By A. Vaidyanathan

THE CONCEPT of interlinking rivers is evidently appealing to considerable sections of the general public and to policy-makers. More than three decades ago, K. L. Rao proposed the linking of the Ganga and the Cauvery. It was followed by Dastur's plan for a garden canal, linking all the major rivers in the country. Both the proposals attracted considerable attention. But due to widespread criticism of their feasibility, desirability and viability, these were shelved.

In the 1990s, the Government appointed a Commission to examine the strategy of water resource development, including the possibility of interlinking rivers. Its report — which is not available to the public — is understood to have given cautious support, subject to a careful examination of all relevant aspects, to the idea of link canals to divert surplus waters from some selected rivers to the water-short basins and regions.

Meanwhile, the Supreme Court, on a public interest litigation, directed the Centre to draw up and implement by 2015 a programme to interlink major rivers. Subsequently, the Prime Minister announced the Government's decision to act on the court directive and appointed a task force to ensure the implementation of the project by 2015. The task force headed by Suresh Prabhu is now active.

The popular appeal of interlinking rivers is based on the understanding that an enormous amount of water of our rivers flows into the sea and that if only this is prevented, and water transferred from water-abundant rivers to water-deficit areas, there will be adequate supply for everyone in every part of the country. At another level, the project is seen as promoting national integration and a fair sharing of the country's natural water wealth. Both these presumptions are far too simplistic.

Whether the linking of rivers will promote integration or generate more disputes and tensions is a moot question. Besides, several obvious, but prima facie important, questions about the concept, and the feasibility, desirability and viability of the proposal need to be clarified before its implementation can be considered seriously. The belief that interlinking is necessary to ensure adequate and safe water supply to everyone and everywhere is wholly misplaced. Domestic use currently accounts for a mere five per cent of the total use of water harnessed through canals, tanks, wells and tube-wells.

The requirements are no doubt growing rapidly but will still be relatively small compared to those of other uses. Interlinking is hardly justified as the solution for this problem. Even if interlinking were justified for other reasons, it will not be possible to reach the water to all the habitations without huge investments in a centralised distribution network. Decentralised local rain-water harvesting, by reviving and improving traditional techniques, can meet essential requirements for domestic purposes more effectively and at a far lesser cost.

By far, the largest user of harnessed water is agriculture. Currently, more than 85 per cent of water from canals, tanks and wells and tube-wells is used for irrigation. The demand on this account is growing and will continue to be, by far, the biggest claimant on available supplies. There is much scope for increasing the efficiency of the irrigation systems in place by reducing waste and through better water management. Measures needed for this purpose — by way of investment in physical improvements and institutional reform — are not receiving due attention.

The need for irrigation arises in regions and seasons when rainfall is inadequate for raising crops and obtaining optimum yields. The total rainfall is adequate to meet crop water requirements in the kharif season over large parts of the country. Irrigation is required essentially to tide over inadequate soil moisture during dry spells within the season. There are, of course, some areas — especially in Punjab, Haryana, Rajasthan, parts of Gujarat, Tamil Nadu — which need irrigation during the kharif season. Practically everywhere, including the northwest, irrigation is essential between November and June. So far, these imbalances have been met by constructing storages to store monsoon surpluses for use in the dry season and by exploiting groundwater. Some areas, such as Tamil Nadu, have exhausted the potential for harnessing the surface flows. In several others, the possibilities for constructing storage are limited. Groundwater resources are already under a severe stress. The scope for expansion is limited. In many areas, the problem is to check expansion and contain the rate of exploitation. It is in this context that interlinking is seen as a way out.

A closer examination of the interlinking idea raises several questions: First, it is based on the presumption that there are large surplus flows in some basins and that the physical transfer is feasible in terms of physical engineering, and can be accomplished economically without creating any adverse impact.

On what basis and who determines the surplus basins and the magnitude of the surplus? The volume of flows during the flood season is misleading as a basis for judging surpluses. Nor can the regions where floods occur be considered water surplus. Most of them may have floods in the monsoon but have inadequate water for use in the dry season. Substantial tracts in these regions do not have the benefit of irrigation. Estimates of surplus made by Central agencies such as the National Water Development Agency are hotly contested by the States.

A more serious difficulty arises from the fact that most of the flow in practically all rivers occurs during the southwest monsoon. Published data from official sources show that 90 per cent of the flow in south Indian rivers occurs between May and November. Data on the Indo-Gangetic and Brahmaputra river basins are classified. Being perennial, the proportion of the total flow occurring during these months may be somewhat smaller but not all that much smaller. For instance, over 80 per cent of the annual flow in the Kosi is between May and November; and almost three fourths between June and October.

The monsoon happens to be the season when rainfall in the aggregate is adequate for crop growth. Of course in some regions, such as Rajasthan and parts of Gujarat and the Deccan, even the kharif rain is far too low and variable for productive agriculture. In some others, more water could help switch to more productive crop patterns. These "deficit" regions are far from those considered "surplus" requiring transport over very difficult terrain and long distances.

Moreover, since the surplus occurs in the rainy season and the demand is in the dry season, it is not enough to merely carry the water from one point to another. Large storages will be necessary. One needs to know the quantum of water to be stored, and whether and where potential sites on the required scale are available, and their likely impact on environmentand human displacement.

All we have to go by are some maps published in the media, purportedly from the Hashim Report, indicating from which rivers and at which locations surpluses will be diverted and to which river(s), and at what points in these rivers the divertedwater will be taken. There is no information on the quantum of water to be transferred through different link canals; the extent and location of the area to be benefited at the receiving end; and the distribution system through which water is to be distributed to this area.

The maps and the sketchy accounts in the media and official pronouncements tell us little on these aspects. If these maps accurately reflect the concept of the interlinking projects sought to be implemented, it will only mean that instead of the surplus flows flowing to the Bay of Bengal via the Ganges and the Brahmaputra and the Mahanadi, they will flow to the sea through the Krishna, the Godavari, the Pennar or wherever!