Is Los Angeles at higher risk for tsunamis? (10/10/2010)
A study just published in Nature Geophysicas by Hornbach et al (2010) reports that the 12 January 2010 strike-slip earthquake that killed over 220,000 people in Haiti also triggered a 3m tsunami. The associated press release (at the end of this item) claims that the study implies that Los Angeles and other metropolitan areas around the world are at higher risk of tsunamis.
The fact that moderate strike-slip earthquakes can trigger submarine landslides is not new. In 1998, an approx 7 magnitude earthquake on the north shores of Papua New Guinea generated a very large tsunami with overland flow depths over 12m, which killed about 2400 people. The cause of the tsunami remained controversial for years, but finally a combination of marine geology, hydroacoustics, field measurements and hydrodynamic modeling proved that the earthquake had triggered a submarine landslide. The key element was a record from a hydrophone (underwater microphone) in Wake Island that had registered the earthquake and the slide generating event, and the timing of the latter fit eyewitness accounts and field observations. I was in PNG ten days after the tsunami, and the tsunami devastation dwarfed even the earthquake damage in Haiti. No matter how obvious the landslide trigger was then, it took years to prove it. I attach the relevant publication in the Proceedings of the Royal Society.
Concerns over the possibility of strike-slip or other moderate earthquakes of magnitude less than 7.0 near the coastline or onshore triggering large tsunamis nearshore was the motivation for the initiation of California's tsunami inundation mapping program in 1999, which eventually got us the tsunami signs identifying risk areas in most state beach areas. The mapping effort started in the aftermath of the PNG event. Even then, the fact that the California maps relied on scenarios of moderate onshore earthquakes triggering offshore landslides triggering tsunamis was controversial. Post the great Sumatran tsunami of 2004, regulator's attention focused on the risk from tsunamis striking California from Alaska or Chile, and a new set of maps was produced that incorporated the risk from tsunami sources far away and near the coast.
Further, the suggestion in the Nature Geosciences paper that the Haiti earthquake triggered a tsunami is not new. Professor Hermann Fritz of Georgia Tech was one of the first scientists in Haiti and reported about the tsunami and the possible landslide trigger in the Ocean Sciences meeting of the American Geophysical Union on 18 February 2010.
The story was covered by Nature News and the BBC, but somehow and despite its official publication in the AGU meeting seems to have been missed by the authors of this study. I am not even sure its senior author ever made it to Haiti.
So what remains is whether, as claimed, landslide tsunamis in the Caribbean are as common as suggested in this paper. I couldn't find sufficient evidence presented to back this up. For one, recall that the Haiti tsunamis - landslide triggered or not- killed less than 10 people out of over 220,000, and the calculations of the authors show that the tsunamis were less than 1m high near the shore. (10 people is 10 too many, but the earthquake toll puts the tsunami toll in perspective, and there is confirmation only for 3 tsunami deaths). Second, finding evidence of landslides does not necessarily imply that they triggered tsunamis. Some of the evidence found may have been older slides, we just don't know. Even if coseismeic with the 2010 event, small or slow landslides are notoriously inefficient in generating even small tsunamis. There is simply not enough documentation presented to back the claim that tsunamis in Haiti are 10 times more likely to be generated by landslides. While I am glad that this paper with provocative inferences made it to press - controversy is the lifeblood of science- the heading in the associated press release that Los Angeles is at higher risk because of what happened in Haiti and the inferences, and in my view diminishes the overall credibility of the work, not to mention the tsunami of press interest it triggered (my phone rang nonstop once a Los Angeles Times story was published).
If indeed nearshore landslide tsunamis are far more frequent than earlier imagined, this is really bad news. However, this paper does not enhance our understanding on this, nor does it provide us with evidence that this is so. Even more urgently than additional research or fanciful press releases advertising the risk at locations around the world that have nothing to do with the location under study, what is needed is public education. All coastal residents anywhere in the world living in seismic zones should that that if they feel an earthquake tremor that lasts over 30 sec, they should evacuate inland or to high ground immediately and wait until they are given official all clear broadcasts.