UW News

October 4, 2018

Q&A with Harold Tobin, director of the Pacific Northwest Seismic Network

UW News

Earthquake expert Harold Tobin joined the UW this fall as professor of Earth and space sciences and director of the Pacific Northwest Seismic Network. While he comes from a faculty position at the University of Wisconsin, he’s no stranger to the risks posed by offshore faults like the Cascadia Subduction Zone, the source of our “big one.”

UW News sat down with Tobin to learn a bit more about his research, experience and plans for the UW-based Pacific Northwest Seismic Network, a coalition among the U.S. Geological Survey, the University of Oregon and the UW that monitors seismic activity from earthquakes and volcanoes in this region.

What drew you to the UW?

HT: To be honest, it was an easy decision. I know the area well. I lived in Oregon for two years when I was in my early 20s, but I’ve been here tons since. I went to graduate school in Santa Cruz, and before then, as an undergraduate at Yale, I did fieldwork for two summers high up in the Olympic Mountains. And before that I had spent a summer volunteering at Mount St. Helens. That got me interested in this region and in plate tectonics. I’ve just moved to this region after being a professor in New Mexico and Wisconsin, but coming to the Pacific Northwest I’m just tremendously excited to be here and I feel like I’m coming back to my research home.

mug shot

Harold Tobin

The challenge of directing the PNSN and taking it to the next level is really exciting. The PNSN already is this fantastic organization that’s top notch in terms of being part of the nation’s data collection system and frontline information system for major natural hazards. The group of people here at PNSN is so talented and dedicated. The scope of expertise required to run 300-plus instruments, of all different flavors, vintages and types, running around the clock with all the data streaming in real time, is enormous. The communications networking system has to be a 24/7, fail-safe operation, and it is. It’s a privilege to be able to come into an organization that’s already functioning at a really high level.

The PNSN has also already established itself as this fantastic link between the scientific measurements and the community at large. The PNSN has to be more than the seismic network; it’s got to be the go-to place for geologic hazards, and understanding how living on top of a subduction zone matters to people in the Puget Sound region, the state of Washington and the whole Pacific Northwest. I’d love to see the PNSN’s reach continue to expand on the public outreach side, and build the research program at the university, too.

What are some current or upcoming projects at the Pacific Northwest Seismic Network?

HT: This is a really exciting time for the PNSN. First of all, the stuff that’s already going on: The advent of the ShakeAlert system and earthquake early warning is happening; some parts are already operational, and much more is coming soon. When our sensors detect an earthquake’s “P-wave,” the first, fast-moving but relatively undamaging wave, it gives you some tens of seconds of warning before the slower moving “S-waves” that really do the damage. Developing a system with the U.S. Geological Survey and university partners that’s capable of detecting earthquakes as they start and alerting civil authorities and the public is ambitious but achievable, and I want to make sure that goal is fully realized.

In terms of new initiatives, the scientific community realizes that if we want to fully understand how the earthquake processes work in the Pacific Northwest, from offshore right up to the mountains and even to the east side, we need even more types of measurements and instrumentation.

We think of the seismic information, which is how the Earth shakes. But the Earth is actually in motion over a much wider range of timescales. Geodesy is the study of how the Earth moves slowly, and seismology is how the Earth moves fast. Really it’s just one big spectrum. We will push to comprehensively measure all these types of motion and unrest.

The next big thing is figuring out how to make more measurements offshore. The shoreline is no boundary, from the Earth’s point of view, for plate tectonics or for the Cascadia Subduction Zone, which is our biggest hazard. Most of the stress buildup leading to the inevitable earthquake is happening offshore, beneath the bed of the Pacific Ocean. We need to be able to monitor that much better than we do now. It’s also a huge challenge because any detector on the seafloor is 10 to 100 times more expensive than on land. But there are new technologies that are emerging to monitor seafloor motion better. Incorporating those into the network is a major long-term goal.

When might ShakeAlert go public – as in, when will our phones warn us that an earthquake is imminent?

