Waves from the past

The tsunami disaster triggers renewed interest in the study of historic earthquakes and ocean disturbances in the Indian Ocean region.

THOUGH the Indian subcontinent is in a seismically active region, tsunamis along the coastline of India have been rare, but not unprecedented. The oldest record of a sea-related disaster on the Indian coast that is interpreted as a tsunami is in 326 B.C. According to a 1974 account by Lietzin, and quoted by the scientists T.S. Murty of Ottawa, Canada, and Arun Bapat of Pune, a large magnitude earthquake in the Indus delta/Kutch region set off massive sea waves in the Arabian Sea, which destroyed the legendary Alexander the Great's Macedonian fleet on its journey back to Greece after Indian conquest.

It is believed by some - but not generally accepted oqing to lack of evidence - that tsunamis may have occurred in 1524 near Dabhol in Maharashtra, in April 1762 in the Bay of Bengal as a result of an earthquake at the Araken Coast off Myanmar, in June 1819 after an earthquake in the Rann of Kutch, and in October 1847 following a quake in the Great Nicobar Island.

The more scientifically authentic, earliest record of a tsunami in the region is the one caused by the Car Nicobar earthquake of magnitude 7.9 on the Richter scale on December 31, 1881. There have been subsequent tsunami events that have affected the Indian shores - in 1883 following the tsunami off the Indonesian coast due to the Krakatoa volcano explosions; in June 1941 on the eastern coast of India as a result of a major earthquake of a magnitude of more than 8.0 in the Andaman Sea; and in November 1945, following an earthquake off Baluchistan in Pakistan, which affected the western coast of India.

The Andaman Sea is a particularly seismically active zone. Though there have been several earthquakes in this area, with magnitudes of up to 8.5, tsunamis have been rare. Not all submarine earthquakes generate tsunamis. There has to be significant vertical displacement of the body of water above the quake region for that to happen. So the nature of the fault and the tectonic movement causing the quake are important factors to consider.

According to Murty, who has studied tsunamis in the Bay of Bengal, earthquakes with magnitudes of less than about 7.2, will not generate significant tsunamis unless the earthquake occurs in a shallow and confined, small body of water. However, local earthquakes of magnitude above 7.2 have been very few - only three after the 1881 earthquake.

The earthquakes of August 1936 and May 1955 had magnitudes 7.3 and 7.25 respectively. They did not generate tsunamis. This is because tsunamis are usually associated with `dip-slip' type of faults, where there is thrusting between the plates and vertical displacement of water. In the case of thrusting earthquakes in the ocean, particularly in what are known as subduction zones that cause major tsunamis, the ocean floor predominantly moves in a vertical direction.

In `strike-slip' type earthquakes, on the other hand, the movement is mainly in the horizontal direction. It so happens that earthquakes in the Andaman Sea are mainly associated with strike-slip type of faulting. This is the reason why earthquakes less than a magnitude of 8.0 in the Andaman fault region do not lead to tsunamis in the Bay of Bengal, while similar magnitude earthquakes in the Pacific do. Also, quakes of larger magnitude are more in the seismically more active Pacific Rim, of which the Sumatra region is actually a part.

The December 1881 tsunami in the Bay of Bengal has been studied in detail by Roger Bilham of the United States and Modesto Ortiz of Mexico. Analysing the data from eight tide gauges surrounding the Bay of Bengal at that time, they conclude that the tsunami generated had a maximum wave height - what is known as the tsunami `run-up' - of 0.8m - 1.0m. Their analysis of the amplitude and waveform of the tsunami indicates that the 7.9 magnitude quake was due to a 2.7m slip of a 150 km long rupture in the subduction front on the Indian/Andaman plate boundary off Car Nicobar, which resulted in a 10-60 cm uplift of the island.

The tsunami induced by the massive Krakatoa volcanic explosions in Indonesia in August 1883, which is the biggest in known history before the present one, and which killed nearly 36,000 people, also had an impact on the east coast of India. A recent numerical simulation of the tsunami by a research group comprising two scientists from Korea, one from Japan, and one from Russia suggests that the waves should have affected the Andaman and Nicobar islands, the eastern coasts of Sri Lanka and India and parts of the western Indian coast as well, somewhat similar to what has happened in the current tsunami.

The results of their simulation appears to match fairly well with the original tide gauge data of the wave heights at various points, which they have been able to retrieve from archival sources. The analysis shows that wave heights reached were of the order of 0.5 m and the tsunami waves arrived about 4 hours after the event. The source of the waves was about 2,440 km from the Indian coast. Unofficial accounts have placed the wave height at Chennai to be about 1.5 m.

