Unprepared

14/09/2007

Conventional thinking has it that floods are caused by excessive rainfall. ARCHITA BHATTA finds out that wasn't necessarily the case this year

While people in the flood-affected areas are struggling to come to terms with the ravages, the political blame game has begun. In a particularly heated debate in the parliament on August 13, members demanded that the government state what it had done to alleviate matters. There weren't many concrete suggestions: only ones to have more embankments.

Aside from the political hullabaloo, state and local authorities have begun to come up with explanations for the floods. "Floods in most parts of eastern Uttar Pradesh were due to heavy rainfall in Nepal,' said Randhir Singh, Gorakhpur's district collector. Assam's water resources minister Bharat Chandra Narah ascribed the floods to "abnormal developments in upstream Arunachal Pradesh. The Ranganadi dam of the North Eastern Electric Power Corporation had probably released excess water without giving us prior information."

Authorities in the state also maintain that heavy rainfall in neighbouring Arunachal Pradesh and Meghalaya caused the floods. Data of the Indian Meteorological Department (imd), however, shows that only one of the three districts of Meghalaya received rainfall that was significantly above normal. It also shows that only two of the 13 districts of Arunachal Pradesh received rainfall far in excess of the normal. In fact, the department's data indicates that the floods were not necessarily caused by heavy rainfall: precipitation in Assam was less than 99 per cent of the 30-year average rainfall in the state, and most of the flood-affected districts in eastern Uttar Pradesh, Bihar and Gujarat received less than normal rainfall before the floods (see table: The skies didn't exactly open up). imd has its own explanation. "The only floods that took place this year were in Bihar. The others can be described as cases of water logging,' said the department's spokesperson B P Yadav.

What ails research
imd's flimsy technicality does not obscure the malaise in the department's workings."imd provides us data that is three-years old. This is of little value in flood forecasting,' says A K Gosain of the Indian Institute of Technology (iit), Delhi who's been working on peak flows in Indian rivers. Officials of the Pune-based National Data Centre of imd say that the centre has monthly rainfall data from 1998 to 2005. But they also said that the centre does not have data for all states, particularly those in the north east. "We don't generate data, only maintain it,' Krishna Kumar of the centre said. So, most research on rainfall patterns in the country is based on old information.

It is significant research that is in question here. Last year, for example, the Indian Institute of Tropical Meteorology, Pune, came out with a study, which showed extreme rainfall events in Central India are rising, while the average rainfall has remained constant. These are vagaries which imd with its reliance on long-term averages seems incapable of predicting.

Last year in February, Gosain authored a Current Science paper that predicted an increase in precipitation, water yield and evaporation in the entire Mahanadi sub-basin. The paper showed that the river's annual peak flow would exceed from the present level of 20,000 cubic metres per second (cumecs)to about 37,000 cumecs because of global warming. (To give a general idea of this means, 37,000 cumecs translates to 3.772 million litres of water flowing through a river per day.)

Melting glaciers
What accounts for such rise in peak flows? Glacial melting? Dulal Goswami, former professor of environmental sciences at Guwahati University, says, "The floods this time were because of the rains but the snow-melt factor also contributed. Gosain, however, says that "currently there is no direct evidence of waters from melting glaciers causing floods. But this cannot be ruled out in future.'

So far, there is evidence that some run off the river (high altitude) hydroprojects are receiving increased discharge due to melting glaciers. But there is no direct evidence of floods due to that. There should be more research on increase in water levels due to melting glaciers. River monitoring in Hurla catchment area near Kulu, Himachal Pradesh by iit Delhi scientists shows that flows increase during February to April due to faster melting (see graph: The melt-down factor?); this does not tally with the rainfall in the catchment. Pratap Chandnani, a hydro-engineer who collaborated with the iit scientists said "rock dams bursting when glaciers melt at a rapid rate are a flood risk'. He cites the example of the Parichu dam in Tibet bursting due to sudden glacier melting in 2004.

In spate of
On June 5 this year, just before the monsoon set in, the South Asia Network on Dams, Rivers and People (sandrp) warned that reservoir levels in the country had risen alarmingly. Was this water from excess glacial melt in February and April? Scientists, as we have seen, do not have a definite answer.

In any case, barely a week after they were issued, sandrp warnings proved true. "Forty one of the 76 dams monitored by Central Water Commission [cwc] had filled up to 20 per cent of their capacity, according to the commission records updated on June 16. Ideally, the storage level should be 10 per cent or less,' says Himanshu Thakkar of sandarp. "In 20 reservoirs, the water level was over a third of their capacity,' he added.

cwc sources, however, say the commission has a guide curve which specifies reservoir levels: they vary according to a dam's objective and location. "Dam authorities should keep in mind this guide curve and downstream flows while monitoring levels of reservoirs and releasing water from them. We give them regular information from satellite about the amount of water likely to flow into dams. They also have their own sources of information,' cwc sources said. Dam authorities maintain that they adhere to the commission's guidelines. But, says K K Sharda, deputy chief engineer of the Bhakra Beas Management Board, uncertain monsoons often militate against this. "For example, this year bad monsoons have kept water levels below the specified levels,' he said.

Representatives of civil society organisations differ. "cwc is very secretive about information. Of the several dams in the country, it releases weekly data on levels in only 76 dams. Water discharge data in the Ganga-Brahmaputra-Meghna and the Indus basins, for example is secret,' Ravindranath of the Assam-based ngo River Basin Friends said.

A statement made in parliament in 2005 by the then member of parliament, Jairam Ramesh, is indicative of the water resources ministry' attitude. "The integrated water resources development plan was developed by the ministry...in 1999. It is so confidential that when I asked for this report...I got a note...saying that