HT: ShakeAlert is here now, really. We have the capability in place now to alert the operators of critical infrastructure like utilities when an earthquake large enough to cause damage has occurred and to expect imminent shaking. That will allow them to take specific actions to protect the public during those critical seconds. In 2018 ShakeAlert has transitioned from research and development to actually using the system to take risk-reduction actions, and pilot programs are expanding. For example, we’re working with the superintendent of public instruction and have identified a number of school districts in areas where the sensor networks are already up to the task that have indicated they would like to develop pilot projects for schools.

There is a big push across all the West Coast states for congressional funding, mostly flowing through the U.S. Geological Survey, for the buildout of the ShakeAlert network. That funding will ensure that there’s the appropriate density of instruments on the ground so that when an earthquake happens, anywhere in Washington, Oregon and California, it’s detected appropriately, can be assessed within seconds for how big of an event it is, and then the alert can go out as the event is unfolding, in real time, to surrounding communities.

But I understand a lot of people are wondering: When will I get something like an AMBER  Alert or a weather alert, where my phone will buzz and tell me the earthquake shaking is on its way? There are still some technological challenges to overcome in order to make that work. The reason is that an earthquake alert has to be real-time down to a couple of seconds. I just moved here from the Midwest, where you might get a severe weather alert, say for a tornado. If it got to your phone 30 seconds after it was issued, that’s no big deal because you typically have maybe 10 minutes of warning. ShakeAlert has a technological need to be issued and reach everyone’s cell phone within just a few seconds in order to be useful, and that is still very much a challenge, for technical reasons surrounding how the phone networks actually work.

The City of Los Angeles is going to be pushing notifications to 35,000 city and county employees first, starting with City Hall, and the mayor would like to push it out to 4 million Angelenos next year. AT&T and other companies are developing apps that will try to do this massive push of data to thousands and then millions of phones.

We’ll let them pilot that in L.A. where there are more frequent, smaller earthquakes, and that will test the system. And then we’ll use what we learn from that and adapt it for our situation here in the Northwest.

How is it that Mexico City and Japan already have earthquake early warning systems? Why is it taking longer to implement in the U.S.?

HT: Mexico City has a fairly uncomplicated geologic situation for earthquake alerts. Part of the reason is that Mexico City is concerned specifically with earthquakes on the subduction fault that are relatively deep, and where the fault line is relatively far away from Mexico City, but still generates strong shaking in Mexico City. That gives them up to a minute or more of warning time in Mexico City before the shaking is strong there. So the alert system worked pretty well for their 2017 earthquakes.

The Japanese system is more like the system that we would aspire to, ultimately, with ShakeAlert. Of course Japan has really high earthquake hazards and very frequent events. They’ve invested enormously in their system, and the country is just blanketed with instruments. In Japan, that system has been fully operational for a number of years now.

Three years ago, I was in Japan in the middle of giving a talk at a scientific conference, and suddenly I felt my cell phone start buzzing in my pocket, and then I realized it was not just my phone, it was every phone in the room, plus the PA system. I didn’t understand, but then the room started shaking. What my phone, and all the phones, were doing was giving a message in Japanese: “Earthquake detected, moderate shaking expected in 5, 4, 3, 2…”

There are some technological differences between the cellular communications networks in Japan and those in the U.S. that we still have to solve here. So they are ahead of us on earthquake warning, but they’re also showing the way in how to do this well.

Japan is also instrumenting their offshore much better than we are. Their main homeland-security issue as they see it is earthquakes and tsunamis, so billions of dollars are being spent on studying the offshore region. We’re not quite there yet to anything like that level, but we’re really pushing for this in Cascadia, and also in Alaska.

How did you become interested in seismology?

HT: As an undergrad I became a geology major, and I became interested in the young stuff, geologically speaking, the active plate tectonic systems and especially subduction zones, like Cascadia. Most of my Ph.D. project was on the offshore Cascadia region. My first project as a doctoral student was diving in the Alvin submarine off the coast of Oregon down to the deep sea trench and literally mapping active faults on the seabed by looking out the window of the submarine. Relating that view to our seismic-wave images of the subsurface geology allowed me to discover new things about how the subduction fault works. I was hooked.

What changes have you seen in earthquake science?

