The disastrous earthquake-triggered tsunami that occurred 11 years ago on the morning after Christmas was huge by any contemporary standards. The epicenter of magnitude 9.3 quake was located in the Indian Ocean near the west coast of Sumatra and it ruptured the 1000 km long Andaman plate boundary, moving the seafloor 20-10 m vertically upwards, thus displacing trillions of tons of under-sea rock. The killer waves radiating from the epicenter slammed into the coastlines of 11 countries from east Africa to Thailand, resulting in 227,898 fatalities.
A lag of several minutes to hours between the earthquake and the impact of the tsunami notwithstanding, both the near and distant coastal communities were taken by surprise. There were no tsunami warning systems in place nor were there any ‘social’ memories of previous tsunamis preserved for most communities to fall back on.
Considered at that time as an unprecedented disaster in its magnitude and transoceanic reach, the 2004 event had also surprised researchers and hazard managers. The research community as a whole had failed to anticipate such events lurking along the eastern seaboard of India. More than a decade after the 2004 event, it is time for us to take stock of what we have achieved in terms of understanding such huge earthquakes/tsunamis in the Indian Ocean.
Advances in Technology
One of the most significant developments in the region has been the establishment of the Indian Tsunami Early Warning Centre (ITEWC)—an in-house unit of the Indian National Centre for Ocean Information Services (INCOIS), operational since 2007, belonging to the Earth Systems Science Organisation, an autonomous body under the Ministry of the Earth Sciences (MoES), Government of India. An important mandate of this Centre is round-the-clock monitoring and warning services for coastal communities on tsunamis, storm surges and high waves. The Centre receives data in real-time from seismic, sea-level and tide-gauge stations along the Indian coast, analyses this data and disseminates it to designated authorities in the event of an undersea earthquake.
The region has thus witnessed technological advances such as deployment of offshore and deep ocean tsunami observation systems and acquired the wherewithal to issue early warnings for future tsunamis. INCOIS, with inputs from India Meteorological Department (IMD), now gathers earthquake data from a real-time seismic network in the region, besides receiving data from about 350 international stations. It also receives data from ocean bottom pressure recorders and tidal gauge networks through satellite connectivity. Currently the warning system is capable of estimating potential tsunami-triggering earthquake parameters in real-time (less than ten minutes after the event).
Guided by a comprehensive Standard Operating Procedure (SOP), the ITEWC—along with Indonesia and Australia, can provide advisories to countries bordering the Indian Ocean.
In the last decade much research has also been done to understand the geo-physics of the 2004 earthquake. Geological, seismological and Global Positioning System (GPS)—based geodetic studies during the last ten years have focused on earthquake source properties, crustal deformation, tsunami potential including modelling of tsunami sources, and long-term forecast of such events.
Today, we know that mega earthquakes-cum-tsunamis had indeed impacted the Indian Coast in the past and have had a rough cyclicity of about 500-1000 years. We know, for instance, that a mega-tsunami similar to the 2004 event had occurred some 1,000 years ago and wiped out the then flourishing port-city of Kaveripoompatinam (Poompuhar, as it was then known) on the south-eastern coast of India, which was then under Chola rule.
But the question remains, as to whether mega earthquakes (scale and magnitude of 2004) occur at predictable intervals and if they are ‘characteristic’ in a classical sense. Yet, the Fukushima nuclear disaster in Japan that was triggered by a 9.0 magnitude earthquake and subsequent tsunami – that occurred in one of the best-prepared countries should continue to remind us against developing an unrealistic sense of hubris as regards such inherently chaotic phenomena.
Anatomy of Earthquakes and Tsunamis
Nature has always been a hidden storehouse of surprises. The March 11, 2011 earthquake was a complete surprise as it challenged the ‘conventional wisdom’ regarding seismic science. The accepted theory that faults would break up in predictable intervals in similar-size earthquakes (characteristic earthquake) proved to be too good to be real for mega-earthquakes. The 2011 Sendai earthquake may have been a ‘composite’ earthquake, but consisted of what seismologists term ‘sub-events’.
The earthquake which was believed to have started as an ordinary event became a runaway process forcing the upper part of the fault zone to break with an extraordinary vertical displacement of the sea-floor, causing a tsunami.
Like the 2011 Japanese earthquake, the 2004 Indian Ocean earthquake, sourced off Sumatra in the south may have also shown some similar runaway processes in its propagation to the north. What happened on the northern part (Andaman side) of the rupture is somewhat shrouded in mystery. It is likely that the speed of the rupture as it approached the northern segment slowed, and the vertical displacement was also smaller on the northern side.
Apart from the strain and energy build-up-dissipation cycles, the constitutive frictional properties between the fault blocks appear to be playing a decisive role in determining whether the nucleated rupture would grow into a mega-earthquake. However, the fact remains that the seismic cycle whether simple or complex must satisfy an energy budget.
Christopher Scholz, one of the most eminent geophysicists of our period, had said that even if the energy release is a periodic, it must be represented by a mean recurrence time because the variance from the mean is a well-defined function of the mean and thus is a measure of the stressing rate despite the spatially variable strength of the fault. From a public perspective, a question that is being asked is about forecasting such massive earthquakes. Have we made any progress in predictive capabilities of such hazards?
