In India, water disputes tribunals are set up as constitutional bodies to resolve disputes over sharing of water from trans-boundary basins and mechanisms are created to enforce the tribunal "awards". The insurmountable challenge facing these tribunals is of assessing the legitimate water demands of the contesting states, and evolving principles and norms for determining their share of water. Groundwater use has increased in many Indian river basins in the past three decades adversely affecting basin runoff. But, this is not factored in while arriving at water allocation decisions by these tribunals, while the hydrological interdependence between surface water and groundwater has long been recognized in international water treaties. Another important factor, which is out of sight, is the likely impact of changing land use and water use on the basin hydrology.
The paper, through a case study of the interstate river basin of Narmada, illustrates how the neglect of these issues challenges the viability of water sharing agreements between the two major riparian states. The paper also suggests ways to increase the effectiveness of existing institutional arrangement for trans-boundary water management in river basins.
The methodology comprised analysis of: 1] changing dynamics of interaction between groundwater and surface water; 2] historical changes in land and water use; and, 3] downstream impacts of water harvesting. The first step included analysis of: groundwater hydrographs for observation wells spread across the basin to examine the long-term trend in the annual rate of change in water levels; the long term change in the net groundwater balance in the basin; estimated flow series for different time periods in relation to rainfall to examine whether the stream flows had reduced for reasons other than rainfall. The second step was analysis of historical changes in green and blue water use in the basin for crop production along with changes in consumptive use for domestic, livestock and other sectors. The third step involved: comparing the primary data collected from two sub-basins of Narmada, on the stream flows and recharge fraction, during the pre and post water harvesting periods; and by comparing the recharge fraction for the treated watershed and untreated watershed in the same basin.
Results and Discussion
There has been significant long-term trend in the annual rate of change (rise/drop) in groundwater levels. However, this trend in water levels does not corroborate with the change in groundwater balance. While groundwater abstraction in Narmada basin had substantially increased--from 1008 MCM in 1985 to 2582 MCM in 1999, the net annual groundwater storage in the basin, which takes into account all inflows and outflows, has become positive. Analysis of estimated flow series for 1965-2000, 1965-1990 and actual flows for 1980-2000 in relation to rainfall, using rainfall-runoff relationship showed that the stream flows had reduced for reasons other than rainfall. This meant that the groundwater outflow into the surface streams had reduced from the time when the level of development of aquifers was low to the present, as a result of the increased pumping and lowering of water levels.
Out of a total water consumption of 23,423 MCM in the basin, 94% was for crop production. Of the total diverted water used from MP part of the basin, 76.7% was used in agriculture, while 30 years ago, it was a fraction of the total water diverted. The use of green water in MP part of the basin had increased slowly compared to blue or diverted water--from 16.58 BCM to 17.52 BCM. This was because the increasing cropping intensity was mainly from expansion in irrigation, and not rain-fed area. The rise in green water use was quite significant when compared to the current blue water use, which is only 4,530 MCM.
The change in use of blue water for agriculture production was rather dramatic, from 600 MCM to 4530 MCM over a thirty year period, facilitated by increased access to wells and surface irrigation. The effective surface water diversion from Narmada River and its tributaries for agriculture was 3.59 BCM in 1999-2000. From this, the crop water use was nearly 57%. The remaining 47% was largely available for reuse. The increase in cultivated land and the increased diversion from rivers and streams can reduce the stream flows.
The Narmada basin also witnessed large-scale decentralized water harvesting, without any hydrological planning. Analysis of primary data from the Hathni and Kundi sub-basins of Narmada showed that after watershed treatment activities, the stream flows in the two sub-basins reduced significantly, while the 'recharge fraction' increased. Further, the recharge fraction for the treated watershed was higher than that of untreated watershed in the same basin for higher magnitudes of rainfall. The reduction in stream flows was therefore found to be due to reduction in runoff coefficient after the watershed treatments.
Four big challenges in trans-boundary water management are: 1] understanding the changing dynamics of interaction between groundwater-surface water due to hydrological stresses, and its effects on stream flows; and 2] understanding how changes in green water use over time affects the basin hydrology; 3] understanding scale of water harvesting, watershed management and private river water lifting by farmers and other water users; and, 4] monitoring the use of groundwater by private well owners and public utilities and direct use of green water for crop production, and enforcing control over the same.
The best strategy for protecting the sanctity of inter-state water sharing agreements in trans-boundary basins is to have an independent regulatory authority to map the changes in land use and well irrigated area in the basin over time so as to arrive at realistic figures of area under well irrigation, groundwater draft, area under rain-fed production, the number of tiny reservoirs etc. The effects of these on surface water flows could be estimated to assess the need for reviewing inter-state water sharing agreements.