HT: I started graduate school at UC Santa Cruz in 1989, and at that time there hadn’t been a magnitude-9 earthquake anywhere on the planet since 1964, since before I was born. And there wasn’t one until 2004, until well after I’d finished my Ph.D. The whole time I was doing my studies we were focused on active subduction zones, and of course there were big earthquakes around the world, but we didn’t have this stuff in the public eye and in the media because there was no direct experience of a major, Pacific-wide tsunami event. People were just getting the inkling in the 1980s that Cascadia wasn’t a quiet zone but was building up the stress for a future, giant earthquake, probably a magnitude-9. Brian Atwater and Kenji Satake were still figuring out the history of the magnitude-9 — all of that was still to come.

But since the 2004 Sumatra earthquake and the devastation in the Indian Ocean, that was 250,000 lives lost, and then not many years later, first in Chile in 2010 and then in Japan in 2011, with the Tohoku earthquake and tsunami, we’ve just seen a massive change. These are the first subduction zone earthquakes that have happened in the modern, digital-instrumentation, satellite-observation era.

person in front of screens

Harold Tobin aboard the research vessel Marcus G. Langseth, conducting a marine seismic reflection survey of the Cascadia Subduction Zone off Washington’s coast.Jackie Caplan-Auerbach

This field of science has been massively transformed in the past 15 years or so, and it’s a tremendously exciting place to be, scientifically. It’s transformed my scientific career. Early on, I was mostly focused on trying to figure out the general geologic processes, thinking about things on the hundred-thousand-year to few-million-year time scale. Now I’m focused on how does the Earth’s geology and the nature of the fault zones directly impact the hazard from the earthquakes. One example is: If a big earthquake happens on the Cascadia subduction fault, it will slip that fault offshore, mostly. We think about the shaking in Seattle, but for the coast the biggest hazard is probably the tsunami. And how the tsunami is generated and how big it is and what areas it affects all depend on whether the fault slips all the way up to the seabed, or whether all the slip is down deeper in the Earth and kind of peters out as you get up to the surface. It changes the pattern of the warping of the seafloor, and that’s what pushes up the water and makes the wave.

We worry a lot about the shape of the faults, and how much of them are locked up or “stuck” so they can generate a big earthquake, whether that locking on the fault is patchy, and how that will affect the pattern of causing a tsunami. So the geological research on these offshore faults is a component of the hazard analysis, which is not the way we used to think about it.

You are currently leading a major research project in Japan, which has seismic risks similar to our region. Can you describe your work there?

HT: The project I’ve been working on for nearly 15 years is an integrated project to study a part of Japan’s subduction zone, south of where the 2011 earthquake happened, called the Nankai Trough, off the coast not far from Kyoto and Osaka. A more than 1,000-year historical record exists for earthquakes in that region, and about every 90 to 120 years there’s a magnitude-8 earthquake with an associated tsunami. We went there long before the 2011 earthquake. The integrated project is imaging the seafloor and the faults below it and then drilling holes using an amazing piece of seagoing technology, a scientific drilling ship that looks like an offshore oil drilling ship. But we’re not drilling for oil — we’re drilling for data.

By drilling down into and around the fault zones we can do a couple of things. One is sample the material in the faults. There are all kinds of geological clues as to what the conditions have been like during the earthquake slips. Literally, during the seconds the earthquake takes, it leaves a geologic record in the rocks. We can then use the boreholes themselves as observatories, and put instruments in them: special kinds of seismometers that go into the fault zone, temperature sensors, pressure sensors. Those let us see the deformation or strain in the rock, how it’s bending and creeping and building up toward a future earthquake.

It’s a huge open question whether fault zones show you anything before the earthquake occurs. From surface measurements, we’ve pretty much established that there is no detectable long-term precursor to earthquakes in general. There’s no good earthquake prediction mechanism today based in science. But there’s a hint that for these especially giant events, some new kinds of data recorded offshore and even down in these holes are showing us that maybe the faults start to pop and creak and strain before the earthquake starts, over the span of hours to weeks, maybe even months to years. That’s a major goal of research now: not predicting earthquakes, but understanding whether there’s even a physical basis for predicting earthquakes — or not.

