Are Earthquakes Predictable? USGS Scientist Annemarie Baltay on ShakeAlert and What We Don’t Know

I was under the impression that there is some kind of prediction, some seconds before the earthquake happens, but you're saying it's not, right? No. Earthquakes themselves are so chaotic. We don't know exactly when or where an earthquake is going to occur. So earthquake prediction is not something that's possible. So there's still so much we don't know about earthquakes. Welcome to Beta Innovators. Our guest today is Anne-Marie Balte, a geophysicist with the United States Geological Survey and the team behind the ShakeAlert system. that provides earthquake awareness along the west coast. In this episode we'll explore how that system actually works, why earthquake prediction itself remains impossible despite decades of trying, and what her research reveals about why some places like Oakland during the 1989 Loma Prieta quake shakes so much harder than others nearby. This is Bay Innovators and I'm your host, Steve Ispas. Anne-Marie, welcome to Bay Innovators. Great to have you here. You're an expert in earthquakes. You work for USGS. That's right. What is the most fascinating thing about earthquakes? What I really like about the earthquakes is that they are really exciting, they affect so many people, and they can be devastating, and yet we really don't know that much about them, right? So think about that an earthquake has to happen inside the earth, and we're just sitting here on the surface of the earth. We have instruments, we have lots of instruments all over the earth, and we even have some instruments down to the crust of the earth. But we've never observed an earthquake like really down in the deep where it starts. So there's still so much we don't know about earthquakes. And they're so powerful. And they're these natural phenomenon that really affect people. And we want to understand more about them so we can, you know, prevent, you know, mitigate the disasters from earthquakes. Of course. So what do we know is that we have the plate, the plates, right? And the plates are running against each other. That's right. And then they break or whatever. trip on each other and then this happens. But even that, even that, we don't see it, do we? So it's like, you know, everything we know kind of about the earth is kind of like putting the earth into like an x-ray or an MRI and we're trying to investigate what's going on. So we understand the bulk properties, right? We know that inside the earth there's a molten mantle and then the crust sits on top and kind of the plates float around on the crust and that's where they bump up against each And then the stress can build up and then they break. So we understand to first order what earthquakes are. We know that physics, that makes sense. But it's so difficult to say exactly when and where an earthquake is going to happen and then exactly the impact of that earthquake and how that's going to affect everyone is still really challenging. And that's what we're trying to do. We want to get the details of that a little bit more known so we can really help everyone be better prepared for earthquakes. I think one point you made before we talked about this is the data points that you have. A large amount of data points inside the earth, probably a lot easier to figure things out. But even to get one data point, you have to drill and how far you drill and all that. Yeah, it's just really difficult to drill down into the earth. And we've done it a couple of times. We've tried and we've done it successfully a few times. It's really difficult, it's challenging, it's expensive, the earth doesn't cooperate. Are there probes like right where the faults meet? Do you have any visuals of that? No, we don't really have like a picture per se, but we do have in Parkfield, which is in central California, we made a big drill hole down and then across the fault and we can see exactly kind of where the fault is right there and so that's as kind of as close as we can get. So do we have a visual of that? Is there a way to show it? Or more like a diagram? Yeah, not really a picture. It's inside the earth. Just a diagram. So what we understand is that the rocks on either side of the fault are different. And so we can sort of get an idea of exactly where that break is, because we have different kinds of rocks with different properties. Now you're part of the team that works on ShakeAlert and Earthquake, so let's talk about it. Tell us how it works, what's actually happening. Sure. So ShakeAlert, Earthquake Early Warning, is a group of agencies that have come together. So USGS, academic partners, Cal OES, state and federal. We have many, many seismometers all over California, Oregon, Washington. And when an earthquake starts... We quickly can detect that the earthquake has started. So we detect the earthquake