On October 25, 2010 a magnitude 7.7 earthquake struck south of S. Pagai island off Sumatra, triggering a tsunami that killed over 400 people. Initial reports related that there was no warning communicated to the Mentawais and that the Indonesian tsunami warning system had broken down. Be that as it may have been, the death toll is perplexing in view of the 2007 earthquake and tsunami in the same region that killed nobody because people self-evacuated. If they did it then, why didn’t they do it now ? This was a question that an NSF expedition as part of the International Tsunami Survey Team strove to investigate in the field survey from November 8 to November 19.
The initial assessment of the TRC can be found at the Wall Street editorial – Waving the tsunami warning flag. After the field survey, the findings largely confirmed the early inferences. The reason that this event was so much more deadly than the 2007 event was that the earthquake was weakly felt, if at all, in most of the Pagais. Locals had been told to self evacuate only if the shaking was stronger than the 2007 event, and hence some didn’t even bother to get out of their beds. The earthquake struck at night, few if any saw the first wave of the tsunami – eyewitnesses who fled reported hearing the sound of the advancing tsunami.
There was no warning disseminated in the Mentawais. Residents in Padang reported that they saw an announcement on television, but didn’t evacuate because it only warned of the potential of a tsunami. Most residents in Sikasap reported not seeing the announcement, and a few were alerted by relatives who telephoned them from elsewhere in Sumatra. In most coastal villages in the Pagais or Sipura there is no electricity or telephone communications, so residents couldn’t had received the televised alert.
The warning system failed, even if it is eventually determined that the TV alert was broadcast before the tsunami arrival in the Mentawais. Currently, the Indonesian tsunami warning system only detects earthquakes, and there should had been provisions for alerting even the most remote coastal villages of the Mentawais, a particularly high risk area for tsunamis. While public education for self evacuation is priority one, there are unusual events which do no allow for direct observation of the common harbinger clues of a tsunami, such as strong ground shaking that lasts over 30 seconds or unusual shoreline motions. A “slow” earthquake (a tremor that produces less felt shaking in comparison with its size) which strikes at night is a nightmare scenario, which unfortunately materialized on 25 October. Even fairly well tsunami-educated residents in self-evacuation failed to take action, exactly because they harbinger signs of the impeding catastrophe were not felt or observed in a manner that would trigger a flee or die response.
The death toll was unnecessarily high. Solar powered sirens connected via VHF or wireless technology to Sikasap are necessary to save lives. Public education efforts must continue to reinforce the knowledge that any ground shaking that lasts over 30sec should trigger self-evacuation. Indonesia needs to install a whole array of tsunami detectors like NOAA’s DARTs to allow for rapid verification of tsunami generation to improve the accuracy of the warnings, but also to allow for informed all clear messaging, an important part of any warning system. As it happened, some local residents stayed away for days not knowing if it was safe to return, others returned too quickly only to flee when they heard the noise from subsequent waves advancing to their villages.
More information can be found in the two post survey articles in the Jakarta Globe.