Wednesday, Dec 29, 2004
Tsunami warning system
By N. Gopal Raj
Joining the international tsunami warning system will help, but even so there will be much that has to be done within the country.
A WARNING of an oncoming tsunami is more than a matter of detecting that an earthquake has occurred under or near an ocean. The experience in the Pacific Ocean, where most tsunamis happen, is that a number of complex steps have to be completed rapidly so that people can be evacuated to safer locations. Yet the system must also be robust enough to avoid false alerts, which would unnecessarily disrupt people's lives, cause substantial financial loss and ultimately lead to a loss of confidence in the system.
Unlike in the Pacific where over 790 tsunamis have been recorded since 1900, Sunday's tsunami was just the second to hit India during the same period. The first tsunami to reach the Indian mainland in the last 100 years was in 1941. Before that, a tsunami was recorded in December 1881.
Moreover, not every earthquake under or near the ocean causes a tsunami. About a dozen earthquakes of over magnitude 5 on the Richter scale have occurred in the vicinity of the Andaman and Nicobar islands since 1973, including two greater than magnitude 6. According to officials of the National Institute of Oceanography in Goa, no tsunamis followed. The magnitude 9 earthquake off Sumatra, which caused Sunday's tsunami, set off several earthquakes in the Andaman and Nicobar islands, including one that exceeded magnitude 7. None of these later earthquakes is known to have caused tsunamis that reached the Indian mainland.
In 1965, the Intergovernmental Oceanographic Commission (IOC) of the United Nations Educational, Scientific and Cultural Organisation (UNESCO), established the International Tsunami Information Center (ITIC) in Hawaii. Three years later, the IOC formed an International Coordination Group for the Tsunami Warning System in the Pacific, with 26 countries in and around the Pacific as its members. The system issues tsunami information and warnings to over 100 places scattered across the Pacific.
Three distinct but overlapping physical processes have to be modelled accurately in order to understand whether an earthquake could have set off a tsunami and then to identify the places at risk. At the Pacific Tsunami Warning Center (PTWC) in Hawaii, computer systems continually monitor data from seismic stations in the United States and abroad, and alert watch-standers whenever a significant earthquake has been detected, says Charles McCreery in a recent issue of the Tsunami Newsletter. "If the earthquake is shallow and is located under or very close to the sea, and if its magnitude exceeds a predetermined threshold, a warning is issued based on there being the potential that a destructive tsunami was generated."
Such computation is based on 30 to 50 independent measurements and, as Dr. McCreery points out, the PTWC's performance improved rapidly as high-quality seismic data from more seismic stations in the U.S. and other countries became available in real time. The time the PTWC needed to issue a warning fell from up to 90 minutes six years ago to the present level of 25 minutes or less.
It is reported that there are about 200 seismic observatories in the country under various organisations. The India Meteorological Department has some 58 seismic stations under it, only 17 of which are digital and networked. More Indian seismic stations must be networked so that their data immediately become available for analysis, says Kusala Rajendran of the Centre for Earth Science Studies in Thiruvananthapuram.
At least 10 parameters about the fracture in the earth's crust that caused the earthquake are needed to predict the tsunami's initial height, according to Frank Gonzalez, Tsunami Research Program Leader at the U.S. National Oceanic and Atmospheric Administration (NOAA). As only the orientation of the fracture and the quake's location, magnitude and depth can be obtained from the seismic data, all the other parameters must be estimated, he said in a 1999 article in the Scientific American. Consequently, the tsunami's height can be initially misjudged.
Computer models can then simulate how the tsunami would propagate in the deep ocean, also taking into account how underwater ridges and mounds would affect the wave. Although the tsunami could be travelling faster than a passenger jet, the wave may be only a few metres high. As a single wave can be more than 750 km long, the slope is so gentle that a tsunami can pass by unnoticed in the ocean.
After the first indication that an earthquake may have triggered a tsunami, "it is necessary to wait until a potential tsunami reaches the nearest sea level gauge to confirm or deny its existence and begin to evaluate its character," says Dr. McCreery. There are currently about a hundred such gauges around the Pacific, most of which can transmit their data via satellite back to the warning centres. But as these gauges are typically located in the harbours and protected bays, the characteristics of the tsunami would be greatly modified by the shallow depth. That severely limits the usefulness of the data from the gauges, according to Dr. McCreery.
Consequently, the NOAA developed the 'Deep Ocean Assessment and Reporting of Tsunamis' (DART) gauge. Each DART gauge has a highly sensitive pressure recorder installed on the ocean floor. From a depth of 6 km, the recorder is capable of detecting if the height of the ocean above it changes by just one cm. This data is transmitted acoustically to a surface buoy that then relays it over satellite to the warning centre. Seven DART gauges have already been deployed, and at least four more are being planned. The DART gauges are sited in deep water so that they can accurately record the tsunami waves as they pass unaltered, remarks Dr. McCreery.
As the tsunami approaches the shore and the depth decreases, the waves slow down but become higher. The last stage of evolution where the tsunami comes ashore as a breaking wave, a wall of water or a tide-like flood is perhaps the most difficult to model, according to Dr. Gonzalez. Wave heights can reach tens of metres, although waves two to three metres high are sufficient to cause damage.
The NOAA has developed a suite of computer models, collectively known as the MOST (Method of Splitting Tsunami), which are capable of simulating the generation of a tsunami, its transoceanic propagation and inundation of dry land. But the NOAA also points out: "The current state-of-the art in tsunami modelling still requires considerable quality control, judgment and iterative, exploratory computations before a scenario is assumed to be reliable. This is why the efficient computation of many scenarios for the creation of a database of pre-computed scenarios that have been carefully analysed and interpreted by a knowledgeable and experienced tsunami modeller is an essential first step in the development of a reliable and robust tsunami forecasting and hazard assessment capability."
In Japan too, "virtual tsunamis" have been pre-calculated for thousands of possible sources for various magnitudes of earthquake from 6.5 upwards. A supercomputer sorts these "virtual tsunamis" when an earthquake occurs and makes the extrapolations necessary when it does not correspond precisely to any one of them.
Issuing a reliable warning is just the first step. It is then up to the civilian authorities to use the warning for evacuations. Plans have to be made and rehearsed so that all the agencies act quickly and in concert once an alert is issued. Sunday's tsunami swept across the ocean and reached India in just two hours. Countries who currently receive international tsunami warnings have found that they do not have the emergency response capacity, and the necessary communications infrastructure. Hence even though a warning may be received, their coastal communities are still extremely vulnerable.
Establishing a reliable and robust tsunami warning system for India is therefore a substantial undertaking. Many Indian seismic stations probably require upgradation and also need to be networked. The seismic station at Port Blair in the Andamans is, for instance, said to be of the old analog type. Sea-level gauges are needed and press reports say that the Government is looking at installing DART-type gauges. Simulation software that model the evolution of tsunamis from generation to landfall may need to be modified to suit the Indian situation. Hazard mapping to identify vulnerable areas would also probably be required. The satellite-based cyclone-warning system can be augmented for disseminating tsunami warning. Most important of all, local-level plans have to be drawn up for evacuating people at short notice. Joining the international tsunami warning system will help, but even so there will be much that has to be done within the country.
Copyright © 2004, The Hindu.