waves on the closest four stations, then we really quickly make an estimate of where that earthquake is and how big it is, and then we can sort of forward propagate to the people who are ahead of the earthquake. So quickly means seconds? Seconds, yeah. Ideally one, two seconds we detect the earthquake. Uh-huh. And... Earthquake waves travel fast, but information travels faster. So electronic waves travel faster. So the idea is we can quickly detect the earthquake and then send the message ahead to the people who are going to be feeling the shaking. Hopefully a few seconds, maybe tens of seconds of warning, depending on the situation. And it goes to people's cell phones. I got it, I think I got it three times. Yeah. First time was a mistake, it's later that it's an error, but the next two times was real, you know, they weren't big ones, but big enough. When does it trigger? Yeah, so we know earthquakes are difficult to understand and to model, and so sometimes you may get an alert and you might not feel shaking, and sometimes you may feel shaking before the alert, sometimes you may get the alert before shaking. That's all normal part of the system operation. We know that it has some variable. Depending on how you're getting the alert, we aim to send the alerts when the earthquake is about magnitude 4 to 5 and larger. And we aim to send them to people who are going to feel the shaking. So if you're farther away, you won't feel the shaking as much, so you may or may not get an alert in that situation. If you're very close to the earthquake, it's more challenging for us to get you the alert before you feel the shaking. So in that case, if it's a local earthquake, you may feel shaking and then get an alert. Again, because we're trying to get ahead of the information, but if the earthquake's close to you, we can't move any faster. So now these alerts, they go to cell phones. So does the system calculate where the person is physically? Like if I have my San Jose number, but I'm in Las Vegas, it's going to count? Yeah. So this is sort of the upstream challenge with the cell phone providers and Google and Apple. So USGS and our partners, we deliver the earthquake information. Um, and And so then the cell phone providers or, you know, Android has a system as well. They have to figure out how to send it. How to send it to, if someone is on the East Coast, the earthquake is here, it's not going to... Yeah, so, but you know, everything's getting better and faster now. So hopefully we can get you in the right, you know, location so that you get the correct alert. Yeah. So you're talking about getting seconds warning with the ShakeAlert. Now I was under the impression that there is some kind of prediction, some seconds before the earthquake happens, but you're saying it's not, right? No. Earthquakes, we talked at the beginning that we understand there's plate tectonics and the stress builds up and then it releases. So we understand the physics of earthquakes. Yes. But earthquakes themselves are so chaotic. We don't know exactly when or where an earthquake is going to occur. Even if we say it's going to happen on the San Andreas, San Andreas is hundreds of miles long. It's so difficult to say when and where. So earthquake prediction is not. Something that's possible and we have looked we have looked for all different kinds of signals We've talked to people whose dogs bark. We've looked for radio signals and chemical signals and We've never found anything any correlation or relation. Yeah, so, you know Sometimes you can see something that looks like it may be a prediction But if we go back and test it scientifically, we've never found anything that is definitively I'm Telling us about the best we can going do is is really be prepared and be aware We live in the Bay Area, we live in earthquake country, to go we know there's going to be earthquakes. So I want to get an exact sense of the speed. ahead So the signal starts traveling when you detect the earthquake, of course going at the speed and get of light very fast. But the earthquake itself is how fast? So there's a couple of different things that are happening. started. So when the earthquake happens, the waves move away from the earthquake. Like when you throw a rock into a pond, the ripples move away. Those waves move from the center outward at a couple of kilometers per second. Okay. But the speed of light is... 300,000 kilometers. Yes, much, much, much faster. So a couple of kilometers per second is very fast. You cannot run that fast. I don't know if your car goes that fast, right? So that's very fast, but we try to get ahead of it with information. And again, you calculate how powerful it is to see how far it's going to propagate. That's exactly right, yeah. So ideally, when the earthquake occurs, maybe we have four stations around it in all directions. And that helps