Be prepared, and not just at dawn
FT readers may enjoy the chapter on HAARP and weaponising earthquakes, but the rest of the book is a one-sided look at a few decades’ earthquake science rather than a new seismology
Earthquake Prediction
Dawn of the New Seismology David Nabhan Skyhorse Publishing Inc 2017 Hb, 240pp, $22.99 ISBN 9781510720978 Prediction is difficult – especially about the future. This remark was first made, not by Yogi Berra nor yet by Niels Bohr, but by the Danish cartoonist Robert Storm Petersen. It’s certainly true about earthquakes. Yet the subject of earthquake prediction has a certain fascination, and there is a thriving subculture of amateur earthquake predictors squabbling, often quite viciously, about who has the better track record. Their methods are varied, but one perennial favourite is the position of the Sun and Moon, and the resulting tidal forces. This is the method favoured by science and science fiction author David Nabhan in Earthquake Prediction.
Do tidal forces affect the occurrence of earthquakes? I think it is certain that they do. The question is how much they accelerate the process. And here we come to that dangerous word, “triggering”. To say that something triggers an earthquake is not to say that it causes one. It may be the case that tides trigger an earthquake next Tuesday that would otherwise have occurred next Friday. But this is not really very exciting. If a fault is not ready to break, no amount of triggering will set off an earthquake. How to tell that a fault is ready to go? That’s the real problem.
The key to useful earthquake prediction is like this. Suppose you have identified some possible precursor X, which you think can be used to predict an earthquake. In any time period, there are four possible outcomes: (1) X is observed, quake happens; (2) X is not observed, no quake happens; (3) X is observed, no quake happens; (4) X is not observed, quake happens. For a system to be socially useful, you need as near as possible 100% outcomes of types 1 and 2. In the case of tidal triggering, obviously, type 3 outcomes will be very common. Many earthquake prediction success stories, once you probe deeper, turn out to be concocted by making public a type 1 result and hiding all the type 3s. Nabhan mentions the celebrated story of the evacuation before the 1975 Haicheng earthquake, but not all the other evacuations that were ordered around the same time, before expected quakes that never happened; nor the disastrous type 4 result in 1976, when an unpredicted earthquake flattened Tangshan, killing around 250,000.
It is certainly not the case, as Nabhan suggests, that earthquake prediction has been a taboo career-killing subject for the last century. In the 1950s it was widely expected that breakthroughs would be imminent. But results have been so disappointing that most seismologists have rather given up hope; indeed, there is evidence now that there is no sort of X for which type 4 results can be avoided; thus the best you can hope for is a partial success rate.
Earthquake prediction is certainly a subject that is very interesting scientifically, but its social application is something else. If I predict a big earthquake to strike Los Angeles next Tuesday, what next? You can’t evacuate the whole city. You can tell people to be aware, be prepared, but if they live in Los Angeles they should be doing that all the time. Credit to Nabhan, he understands this, and his final chapter on earthquake preparedness is excellent. But he is over-optimistic about being able to limit the need for care to specific time windows.
He makes much of his conclusion that six fatal earthquakes in South California occurred at dusk or dawn near a full moon. But this is a small sample, and an earthquake doesn’t know if it is to be fatal or not. So let’s take all earthquakes ≥ 5.5 M within five degree longitude of Los Angeles and north of the equator, since 1960 (about 400 events) and look at the times of occurrence. A very quick analysis suggests that there is a bias towards dawn events, but they occur at other times of day as well. To be precise, about a third of events occur between 4am and 8am local time (12.00–16.00 UTC), twice what you would expect from a random distribution. There’s no peak around dusk. So relying on time windows of increased probability runs the risk of being surprised by an untidally-triggered earthquake at midnight.
Most seismologists would agree that the key to saving lives is to invest in better buildings, and encouraging general preparedness. I remember checking into a hotel in Japan and finding a flashlight clipped to the head of the bed. “Aha,” I thought, “Here is a country that takes earthquakes seriously!” Comes an earthquake in the middle of the night, the electricity is out, you have to evacuate in total darkness, that little flashlight makes a big difference to your safety.
What Nabhan has written is not a bad book, and not a stupid one. He won’t win any friends by calling it ‘Dawn of the new seismology’, when it mostly consists of a one-sided view of stuff that has been kicking around for the last 20–30 years. His writing is somewhat strident and slapdash – faults do not “explode”, neither do they “erupt” – and I always feel that making assumptions about the nationality of your reader is a sign of amateurism. His research is a bit careless as well. I am delighted to read that Charles Davison (1858–1940, Birmingham schoolmaster) was one of the great seismologists of any era, though I may be the only person on the planet who would agree with him on that. Nevertheless, it is strange to read that what Davison thought about the 1923 Tokyo earthquake can only be imagined, when he published an entire book on the subject.
Of interest to readers will be the chapter on weaponising earthquakes; inevitably HAARP makes an appearance, and no, I don’t believe it for a moment.
A more nuanced approach would have made for a better book, and it lacks the insights that can come from a professional engagement with the subject. The general reader will certainly learn a good deal from it, but as a popular guide to seismology I would rather recommend the books by Sue Hough (USGS), or (ahem!) my own contribution to the genre. Nabhan’s book is heavy on self-promotion, so I think I’m justified in indulging in some myself.