In Assam, heavy rains in the fourth week of April caused the first wave of floods that continued till August, leading to algal growth in the flood water at several places (Photo: Parikhit Saikia)
In Assam, heavy rains in the fourth week of April caused the first wave of floods that continued till August, leading to algal growth in the flood water at several places (Photo: Parikhit Saikia)

Deluge despite deficit

This year, floods have occurred even in areas that are arid or have received less than normal rainfall. As erratic weather events become more common with climate change, India urgently needs to formulate a national flood management policy
1.

Update (14/8/2017): India’s meteorological department forecast a normal monsoon in 2017, but excess rain over short periods has flooded some of the country’s driest regions (like Rajasthan and Gujarat) this year. Most recently, eastern states of Assam and Bihar have been hit by fresh floods and army has been called for help. In Assam, 21 of the 22 districts have been affected. Some 80 per cent of the Kaziranga National Park, home to India’s one-horned rhinos, has been flooded. Four districts in Bihar have been flooded as Tapti and Mahanadi rivers swell.

Floods are a regular in India. But there is something peculiar about them this year: they not only came early, but have stayed longer. And worryingly, they have not been proportionate to the rainfall. Madhya Pradesh and Rajasthan were in a drought-like situation when floods occurred in July-August, while Bihar and Assam—two states that are facing their worst floods in almost three decades—have had deficit rainfall this monsoon. These patterns seemed to have escaped attention.

Nobody took notice when floods hit Assam this April. A new government was just sworn in and the media did not pay much heed to the unusually high pre-monsoon rain. Everyone thought the rising waters of the Burhidehing and Desang rivers would subside. It did not happen. The flood continued till the end of August, covering more than 90 per cent of the state and affecting over four million people.

“The pre-monsoon rainfall this year has been very strong, both in terms of the absolute rainfall and its geographical coverage,” says D C Goswami, former head of the Department of Environmental Science, Guwahati University. According to the National Remote Sensing Centre, Hyderabad, heavy rains in the fourth week of April caused the first “wave” of floods in Assam. On April 25, the Burhidehing and Desang crossed the danger mark and caused floods in six districts—Dibrugarh, Sivasagar, Jorhat, Tinsukia, Cachar and Charaideo. Experts say the heavy pre-monsoon showers have provided the clinching evidence of the increasing frequency of western disturbances that bring the rain. According to the Regional Meteorological Centre, Guwahati, rainfall in the state was 21 per cent above normal in March-April.

What is frightening is the fact that such unusual rain will become more common with climate change. “It is the pre-monsoon showers, rather than the monsoon, which will increase because of global warming,” says Subashisa Dutta, professor at the Department of Civil Engineering, IIT-Guwahati. In 2012, Dutta studied the impact of climate change on the hazard of floods in the middle Brahmaputra reaches and found that the peak discharges from the river, the average duration of a flood wave and the intensity of pre-monsoon showers are likely to rise during 2071-2100.

It is quite ironic that Assam witnessed floods when its total monsoon rainfall between June 1 and August 16 was 25 per cent below normal. So, what caused the flood? Data suggests that most floods were preceded by extreme rainfall events (see ‘Under water’).

Most of the floods this year were preceded by extreme rainfall events

The India Meteorological Department (IMD) divides extreme rainfall events into two categories: rainfall of 124.5-244.4 mm in 24 hours is “very heavy” and rainfall more than that is “extremely heavy”. In July alone Assam recorded six “very heavy” rainfall days. By July 25, the Central Water Commission (CWC) had declared moderate floods in three districts of the state where the Beki, Sankosh and Brahmaputra rivers were flowing above the danger mark.

Arup Kumar Sarma, professor at the Department of Civil Engineering, IIT-Guwahati, says the floods this year were unusual in another way: they seem to have caused more damage. “This is because the direction of the storm and the direction of flow of the river was the same. In such a situation, rains directly contribute to the rise in water levels,” he explains.

However, there is not enough long-term data to claim that extreme rainfall events have increased over the years. For instance, in Assam, no significant trends in 24-hour rain were observed between 1951 and 2010, says the state’s action plan on climate change. The report, however, states that extreme rainfall events in Assam may increase by 5-38 per cent and incidences of floods by 25 per cent during 2021-2050.