We first proposed the project, mostly led by U.S. and Japanese scientists, as a coalition effort to comprehensively study this part of the world as a kind of a case study that applies to all subduction zones. Nankai and Cascadia are sister subduction zones, because they’re similar in many ways.

See the welcome announcement from the College of the Environment

Read a recent profile of Tobin in EARTH Magazine

Watch a video about the Japan drilling project

The drilling ship is heading back out to our main site off Japan very soon in October. At the site, beneath 2 kilometers (more than a mile) of seawater at the bottom of the ocean, lies the top of a hole we have already drilled 3 kilometers down from the seabed. It’s lined with steel pipe, just like an oil well is “cased.” It’s all in place down to 3 kilometers, but our fault zone lies at 5 kilometers depth, so we need to extend this borehole from 3 kilometers to 5 kilometers to reach it.

Our goal from October to March is to finish drilling that hole, the culmination of the whole project. We started our drilling in 2007, and it has proceeded in many stages as we’ve been working our way to build a deep observatory borehole. I’ll be going to sea from just before Christmas to early February with a huge team of almost 200 people, trying to make sure that our borehole gets drilled safely and makes it to our target zone.

Once all the instruments are connected we’ll be studying the fault during the period between earthquakes to understand the forces and stresses that accumulate in the fault to create an earthquake. Of course, if an earthquake occurs during the experiment, and these instruments will be in place for decades so that’s possible, these instruments will record right up to the earthquake, and it will be a unique and valuable study of how a fault works. But even without an earthquake, it would be the first time in the world that we have that level of instrumentation on a major fault like this.

What about clustering of earthquakes? Should we be worried that the Pacific Rim has been pretty active, or when we hear about local swarms of earthquakes?

HT: One thing we can say is when little earthquakes occur, and even swarms of earthquakes, there’s no obvious direct link that means that we’re about to get a big earthquake. When it’s in a volcano, people pay a lot of attention to those earthquake swarms, because it can mean that magma is moving inside the volcano and there’s potential for eruption. Volcanic eruption is a bit of a different story than the crustal faults. There is always activity somewhere around the Pacific Rim, so when a few faraway earthquakes make the news in a short period of time, sometimes people wonder if that means the Big One is more imminent here. At long distances there’s just no evidence that major earthquakes are any more or less likely after other ones. That’s been studied exhaustively, so I don’t think people should worry about that.

Broadly speaking, small events nearby shouldn’t make people start freaking out, either. We just don’t see patterns like that in big subduction zones. Of course, if we start to see something really unusual happening, then the PNSN folks will be very focused on it, and we can imagine a scenario where we might start talking to the public, but that’s a long way out.

So should people be worried? For the Cascadia Subduction Zone we absolutely expect a very large earthquake to happen someday. But right now, that day could be later this afternoon, or it could be 150 years from now, and there’s not much difference in the likelihood of either of those.

And the Cascadia Subduction Zone is far from the only hazard in this region. There are a number of faults below our feet that could have effects that are different from the so-called ‘Big One.’ Some of those present significant earthquake hazards because, while they might happen relatively infrequently, they could have dangerous effects. We could have a magnitude-7 slip on the Seattle fault, which is closer to major cities and relatively shallow, rather than something bigger that’s much farther away, and that could create really strong shaking in this region. It would be more like the Kobe earthquake in Japan, or the Christchurch earthquake in New Zealand. That’s a very serious concern but fortunately there’s also evidence those earthquakes are quite rare, with perhaps thousands of years between them.

My attitude is: Don’t be afraid that it’s imminent, but just recognize that the hazard is there and be prepared. You should know what your game plan is for your family and your home, and take the steps that are recommended for making sure you’re as safe as possible.

Here’s one perspective: I think about earthquake hazard all the time, and yet I just moved here. I have no hesitation about living in this region despite knowing the risks, because it’s something to be prepared for rather than be freaked out by. That’s my vote of confidence that it’s not too scary.


For more information, contact Tobin at 206-543-6790 or htobin@uw.edu.


For general questions and tours of the PNSN facility, contact communications manager Bill Steele at 206-685-5880 or wsteele@uw.edu.