The Indian Ocean tsunami, which followed the submarine earthquake off the north-west coast of Sumatra on 26 December 2004, attracted instant worldwide attention. The horrors of the event, the tragic loss of lives, the extent of the devastation and the magnitude of the job of reconstruction were projected by television into homes around the world, and raised the questions of how to anticipate and deal with such events in the future.
It was to try to answer these questions that the ICE Maritime Board, together with the Society for Earthquake and Civil Engineering Dynamics (SECED), organised a half-day meeting in London on 7 February 2005. The meeting was chaired by Professor John Burland, of Imperial College, and addressed by a number of specialists in oceanography, earthquake engineering and disaster relief.
Professor Burland explained that the purpose of the meeting was to bring together the experience and expertise of engineers and other specialists to share their knowledge and to educate the wider community. The information collected from this meeting would help to inform governments and other public bodies, would demonstrate the capacity of the civil engineering profession, in liaison with other specialists, to deal with these disasters and would raise the public profile of civil engineering.
The meeting was opened with a startling eyewitness account by Steve Magenis, of Posford Haskoning, who was on holiday in Phuket in Thailand and was actually sitting on the beach when the waves arrived. He ran back to safety in time. Interestingly two hours earlier he had felt the local earth tremor from the original earthquake some 400 miles away.
Initial reports from the Earthquake Engineering Field Integration Team (EEFIT) were presented by Dr Sean Wilkinson, of the University of Newcastle Upon Tyne, and Dr Navin Peiris, of Arup, who had visited Thailand and the south-west coast of Sri Lanka in January 2005. Their observation was that the scouring of foundations was a serious problem and had caused the collapse of many buildings, railways and bridges. The collapse of bridges also caused the failure of service pipes attached to them, such as water, telephone and electricity. Damage extended from 400 m to 1·5 km inland, caused by waves from 2 m to 5 m in height.
Dr Clark Fenton, of Imperial College, said the submarine earthquake off the coast of Sumatra was magnitude 9 (on the Richter scale) and the fifth largest within the last 100 years. More severe earthquakes around the Pacific rim have occurred in Chile in 1960, magnitude 9·5, and in Prince William Sound in 1962, magnitude 9·2. Tsunami waves have also been created by submarine earthslides. Dr Fenton emphasised the need to understand the underlying science of submarine earthquakes. He noted that ‘higher probability’ regions could be identified along subduction zones, and identified example data that could be used to estimate whether a major movement was more or less likely. Dr Fenton also discussed the difficulties in making predictions and, hence, the limitations of predictions.
Professor William Allsop, of HR Wallingford, explained that all key parts of the tsunami generation, propagation and inundation could be modelled. He noted previous major tsunamis had followed the earthquakes in Lisbon in 1755, Krakatoa in 1883 and Chile in 1960. On 27 December 2004, the day after the Sumatra earthquake, Japanese engineers had used ground movement data to model the generation of the Indian Ocean tsunami. This model showed that major wave fronts propagated to the east, towards Thailand, and west to Sri Lanka and India. Waves to the north, towards Bangladesh, and to the south were substantially weaker. The westward wave had a speed of propagation of about 200 m/s and wavelength of about 400 m. Professor Allsop explained that the wavelength was more important than offshore wave height in predicting the potential damage which the tsunami would cause when the waves shoal up at the shore.
Tsunami generation models require predictions of vertical ground displacements. Propagation models can use shallow water equations to give practical guidance on general tsunami propagation to identify safer or more exposed areas. At Phuket Bay the wave heights (troughs and crests) had been measured by a Belgian yacht which happened to be anchored 1 km offshore in 12 m of water. Professor Allsop noted that a tsunami wave is not always preceded by a trough, although if such a feature is seen, ‘run for high ground.’
Dr David Ingram, of Manchester Metropolitan University, discussed more complex aspects of the wave modelling and identified areas for future research in tsunami and inundation modelling. Different stages in wave propagation need to be considered and modelled separately. Oceanic propagation is well understood but predicting the run-up wave is more difficult. The surface slope, roughness and degree of saturation of the surface of the coastline or land need to be modelled accurately.
Dr Ingram considered the Boussinesq equations were more accurate initially and should be used for the initial stages of wave propagation, and the shallow water equations could be used for the later stages, but cautioned that Boussinesq models could be very computationally expensive and slow.
The contribution from the British Geological Society was given by Dr Lars Ottemoller, who explained how a tsunami early warning system could be established to respond to the detection and location of an earthquake. He emphasised that an early step in the development of such a system must be to conduct an initial hazard assessment. In specifying a warning system for the Indian Ocean it will be essential to develop regional cooperation to improve the infrastructure and to tailor the system to meet the needs of the participants. The worldwide net of seismic recording stations could detect the earthquake and locate its centre. This might be done within 10–20 minutes of the earthquake. It would then be a separate action to predict the magnitude, speed and direction of any tsunami wave. This would ensure the local warning could be issued in time.
Jeremy Larken, Managing Director of OCTO, explained the practice of crisis and emergency management. A disaster is different from an emergency and requires performance under pressure. The first criterion is to save life and limb, even recognising that some lives may be lost. An important factor is to get information quickly. A disaster situation worsens rapidly, during which time the options for crisis management reduce. Lack of good quality information will also limit time available to make a decision and reduce the likely success of the decision made. Leadership is also important and needs to replace consensus management in a disaster situation. Communication with the public is essential and need not be high-tech.
The many contributions from the floor included an offer by Professor Robin Spence, of the Department of Architecture at Cambridge University, for eyewitnesses of the tsunami to send him their reports. He said these accounts would be included in a research project being undertaken by his department which will be published in due course.
