USGS slams study’s claim of 99.9% chance of large L.A. earthquake, Los Angeles Times
“The U.S. Geological Survey is raising serious doubts about a recent study that calculates a 99.9% chance of a large earthquake in the Los Angeles area in the next three years. The USGS took the rare step of issuing a statement raising questions about the study, recently published in the journal Earth and Space Science and coauthored by a research scientist at NASA’s Jet Propulsion Laboratory in La Caada Flintridge. “I have serious doubts that the conclusions of the paper are supported by the analysis that’s presented there,” Robert Graves, a USGS seismologist and Southern California coordinator for earthquake hazards, said in an interview Tuesday. The study lists Graves as a consultant who helped with calculations for the study, but Graves says he has deep concerns about the study.”
NASA Study Improves Understanding of Los Angeles Earthquake Risks
“A new NASA-led analysis of a moderate magnitude 5.1 earthquake that shook Greater Los Angeles in 2014 finds that the earthquake deformed Earth’s crust across a broad region encompassing the northern Los Angeles Basin and northern Orange County. The shallow ground movements observed from this earthquake likely reflect strain accumulated on deeper faults, which remain locked and may be capable of producing future earthquakes.”
Potential for a large earthquake near Los Angeles inferred from the 2014 La Habra earthquake, AGU
Regardless of the merits or not of the paper, this is the problem with any earthquake prediction. There is so much at stake for people living in the area that there’s no way you can make people happy.
Whether you’re right or wrong, there will be a lot of people (usually in well-funded organizations) who will come down on you like, er, a collapsing chimney, if you open your yap to predict anything.
Better to say nothing.
And if you do live in a quake-prone region, I recommend Peter Yanev’s book Peace of Mind in Earthquake Country, http://www.theearthquakeboo… (And no, I’m not related to the guy.)
Let me add to my earlier comment. This is a perfect example of why individuals should look at the worst-case scenario proposed by the paper and, ignoring government denials, make private plans or preparations on the assumption that the prediction is accurate.
There’s a downside: disruption of your family life in all its aspects in the LA area, which is not negligible. It’ll cost you significantly to uproot and move elsewhere. But there are definite upsides, and I don’t need to list them.
In the bigger picture, we’ve all had too many years of too many carefully phrased denials by word-parsing lawyer officials to trust anything coming out of a government agency, no matter who sits in the White House.
Please don’t say that. If you do, I’ll have to invent some sort of terrible danger to inhabitants of Colorado. Otherwise, all the people who believe this sort of pseudo-logic would move to here, and I’d really rather not have them around.
Seriously. it is worth mentioning the six Italian scientists who were tried, convicted and later acquitted because they predicted a major earthquake in L’Aquila was unlikely. That prediction was published shortly before a 5.9 earthquake hit that city in 2009.
Three-fourths of the lawyers on planet Earth live in the United States (or some high percentage like that). There are many take-home thoughts regarding that fact, but one biggie is that certain situations have no good outcome for those who put out a warning.
I wish this weren’t so, but society is what it is. Se my note below.
To be fair, the dispute was over the specific probability stated in the NASA paper, which the USGS commenter noted was not supported by the data and analysis. No one disputes the obvious risk of major earthquakes in Los Angeles.
Technically, yes. But specifically, it was a claim that there was a 99.9% probability of a magnitude five or greater earthquake. This is regarding a topic which, I think, most people in the field agree that predictions are extraordinarily difficult. Claims about 99.9% probability are either absurd (claimed accuracy is not credible) or trivial (like predicting the Sun will rise within the next 24 hours, at latitudes between the arctic circles.) An 80% chance of a magnitude five earthquake would be interesting and credible. A 99.9% chance of a magnitude three earthquake would be credible but obvious and uninteresting. But 99.9% chance of a magnitude five seems to push the limits of credible prediction.
The prediction is not based on calculation of actual future movements of rock, but is based on the statistical distributions of large numbers of earthquakes under circumstances of slippage and stress accumulation. A 5.0 earthquake causes only minor damage and is fairly common in this area. The main assumption is that the number of earthquakes that will occur over a 3 year period is a random variable.
“The B value shown here is consistent with B values for Southern California determined by Mori and Abercrombie [1997] for earthquakes > 9 km depth. For the Gutenberg-Richter relation to be completed, this deficit must eventually be filled with large earthquakes, up to M6.2, which is consistent with the above analysis. We assign a probability to these large earthquakes using a Weibull distribution [Weibull, 1951] and the assumption that over long times and large regions the Gutenberg-Richter magnitude-frequency relation is linear [Rundle et al., 2012; Holliday et al., 2014; Rundle et al., 2016]. The calculated probability for a M ≥ 6 earthquake within a circle of radius 100 km, and over the 3 years following 1 April 2015, is 35%. For a M ≥ 5 earthquake within a circle of radius 100 km, and over the 3 years following 1 April 2015, the probability is 99.9%.”
I should have known the press summary was misleading. A 99.9% probability of a magnitude 5 earthquake is along the lines of predicting sunrise. A 35% probability of a magnitude 6 is plausible and much more interesting. I’ll have to check and see what the objection from the USGS is.
There wasn’t much to that either. It sounds like the newspaper called a USGS geologist and he told them the prediction given in the article was not supported by the analysis given. Which is true, the article just quotes a method and does not provide the full statistical analysis, which might be too long for a paper publication. IMHO the publisher of the paper should have asked that the actual statistical data and analysis be included in an online supplement.
I agree the press article was very misleading. So Keith should not have positioned the article as an attack on NASA, but rather as an attack on the press. But while the NASA article may be in error, I would not characterize as BS and I don’t think the USGS did so.