Prof. J R Kayal

The Indian plate, separated from the Antarctic, started moving to the north northeast about 180 million years ago. The present day movement of the Indian plate from the Carlsberg spreading ridge results in collision in the Himalaya and subduction in the Andaman-Sumatra. These plate margins, therefore are the major seismic belts of the moving Indian plate.

The concept of plate tectonics is the most satisfying explanation for a majority of earthquakes. The basic idea of plate tectonics involves earth's outermost part, the lithosphere (100-200 km thick), which consists of several large and fairly stable slabs - the plates. Boundaries of these plates are the seismic belts of the world. At the mid oceanic ridges, up welling of lava is a continual process. This molten rock creates new sea floor on either side of the ridge and these mid-oceanic ridges thus constitute the spreading zones of the earth or divergent plate boundaries.
Since the earth's size remains the same over a long period of geological time, the moving plates must be absorbed at some places. The burial grounds of plates - the convergent plate boundaries, are believed to be the ocean trenches, where the plates plunge into the earth's interior. This process is known as subduction - as happens along the Andaman-Sumatra trench, the Japan trench, the Chile trench and so on (Fig 2a). The other type of convergent plate boundary forms the continent-continent collision zone - as happens in the Himalaya, where the Indian plate is on a head-on collision with the Eurasian plate (Fig 2b). A third type is the transcurrent boundary, where the plates move past one another - as happens along the San Andreas (California) fault between the Pacific and the North American plate.

Indian plate movement
The Indian plate, separated from the Antarctic, started moving towards the north northeast about 180 million years ago. About 55 million years ago it made contact with the Eurasian plate, and the head on collision started (Fig 1). The present topography map shows the effects of this head on collision with lofty, still rising Himalaya and the abyssal Andaman-Sumatra trench in the Indian oceanic plate. The present day movement of the Indian plate from the Carlsberg spreading ridge results in collision in the Himalaya and subduction in the Andaman-Sumatra. Understandably these plate margins are the major seismic belts of the moving Indian plate.

Indian plate earthquakes
Seismic network
After the devastating 1897 great Shillong earthquake, the first seismological observatory in India was established in Alipore (Kolkata) in 1898 by the India Meteorological Department (IMD). Substantially precise epicentral earthquake data became available from 1964 onwards with the inception of the World Wide Seismograph Station Network (WWSSN) and more seismograph stations (about 15 by 1960) in the national network. The WWSSN was upgraded to the Global Standard Network (GSN) with digital instruments in the 1980s. These data are available on the United States Geological Survey (USGS) website almost in real time. Post 1993 Latur earthquake, the national network was further upgraded with a denser and digital seismic network. Now about 100 permanent stations and several telemetric networks are run by different organisations, institutes and universities in the country.


General seismicity
The general seismicity map of India shows intense seismic activity all along the Himalayan collision zone, Indo-Burma ranges and along the Andaman-Sumatra subduction zone. It is argued that the Andaman-Sumatra subduction zone is extends beneath the Indo-Burma ranges. The meeting zone of the Himalayan and the Indo-Burma arcs is named Assam syntaxis. The earthquakes in the Himalayan collision zone and in the syntaxis zone are shallower (< 80 km), whereas the earthquakes in the Indo-Burma-Andaman-Sumatra subduction zone are deeper, down to 300 km within the subducted Indian plate (Kayal, 2008). The earthquakes in the middle of the plate, away from the plate margins, are called intra plate earthquakes; these are infrequent and much shallower (< 50 km).


Large and great earthquakes
Locations of the large (M~7.0) and great earthquakes (M~8.0) in the continental part of the Indian plate are shown in Fig 3. These earthquakes follow the Himalayan mountain belt and the Indo-Burma ranges; except one great and one large intra plate earthquakes in the Kutch area of Gujarat. Among the few intra-plate damaging strong earthquakes (M~6.0), 1923 Satpura, 1967 Koyna, 1993 Latur and 1997 Jabalpur earthquakes are worth mentioning. Two more large or strong intra plate earthquakes, 1720 Delhi and 1919 Gujarat, are reported in the historical catalogue, but their magnitudes are not well ascertained.


Seismic hazards and risk mitigation
Seismic hazards still fresh in their crescendo are the 1993 Latur (M 6.3) and the 2001 Bhuj earthquake (M 7.7), with an enormous loss of lives of over 10,000-20,000 persons. The loss of so many lives in the Latur earthquake was also attributable to poorly built houses made of boulders and mud. On the contrary there were no casualties among those who lived in the bamboo-thatch and in the well built concrete houses. The lesson to be learnt is that technique and material used play a significant role in withstanding the impact of an earthquake.