us to figure out where the earthquake occurred. And then based on the amount of shaking at those four stations, if it's a lot of shaking, we say, well, it's a larger earthquake. If it's less shaking, it's a smaller earthquake. And exactly what you said that. If it's a larger earthquake, we want to warn people farther away. If it's a smaller earthquake, maybe we warn people, you know, not as far away. Now we have a diagram here that we wanted to look at, so let's take a quick look. Okay, so what are we looking at? Yeah, so this is a map of the Bay Area, right? We're sitting around here. Mm-hmm. And I guess no earthquakes right where we are. Yeah, that's right, yeah. So we can see a couple of things on this map. One is that the faults that we know about, the major faults, are in black. This is the San Andreas fault. The very famous one. Yep, the famous San Andreas fault. This is the Hayward fault here, Calaveras, and then other faults you can see that are named we know about. And then there's some little faults, you know, that are less famous but still have earthquakes. The second thing on here is the earthquakes themselves, and they're colored by the decade in which they occurred, and the size of the earthquake indicates, the size of the dot indicates the size of the earthquake. So we can see, this is where the famous 1906 earthquake, 1907.8, started, and it ruptured along the San Andreas here. And of course, it's surprising how close the center of it was to the city itself. Absolutely, yeah. Well, it caused devastation in San Francisco and, of course, caused the fire following 1906. So really impactful on San Francisco. We can also see down here the Loma Prieta earthquake in 1989 in the blue dots and many aftershocks of the Loma Prieta earthquake. So after a large earthquake we know that stress is still being released and things are resettling and we have many aftershocks that occur after a large earthquake. Now what's surprising here is is it most of these are from the 89? All this blue stuff? Yeah so the the 89 earthquake started here and kind of ruptured this whole length Um, the fault here after that was all of these blue earthquakes. So after, again, after a large earthquake, um, things kind of resettle and we know that there will be aftershocks. Yeah. Um, and there will be. One thing I noticed here, some of these are like, if this is the fault, some of them are a little bit off. That's right. Yeah. So a couple of things are happening. This is just a surface map. Most of these earthquakes are starting maybe five, 10 miles deep in the earth. And we don't know exactly. This, how this fault cuts down into, yeah. So Stevie picked up on a great point. This is an area of great debate of exactly where this one started. Did it start on the San Andreas? Did it start on a slightly different fault? And you can see here, there's all these little fault pieces here. And again, how they interact at depth is something we're still really studying and trying to figure out. Very interesting. It's amazing how much we know, but still how much we don't know. Yes, that's right. And then earlier this year, maybe in April 2026, we had a Boulder Creek earthquake that occurred right around here in a location similar. to the Loma Prieta earthquake, so that's been interesting for us to understand the earthquake as well. I think what's probably most of, to me, fascinating thing is that this whole, what is, is this Point Reyes? This is Point Reyes, yeah. Is it the result of the fault, that water? Exactly, yeah. That's so amazing. So what we know, we talked about, you know, broad scale stepping back plate tectonics, is that this is the North American plate over here, this is the Pacific plate. And so generally you are sitting in North America, I'm sitting in the Pacific. Generally the Pacific plate is moving to the north and a little to the west. And that motion is accommodated across all of these faults. So you can also notice that these faults are what we call sub-parallel. So they're all kind of running generally in the same direction and they're all accommodating the same sense of motion. So anytime you step across the fault, you'll step to the right. We call that a right lateral. And so this whole area is moving north and that's exactly what happened here when you have a fault. Actually the San Andreas Reservoir on the peninsula is a similar feature. You can see it here. So tell us what's happening, how these actually two earthquakes have progressed. Yeah, so we're gonna compare the 1989 Loma Prieta earthquake magnitude 6.9 that occurred near Santa Cruz with the 1906 magnitude 7.9 earthquake. So a whole one magnitude unit bigger, and I want to impress on you just how much bigger that earthquake is. So we'll look at what happened. Now when an earthquake occurs, it's not really just a dot. We showed it as a dot on a map, but that's not