Though Bihar received 15 per cent deficit rain this monsoon, 3.8 million people in 23 districts faced floods (Photo: Chinmaya N Singh)
Apart from Assam, floods impacted nearly five million people in 19 states by the third week of August. Most of these were triggered by the first monsoon showers. Remarkably, just before the floods, some of the states were hit by drought. In Madhya Pradesh, floods came after a drought in April-May. Until June 8, the monsoon had hardly covered the state and IMD had forecast severe to very severe heat wave in some parts. But on July 12, the 24-hour rainfall in Burhanpur and Betul districts was an astonishing 1,135 per cent and 1,235 per cent above normal respectively. By July 18, more than 35 lives were lost and nine people went missing as 23 districts in the state faced flood-like situation. Again, on August 20-21, Bihar’s 12 districts witnessed extreme rain events. In Sagar, the Khurai meteorological station recorded 187.2 mm rainfall on August 20. Ken and Tamsa rivers were flowing above the danger mark due to heavy rains in Panna and Rewa districts. The total death toll in the state due to rain-related accidents from the beginning of the monsoon season to August 22 stood at 102.

Similar was the case with Bihar where the cumulative rainfall between June 1 and August 22 was 15 per cent below normal. However, 3.8 million people in 23 districts, including Patna, were facing floods by August 21. The Ganga was flowing above the danger mark in Buxar, Munger, Khagaria, Siwan and Katihar. Bihar Chief Minister Nitish Kumar said silt deposition in the river after the construction of Farraka barrage in West Bengal caused the rise in water levels.

However, D K Mishra, convener of Bihar-based non-profit Barh Mukti Abhiyan, says, “Since 2007, the cumulative seasonal rainfall in the state has been abnormally low and the flood this year was caused by heavy precipitation in Nepal” (see ‘Floods no longer a rural problem’ on p29). Regardless, the state did see extreme rainfall event on 11 days in June-July. On July 31, the Madhwapur meteorological station recorded 200 mm rain.

In the adjacent state of Uttar Pradesh too, the Ganga was flowing above the danger mark in Mirzapur, Ballia and Varanasi districts while parts of Allahabad were inundated on August 21.

In Rajasthan, conditions were under control till July. “In August, the intensity of rainfall shot up like never before,” says Y P Mathur, professor at the Department of Civil Engineering, Malaviya National Institute of Technology, Jaipur. On August 11, the rainfall in nearly half of the districts was more than 100 per cent above normal. Pali and Sikar districts received a phenomenal 1,059 per cent and 1,100 per cent above normal rain. According to local newspapers, rainfall in Jodhpur was the heaviest in the past 90 years. The gates of the Bilaspur dam had to be opened to drain water into the Banas river to prevent flood in Tonk district. Extreme rainfall events were again recorded in 16 meteorological stations of the state on August 20-21. According to Rajasthan’s draft action plan on climate change, published in 2010, the frequency and intensity of extreme rainfall events in the state is likely to increase by 20 mm a day during 2017-2100.

Uttarakhand faced floods on May 28 itself, when a cloudburst hit Tehri district. Two people were killed in the flood that followed and nearly 400 houses were reported to have been damaged till June 4. By July 25, flood had occurred in 13 districts and 49 lives were lost.

Scientific studies point out that while monsoon rain is decreasing in India, extreme rainfall events are increasing, particularly in July-August. A paper published in Nature Climate Change in 2014 states that the intensity of extremely wet spells and extremely dry spells during the South Asian monsoon season have been increasing since 1980. “We are looking at rainfall extremes that only occur at most a few times a year, but can have very large impact,” says Noah Diffenbaugh, senior fellow at the Stanford Woods Institute for the Environment, under whom the study was conducted. The team compared data collected by IMD and other sources over the past 60 years for July and August. It found that the average rainfall over Central India has reduced from 10 mm a day to about 9 mm a day. However, the day-to-day variability in rainfall in these months has increased, leading to periods of heavy rainfall or prolonged dry spells.

Apart from causing loss of life, floods are a huge financial drain. They are the biggest threat to agriculture, impacting 7.5 million hectares (ha) of farmland every year, particularly in agrarian states like Uttar Pradesh, Bihar and Assam. In Assam, close to 0.1 million people move to relief centres every year. Same is the case with Bihar, where thousands spend four to five months in camps. According to an analysis by Down To Earth, state governments of Assam and Bihar are officially involved in flood relief and rehabilitation works for seven months every year. And it is likely to get worse. Data shows that flood-affected areas in the country have increased by 162 per cent since 1960 (see ‘Uncontrolled surge’). The average annual flood damage has also increased by over 160 per cent—from R1,805 crore a year during 1947-1995 to R4,745 crore a year during 1996-2005, according to the National Disaster Management Authority. Though there are no recent estimates, it would not be wrong to assume that with growing urbanisation, the damage would have only gone up in the past decade.

