 Hello everyone and welcome. Thanks a lot for joining us today. My name is Elizabeth Aida and I direct the Board on our Sciences and Resources and the Water Science and Technology Board at the National Academies of Sciences, Engineering and Medicine. I have the real pleasure of serving as your moderator for today's webinar. We're delighted that Dr. Richard Allen and Dr. Eric Fielding were available on rather short notice after these earthquakes to share their insights on these events that began actually almost precisely two weeks ago. Now before turning the microphone over to Richard, I wanted to review the agenda and how you can provide questions for either of our speakers and you'll see something on the screen momentarily to help you with that. So Richard and Eric will each speak for about 20 minutes, after which we'll share your questions with them for that final quarter of an hour or so of our webinar. So to ask a question, you can use the question and answer box which you can locate by hovering your mouse near the bottom of your screen and you should see a button there labeled Q&A. So you can then type in and send your questions there and we'd ask you to use the Q&A box and not the chat feature. And you can send those questions at any point during the webinar. Dr. Deb Glickson, who oversees the Committee on Seismology and Geodynamics, and I will be making note of your questions as they're entered throughout the hour and then we'll share them with Richard and Eric. And while we're really excited to have so many of you interested in this webinar, I'll ask you please to realize that we won't be able to get to all of your questions in that final quarter of an hour, but we'll certainly do the best we can. And should you have any technical issues, please use the Q&A feature or email Rami directly. You see his email address up there on the screen, archipeta at nas.edu. So with that said, let me introduce Richard and Eric and get us underway. Richard is the Director of the Berkeley Seismological Laboratory and Professor in the Department of Earth and Planetary Science at the University of California, Berkeley. He's focused significant effort in earthquake alerting systems, developing methodologies to detect earthquakes and issue warnings prior to shaking. And obviously we'll be hearing more about that today. Eric is a principal scientist at the Jet Propulsion Laboratory, the California Institute of Technology, and his research interests include active tectonics and fault interaction by the transfer of stresses with applications to earthquake assessment. So with that said, I'd like to turn the floor over to you now, Richard. So whenever you're ready, you can start sharing your screen and get underway. All right. Well, thank you, Elizabeth. You bring up my slides here. Can you see my slides now? Yes, we can. Perfect. Great. All right. Well, thank you, Elizabeth. So I'm going to focus, as Elizabeth just said, on the earthquake early warning aspect of the Ridgecrest earthquakes. Earthquake early warning for those who are not familiar, it's a relatively new technology, it's a relatively new approach to reducing hazard. And it's all about very rapidly detecting the beginnings of an earthquake, assessing the size of the earthquake, the scale of the earthquake, and then pushing out a warning to people before they feel the shaking. And we're talking about a few seconds to a few tens of seconds of warning. That's the goal of earthquake early warning systems. And we've been working here in the US, many of us have been working towards developing an early warning system that is now operational. It's called ShakeAlert. And so I'm going to focus in my 20 minutes on talking about how ShakeAlert performed during the course of this earthquake, the first real big test for the early warning system. And some of the lessons that we can learn and take away from this particular earthquake. Okay, so first of all, what exactly is the status of ShakeAlert right now and at the time of these earthquakes? ShakeAlert has been under development. By very large group of people, you can see some of the institutions involved in its development over the course of the last decade. The USGS is the lead agency and they have the statutory responsibility for delivering alerts. I mean, it's also in partnership with state agencies, in particular Cal OES has been very involved in building the infrastructure, supporting the infrastructure and for the system. And then of course the universities on the West Coast that run the seismic networks Berkeley, Caltech, Oregon and the University of Washington. And I also want to include the Gordon Betty Moore Foundation who made a significant investment in the development of ShakeAlert over the course of the last years. Okay, so it's been under development for a while, but the big switch was flipped back in October of 2018. And in October of 2018, it was officially declared ShakeAlert was open for business. It was generating alerts that were available for use. And what exactly does that mean? ShakeAlert is operating in California, Oregon and Washington, so the three West Coast states. It's available. The alerts are available for use by technical and industrial users. So think of these as being expert users who can receive the alerts. They can digest them. They have a real understanding and then they can use them to automate