really right. It actually ruptures along the length of a fault. So it's like a fire truck driving down your street. It starts at one end, it turns on its sirens, it drives down the street, it turns the sirens off, and it stops. Everybody on the street is affected by that loud Sirens and even some, you feel a truck come by sometimes, right? Yes, yes. A larger earthquake, the fire truck is driving for a longer distance all the way down the fault, and then it stops. And that actual part of the fault is moving, is rupturing past it. Now, the second thing that happens when an earthquake occurs is the waves move away from the earthquake. So, again, when you throw that rock in the pond, the rock. sinks immediately the ripples move away from you. So we're going to see both effects here. One is that the earthquake is occurring for a long time and the second is that the waves move away. Now let's look at the Loma Prieta earthquake. Yeah we see it right here. Yeah we see it here rupturing and it's still going. The truck is still driving. Okay now the truck stops about 10 seconds. The waves are going to move away and so people are still feeling the effects of this earthquake but the earthquake itself has stopped rupturing and you can see that the red color dies down. It's not as intense and eventually that will stop shaking and go away. Now let's compare that. This is the same picture, kind of side view of the San Francisco Bay Area. Right. Let's compare that to what we see for the 1906 magnitude 7.9 earthquake. And some mention it 8.1 or 8.2. I guess it's always an estimate. All right, so the earthquake starts off of San Francisco. That fire truck is driving now in both directions. It's two fire trucks driving north and south, still going. Much bigger intensity much redder colors that means more shaking at the source and just much longer distance over which the earthquake is occurring and longer rupture time so it's still driving still driving to the south you can see still generating these really strong shaking waves to the south still going so that means it ruptured this much that's right yeah maybe 400 kilometers wow and now the rupture is just starting to die out maybe 40 seconds later So if you were anywhere along the San Andreas Fault during that, you would have felt a lot of shaking. In addition, people who were off the fault are now feeling the effects of those waves traveling away. Of course, of course. So that's the difference between a 6.9 and a 7.9 earthquake. So a huge difference there. There was no tsunami on this one. I guess there's other factors that make a tsunami happen. Yeah, that's great. So a tsunami occurs when Water is disturbed. And so generally the water needs to be uplifted. So if you're in the pool with your kids and you sink the kickboard and then they let the kickboard go, it goes and then the water kind of goes like that. That's a tsunami. So we need something under the water to lift up or down. Either way is fine. What you can see about... So this was too close to the shore. These faults, right, are not in the ocean, really. The other thing about it is that we talked about these are right lateral faults. So they're surface faults moving past each other this way. They're not moving up or down. So even if these were under the ocean, they're just moving next to each other, and that's not going to cause a tsunami wave. We get tsunamis when we have subduction zones. So in other parts of the world, the plates are. subducting underneath another plate and so that actually causes some vertical change in the height of the seafloor very quickly right on the order often you know 40 seconds or something is a pretty quick disturbance for an ocean wave and that causes the tsunami. Let's talk about what should one do when either they get the message or they actually start to shake. I guess there's a lot of education about this, but I want to ask you, what do we do? Hopefully, we don't have to do anything. I mean, it doesn't happen, but if it does happen. No, hopefully it doesn't happen, but like I said, we live in earthquake country, and we know that earthquakes occur here. Okay, so I suggest everyone take a look at where they are right now in your house or in your place of work, maybe at your kid's school or something like that, and where would you actually drop cover and hold on right now? Right now. So if you don't have something to go under, you can go next to something. Think about things that might fall on you. And so if something fell, it would maybe hit the chair. And if you were lower than that, you would be protected as well. Ideally, you can get under something. But if you can't get under, you can get next to something. If you're in bed, you should protect your organs, your vital organs, and your face. So roll over, you can put a pillow on top of your head, stay protected. Ideally you're spending