For a reliable system

In the matter of establishing a tsunami warning system India plans to go on its own despite the preference of experts for a regional network.

R. RAMACHANDRAN in New Delhi

A GREAT deal has been said and written about how lives on the Indian east coast could have been saved if India was a part of the Pacific Tsunami Warning System (PTWS) network. Twenty-six countries from the Pacific region are members of the network, which was established by the Hawaii-based Pacific Tsunami Warning Centre (PTWC) in 1968. This was following the creation of the International Tsunami Information Centre (ITIC) in 1965 by the Inter-governmental Oceanographic Commission (IOC) of the United Nations Educational, Scientific and Cultural Organisation (UNESCO).

It is a system meant for the earthquake and tsunami prone Pacific region. India, naturally, is not a member of this network as it does not make logical sense to be part of a system that seems irrelevant for India, which is situated over 3000 km from the Pacific and with almost zero risk from natural hazards originating from that region.

Consider the following. Over 790 tsunamis have occurred in the Pacific since 1900. In contrast, the December 26, 2004, tsunami that hit the Indian mainland was only the second in this region during the same period. The last one was in June 1941 as a result of an earthquake of magnitude (M) 8.1 in the Andaman Sea. However, as reliable scientific records of that period are not available, presumably owing to the Second World War, its real impact is not known. Tide gauge data of the 1880s indicate that a tsunami occurring off Car Nicobar coast in the Andamans, owing to an M7.9 earthquake in December 1881, did affect the Indian coast. Similarly, the massive tsunami set off by the Krakatao volcanic explosion of August 1883 also seems to have reached the Indian shores.

Being close to a subduction zone between the Indian plate and the Burmese micro-plate, the Andamans is a highly earthquake-prone region where earthquakes of a magnitude greater than 5.0 do occur periodically. But not all ocean earthquakes cause tsunamis. To generate a significant tsunami in the Bay of Bengal region that can propagate large distances, the earthquake should have a magnitude greater than 7.2, according to Tad Murty, the Internationally renowned tsunami expert of Indian origin from the University of Manitoba, Canada. Moreover the fault characteristic should be of the dip-slip or thrusting type, rather than strike-slip type which apparently is more characteristic of the Andamans (Frontline, January 28). But such tsunamigenic quakes have been rare. It must also be pointed out that several large aftershocks, including one that measured over M7.0, have occurred in the Andaman region following the M9.0 Sumatra earthquake that caused the biggest tsunami in recent history. But none of them generated any significant tsunami that reached the Indian mainland.

So, the probability of a tsunamigenic earthquake in the Andaman region is very low. Indeed, past history would suggest that the probability of a tsunami from earthquakes in the Sumatra region propagating up to the Indian coastline is also very low, even though tsunamigenic earthquakes that result in significant local tsunamis, are common in this region. The last known major tsunamigenic quakes off the Sumatra coast were in 1833 and 1861, of magnitude 8.7 and 8.5 respectively. Though these were large earthquakes, since it was before the arrival of tide gauges, their impact on the eastern Indian coast is not known (Frontline, January 28).

Even if India were part of the PTWC network, a warning such as the one that was sent out by the Centre soon after the detection of the Sumatra earthquake would have been of limited use because there was no threat assessment to the Indian Ocean region. In fact, such an assessment could not have been made because historical earthquake and tsunami data of the region would have indicated little about the impending disaster.

The occurrence of the Sumatra tsunami will not alter the risk probability significantly; it will continue to be very low for the region as such a large earthquake is unlikely to recur for a long time, perhaps a couple of centuries. However, the event has drastically changed the perceptions of tsunami threat and hazard mitigation in the region. The enormity of the disaster has drawn the attention of the United Nations as well as individual technologically capable nations to the problem of establishing a tsunami early warning system in the region. At the recently concluded United Nations Conference on Disaster Reduction at Kobe, Japan, the U.S., Germany, France, India and Indonesia made independent proposals for establishing a tsunami warning system in the Indian Ocean, even as the UNESCO mooted a system as part of a global network for disaster reduction under the U.N. umbrella.

According to UNESCO Director-General Koichiro Matsuura, the regional network, together with the related equipment and a regional centre, would be established by mid-2007, with an initial cost of about $30 million and an annual maintenance cost of $3 million. The plans for establishing this would be finalised at the June meeting of the IOC, he said.