systems or take other kinds of responses. The alerts were only available to a limited cross section of the public. Specifically, they were available to the population of Los Angeles and that's through an app. And I'm going to talk quite a bit about the app that was available and to issue the alert across the county of Los Angeles. So that was the status of ShakeAlert at the time of these earthquakes. Okay, and so just one slide to introduce the earthquakes. Of course, there were two main events that were part of the Ridgecrest sequence. It really started with a magnitude 6.4 earthquake. Hopefully you can see my pointer. A 6.4 earthquake up here in the Mojave Desert in the morning of July the 4th. And then there was a magnitude 7.1 fairly close to the first event that came a day later at about 8.20 in the evening. And what matters here for early warning is the distance between the people the people were warning and the earthquakes. The population of course close to the epicenter of these earthquakes is pretty sparse, but the population is really focused in Los Angeles. And so that's why I've labeled this Los Angeles is about 200 kilometers away from these earthquakes. Of course, these are the two main events, but there have been thousands of aftershocks associated with them as well. This plot comes from Doug Gibbon and the USGS and just shows the time history of the events with the magnitude. And you see just as we expect the main events followed by aftershock sequences for this what's been a very, very active few days. Okay, so that's the earthquakes. So how did ShakeAlert perform? How did the early warning system perform in these events? So I'm going to show you a video now. I'm going to play a video. It's a replay of a tool that we call the user display. This is a desktop application that many people run to receive warnings from the ShakeAlert system. A typical use for this kind of application is in a kind of emergency operation center so people can actually see what's going on and see earthquakes occurring in real time. And what I'm showing you here is a replay of the warning that they had at LA City Hall. And so that's what the little house symbol here is showing that this is a warning for downtown Los Angeles at LA City Hall. So let's play what happened. So as soon as the earthquake is detected, you start to get the warning. And you can see the yellow and the red circles here. That shows the P wave and the S wave radiating out from the earthquake. In the lower left here, it's telling you the warning information for your location. This is a countdown, 30 seconds, still 30 seconds, until the S wave, so most of the strongest shaking, arrives in downtown LA. Then down here, this is the intensity expected in downtown LA, intensity 3. And this shows the estimated magnitude for the earthquake. So still we have 13 seconds, 12 seconds, until the S wave arrives in Los Angeles. So you probably noticed at the beginning, there was about 48 seconds of warning until the S wave arrived, right about now. And so this gives you a sense of the warning that Shake Alert provided, again, about 48 seconds of warning for downtown Los Angeles, with a shaking, expected shaking intensity of 3 for the downtown LA area. Okay, so just in case you didn't capture that, or as it flew by, here's the snapshots of the key pieces in time here. So this is the first alert that was pushed out, 48 seconds of warning. Initially, the magnitude estimate was 5.7, and so the shaking intensity in downtown LA was estimated to be intensity 2. About two seconds later, the estimated intensity goes up, we're up to intensity 3 at this point. It then remains at intensity 3 for the duration of the event. The estimated magnitude does gradually increase reaching a magnitude of 6.3, some tens of seconds later. So that was the alert for the 6.4. Now what actually was the ground shaking in the magnitude of 6.4? Well, this is the USGS Shake Map for this earthquake. So obviously the epicenter is up here, and the colors show strongly shaking closest to the epicenter. So further from the epicenter, of course, the shaking intensity decreases. This blue line here is the threshold for intensity 4. So down here in Los Angeles area is all in sort of intensity 3 region based on these contours. But I will point out, if you actually look at all of these triangles or stations, and so we have actual observations of what the shaking intensity was, many of them are intensity 3, of course, and a few intensity 4s down here in the downtown region as well. Okay, so that was the performance for the magnitude 6.4 earthquake. Now onto the magnitude 7.1 earthquake. Same thing, I'm going to show you the video for the alert in downtown Los Angeles. About the same initial warning, obviously 49 seconds in this case. Intensity very rapidly becomes intensity 3. The magnitude estimate creeps up with time, still 30 seconds until the warning arrives. So of course the idea behind warning is that you use this time to take protective measures. So right now everybody in front of their computers could be getting under the sturdy desk that they're probably sat in front of. So that by the time this S wave arrives, if you were in downtown Los Angeles, you could be in a safe location to prevent things falling on your head, things like that. Okay, so that's the same warning video for the magnitude 7.1. So then