eight hours a day in bed, right, so it's good to think about what you would do. It's a third chance. Yeah. Exactly. The natural instinct, at least for me, because I've been through two major earthquakes, one was in Romania. Yeah. In 1977, it was a 7.2, but for those constructions over there, that's disastrous because they're not built to any code, so there was a lot of... casualties and the 6.9 here. Now, at least my first instinct is to run outside if I'm close to the door. What are your thoughts on that? That's not the recommendation here, especially in California, where for the most part, our construction is good. And the time it would take you to exit where you are, you're more likely to get injured doing that. Even though you're I guess it also depends how close you are. Like if you're right next to the door, maybe. Yeah, no, we recommend that people shelter in place. That's why I say find a good place right now where you would be safe during the shaking. Once the shaking stops, then assess your situation and you can get up and evacuate. So, in fact, at home, we recommend that you have a pair of extra shoes underneath your bed, for example, so that when you do go to evacuate, you don't have to step on all the glass and things that have fallen on your floor. And there's, of course, the bigger issue of what happens afterwards, right? With maybe gas lines being ruptured or power lines down. So we recommend that you have three days of emergency supplies in your house in case you need to shelter in place. So food, water, medications, things for your pets, communication plan, and even have your important documents ready to go. I think we actually saw this more vividly during the wildfires. But we can think about the same kind of preparedness for any sort of natural or unnatural disaster. Thank you. Thank you. And I don't recommend that you go out Thank you. Thank you. and buy an emergency kit necessarily. Thank you. Thank you. But look in your pantry. Thank you. Do you have some rice, some pasta, Thank you. Thank you. some canned beans, something that you like to eat? Thank you. Thank you. Just have an extra backup. Thank you. Thank you. And again, I think we saw that, you know, during the pandemic, Thank you. Thank you. it was always good to have an extra meal in your house. Thank you. Thank you. Thank you. Now, we haven't had, you know, again, Thank you. Thank you. 89 was a moderate, I guess considered moderate, right? Thank you. Thank you. Thank you. Thank you. The 89 one. Thank you. We haven't had a big one since 1906. Thank you. That's right. Thank you. People keep talking about the big one. What is What does that mean? What does the big one look like? Tell us about that. The big one is a little bit of a funny word, right? What does it mean? Usually, we say the big one is the 1906 earthquake. And we think that the 1906 type earthquake maybe will happen every 300, 400 years. We don't know that much about it, like we were saying. So in the meantime, in the Bay Area, we should be more prepared for a 6 point something, 6.7, 6.9. So actually, we have a forecast for the Bay Area. Right now, we're in a 30-year period from 2014 to 2043, where there is a two-thirds chance. So more likely than not, two-thirds, that we will have a 6.7 or larger in the Bay Area. So that's really, for the Bay Area, the earthquake we want to think about. We saw that illustration of the shaking during the 1906, so we may see, you know, damage, widespread damage, but not total devastation, probably, like the 1906 earthquake. And what are the chances of 1906 earthquake or bigger? I don't know. I don't have that statistic for 1906. But I guess the same precautions would apply for whatever. Okay, so this is what you mentioned, the 72%. Yes, so right now we're in this time period from 2014 to 2043, so a pretty large time period, and we see that there's 72% chance, which is two-thirds probability to have a 6.7 or larger someplace in the Bay Area. And what this map shows us is the probability of that earthquake occurring on each different fault, right? So here we see the San Andreas, 22% chance of hosting that. Hayward Fault, 33% chance. And those are actually kind of similar chances, so we say two-thirds, maybe one-third chance it's on the peninsula, one-third chance it's in the East Bay, one-third chance it doesn't happen. But that's still more likely than not to experience a 6.7. So at the USGS, we've done a lot of preparedness for what could that look like. What's a scenario that might happen? How can we get people sort of visualizing possibilities? And we don't want to scare anyone, but to think about being prepared. So you're going to have your food, your water, your plan. How are you going to contact your family, your loved ones, people you live with, all these kinds of things. We want you just to be thinking ahead a little bit. The movie 