However, the U. S., which announced its plans to expand the Pacific network to include the Caribbean and the Atlantic with 36 more DART-gauges after the Sumatra event, expressed its scepticism about the U.N.'s capability to implement such a system. Instead, it proposed that the international Global Earth Observation Satellite System (GEOSS) programme, of which 55 countries are members, could be used to establish the Indian Ocean system under the U.S. leadership at a cost of $37.5 million.

Amidst jockeying by individual nations, the draft statement at the U.N. Conference called for international cooperation in establishing a tsunami early warning system for the Indian Ocean "tailored to the specific circumstances of the Indian Ocean and the individual requirements of countries, under the coordination of the U. N." Towards this, a total of $8 million has been pledged by some advanced nations, of which Japan alone has committed $4 million.

However, as was made clear by Minister for Science and Technology Kapil Sibal at the Tsunami Brainstorming Conference held in New Delhi during January 21-22, India plans to go on its own, even as Indian and foreign experts underscored the importance of being part of a regional network that included all the countries of the Indian Ocean region. The broad framework for the system, which is estimated to cost Rs.125-130 crores, was unveiled at the Delhi meeting, which was organised by the Indian National Science Academy (INSA) in association with the Department of Science and Technology (DST) and the Department of Ocean Development (DoD).

The key elements will be in place by March 2006 and the system is expected to be operational by September 2007. "India had the technological capability to put in place an effective warning system," Sibal said. It was, however, stated that India would be willing to contribute to the Tsunami Early Warning System for the Indian Ocean region. The active institutional mechanism developed by IO-GOOS, the regional Indian Ocean alliance for Global Ocean Observations System of the IOC, could be effectively used to disseminate the tsunami warnings to the countries in the region in real-time, India has suggested.

In the proposed form, it is more an Indian Tsunami Warning System than an Indian Ocean Tsunami Warning System. It will be similar to the Pacific network based on DART-type Ocean Bottom Pressure Recorders (BPRs) coupled to moored ocean buoys and integrated with the existing tide gauge network of the Survey of India (SoI) and seismic station network of the India Meteorological Department (IMD), both of which will be significantly improved and expanded.

India has about 212 seismic stations under various organisations. As many of them are in research mode, they are not taking observations continually. The operational network is the one that is operated by the IMD comprising 51 stations, 17 of which are digital and networked. With the help of the digital network, it is now possible to determine location and magnitude within 15 to 30 minutes of an earthquake. Many more stations need to be made digital and brought into this operational network to improve accuracy and reliability. The IMD plans to expand its network by connecting 177 stations as part of the early warning system so that earthquake parameters can be determined in 10 minutes.

Likewise, the SoI's network of coastal tidal gauges will be made robust to withstand tsunamis and be able to measure them as they shoal up. At present, these can measure waves of heights up to 3 metres only. This network will be expanded by about 50 or more gauges and all of them will be made on-line. The SoI and the National Institute of Oceanography (NIO) will be undertaking high-resolution bathymetric mapping of the Indian region. This data will be required for model studies and numerical simulations of coastal run-ups.

A fundamental policy shift in the internal reporting of earthquake information is also being brought about. Earlier, only an earthquake of magnitude greater than 5 within the country was reported to the agencies concerned. Now it has been decided to report all earthquakes above M5.0 irrespective of whether they are within the country or not. It was due to this directive that the IMD, even though it had recorded and measured the Sumatra quake immediately, did not inform other departments concerned. Global sharing of seismic data in real-time is, however, still regarded as sensitive for inexplicable reasons.

IT is, however, important to emphasise that, though operational, the Pacific system is in a state of evolution and refinement. In particular, the DART-gauges are still in the research mode with effectively only the U.S. (out of a total of 26 countries) having installed these devices. So far, the DART system has been used in situations when the data from the BPRs indicated that tsunamis will not result and the predictions have turned out to be correct. The last such instance was in November 2003 off the Hawaiian coast. The system's performance is still to be tested in a situation when a tsunami does occur.

