding when one realises that this amounts to more than the height of a six-storeyed building. By contrast, Pokhara Valley, which is said to have the most rainfall in the entire Nepal Himalaya, records an average annual rainfall of only 3800 mm. It is much less for regions closer to the Ganga-Brahmaputra delta, such as Kathmandu Valley and Darjeeling.
Comparing Cherrapunji rainfall data with flood occurrence in Bangladesh proves instructive. The rainfall in Cherrapunji was far above normal from July to September in 1974 and 1988. These were also two major flood years in Bangladesh in the last two decades. In 1978 and 1986, which were years of very low flood in Bangladesh, the rainfall in Cherrapunji was also below average.
The Meghalaya Hills are located adjacent to the vast floodplain of northeastern Bangladesh. They form a first topographical barrier for the humid monsoon winds on their way up from the Bay of Bengal to the Himalaya. It is natural, therefore, for these hills to get very high rainfall in the summer. The hills have shallow soils and rocky surface, so the runoff is immediate when it rains. As a result, and in spite of the comparably small area, a considerable volume of water pours down from Meghalaya into the Meghna floodplain to the south, and to the Brahmaputra in the north, which too has a very short distance to go through Assam before it enters Bangladesh.
The high average hydrological flow from the Meghalaya Hills and the obvious correlation between high rainfall in Cherrapunji and high flood in Bangladesh indicate that deluge in Meghalaya is an important, perhaps even decisive, cause of flooding in Bangladesh.
The collected data indicate that floods in the Brahmaputra Valley of Assam may have a connection with floods downstream, but floods in the Ganga plain of India do not seem to be related to floods in Bangladesh. This is seen in the flood figures for 1988 and 1978.
Both Assam and downstream Bangladesh were hit by severe floods in 1988. The fiercest flood wave in the Brahmaputra Valley coincided with peak flooding in Bangladesh. A satellite image of 16 July of that year, which was not even taken at the height of the flood, shows an almost constant line of flood patches along the Brahmaputra through Assam down to Bangladesh.
In the 1978 monsoon, the inundation in the entire Ganga plain of India was far above average: at least 40 million people were affected by the catastrophe. In Bangladesh, however, flooding remained far below average; only a few, regionally limited flood events were reported from the western part of the country. Flood on the Ganga seems significant for all of Bangladesh only when it peaks simultaneously with the Brahmaputra flood.
The statistics show that, climatologically, the dry years in the west (the Ganga catchment) often coincide with humid years in the east (the Brahmaputra/Meghna catchment), and vice versa. The extent of flooding in Bangladesh in a specific year corresponds to the climatic conditions in the east and to the level of flooding in Assam rather than to the climatological conditions in the west or the flooding on the Ganga.
There also seems to be a historical and sociological reason why the Brahmaputra´s significant flooding as it relates to Bangladesh has been neglected while the Ganga´s has been highlighted in general writings and in the media. The Brahmaputra, flowing as it does through only a brief tract of the Indian Northeast before entering Bangladesh, is not as widely known as the Ganga. The latter, on the other hand, flowing as it does through the heartland of historical Hindustan, is fully part of the South Asian cultural pantheon. This may serve to explain why, when a flood hits Bangladesh, the public looks to the Ganga and its Himalayan tributaries as the cause rather than the Brahmaputra. This, then, can partly explain why the hill peasant of the Himalayan foothills is blamed whenever Bangladesh is inundated.
The role of rainfall within Bangladesh in triggering or exacerbating floods also seems to have been overlooked as experts and lay persons alike seek answers in upper catchments. High rainfall in Bangladesh, after all, is a typical feature before and during heavy floods and without doubt contributes significantly to inundation.
Again, the extraordinary flood of 1988 provides an interesting case in point. As a result of the position of the monsoon trough over northern Bangladesh and Assam, the northern part of the country got heavy rainfall that year, while the south remained almost free of precipitation. Rainfall in Sunamganj in northeastern Bangladesh was extraordinary: over six days, from 24-29 August, 960 mm of rainfall was recorded at this station, which is almost 20 percent of the average annual rainfall in this area. The period of nationwide flooding in 1988 was between 20 August and 6 September, with the peak recorded around 31 August.