2012, how realistic is that? Not so realistic. Maybe the most realistic part is that people kind of came together and the family worked together to get out of a disaster. But the science behind it, it's not super realistic what's going to happen here. So plates don't fall into the ocean. Another thing people often think is that the plates are going to open up and there's going to be a giant hole like down into, you know, Hades or something. That's not realistic. So as we discussed, like these plates are pushed up pretty tightly against each other and they just pop next to each other. There's no way for anything to open. There's no physics that makes anything open or fall off. There's nowhere for anything to fall, right? The whole Earth is like a giant puzzle that's really quite tightly squeezed together, and it's a lot of pressure that pops these plates next to each other. When we see these mega earthquakes like in Japan or Sumatra, Chile, Alaska, those also are the plates pushing into each other, right? There's a lot of pressure pushing the plates together. There's really not very much pressure opening things. So it's unrealistic to see things fall off or have these giant holes open up due to earthquake activity. That's good to know from a scientist. So, So it's not like a domino effect, in the sense that the more it pushes, the more the chance that it's going to crack more. It doesn't work that way. No, so it's an interesting question you asked about domino effect. But we do know that, like, Earthquakes can sort of communicate a little bit with each other. We talked a lot about how when the earthquake occurs, the waves radiate away from each other and the shaking goes away. Sometimes that can trigger other earthquakes. So, you know, if there is like a puzzle, it's all pushed together. Sometimes... One little thing can pop something off. So it's chaotic. We don't really know exactly what causes an earthquake to occur. So we have seen cases when we have a global, like, 2004 Sumatra earthquake, 2011 Tohoku earthquake. We've seen when those waves travel around the earth, they actually trigger very small earthquakes in other places. So they do sort of communicate a little bit, but it's on a much smaller scale than what we like to see in the action movies. Yeah. Another thing I'm hearing is that it's good to have smaller earthquakes because they release pressure versus not having anything. And then at some point in time, it's going to rupture the whole thing. Is that true? So that's a common thing that we hear. And the idea is sort of It's valid, but small earthquakes are so much smaller than big earthquakes. So if we think back to that picture of the 6.9 and the 7.9, the 7.9 actually released much more energy than if it were 10 of the 6.9. So let me back up a little bit. We know when we see earthquakes occur on the map, we know that every time we have a magnitude 6, for example, we see 10 magnitude 5 earthquakes. And every time we have 10 magnitude 5 earthquakes, we have 100 magnitude 4 earthquakes. So we already see this kind of pyramid effect happening, right? And you can see that on the map. There's lots and lots of little earthquakes. So every time we increase a magnitude unit, there's a decrease of 10 in the number of earthquakes. The larger earthquakes are more than 10 times bigger than the smaller earthquakes. So the math just doesn't quite math to make that sentiment work. So we would have to have something like 32 smaller earthquakes to relieve the pressure of one larger earthquake. So just to be specific, we would need 32 magnitude 5 earthquakes to release the pressure of one magnitude 6. What we observe, we only have 10 instead of 30. So it's a nice idea, but it just doesn't quite work out for what we see naturally in there. And that's why we continue, unfortunately, to have these really larger effects. So let's talk about the 1989 Loma Prieta earthquake, very famous because it was recent, you know, I just moved to California three months after, no, before, so I got it, you know, I was here. That's crazy, yeah, scary. And there seemed to be kind of different damage done in different areas of the Bay Area, so can you talk about why different areas seem to have different degrees of damage? So we know kind of two things. One is that we have more damage close to the earthquake. Which makes sense. Makes sense, right? The other thing that we know now and that has became more apparent after Loma Prieta is that the kind of soil you're sitting in really affects how you feel and how your building experiences are quite damaged. So let's imagine a rock and a bowl of jello and if you shake the bowl of jello and you take your hands away it kind of keeps jiggling and in fact sometimes it kind of even maybe it makes a mess it slops out of the bowl. If you take that rock and you shake the rock No, nothing happens. It's not