The great earthquakes (M~8.0) of the Himalayan region, 1897 Shillong, 1905 Kangra, 1934 Bihar/Nepal and 1950 Assam syntaxis and plateau resulted in the loss of about 30,000 lives - but if such an event were to occur today, it would lead to much higher casualties. Unprecedented growth of population in the Himalaya coupled with earthquake non-resistant housing are the chief drivers of this situation. For example, a large number of earthquake non resistant multi-storied brick houses are being built in and around Shillong, which has already experienced a devastating earthquake (M~8.7) in 1897. In fact, it is not the earthquake, but the poorly built houses and ignorance that kills people.
Crustal deformation studies through improved instrumentation show that the Himalayan segment is ready for a large/great earthquake at any time. It may be mentioned here that about 30 years ago loss of lives in the developed and developing countries was almost of the same order. But today the loss of human life due to a large earthquake in a developed country like Japan has been minimised drastically, whereas it has been enhanced over 100 times in countries like ours.
Seismic hazard or risk mitigation is a challenging task in our disaster mitigation programme. Since successful prediction of an earthquake with specific time, space and magnitude is yet to be achieved or understood, the first and foremost task to mitigate disaster should be to follow the building code based on the seismic zoning map of India (Fig 4) and using available maps on microzonation in the urban cities (Fig 5). Such maps identify the most vulnerable pockets of seismic hazards/damages, susceptible to ground amplification or liquefaction. Therefore older buildings in such pockets need retrofitting and newer ones need special construction designs (Fig. 5). Also authorities should decommission permits for making new habitations in danger proneas.

Large and great earthquakes in the ocean
A large part of the oceanic plate of India, subducting beneath the Andaman-Sumatra trench has produced several large and great earthquakes in the past, some of which generated destructive tsunamis. Largest among them are the historical earthquakes that occurred in 1833 (M~8.7); 1861 (M~8.5); 1881 (M 7.9) and 1941 (M 7.7) (Fig 6). While these large earthquakes ruptured only a few hundreds of kilometres (~200-300) of the plate boundary, the 2004 Sumatra mega thrust earthquake (M 9.3) ruptured more than 1300 km of the arc, stripping the regions that were ruptured in the past as well as the intervening unbroken patches, and generated the devastating tsunami that snuffed out the lives of nearly 2 lakh people living along the southern coasts of India and southeast Asia. The energy release of M 8.0 is equivalent to about 100 million atom bombs, while the energy release of an earthquake of M 9.0 is equivalent to the occurrence of 30 great earthquakes of M 8.0 at one time.
The other tsunamigenic zone is in the Arabian Sea where there is a record of a great earthquake (M~8.0) south of the Makran coast in 1945, at the Makran subduction zone. This event generated a destructive tsunami killing about 4,000 people along the coast of Pakistan, Iran, Oman and north western coast of India.

Tsunami risk and hazard mitigation
A tsunami warning system monitors the occurrence of any tsunamigenic earthquake in the sea, and can predict the arrival of the tsunami to the coast. The time interval between the occurrence of earthquake and the arrival of tsunami depends on the distance from the source to the coast, which may vary from couple of minutes at the Andaman-Sumatra islands to a few hours at the east coast of India. A tsunami warning system is now established by the Indian National Centre for Ocean Information Services (INCOIS), Hyderabad, that records the real time telemetric-observations of the tsunamigenic earthquakes in the sea. The tsunami warning system of the INCOIS is working well. The other steps for the tsunami hazard mitigation could be to avoid habitation within 500 m of the coastline, and also by mangrove plantation at the coast to break the sea waves.

In Conclusion
The seismicity and seismic source zones in and around the Indian plate are well understood with available seismological data. However the data source is too limited for accurate space, time and magnitude prediction of earthquakes. Although high precision instrumental data being recorded since the last few decades will enable future understanding of the recurrence period of a large or great earthquake for 100 to 1000 years depending on the source zone and tectonic stress accumulation - its present window period is too narrow for prediction.
Coastal zones of India, a long stretch of the east coast and a small stretch of the west coast, are prone to tsunami hazards. These hazards can be mitigated efficiently with the tsunami warning system. Also general awareness and preparedness is vital for natural hazards like earthquakes and/or tsunamis. A case in point is a young girl from UK holidaying in Phuket (Thailand) who interpreted the abnormal ebb in the sea water accurately on 26 December 2004 and raised the alarm to save herself and hundreds of others. On the contrary people along the Indian coast waited to watch the sharply receding waters and lost their lives.
Natural hazards, particularly earthquakes, can neither be stopped nor be precisely predicted. In case we do predict that a large earthquake would occur in a heavily populated city such as Delhi or Kolkata within the next 15 days or say within a month - would that entail total evacuation for a month? Further, even if the loss of lives were minimised would damage to habitations be taken care of? We, therefore, should learn to live with earthquakes and combat them with disaster proof structures coupled with preparedness.

.

The author is CSIR Emeritus Scientist, Jadavpur University, Kolkata, jr_kayal@yahoo.com