The Ganges-Brahmaputra-Meghna mega-basin is the second largest deltaic region in the world, in which rivers flow in one of the most densely populated regions.

Rivers know no territorial boundaries. They flow as an individual unit depending on the nature of terrain. The use of rivers is of two types: (a) non-consumptive and (b) consumptive. For example, navigation is a non-consumptive use because water is not depleted or reduced through navigation, although water in rivers can be polluted through navigation. Consumptive use reduces water because it is mostly used for industrial, irrigation in agricultural lands and other purposes including domestic use.

Bangladesh depends on the Ganges and its tributaries. Water in rivers cannot be diverted without the consent of co-riparian countries (countries through which rivers flow) because international law recognizes the right of co-riparian countries to enjoy all the advantages, deriving from water for welfare and economic progress of the country. Historically, culturally and economically, Bangladesh has always been rich in the use of water in various purposes.

At least 54 rivers flow from India to Bangladesh and India has diverted water from many common or trans-boundary rivers without the consent of Bangladesh. For example, in 1974 Farakka Barrage across the river, the Ganges, was built to divert without the consent of Bangladesh to flush silt of the Hoogly river. The Farakka Barrage has been the key cause of the water crisis inside Bangladesh.

If farmers of Bangladesh do not seek alternative sources of water, they may find it increasingly difficult to grow crops, but alternative sources of irrigation are often too expensive and will drive farmers to the point of bankruptcy.

The 1996 Ganges Treaty between India and Bangladesh is an agreement to share surface waters at the Farakka Barrage near their mutual border for thirty years. But the treaty divides water flow without sharing the value and uses of the river between the two countries. It does not even take into consideration the uppermost riparian, Nepal, meaning that it takes neither a whole-of-basin approach to river management, nor does it factor in the effects of upstream use of the Ganges on water availability at the Farakka Barrage.

The Ganges Treaty illustrates that a legally binding agreement is not the same as meaningful cooperation between the parties. The Treaty does not create a community of interest in the shared management of the river and has left Bangladesh with numerous concerns and unresolved issues.

Some analysts take the view that the Ganges Treaty is a water sharing agreement in its most undeveloped state because it is solely an arrangement for the volumetric allocation of river flow in the dry season. It does not concern benefit sharing, nor does it purport to be a comprehensive river sharing and management treaty. Instead, the Ganges Treaty establishes India's right to withdraw up to 40,000 cusescs of flow at the Farakka Barrage between 1 January and 31 May every year. If availability at Farakka falls below 70,000 cusecs, the flow will be divided equally between the two countries, while guaranteeing a minimum of 35,000 cusecs to each over alternating 10-day periods between 11 March and 10 May annually.

A further problem with the water allocation is that it is based on averages from flows between 1949 and 1988. Since then, the dry season discharge of the Ganges at Farakka has declined due to increased upstream uses for agriculture and other purposes, and this has led on several occasions to Bangladesh not receiving the Treaty-apportioned flow. By focusing exclusively on dividing water flow, the Treaty has not taken into account how current and future uses further upstream are likely to affect availability of water at Farakka.

The difficulties that arise between India and Bangladesh arising out of this Treaty are not limited to simply reduced flows below 50,000 cusecs. The issue of flooding is of particular concern to Bangladesh. As a deltaic floodplain, Bangladesh already faces even greater risk of flooding than drought. Because the Ganges Treaty allows India to withdraw a maximum of 40,000 cusecs irrespective of the overall flow of the river, when the flow reaches 2,000,000 cusecs - the point at which the river breaches its banks - there is no recourse in the Treaty for flood alleviation. The flood- risk is compounded by the rapid rise in silt deposition rates occurring since the Farakka Barrage was erected.

Some analysts are of the view that the Ganges is not merely the flow of waters of the river but has multiple values as a multifaceted ecosystem. It requires a basin-wide approach and consequently, river-sharing - in the sense of creating mutual benefits - is more likely.

Any 'water sharing' agreement based on volumetric allocation of river flows in the absence of a basin-wide river management regime is bound to exacerbate underlying issues (such as unaddressed concerns over flooding, lack of adequate data sharing of rivers) that can only be addressed from a place of a genuine community of interest and political will for 'river sharing', which take a positive sum approach rather than a zero-sum one.

Another concern is reported to be China's proposed diversion of water from rivers (such as Brahmaputra) of the Tibetan plateau for its agricultural purposes. One of the Chinese projects calls for, in the first phase, building 300-kilometre of tunnels and channels, through which the Brahmaputra river would be re-routed northward toward China.

There is another global dimension on the scarcity of fresh water. The construction of dams and barrages on the 60% of the largest rivers has reduced water. According to UNESCO, availability of fresh water is expected to drop by one-third within 20 years. As a result water experts believe that water-disputes between states would likely to be on the increase. Furthermore South Asia is partly to be affected due to melting of the Himalayan glaciers due to global warming.

Barrister Harun ur Rashid, Former Bangladesh Ambassador to the UN, Geneva