Source: Central Water Commission data; Rashtriya Barh Ayog report, 1980; Planning Commission working group on flood management for the 12th Five Year Plan

According to a working group on flood management, set up by the erstwhile Planning Commission in 2011, R1,26,000 crore have been spent on flood control in the country since 1953 and the economic loss caused by floods has been R8,12,500 crore at 2011 prices.

The working group said that this cost would go up by 10 per cent every year. This means that in 2015, the total cost would have been R11,46,940 crore, or 7.6 per cent of India’s estimated GDP for 2016-17. To make matters worse, the states do not take the problem seriously since the Constitution does not mention flood management under either the Union, State or Concurrent lists. “States give damage estimates only to avail relief packages from the Centre,” says a CWC official, requesting anonymity. During the tenure of the working group only Bihar, Assam, Rajasthan and Kerala took part in the deliberations. Of these, two states did not attend the deliberations after the first meeting. Even five years after the Planning Commission gave its estimates on the total flood damage in the country, the states have not confirmed or denied their accuracy.

After 40 years
 
Experts differ on the relevance of the recommendations made by Rashtriya Barh Ayog in 1976 in today's time
 
KAMTA PRASAD
MEMBER, RASHTRIYA BARH AYOG

Recommendations of the Rashtriya Barh Ayog (RBA) were quite comprehensive and received praise. Even the erstwhile Planning Commission's working groups on flood management acknowledged this. But, sadly, these have not been implemented even 40 years after they were made. There is a lack of political will to implement them, engineers are reluctant to modify their methods and the bureaucracy is not prepared to involve communities. RBA had also suggested short/medium-term measures, such as construction of embankments, and long-term strategy, such as training communities in disaster risk reduction. But the government has only focused on building embankments. That is the problem.”

R RANGACHARI
CHAIRPERSON OF THE COMMITTEE TO REVIEW THE IMPLEMENTATION OF RASHTRIYA BARH AYOG RECOMMENDATIONS

We found that states were not sincere in the implementation of Rashtriya Barh Ayog recommendations. They cited inter-state issues and lack of funds, but wasted money on unplanned relief and rehabilitation works. There was hardly a law against encroachment on the floodplains, and where laws did exist, they were not enforced because politicians had a stake in developmental works on the floodplains. States need to understand that rivers have no political/geographical boundaries. The Centre needs to play a more proactive role in flood control. Long-term planning, as opposed to short-term solutions, is key.

NIRMAL SENGUPTA
CHAIRPERSON OF THE WORKING GROUP ON FLOOD MANAGEMENT FOR THE 12TH FIVE YEAR PLAN

While there is no denying that the Rashtriya Barh Ayog report was an excellent assessment of the flood situation in India, some of its recommendations are not applicable today. For instance, floodplain zoning caught the fancy of policymakers in the 1980s primarily because of its popularity in the US. But flood-prone areas have increased substantially since then and moving people out of these zones is not feasible. Flood is a complex problem that requires the attention of a number of departments at the state and national level. At the same time, we need to learn to live with floods and develop the coping capabilities of communities.”

Even 40 years after India’s first and last commission on floods was constituted, the situation has not improved. There is no national-level flood control authority in the country. Rashtriya Barh Ayog (National Flood Commission or RBA) was set up by the then Ministry of Agriculture and Irrigation in 1976 to study India’s flood control measures after the projects launched under the National Flood Control Programme of 1954 failed to achieve much success. In 1980, RBA made 207 recommendations and four broad observations. First, it said that there was no increase in rainfall in India and, thus, the increase in floods was due to anthropogenic factors like deforestation, drainage congestion and badly planned development works. Second, it questioned the effectiveness of the methods adopted to control floods, for instance, embankments and reservoirs, and suggested that the construction of these structures be halted till their efficacy was assessed. However, it did say that embankments could be constructed in areas where they were effective. Third, it said that there has to be consolidated efforts among the states and the Centre to take up research and policy initiative to control floods. Fourth, it recommended a dynamic strategy to cope with the changing nature of floods.

An analysis of the report suggests that the problem begins with the methods of estimating the flood-prone area of the country (see ‘Uncontrolled surge’). An accurate estimate is crucial for framing flood management programmes. RBA estimated that the total area vulnerable to floods in 1980 was around 40 million hectares. The figure was reached by calculating the maximum area affected by floods in all the states in any one of the years between 1953 and 1979, and adding to it the area where flood protection works had been undertaken. The areas where protection works failed were then subtracted from this total.