The next major earthquake that resulted in a tsunami in the Bay of Bengal was of magnitude 8.1 earthquake in the Andaman Sea (12.9N, 92.5E) in June 1941 and the tsunami did hit the east coast of India, damaging masonry structures and property in places like Chennai. The quake ruptured the region near the Andaman island. The number of dead in this event is not known. While some believe that there may not have been any dead as the magnitude has probably been overestimated, C.P. Rajendran, a scientist at the Centre for Earth Science Studies (CESS), Thiruvananthapuram, an expert in the seismology of the Andamans, believes that the value may actually be an underestimate. According to him, the rupture length of over 800 km suggests a magnitude higher than what has been calculated on the basis of seismic waves.

The Andaman and Nicobar islands were under Japanese occupation at that time, so data from there was not available. However, according to Rajendran, for some reason, reliable tide gauge data of the east coast for events of that period are also apparently not available with the Survey of India. In fact, no records of the 1941 tsunami seem to have survived, although unofficial accounts put the toll as 3,000 dead. On the basis of non-scientific and journalistic sources, Murty and Bapat suggest that the height of the tsunami wave was of the order of 0.75-1.25 m.

On the western side of India, though anecdotal evidence suggests that tsunamis may have occurred following earthquakes in the Kutch region in 1524 and 1819, the first authentic evidence of a tsunami is of one caused by an earthquake of magnitude 8.25 off the Makran coast in Baluchistan in Pakistan (24.5N and 63.0E) in November 1945. This is supposed to have caused tsunami waves as high as 11.0 to 11.5 m in the Kutch region on the west coast of India and the waves are supposed to have reached Mumbai as well. A total of 4,000 people are supposed have died as a result of the earthquake and the tsunami. This information is again based on non-scientific sources and anecdotal accounts, as tide gauge data during the 1940s seem to be particularly lacking, perhaps owing to the Second World War. In the context of the present earthquake of 9.0 magnitude off the northwestern Sumatra coast and the spread of the tsunami caused by it throughout the Indian Ocean, it is of interest to know the consequences of the large earthquakes of 1833 and 1861 off the Sumatra coast. In size and location, the 1833 quake was quite similar to the present one. Its magnitude is estimated to be 8.7-8.8, even though the rupture area was much smaller (about 300 km) as compared to the present one of over 1,200 km. In fact, a recent estimate of the uplift of the Island, based on growth-ring record of coral-atolls, seems to suggest that the magnitude may have been as high as 9.2. The location, however, was in the central region of the western Sumatra coast. The 1861 earthquake was of a lower magnitude and is estimated to be 8.3-8.5 magnitude but again originating from the central region of the coast.

Tsunamis from both these Sumatran earthquakes occurred before the introduction of harbour tide gauges in most parts of the world. However, their sizes seem to suggest that the tsunamis would probably have affected the entire Indian Ocean basin. The 1861 tsunami was perhaps not strong enough to cause damage on shores across the ocean. Geoscience Australia has, however, been carrying out tsunami modelling studies on the 1833 earthquake from the point of view of estimating the threat of tsunamis from earthquakes in this region. North-west off Indonesia, the Australian Plate subducts beneath the Sunda Plate (which forms part of the larger Eurasian Plate in the Pacific). This subduction zone is called the Sunda Arc or the Sunda Trench.

The Sunda Arc comprises two distinct zones. The eastern part is where relatively old (over 100 million years) oceanic lithosphere subducts offshore Java. Very few classical subduction zone earthquakes - such as the present one - occur in this region. On the other hand, further to the north-west on the Sunda Arc, young (40 million years) oceanic lithosphere subducts offshore Sumatra. The subduction of such young oceanic lithosphere can cause the most massive thrust earthquakes that generate huge tsunamis. However, this subduction zone is not as active as the ones in the Pacific Rim, which cause frequent tsunamis. Also, the thrust earthquakes that do occur as a result of this subduction and the propagation characteristics of the tsunamis that they do generate have been such as to cause local effects but not affect coasts thousands of kilometres away in the northern part of the Indian Ocean.

The modelling of the 1833 quake by Geoscience Australia has thrown up interesting results. The modelling is for open-ocean or deep water propagation only and does not estimate the effects of the shoreline topography and the consequent run-up. Though the bulk of the propagation is away from the Indian coast and into the central Indian Ocean basin towards the African coast, the simulation shows that the tsunami should have been strong enough to reach the Indian eastern and southwestern shores .

Clive Collins, Project Director at Geoscience Australia, has been seeking information on the Internet about possible tsunami related events of November 1833 from any possible source. So far, however, the organisation has not received any pertinent data from any source that could confirm the results of their model. The 2004 tsunami data from the affected Asian countries would, however, be of immense use in validating the model, Collins says. Indeed, a mega thrusting of this kind, resulting in a 9.0 magnitude quake and a vertical displacement of the fault plane by as much as 15 m, may not recur for hundreds of years.

The probability of a killer tsunami on the Indian coast may continue to be negligible but the current mega disaster has thrown new light on threat perceptions in the region from distant tsunamis in the Indian Ocean. This should trigger a renewed interest in the study of historic earthquakes and ocean disturbances in the region, so as to generate an appropriate database. There should also be study as to whether, buried in the past records, there was ever a hint of an impending disaster of such magnitude.