again, let's capture the key points here. So the first alert was 49 seconds before the S wave reached Los Angeles. Initially, magnitude 5.5 and an intensity of 2. After three seconds, the magnitude has increased and we're up to intensity 3. And the warning remains at intensity 3 for downtown LA for the duration of the event. The final magnitude estimate, it gets up to a magnitude 6.3. So frankly, that's not great. The actual magnitude is 7.1 and I'll come back to that in a few moments. So the alert intensity was 3. What was the observed shaking intensity in downtown LA? So again, the shake map. This is the bigger event, of course. And so the contours are all pushed further out. And this console here is the 3.5 console, meaning that this region, basically downtown Los Angeles actually experienced intensity 4 during the course of this earthquake. Okay, so that's the performance. We're obviously focused on downtown LA. And so the question comes, was this a success for Shake Alert? This was the first significant event. Unfortunately, all your microphones are turned off. So I'll tell you what I think. I would say that this was a success for Shake Alert. It was the first major event. It pushed out a warning. And I would say that the warning was pretty accurate to what was actually experienced. So that would be my take. Okay, but how was it received by people in Los Angeles? Well, they see it very differently. So here's a few newspaper articles. This one in the Los Angeles Times immediately after the magnitude 6.4, why LA's early warning system didn't send an alert before the magnitude 6.4. So LA's early warning system didn't push out alert. New York Times is after the second earthquake. California's alert apps didn't sound for the two big earthquakes. Why not? And then finally, as time went on, I think that we sort of shifted from the understanding being that the early warning system didn't work because people didn't get an alert to people actually understanding the nuances of the system that in fact, the system worked exactly as planned. And as this third article headline says, that's the problem. It worked exactly as planned. So why is that this disconnect? I would say I think that geophysics community seismology community would say that the system did very well, but clearly the population in LA is very dissatisfied with the system because they didn't get a warning. So why is that? Okay, so here's the reason. So as I mentioned at the beginning, the city of Los Angeles is the only region right now that is making the alerts public. And the reason for that is that the city of LA developed an app that's called ShakeAlert LA. It was developed by the city as you can see in 2019. And it was decided that the threshold for the alerts when this app would alert people is when the magnitude of the earthquake is greater than or equal to magnitude five and the intensity is greater than or equal to intensity four in the county of Los Angeles. So this is only this only works in the county of Los Angeles. And you have to satisfy those two conditions before an alert is pushed out. Okay, so let's compare that then to what actually happened with these two earthquakes. First of all, in the July 4th earthquake, as you saw a moment ago, the alert intensity in LA was intensity three. And so of course, no alert was pushed out to the app. The observed intensity was intensity three. And there were a few observations of intensity four. So did it work? Did it work as expected? I would say yes, this was correct. Mostly I say mostly just because there were some observations of intensity four. The most observations in LA for this earthquake were intensity three. And then for the July 5th magnitude 7.1 earthquake, the alert intensity was again intensity three and therefore no alert was issued. But the observed intensity was most definitely an intensity four. And so in that sense, it did not work correctly because it did not push out an alert, but there was shaking intensity four. But as you can see, it was pretty close. So the threshold is intensity four. The alert estimated alert intensity was three. So it didn't quite reach the threshold. And so no alert was pushed out. So that's why I say I think that the system worked pretty well. But what is very clear is that from a user perspective, the system completely failed. And there's no sugarcoating of that. Okay. So what are the key takeaways that we should take from this? What are the things that we can learn from this? I have sort of three key points. The first is that people want lower threshold alerts. I think that is loud and clear in terms of what we're hearing from people. They don't just want warnings for damaging earthquakes. They want warnings for experienced earthquakes. What I mean by that is that the reason that the thresholds were set at intensity four is because that's when you start to see damage. And so the thinking was that we should be pushing out warnings when people are going to see damage. And what is very clear that that is that people don't expect that. People expect to get a warning if they're going to experience the earthquake, i.e. they're going to experience, they're going to feel significant shaking. The other thing is it's clearly better to over alert than under alert. And I think that's part of a sort of generational change that is going on right now, that people would rather have