We asked her about her work — and how this month’s big quakes is helping scientists refine California’s fledgling earthquake alert system.
Q: Tell us about your role with helping to develop the ShakeAlert system.
A: At the moment one of my projects is to ask questions about what the best possible earthquake early warning system could do based on how fast seismic waves travel so that we can measure ourselves against what the perfect ideal would be. It’s important to know what’s the best you can possibly do to know how close you are to getting there.
Q: Are there good systems already out there?
A: We’re not the only early warning system. Japan has been running an earthquake early warning system for many years now. Of course, all these different countries have all these different seismic hazards. The question is, what are they doing and would that apply in the U.S.?
Q: What did you learn from Japan?
A: In Japan, they alert for a much higher level of shaking than had even been discussed for California because of the difference in how frequently they happen there. When we were talking to them about what level shaking we think we should be warning about, in Japan that is a ridiculously low threshold. Alarms would be going off all the time.
Q: What was significant to you about the Ridgecrest earthquake?
A: The chance to interact with people. This was the first earthquake that has been significant, that people noticed, since the earthquake early warning system started. This is our chance to actually get feedback. Is this what people wanted?
Q: The ShakeAlertLA mobile phone application did not activate in the Los Angeles area because the quake was in the neighboring county and the shaking in LA didn’t reach the intensity threshold to trigger an alert. How did people respond to that?
A: The system worked, it picked up the earthquake, it picked up the shaking. The City of Los Angeles distributed the early warning app, and the system correctly identified that there would not be damage in LA. The app didn’t put out an alert because there was no damaging shaking. But there were people who were unhappy, saying ”I would have liked a warning” even though there was no damaging shaking. Different people may have different preferences. There’s no right or wrong answer here. But just as with hurricane forecasting there’s some uncertainty, where will it make landfall, how strong will it be? I do hope we can find a way to communicate any kind of forecast range of possible outcomes.
Q: Will that lead to changes?
A: The thinking was that the thing people would enjoy best is being warned before damaging shaking. Do people want alerts for any shaking they might feel including shaking they don’t end up feeling, or for other types of shaking? Those conversations are already happening now and there may be an announcement of changes.
Q: Where are we at in rolling out the ShakeAlert system statewide?
A: We have a finalized plan. The plan is to go public California-wide sometime this fall.
Q: What are the challenges?
A: There are two limitations that are very important to understand. There’s always going to be a region that is not warnable and that’s from right where the earthquake begins to the nearest seismometer. The farther seismic waves travel is time you’re losing to get a warning out. The target I believe is a minimum of one seismometer every 20 kilometers, with 10 kilometer spacing in densely populated areas, and within 5 kilometers of all mapped fault traces. We have been working for some number of years now to put sensors in the ground, but that requires getting permits approved.
This is a shaking forecast. Just like when you see a thunderstorm that looks like it might issue tornadoes or a hurricane and track it, there’s a range on it because you’re attempting to forecast the future based on information we have now. But what’s interesting with earthquake warnings is that because the time involved is so short, the system has to be entirely automated. There’s always going to be inherent risk that something incredibly stupid might happen that no human would allow.
Q: Ridgecrest was hit by a significant magnitude 6.4 earthquake July 4 that’s now being called a fore-shock to the much stronger 7.1 magnitude quake a day later. How unusual is that?
A: It looks a lot like the Elmore Ranch and Superstition Hills earthquakes in 1987. (On. Nov. 24 1987, two significant earthquakes struck along the southern San Jacinto fault near the Mexican border in Southern California, the magnitude 6.2 Elmore Ranch earthquake and the magnitude 6.6 Superstition Hills earthquake a few hours later.) It’s always fun when history repeats itself. In general, the rule of thumb is that if you experience an earthquake, the chance of another earthquake in the same week of larger magnitude is about 5 percent on average, or 1 in 20.
Q: These earthquakes struck pretty far inland, we normally think of earthquakes along the San Andreas fault line. Are these related to the San Andreas?
A: It’s not part of the San Andreas fault but part of the same tectonic plate boundary. There’s a much broader boundary zone in southern California, and these faults are part of that very broad zone. There are faults all across the state. In Northern California, there’s the San Andreas, Hayward, Calaveras, a much narrower zone. In Southern California it’s almost a sort of continuous transition from big faults along the coast all the way to faults in Nevada.
Q: Do these big quakes affect the probability of earthquakes along the San Andreas?
A: They’re so far away from the San Andreas fault we don’t think it affects the probability of anything happening on the San Andreas.
Q: Even though the damage was relatively light, it still has folks in Ridgecrest rattled.
A: One thing that’s interesting is how scary they are, how impactful and stressful earthquakes can be. Earthquakes can cause PTSD-like symptoms, and the continued shaking from aftershocks can be traumatizing even if they aren’t damaging.
Five interesting things about Sarah Minson
- Raised in Mendocino County, California
- Earned her bachelor’s degree in geophysics at UC Berkeley in 2003.
- Earned her master’s and doctorate degrees in geophysics at the California Institute of Technology in 2005 and 2010, with a doctoral thesis on “A Bayesian Approach to Earthquake Source Studies.”
- Expertise in kinematic rupture models, earthquake early warning, Bayesian analysis, earthquake source mechanisms, crustal deformation.
- Recent awards include 2018 Kavli Fellow (National Academy of Sciences and The Kavli Foundation), 2016 U.S. Geological Survey STAR Award, 2014 Presidential Early Career Award for Scientists and Engineers (PECASE).