even interesting, right? So if your house is sitting on a rock and the earthquake waves come and the shaking comes, it's relatively less impact. But if you're sitting in a bowl of jello, or what's more realistic, if you're sitting near the edge of the San Francisco Bay on some infill, You know, you know, the San Francisco Marina or what's now the Dog Patch is a lot of infill. Over in Oakland, Alameda, some of that is some very soft bay mud. Those muds, we know, they shake more, they shake for longer, and they amplify, they increase the ground motion. So what we saw from Loma Prieta earthquake was actually damage in San Francisco and in Oakland, quite far away, farther away than we would have expected from this earthquake. But the ground motion was amplified due to these very softer soils compared to sort of the hills on either side, right? The Peninsula Hills and the East Bay Hills saw relatively little or almost no damage from this. Fascinating. Any other projects you're working on within this? ShakeAlert, of course, is a big one. And I understand there's so much more than just getting the data. It's communicating with the different providers to get the data to the people, right? It's a lot more. So my interest and my personal research is really about these effects of how the ground shaking affects different places. So what we really try to understand is that if an earthquake happens and then the shaking goes out, how are different places going to be affected? by the shaking and how can we best model that so we can help you understand you know how to build your house maybe potentially choose where to live where we should build trains how to build bridges better in those places and so we're always just looking for these earthquakes that are recorded in different places and how um it affected different different locations you know even if you could be 10 miles from the earthquake on one side versus 10 miles on the other you may have a very different experience and that's what we're trying to understand and model What is most satisfying about your job? What is the thing that you look forward to? Well, it's not earthquakes. Well, maybe you are small ones. Oh, yeah. Well, I think what's satisfying is that it's interesting. Like I said, it's a lot of unknown. So it's, you know, an interesting intellectual problem. But then every time we discover something, it's taking us one step closer to helping people really mitigate the effects of earthquakes. So we can build stronger. We can be more prepared. We can be more aware of where we are and where we're living and hopefully reduce the effects of earthquakes. So it's a nice combination of this, like, science. With, you know, helping society. I can ask you about AI. We live in an age where AI is everywhere, in a toothbrush, tool, whatever. Whatever. Yeah. Do you see, and there's also a lot of controversy, you know, if it's useful or not, or how it can damage. Do you see it helping your research or your studies? A lot of AI tools have been very useful for us, right? So AI tends to take a lot of data and distill it down and help us understand the patterns. And we have earthquakes all over the place. We have sensors all over the place. So we're constantly getting streams and streams and tons of data. So just as an example, one seismometer can record 100 times per second. And maybe at one station, we have three or six of these seismometers. So we're getting 600 samples per second constantly coming in all over the place. And we have hundreds of stations, thousands of stations across California. So AI can really help us sort of digest all that information and find patterns and understand what we're seeing. So it has been very helpful in that. But what we're doing is really trying to still uncover the unknowns of what's happening with an earthquake. And that still really takes creativity, asking the right questions, looking at the data from different angles. So we need a combination of our own human creativity, coupled with some AI is a great tool that can really help us. Talk to us a bit about patterns, and not patterns kind of what happens after the earthquake, but earthquake patterns. Are there any patterns? Yeah, definitely there are patterns. That's a great question. Again, we know earthquakes typically occur on faults, and we understand that the faults are kind of loaded at a constant rate. So from the external... Because the whole thing is moving, right? Yeah, because the whole Earth is kind of moving and these plates are very large. So we have the North American plate, remember you were the North American plate, I was the Pacific plate. We know that they're being pushed together and that the Pacific plate is moving north kind of at a constant rate, but it's not slipping smoothly, it's stuck. And so it jumps forward, stuck, jumps forward, and stuck. We know, though, that