more information than less information and then be able to decide what to do with it. The second one is that I think ShakeAlert performed very well. I'm quite pleased, I have to say, with the performance of ShakeAlert. In the case of the magnitude 6.4 earthquake, we got a very good magnitude estimate, a very good location and intensity estimate. So we did a really good job. In the case of the 7.1, I would characterize it as a disappointing magnitude estimate. It was an estimate of 6.3, so it was low by 0.8 magnitude units. The rule of thumb that I use in my head is that we expect the magnitude estimate to be within half a magnitude of the true magnitude. We were a little more than that, 0.8 off. However, as the earthquakes get bigger, estimating the magnitude just become more challenging. So this was disappointing, but I think we need to understand however much great work we do in the future to improve the magnitudes and I'm sure we will. Whatever system we design has to understand and has to recognize and the users have to recognize that we can have areas of this kind of size. The location was very good. The intensities were okay. It was off by one intensity unit and that was because of the magnitude estimate. And then finally, the third piece that I think is sort of largely forgotten but is really critical here is that we have to take note of the fact that the seismic networks, they performed really great. What I mean by that is that they provided data. They've been reporting on more than 15,000 earthquakes over the course of the last few days. All of the infrastructure worked as expected. We didn't have any failures to the system. Obviously the system generated shake alerts. It generated all of the other products, moment tensors, shape maps I showed you. All of these products were generated successfully. And then of course the websites and the social media outlets that we use to distribute this information have been used by millions of users. And so I think this is a huge, a great success that we shouldn't forget about for this particular earthquake. And then so in summary, the question of how we push out the alerts, we really didn't do a good job here and we clearly need to improve our understanding of the needs and the expectations of the users. That's a key lesson to be learned here. Shake alert I think did well, but clearly there's room for improvement and there are really many people working on what those improvements might look like. And finally the networks did great. I just want to mention our colleagues at Caltech and at the USGS in Pasadena who are of course operating the networks in Southern California. They've really done a fantastic job over the course, both building up to these earthquakes because the system then worked and then of course since these earthquakes occurred keeping the system operational and pushing out all of that information. And that's the point that I really want to end on. This is my last slide is that we're really in a very fortunate situation now where we have good robust geophysical networks up and down the West Coast. This is the network that is feeding data which is Shake Alert today in California, Oregon and Washington. Shake Alert started out with just the blue stations and then expanded of course up into Oregon and Washington and then has been adding stations in California, Oregon and Washington. There are currently 917 stations. The gray dots here adds in the stations that are planned. A total of 1675 stations are planned for Shake Alert. In fact the majority of these have now been funded in California. And so I think this is something to really take away from this that we now have these robust geophysical networks across the West Coast that are providing data both to reduce hazards in event sequences like this and to understand these earthquakes and also for the fundamental science that follows from events like this and of course can be making use of the data at all of the times. One thing to mention is that approximately one third of these stations are broadband stations and about two thirds of them are strong-motion stations. Okay, I'm going to stop there and I will pass over to Eric. Great, Richard. Thank you so much. And Eric, when you can share your screen and just you're good to go. Eric, I think you're muted. Hello. Yes. It told me the host had muted me so it wouldn't let me unmute. So I'm a geophysicist at the Jet Propulsion Laboratory. My name is Eric Fielding and the Jet Propulsion Laboratory is operated by the California Institute of Technology for NASA. So we are supported by NASA but we actually work for Caltech. I'm not going to go into detail for this. I'm more eye-candy picture on the first slide here but so I'm going to go back to the next slide. I'm going to be talking about the surface deformation aspect of this earthquake. This is a map of the long-term deformation that's measured by the GPS network that's operated by Yanavco. There's a whole lot of GPS stations in California and we can see that the Pacific plate is moving to the northwest relative to North America and there's this wide zone here where the arrows get shorter and shorter and that's showing that there's a distributed deformation across California. The yellow line on this map is the San Andreas Fault where the biggest part of the deformation is located but there's also a significant amount