However, this is a flawed methodology and RBA itself acknowledges this fact. “It is clear that while the maximum area flooded in any one year may broadly indicate the degree of the flood problem in a state, it does not strictly indicate the area ‘liable’ to floods as different areas may be flooded in different years,” states the report.

There is another problem. The very definition of flood-prone area does not reflect the effectiveness of the flood management works undertaken, wrote G S Purba, former chief engineer at CWC, in a paper presented at the Indian Disaster Management Congress in 2006. “One reason for this could be that we are still unsure of the performance of flood management measures,” says V D Roy, member, CWC.

Despite its flaws, the method continues to be used. “We have begun collecting data for a more scientific assessment—one that relies on frequency-based climate modelling— but adopting it in full force will take time,” says Roy. In a 2011 meeting of the working group on flood management for the 12th Five Year Plan, C Lal, Director of Flood Management Programme, CWC, acknowledged that scientific criteria need to be adopted to assess flood-prone areas. This should be based on frequency of flooding and period of inundation as gauged by contour maps and satellite imagery, he suggested.

The RBA report also recognised the need for timely evaluation of flood management projects. It entrusted state irrigation and flood control departments, CWC, Ganga Flood Control Commission and the Brahmaputra Board with the task of adopting or discarding them on the basis of their performance. But this has not been the case. “In 2001, we found that state governments had hardly done any assessments of flood control projects,” says R Rangachari, chairperson of the committee and former CWC member.

Even when flood management projects are evaluated, the reports are not credible. “Most post project evaluations that we laid our hands on did not have enough data,” says Rangachari. Moreover, the evaluation is generally done by departments that undertake flood-control projects. “I haven’t come across any third-party evaluation of flood management projects in Assam,” says Goswami.

A major problem is the inaction on recommendations of evaluation reports. For instance, in 1978, RBA asked the Programme Evaluation Organisation of the Planning Commission to review the Kosi embankments. The study, published in 1979, concluded that embankments had, in fact, enhanced the flood problem. But still the embankments were raised by 2 m in 1987-88.

Another key recommendation of the report was that state governments should legislate on floodplain management, particularly encroachment and floodplain zoning (demarcating a floodplain into zones and specifying the level of development allowed in each). But apart from Rajasthan and Manipur, no state has enacted laws on floodplain zoning. Same is the case with legislations on encroachment. Only Bihar, Punjab, Andhra Pradesh and Uttar Pradesh have laws against encroachment of floodplains. Even Assam, a state that gets flooded almost every alternate year, does not have specific laws against floodplain encroachment. State governments argue that such laws would hinder development.

The RBA report also said that state governments must focus on maintenance of completed works rather than constructing new structures. In Assam, for instance, over 4,470 km of embankments were constructed till the 1970s, says Ravindranath, founder of Assam-based non-profit Rural Volunteers Centre (see ‘Train people in disaster risk reduction’). Over 90 per cent of these have outlived their life and need maintenance. During floods this year, breaches were reported at 28 places, stated water resources minister of Assam Keshab Mahanta in the Assembly on August 11.

In 1981, the government accepted all RBA recommendations and circulated them to the states and concerned ministries. This was the first attempt in the country to evolve a national response to floods. However, between 1987 and 2001, CWC, four regional task forces, Planning Commission’s working groups on water and flood management and an expert committee constituted by the Ministry of Water Resources reviewed the implementation. They found that the recommendations had not been implemented. The expert committee, the last major effort to track the RBA recommendations, had 10 meetings with various groups and state governments and undertook field trips to flood-prone areas. It concluded that while international and inter-state issues, fund constraints, population pressure and lack of infrastructure were the main difficulties in implementation of some of the recommendations, there were several others which could have been implemented. “It has been 12 years since we pointed out the loopholes in implementation, point by point, but little action has been taken by the states and the Centre,” says Rangachari.

Floods no longer a rural problem
 
Earlier, floods were mainly a rural problem and would last for a couple of days. This has changed—floods now also hit urban areas and last for months.

As flood-prone area in the country continues to increase, there is unprecedented pressure on the government to find a permanent solution to the problem. This would remind some of us of the early 1950s, when the government released its first Flood Policy.