More study is required to pinpoint the exact correlation between localised rains and heavy inundation, but the connection seems clear.
In summary, the data shows that the Bangladesh floods are influenced by a combination of regionally differentiated rainfall patterns. The influence of the Ganga Himalaya is not important because the catchment is at a distance from the Bangladeshi floodplain. As a result, the peak discharge from heavy rainfall or cloudbursts which are carried down by the Himalayan tributaries such as the Kosi or Gandaki level themselves out by the time they join the “base flow” of the Ganga and subsequently arrive in Bangladesh.
The contribution of the Arunachal Himalaya to a flood on the Brahmaputra cannot be assessed due to lack of data. The Meghalaya Hills, however, seem to play a decisive role due to their location close to the floodplain of Bangladesh and because of the high rainfall and rapid runoff. Finally, it can be said that the rainfall within Bangladesh itself is a significant factor.
The Himalaya, at least the part located in the Ganga catchment, seems to have negligible impact on floods in Bangladesh. If this statement is accepted, then the habit of blaming mountain inhabitants for the flood catastrophes far downstream must be abandoned. This would also considerably improve the political climate in the area. This does not relieve the mountain people of their responsibility to use their environment in a sustainable manner, however.
A lot of questions about rainfall, flooding and mitigation measures remain to be answered. The present study has merely added some pieces to the complex puzzle as to why Bangladesh floods. Also, it has contributed to the further devaluation of the Himalayan Degradation Theory. Now, researchers are required to move ahead with the task of sharpening their understanding of the hydrological regime of the Ganga-Brahmaputra region. This can only be achieved through the permanent and free exchange of scientific data among the states of the region. A
T. Hofer is physical geographer at the Department of Geography, University of Berne.
A study was conducted to see how the views on flooding differed between farmers, politicians, journalists, engineers and donor agency representatives in Bangladesh. The result showed that farmers were the least concerned of all. For them, floods were a part of life which their ancestors and they had learned to adjust to. They also knew the benefits that floods brought in terms of fertility to the soil.
“People do not die if there is flood, but people die if there is no flood,” goes a local saying in Sirajganj District. One farmer in Simla village in the Brahmaputra floodplain told researchers, “If there is no flood there will be no crop, the soil will turn into a desert.” According to the peasants interviewed, they feel a much bigger threat from what scientists call ´lateral river erosion´. Whereas the land is always there after a few days when the flood recedes, nothing is left when your household and land have been carried away by a shifting river.
Politicians, the researchers learnt, believe that floods are a major problem for the suffering they supposedly bring to the populace. As far as the politician is concerned, therefore, the flood problem has to be solved by eliminating it. This will potentially involve large projects and expensive packages of foreign aid.
Engineers, too, see flooding simply as a problem of high water volume, to be resolved by implementing technical measures. According to one expert, “There is a tendency among engineers to perceive water as a burden. Therefore, water should be controlled and conducted to the ocean as soon as possible.”
For journalists, flooding is the season for catchy headlines and over-statements. Press analysis in Dhaka is still dominated by reports on the floods´ origins in disturbed Himalayan ecology, and the unquestioning acceptance of the theory that flood intensity has increased over recent decades. Meanwhile, by its exclusive coverage of the floods, the Western media has helped define monsoon flooding as the main problem of Bangladesh, which is hardly the case. The problem of river erosion is addressed by the Dhaka press but ignored by the Western reporters.
Influenced as they are by the politicians, engineers and journalists, the donor agencies have tended to focus their attention on floods and the search for a technical solution. Thus, while the farmers may be quite happy to live with floods but want assistance to tackle river erosion, landlessness and economic problems, the donors tend to come up with more aid for flood mitigation. The great flood of 1988 galvanised the international community on this score, with the G-7 summit of industrialised (aid-giving) nations in July 1989 calling for, “effective, coordinated action by the international community in support of the Government of Bangladesh in order to find solutions to the major problem (flood) which are technically, financially, economically and environmentally sound.”
When it comes to perception of floods and their danger, therefore, few heed the wisdom of villagers, even though it is they who have to live with the flood. It is clear that while making decisions relating to flood mitigation and related development activities, the views of the scientist, engineer, politician and journalist must all be taken into account, but without neglecting the most important perception of all that of the inhabitants of the floodplain.