of deformation that's located further to the east primarily in what we call the eastern California shear zone which is the location of this earthquake. This is a map showing the earthquakes that have happened in Southern California in the last 150 years. The 1857 earthquake was on the main San Andreas Fault but since 1857 has not had a significant earthquake. The eastern California shear zone here in eastern California has had major earthquakes in 1872, 1992, the Landers earthquake and 1999, the Hectermine earthquake followed by the earthquakes this year. We also had a major magnitude 7.5 earthquake in South Bakersfield called the Kern County earthquake, 1952. In fact, since 1857, the major earthquakes in Southern California have been off the San Andreas Fault. One of the ways we can quickly get information about which fault moved in an earthquake, this is not real-time information. This is developed in the days afterwards. It's to use GPS stations. This is a map made by my colleague, Jing Kang Shen at UCLA, showing the deformation of the magnitude 6.4 earthquake. And we can see that stations moved up to, the stations closest to the fault moved up to about 8 centimeters, 80 millimeters. And the pattern of deformation shows us that the main rupture from the magnitude 6.4 earthquake was on this northeast Southwest trending fault. Then on July 5th, California time, it was actually July 6th in a universal time, there was a much larger magnitude 7.1 earthquake. And again, we have this map of the GPS displacements due to that earthquake. And we can see that the pattern of deformation is different because in the 7.1 earthquake, the fault ruptured on this northeast, a northwest Southeast trend, roughly perpendicular to the fault that ruptured in the 6.4 earthquake. That's also shown by the aftershocks that are shown on this map. So, the main data that I work with is what we call radar interferometry. It's a way of taking two radar images from before and after the earthquake. Because these earthquakes were only 33 hours apart, we didn't get a radar image in between the two earthquakes so we can only see the overall deformation from the two earthquakes together. This is a map showing the overall deformation from a satellite called Sentinel-1 that's owned by Copernicus, which is a European Union organization. It's operated by the European Space Agency. And we process this data over this entire area. You can see the LA is down here at the bottom. And we can see immediately that most of the deformation is on this northwest Southeast trend, which is the main rupture of the 7.1 earthquake. So we also process data from a Japanese satellite, the Japanese Aerospace Exploration Agency, ALOS-2 satellite, acquired the first image over this area on July 8th. So this is a map showing the deformation over just a small area concentrated on the ridge crest because of the way they acquired this data. And with aftershocks that were estimated by... These are aftershocks relocated by Professor Zachary Ross at Caltech. And again, we can see the main deformation on this northwest Southeast trend, but we can also see that there's a secondary trend of deformation where there's a sharp color change that extends towards the southwest closer to the city of Ridgecrest, which is this star. And that shorter fault rupture to the southwest is the one that ruptured in the 6.4 earthquake. And in fact, we're getting reports now that as they do a more detailed investigation that the damage in the city of Ridgecrest was actually more severe from the 6.4 than the 7.1 because the fault ruptured closer to the city than it did in the 7.1. We've also did a type of analysis that's called pixel tracking. This is a different way to take the difference between the two images and it's by doing cross-correlation between the radar images. And because it's working with the... It can actually extract two components of deformation. My colleague, Mong Hong Wong, at the University of Maryland was able to send me this file last night. It's not completely nice presentation because it's very hot off the presses, but he was able to combine data from two different radar tracks to make maps of the full three-dimensional deformation of the ground surface. So here on the left is the vertical deformation. You can see because this fault moves primarily sideways, the vertical deformation is small, except for this one location here where the ground surface moved down by more than a meter. And that is a place where the fault... There's two separate faults that moved and caused a block of the Earth to drop downward. It's what we call a pull apart. And the field geologists were actually able to map some of this vertical fault motion in the field, although this... almost all of the rupture from the 7.1 earthquake is inside a navy base. So the access is highly restricted and only government personnel so far have been able to look at the faults in the field. But they did see this vertical block motion. And now that we have this radar imagery, we can see what the limits of that block motion are. Then here in the middle, this is the east-west motion. This is showing the east-west component, which shows the slip on the magnitude 6.4 earthquake, primarily on the northeast southwest fault. And then over here on the right is the north-south component of motion, which shows the very large slip of the 