Reading out the policy in Parliament on September 3, 1954, Gulzarilal Nanda, the then planning minister at the Centre, said that flood control measures would be rolled out in three phases: the first two years would be devoted to construction of revetments and spurs at selected sites, the next six-seven years would be devoted to medium-term flood protection measures such as construction of embankments and improvement of drainage channels, and another three to five years would be devoted to long-term measures such as construction of storage reservoirs. But things did not move the way the government desired. The gap between planning, implementation and benefits remained elusive and the government set up Rashtriya Barh Ayog in 1976 to review the performance of flood control measures. The Ayog published its report in 1980, and made 207 recommendations. Many of these involved other countries and, as a result, no unilateral steps could be taken.

As the situation stands today, the states have constructed 35,200 km embankments along rivers, dug 39,710 km of drains and raised 7,713 villages to protect themselves from floods. Despite this, the flood-prone area of the country is around 50 million hectares—almost double of what it was before any flood protection work. It seems the investment in the flood control sector has done more harm than good.

D K Mishra is convener of Bihar-based non-profit Barh Mukti Abhiyan

So what is the way out? Sarma says that in view of climate change and varying precipitation and river flow patterns, hydraulic structures, embankments and reservoirs must be designed to accommodate worst case scenarios. Dams should be designed with a cushion to absorb flood runoff. Till 2011, India had 4,728 large dams (height greater than 100 m and capacity more that a cubic kilometre) and 397 were under construction. However, of the 78 large dams assessed by the 2011 working group on flood management for 12th Five Year Plan, dedicated flood cushion was provided only in 10. “As a policy, a minimum flood cushion of 10 per cent of the storage capacity should be provided in all new dams and, if affordable with respect to other purposes, providing even a flood cushion of 20 per cent could be considered,” the group recommended.

Similarly, embankments, which are currently designed to withstand floods once in 100 years in urban areas, need to be designed to withstand floods once every 10 years. Ravindranath says that most embankments in India are earthen and vulnerable to breaches. “Railway engineers build concrete embankments along railway lines. Why can’t the state irrigation departments follow their lead?” Ravindranath asks.

Lack of communication between states and among countries is another reason for devastation caused by floods. When one state releases water from a dam it must inform the states that fall in the lower reaches of the river well in advance. This would give time for evacuation. The same holds true for countries that share rivers. Ravindranath says that there are enough laws on these matters but they are rarely followed.

There is also a need to improve our forecasting capacity. At present, 175 cwc stations measure river levels and 1,289 automatic rain gauges of IMD measure precipitation. While CWC and IMD have begun modernising, the coverage of rain-gauge stations in hilly areas is still inadequate.

H S Dhami, vice-chancellor of Kumaun University, says that the weather changes every two to 5 km in the hills, as opposed to the plains, where it is uniform for even a 100 km stretch. Climate change will only aggravate these variations and we need to be prepared. Implementation of the RBA recommendations could have given a major boost to India’s flood mitigation efforts, but we have already lost 40 years.

Train people in disaster risk reduction
 
Carrying more water per unit area than any other river in the world, the Brahmaputra basin is one of the world's most dynamic and complex fluvial systems. While Assam has seen floods since antiquity, unregulated construction on the hills, failure of embankments to check water flow and, now, extreme rainfall events have aggravated the flood problem.

More than 4,000 km of earthen embankments were built along the river between 1953 and the 1970s. But these get breached every year. In 1998, there was a fatal flood, aggravated by 358 breaches on the Brahmaputra dyke. To add to it, there's a new construction project being declared everyday. There are two big projects under construction. The 1,700 km trans-Arunachal highway between Tawang and Kanubari is expected to be completed by 2018. Construction of the Asian highway began in 2012, and, once completed, the road system will stretch 3,200 km, from Moreh in India to Thailand's Maesot via Myanmar's Tamu, Mandalay and Myawaddy. Aggregates used in road construction for all these projects are mined locally. Mining loosens the soil and contributes to the rise of the river bed.

The only way out for communities that face floods every year is to get trained in disaster risk reduction. The National Disaster Relief Force helps in rescue, rehabilitation and trauma-counseling post-floods. But relief is a mela—the government sanctions month-long supplies of rice, dal, oil, candle, matchbox and fuel for flood victims but they get supplies only for two-three days. The saddest part is that the corrupt can be taken to court under the Disaster Management Act of 2005, but people are not aware of their rights. The government must focus more on non-structural solutions. Non-profits and private actors cannot bear the entire load of training communities. Ravindranath is the founder of Assam-based non-profit Rural Volunteers Centre
Down To Earth
www.downtoearth.org.in