'The Worst Tsunami Could Trigger A 15m Wave In Gujarat'

Will the Tsunami Warning System (TWS) proposed to be installed inHyderabad be a part of the larger Indian Ocean TWS?

From the media reports I have read, my feeling is Mr. Kapil Sibal (UnionMinister for Science and Technology) is talking of an Indian TWS, which willcollaborate with the international one for the Indian Ocean. So the September2007 goal is for the Indian TWS, and I think they are well on target.

The Atlantic Ocean TWS is supposed to be in place by this year-end. And isit not necessary for the Indian TWS to be part of the larger Indian Ocean TWS?

The Atlantic Ocean TWS involves only two countries, the US and Canada. Theyare already partners in the Pacific Ocean TWS since 1948, and now they need toinvolve Europe. Besides, tsunamis in the Atlantic Ocean are local, minoroccurrences. However, India has to work with 37 other nations in the IndianOcean rim.

You have said that the Pacific, Atlantic and Indian Oceans arestructurally different and the tsunamis they generate also differ vastly.

The Pacific Ocean triggers 6 to 9 tsunamis a year. However, most of these(but not all) are small with little or no impact. We don’t generally issue awarning when the tsunami is extremely small. Both the Indian and Pacific oceanshave tectonic converging plates which generate tsunamis. But the numericalmodels (on which the warning systems are based) for both oceans have to be verydifferent. After the December 26 tsunami, the Americans, Japanese andRussians-who have worked only on the Pacific system-want to transplant thePacific Ocean TWS on to the Indian Ocean. That’s okay for the instrumentation,but not okay for the numerical models which are the brain of the system.

The Pacific Ocean is the largest ocean. A tsunami here travels largedistances and goes to every corner of the Pacific Ocean; it gets reflected andcomes back. But by then, the action is all over. Whereas the Indian Ocean isvery small, and is cluttered with islands. Even as the tsunami is under way, thereflected waves begin coming back. For instance, in Kollam, Kerala, the directtsunami waves alone did not cause all the damage. The waves reflected fromLakshadweep islands also contributed. The Indian Ocean is a semi-closed systemwhere the reflected waves are significant. The Pacific is wide open, where we donot worry [in the warning system] about reflected boundaries. A closed system isvery difficult to model-mathematically and physically. We need all kinds ofboundary data-about estuaries, rivers, backwaters-in the Indian Ocean. That iswhy the Indian Ocean TWS takes more time. You can buy instruments and put them,but development of computer models will take some time.

What instruments?

The seismometers that record the earthquakes. The IMD and NGRI may upgradeand enlarge the existing network . I am not aware of these details. Same withthe existing tide gauge network which records the tides regularly , but alsowill record the tsunamis when they occur.

Are you officially consulting with the Department of Science andTechnology or the Government of India?

No, I have no official status in India. But I am working with various groupshere, as an individual scientist.

Other countries have tapped on your knowledge though…

Yes, Canada, Australia and the USA. In India, I was invited for the firstthree brainstorming sessions; I am working with the IITs in Delhi and Kharagpur.But I don’t have any official status.

You have spoken about the need for detailed coastal inundation maps to begenerated by all the coastal states, and how the TWS would be incomplete withoutthese maps. Of India’s nine coastal states, how many are working on thesemaps?

I am not aware of the status in every state. The only state I am aware of isGujarat where I met chief minister Narendra Modi on September 23 at a tsunamiworkshop in Gandhinagar. I felt that Gujarat is serious in developing proceduresto protect its population and coastal infrastructure from tsunamis. I also metthe Chief Minister and relief commissioner of Pondicherry on October 25 at theirinvitation and they also are quite serious about tsunami preparedness. I alsowas invited by the Indira Gandhi Centre for Atomic Research in Kalpakkam onOctober 23 where I gave a talk on the numerical models that should be developedto prepare vulnerability maps.

What is the scenario in Tamil Nadu?

No official from TN ever contacted me. So I cannot say what is happening.

But do affected states like TN and AP have the infrastructure to generatecomprehensive coastal inundation maps which you have argued are most crucial incountering the damage unleashed by a tsunami?

I do not know… But I am sure they must be doing something about it. I havesome idea of what the central government is doing through media reports and bytalking with colleagues-they have designated INCOIS in Hyderabad as the TsunamiWarning Centre, and NIOT in Chennai is involved with the instrumentation andmodelling; the National Institute of Oceanography in Goa is also involved in themodelling. But I have no idea about what the state governments are doing exceptGujarat where I am working closely with Dr. Arun Bapat who is an advisor toGujarat on earthquake and tsunami issues.

Indonesia has talked about an early warning system which includes seabedsensors to be in place by end-2005. The reaction time will be 10 minutes to thedate centre, and from there in five seconds it will reach people. Is thisindependent of the larger Indian Ocean TWS?