Unlike the surface process, earthquakes nucleate at depths of tens of kilometres below the surface. We still lack a clear understanding of the properties of rocks including the pressure, temperature and fluids and the attendant complexities at those depths. It is also not possible to observe the earthquake nucleation processes at those depths, unlike processes that are amenable to direct monitoring.
However, the combined use of satellite based measurements and numerical modelling of crustal deformation and fault-specific geological studies have made long-term forecasts possible in some regions. A game changer for short term prediction would be to isolate any precursory signals that can be captured years, months, days or hours before the occurrence of a big earthquake. Ultimately, the success of earthquake prediction depends in large part on resolving this question.
Recent Insights into the 2004 earthquake and tsunami
Recent insights into the 2004 earthquake allow us to cautiously suggest that this earthquake could have been anticipated probably months before it had actually occurred (Paul & Rajendran, 2015). These insights have come from the measurements made at Port Blair on the movement of the earth’s surface using space based GPS systems (Fig. 1). The GPS measurements have now developed into an accurate way to record subtle fault related movements during pre-earthquake slow crustal movements as well as rapid motion that occur during earthquakes. Although some GPS measurements were made prior to the 2004 earthquake, these data were properly evaluated only after a decade since the earthquake.
In an analysis published in 2015, the researchers show that the GPS monitoring site in Port Blair started to slip down from its original height between 2003 and 2004 suggesting a pre-earthquake slow downward movement (Paul & Rajendran, 2015). The tide gauge data from Port Blair (analysed much after the earthquake) also showed a consistent change in sea level indicating a pre-earthquake slow downward movement (Catherine et al., 2014), adding further validity to the GPS derived pre-earthquake ground level changes. Visual observations made 15 months prior to 2004 had also suggested that the near shore coral colonies were emerging near Port Blair and elsewhere. It is thus, now concluded that Port Blair, which initially emerged, started subsiding one or two years prior to the earthquake. Thus, there was a slow emergence coupled with a slow subsidence (sinking) just 1-2 years before the earthquake, followed up by the sudden slip during the earthquake. The pre-earthquake slow downward movement is estimated to be equivalent to an earthquake of magnitude 6.3 or more. The magnitude of the slip could be higher towards the plate boundary, located 200 km west of Port Blair. Unfortunately, these revelations have come ten years after the earthquake.
So what could have happened during the earthquake rupture? A likely scenario is that the fault along Sumatra and North Andaman, that had reached its peak in accumulating stress and was ready to rupture, may have initially failed in the Andaman part much before the actual earthquake (December 26 event) in the form of a slow slip sometime in 2003. This partly addresses my earlier rider on a possible slow speed of the northward rupture. Unlike the sudden slips during the earthquakes, the slow slips are not felt nor easily detected by conventional earthquake monitoring equipment. When the sudden release of energy took place on December 26, 2004 off Sumatra—the epicenter of the quake, the fault was propagated to the Andaman side in the north. This part of the plate boundary, which had already released accumulated stress in a pre-earthquake slow slip, caught up with the earthquake fault, but with a low release. This is borne out by the low amplitude tsunami waves on the Andaman coast, compared to the southern Nicobar Islands and Sumatra on December 24, 2004.
Thus, in all probability, it can be surmised that the satellite based geodetic data prior to the December 2004 earthquake revealed a deformation signal, the significance of which was not understood at that time.
It is thus pertinent to monitor pre-earthquake ground level change with greater instrumental coverage and appropriate theoretical tools. From that perspective, one can conclude that the 2004 December earthquake was a missed opportunity to gain further insights into the pre-earthquake process and earthquake science.
In future, we have to do well to focus on other trouble spots that have the potential to produce huge tsunamis such as the Makran Coast of the north-western Arabian Sea, and the Myanmar Coast in the northern Indian Ocean.
The Makran Coast of the north-western Arabian Sea, spread along Iran and Pakistan, holds the potential for a major tsunami that could direct its massive energy onto the western coast of India. This is an area where India, Iran, Pakistan and the Gulf countries can come together and collaborate so that a potential cross-border transoceanic phenomenon could be understood in greater detail to plan better strategies. We must also remember that these coasts host a number of critical facilities like nuclear reactors, which are vulnerable to tsunami hazards.
Taking a leaf out of the Fukushima experience in Japan, we need to be relieved that the 2004 tsunami could only partially impact the Kalpakkam reactor on the eastern coast, precluding a full-blown catastrophe. As the dictum goes, the past is the key to the future. As Philosopher George Santayana had said: ‘those who cannot remember the past are condemned to repeat it’. Therefore, the lessons learnt from 2004 should not be forgotten because it is the only modern benchmark event in our memory that should serve in guiding us, henceforth, in the event of similar future disasters.
Catherine, J.K., Gahalaut, V.K., Srinivas, N., Kumar, S., & Nagarajan, B. (2014). Evidence of strain accumulation in the Andaman region for the giant 2004 Sumatra-Andaman earthquake. Bull. Seismol. Soc. Am. 104(1), 587–591. Retrieved from http://dx.doi.org/10.1785/0120130141.
Paul, J., & Rajendran, C.P. (2015). Short-term pre-2004 seismic subsidence near South Andaman: Is this a precursor slow slip prior to a mega thrust earthquake? Physics of the Earth and Planet. Retrieved from doi: 10.1016/j.pepi.2015.08.006.