7.1 earthquake was actually very close to the epicenter of the earthquake. We also have processed the GPS data over a longer time interval that includes the two earthquakes at JPL. And this is a map showing the overall deformation of the two earthquakes. This one station closest to the fault rupture moved over 65 centimeters. That's more than two feet between the two earthquakes, most of that deformation in the 7.1. And by taking this GPS information plus the radar maps, we're able to make an estimate of how much the fault moved and where it moved. So this is what we call a fault slip model. This is a very preliminary model that was actually made last week by a grad student at Caltech, Benjamin Edini. Just to give you an idea of what we use this, we can use this ground surface deformation to do by making a computer model of the elastic structure of the Earth plus with some kind of assumption of what the fault planes are. We can then estimate how much each part of the fault moved at depth just from the measurements made on the surface. So this will be updated much more with a future analysis. But it gives you an idea of what we use this first deformation for. Another thing we can do with the radar imagery is look at how much the surface changed in between the two radar images. We make these products that they're called damage proxy maps. It's a proxy for damage because we only know that something changed at the surface. We don't know what exactly changed. But the interferometry measurements actually include a measurement estimate of what we call interferometric coherence. And we can take the interferometric coherence that includes the earthquake and subtract out the interferometric coherence variations that were present before the earthquake and get a map that shows where the ground surface changed significantly in between the two earthquakes. So the red areas are the areas that changed strongly and we can see very clearly the long northwest southeast fault of the magnitude 7.1 earthquake and the shorter southwest northeast fault of the 6.4 earthquake. There's also a large amount of change over here in Searle's Lake. This is a large area next to the town of Trona. It was close to Searle's Lake. It's an area of active borax mining for over 150 years. And some of this material has slid out into the lake and they had caused a huge amount of damage to the small town of Trona that's next to Searle's Lake. And we can also see that in some places there's several different fault lines here. So it's clear that this fault, this earthquake rupture, just not only the main fault, but actually a number of other faults during the earthquake process. And we want to, we provided this information to the field crews that were able to go to the field and know where to look for the fault ruptures in the field. So in conclusion, we were able to use the surface deformation data to see that the magnitude 6.4 earthquake was on this northeast trending fault and with a left lateral motion where the far side moved to the left and the main rupture was on the northwest trending fault right lateral motion where the far side moves to the right. We see that complex pattern of fault ruptures and we can use the surface deformation with the radar satellites and the GPS data to constrain a model of how much the fault slipped at depth and then use that to estimate how much the stress has changed on nearby faults to see what change in probability of earthquakes would be on those other faults. In addition, we can use the radar data to map change in the ground surface which is often due to the fault breaking or damage to buildings and roads. And because I work for NASA, I will put in a plug for the NASA ISRO SAR mission. That's the Indian Space Research Organization cooperating with NASA to build a radar mission that a large part of that's being built here at JPL and we'll be ready for launch in January of 2022. So we'll be able to get more of this type of data available in the future. Thank you. Terrific. Thank you so much, Eric and Richard. And those were excellent presentations, very clear. And I think you reached a lot of different members of our audience with that. We did get quite a large number of questions and so we'd like to spend some time here sharing those with you just to review if you haven't had a chance to send questions that the instructions are up there and they'll remain up there throughout this last sort of a quarter of an hour that we'll be talking with Richard and Eric. So just use the Q&A box if you'd like to send us a question. We'll keep tracking those as they come in. So Richard, I think I'm going to turn the first question over to you and that is related to the plans for delivering public alerts across the Shake Alert region. We actually had a lot of folks from different parts of the entire region asking, well, what about San Francisco? What about San Bernardino County and so forth? So I wanted to hear a little bit more about that delivery of the public alerts. Sure. Yeah, I mean, that's really the key question and we've known that. We didn't need to have this earthquake to know that was the issue. It's the last mile problem. Now that Shake Alert is generating these alerts, how do we push them out and get them out to everybody who wants them? And there are a lot of people working on this. As I mentioned, it's the Shake Alert app