over many years, it has to move a certain amount. So we understand that there has to be a certain number of earthquakes that occur. And so we can sort of say, all right, they're not going to occur really regularly, but over a very long time period, we can expect this many earthquakes. So that's certainly something that goes into all of our models. And the other thing we know is that when a large earthquake occurs, Smaller earthquakes occur after. So those are called aftershocks. We're all familiar with the idea of aftershocks, but we know that there are certain patterns. So the largest aftershocks occur right after the main shock, typically, and the number of aftershocks will decay as we... with time after the main shock. So we can expect the largest aftershocks quickly afterwards and then decay. So we know all these statistical patterns. How fast are the plates moving? It's like some centimeters per year, per year, right? Is it fair to say that let's say you know in 10 years maybe it moves whatever 20 centimeters and to say that it takes this amount of energy that needs to be released in these 10 years. And then there's going to be this many earthquakes. That's exactly right. That's a fair statement. Okay. Yeah, so we can say that. That's exactly right. We can say that on the whole, we need to release this much energy. Yeah, yeah. And so that helps us to build these models, physics models, large scale. But we can't tell you today or tomorrow there will be a five or a six. Of course. So we can't divide out that total amount of energy into little pieces and say. But I even wonder, even if you have two rocks kind of sliding against each other, sometimes they may go smooth. That's right. And do the distance smooth, and sometimes they will get stuck, right? That's right. Yeah, actually, it's great. In the Bay Area, the Hayward Fault is notorious for what we call creeping, which means exactly what you said, Steve, that they're moving slowly. Maybe someone came in and put some oil in there. You can go to places in Fremont and in Hayward, and you can see the curb is offset. Wow. Yeah, and that's just where it slowly creeped by. Yeah, it's moving slowly. And in those places, we don't know exactly, but we wonder if that's accommodating some of the motion and maybe we'll have fewer violent earthquakes. So actually right now, today, in fact, there's a big study going on on the San Andreas Fault south of here, south of San Jose, where they cut a big hole across the fault and they're trying to see if there were violent earthquakes or if there were these slow moving earthquakes and is there a balance between these two. And so they're trying to look back into history and prehistory and understand the pattern of when these earthquakes occurred and how they occurred. So if it moves, let's say, 20 centimeters in 10 years, for example, right, those 20 centimeters on a highway or on a bridge or even on a sidewalk would be like this much of a gap. Do we see that? Yeah, we do see that. Yeah. So in places where it's slowly creeping, I mentioned Hayward, you can see these curves offset. Actually, the UC Berkeley Stadium. is built right on the Hayward Fault and you can imagine a stadium looks like this, right, goalpost to goalpost. They built it across the fault. They knew the fault was there. So when a fault is there, it actually makes a nice flat spot because it kind of shakes the earth a little bit and you get this kind of like bench a little bit. So they put the stadium there, they lined it up, fault goalpost to goalpost, they glued the top of the stadium together, but they allowed the bottom to rotate and you could unfortunately For seismologists, they rebuilt the stadium a few years ago. But until then, you could go up to the top behind the goalpost and look down, and you could see that the stadium had moved. Yeah, so we do see that. We also know that it can be dangerous to build on faults. And so I guess we stay away from building. So there is a zoning rule called Alquist Prieto or AP zones, and you're not supposed to build on faults. Now on the peninsula, a lot of things were built before we had these rules. So there are unfortunately houses that are built very close to faults. But right now we don't allow that building to occur anymore for exactly the reason that you said. We don't want your house to be torn apart or moved. 10 centimeters. 10 centimeters. Yeah, that's too much. Yeah. Yeah. Great. Annemarie, thank you so much for your time. Good. Really appreciate it. Thanks for keeping us safe. Thanks for having me. Keep doing so and let us know what's happening. Great. Thank you. Thank you for watching another episode of Bay Area Innovators. If you enjoyed this episode and you know someone who should be on our show, please send us an email at sfbayatepochtimes.com. We would love to hear from you. Thank you and see you next week.