Indonesia is the most affected country in the Indian Ocean.and most tsunamisin the Indian Oceean would start there. So there’s political pressure on theIndonesian government to do something. But when we have 37 countries in theIndian Ocean, we cannot do things overnight.

Should India have seabed sensors?

NIOT and others will decide these issues. The central government institutionsare on top of the things. Even if you have all the money in the world, thesethings take time. The updating of the Pacific Ocean TWS for Canada (that is, thebrain of the warning system), which I oversaw, took three years [in the 1990s].The evidence is that India has moved very swiftly on the TWS, probably fasterthan anybody else. Others are yet to catch up.

Since the Indian Ocean is more complex than the Pacific, will it take morethan three years?

No. As we learn from the past we grow faster. We also have much fastercomputers now. The generation-propagation models, which I recommend the centralgovernment institutions adopt, should take no more than a year from now. And ifevery state starts today on inundation models, they could all be completed in 18months.

What does an inundation model tell us?

It tells you the heights of the tsunami waves at each selected coastallocation. The public wants to know mainly two things: how high would the wavesbe, and how far inland would the water come.

In the December 2004 tsunami what was the maximum wave height in India andthe maximum reach inland?

I was not directly involved in any tsunami coastal surveys in India. But fromwhat I read in the reports, it did not go beyond 1.2 km inland anywhere. Even inan undocumented case, the maximum it could have gone is 1.5 km. And the maximumwave height was 8 metres.

What could be the worst tsunami waves the Indian coast could witness?

I have examined, irrespective of the size of the earthquake, what amplitude agiven coast can support. It varies from place to place. When we look at some ofthe scenarios for India, the highest amplitude can occur in Gujarat. Even in theworst scenario, the maximum height of waves will be 15 metres in the Gulf ofKutch. It cannot ever go up to 100 m since no coast-given the coastal geometry,ocean bathymetry, land level etc-can support such levels of water, howeverstrong the quake.

To give an analogy, a bucket, however fast the water gushes from the tap, cancollect only a specified volume. Similarly, no matter how big the quake, a givenarea of the coastline can only support a certain maximum volume. The Gulf ofCambay will see a wave of 13 m at best. In Karnataka, Maharashtra and Kerala itcan rise to 7m or so. In southern Tamil Nadu, Nagapattinam and Cuddalore, themaximum a wave will rise is to 9 to 9.5 m. In Orissa, the Paradip coast cansupport a high amplitude (owing to a degenerate semi-diurnal tidal amphidromicpoint there) of up to 10 m. In Bengal, up to 8 m. In the Andaman, it could go upto 11 m. These are rough calculations I have made; the numerical models willgive us a better picture.

How do we factor future tsunamis with data from inundation and numericalmodels?

You cannot run the models after an earthquake and tsunami happen. All themodelling has to be done beforehand. You model every conceivable possibilityfrom minus to plus infinity. Every parameter-the intensity of a quake, itsvarious epicentral locations, the state of the tide when a quake happens-has tobe computed. It will be one model, but you run countless scenarios on it. Yousynthesise these results and put them into the warning system. So when the realtsunami happens, all you need to do is punch a few numbers-the size of thequake, the seismic moment, epicentre-and the computer will pull out the closestmatch. Then you use this to warn hundreds of places on the Indian Ocean coast onwhat the amplitude and inundation would be. I have estimated that a minimum of87 inundation models will cover the east and west coasts of India as well as theAndaman & Nicobar Islands. A breakdown of this , for example, is 12 modelsfor Tamil Nadu.

How much of the damage to life and property will be reduced with aneffective TWS?

Over 95 percent of human lives can be saved, but loss of property-thephysical damage-will be there. You cannot stop the tsunami from happening. Youcan only improve the natural environmental barriers which may help somewhat toreduce the damage.

Can you tell us about the travel-time atlas that has been devised inIndia?

Many people think the bigger the quake, the faster the tsunami travels. Nottrue. It only depends on the depth of the ocean. The amplitude of the tsunamidepends on the size of the quake, but not its speed. Because we know the oceandepths, we can compute the travel times once and for all. We don’t need towait for the earthquake to happen. And this is what we have done with IITKharagpur as the lead agency. IIT Delhi was involved and so was I. AvijitGangopadhyay from the University of Massachusetts was also part of the team. Wehave completed the travel-times for all the 37 countries in the Indian Ocean for250 locations. On the Indian coast, 49 locations have been chosen.

Travel time from which point of the tsunami’s occurrence?

The travel time atlas works backwards. You choose a point on any coast andwork backwards. A tsunami may occur anywhere. Since the contours [of the atlas]are in hours, no matter where the tsunami occurs, you can determine the traveltime for a specific point. This atlas should be printed and ready by end ofNovember.