One of the main reasons for other nations in the Pacific Rim not installing DART-gauges is lack of money, points out Costas Synolakis, head of the tsunami project at the University of Southern California. Even though the U. S. itself needs 12 of these, only six have been funded so far since 1997. A single such system costs $250,000 and annual maintenance cost per system is $125,000 (at 2001 prices). This does not include the cost of the ship time needed for monitoring and maintenance cruises. Significantly, the lifetime of the system - given the extreme conditions in the oceans - is only two years, according to the NOAA. The question of cost, therefore, becomes important, if India wants to implement the system autonomously.

This also raises the issue of where will India source the BPRs from. As regards the buoys, the existing network of data buoys (costing about Rs.50 lakhs) of the DoD could be augmented to receive data from BPRs. Though the exact number of BPRs needed for the Indian system has not been firmed up yet, it seems to be of the order of 8-10. According to Synolakis, the U.S. manufacturer's present capacity is only about 3-4 gauges a year and the U.S. demand itself should occupy it fully. Apparently the Germans have their own version of the pressure recorders and, if need be, these can be procured, the Minister said.

However, given the Indian Space Research Organisation's (ISRO) expertise in developing transducer-based pressure recorders, U. R. Rao, the former Chairman of the Space Commission, made a strong case for ISRO implementing the entire system. According to him, R&D for developing these is already 50-60 per cent complete. He also said that the cost of these would be nearly half the cost of the NOAA's DART systems. More important, INSAT satellites can be used to transmit data from these to the network, Rao stated.

"The critical thing is not so much the instrumentation itself and India's capability to make these," pointed out Synolakis. "It is in testing and validating the system. The crucial aspect is their frequency response at very low frequencies that are characteristic of tsunamis, which is quite different from pressure transducers in normal situations. To get a good signal-to-noise ratio takes a very long time. It has taken the NOAA 30 years to perfect this. Whatever you do should be a credible and tested system. Similarly, current computer modelling of tsunamis often underestimate wave heights by a factor of ten. Improved forecasting will be possible if India is part of a regional network collaborating with other organisations like the PTWC," he said.

Tad Murty also recommended the creation of a regional tsunami warning network under the U.N. system, rather than an Indian one in isolation. In fact, in the system that he proposed at the meeting, the network's regional centre would be located at Visakhapatnam with Indian institutions providing the technical support to such a system. "India can do it alone but that would not suffice to give adequate warning times and there is the likelihood of too many false alarms," he said. For the DART-type system to be reliable and robust and issue timely warnings, monitoring and tracking over a larger geographical area through a wider network of sensors spread throughout the Indian Ocean region, as part of a regional network, would be essential.

According to foreign experts present at the meeting, given India's capability, countries of the region are ready to give India the responsibility to implement and operate the regional network. But this would require India to share data, particularly from its seismic network, in real time globally. But India has been reluctant to share its seismic data and put it on the global network (see picture) as due to some unspecified security considerations, even as global networks are becoming increasingly sensitive enough to pick up the tiniest of seismic signals anywhere around the globe. "I had thought that Western nations had been self-centred and narrowly focussed by not studying the hazards of the Indian Ocean region sufficiently. I am afraid, India is equally self-centred in its attitude to the tsunami threat that affects the entire region," Sylonakis remarked.

But even from India's own perspective, it is important to become part of the global network, points out Kusala Rajendran of the Centre for Earth Science Studies (CESS), Thiruvananthapuram. "For determining earthquake parameters in large magnitude events, nearby seismometers alone are not good enough. You need data from distant detectors. "For an early warning system you would need this in real time and unless you are part of the global network you will not have access to distant data," she says.

Even while admitting that this policy of not sharing data globally in real time needs a thorough relook, the DST and the DoD are veering to a policy change that will make available seismic data in near real-time if the magnitude is above a certain threshold, say 5 or 6. This means that, in cases of tsunamigenic earthquakes, India will make available its seismic data globally after a lapse of 10 minutes or so. But 10 minutes can be vital for a tsunami, especially given the short lead times available in the Bay of Bengal region.

"DART-system or any other is not the panacea," points out Synolakis. "Statistically speaking, in the Pacific, more people have been saved by awareness and self-evacuation than through warnings issued by the PTWC," he added. Indeed, even as India gears up to establish an early warning system, education of people at risk should assume top priority. Equally importantly, many more than the present handful of scientists working on the subject of tsunami would be required if instrumentation-based warning system has to be effective, particularly when the region does not have a history of tsunamis for modelling studies. Sadly, the number of skilled earth scientists in the country is declining rapidly.