already for the city of LA. There are various other apps that are under development. Actually, before I talk about the apps, there's also an effort to use WEA. WEA is the wireless emergency alert system. That's the system that pushes out presidential alerts and amber alerts. The challenge with that system is how much time it takes. In fact, that's the challenge with many of these systems is how do we ensure that the system that we're using pushes out the alert in a second or two, which is what we really need in order to make use of it. So WEA system testing underway in Oakland and testing in San Diego seems to suggest a few seconds, seven seconds kind of delays in pushing the alerts out through WEA, but there's the hope to use WEA later this year. There are several groups who are developing apps as well. In fact, I'm going to share my screen again very briefly just to show you one of the ones that we're actually working on. So there are several apps with several groups. At Berkeley, we've also been working on using apps to deliver alerts with the My Shake project. And so this project actually started out. In fact, it still does. It's intended to do two things. It's intended to use cell phones to detect earthquakes and also, of course, you can use cell phones to then deliver alerts. And so there's a new version of My Shake out actually right now. It's available for Android and for iPhones. And it's intended to give people what they need before, during, and after an earthquake. So there's safety information, there's preparedness information built into the app. There's also, you can look at recent earthquakes. As you can see the example here on the screen, recent earthquakes. People can then use the app to report their experiences in an earthquake. This is a brand new app. It was actually 25% rolled out at the time of the Bridge Crest earthquakes. It's now 100% rolled out. In the center here, you're actually seeing the reported damage across the San Francisco Bay Area in a magnitude 4.3 earthquake that happened just a few days ago. And so people after the earthquake can report the shaking, the shaking intensity, very much like did you feel it? Many people reported, did you feel it? Which is great. In addition to that, they can report damage. And that's what you're seeing. All of those blue hexagons are showing that there was no reported damage, given it was only a magnitude 4.3. But then to get to your question, Elizabeth, again, we're also testing the delivery of shake alerts using the My Shake app. So the public version of the app does not receive the alerts to be very clear. So you can download the app for free in the store today, but you will not receive shake alerts today. But we're testing that. And this is a project funded by the California Office of Emergency Services. And they're supporting this to really figure out if we can use this to deliver alerts across the entire region, across California. And so it's in testing right now, but there's a hope that later this year we'll be in a position to be able to start delivering alerts across California and perhaps beyond that. So to all of the regions that you just mentioned, Elizabeth. That's terrific, Richard. Thanks so much. There was a lot of, as you pointed out too, interesting from the standpoint of the public and being able to have lower levels of alerts. And in that sense, is there any plan along the way to allow users to pick their own thresholds for alerts that may be more complicated? Yeah, no, absolutely there is. So it's sort of been interesting. This has swung backwards and forwards. The original versions of the, we had an early version of the app that was used to demonstrate shake alert capabilities. And it actually showed you the shaking intensity, estimated shaking intensity at your location and also a countdown until the time that the shaking started. And we got feedback on this and talking to people in Mexico and getting the early warnings in Mexico and in Japan. And the conclusion was that that was too much information and instead we needed to have something very simple which was just saying earthquake, drop cover and hold on expect shaking. And so that's what we've moved to. That's what the ShakeAlert LA app does. That's what the MyShake app is currently doing in testing as well. I think one of the key takeaways from this earthquake is that that's not sufficient. There's actually a requirement for more information. There's a lot of discussion going on and of course there's a lot of work to be done to really gather input and lessons learned from this earthquake. But we seem to be going in a direction where maybe there are two levels of alert. So when the earthquake is significant, grade maybe greater than magnitude five for example, then you push out an alert and there'll be two levels of alert. There'd been alert to the region where people are going to feel shaking but we're not really expecting damage. So something like intensity three, four kind of range. And then there would be a much more urgent alert, urgent sounding alert to expect strong shaking. And I think that this would actually address a lot of the concerns that we're seeing in terms of people's feedback. So this is an interesting question. This is a social science question. Really a seismology question. But what we need to do is effectively couple that social science with the seismology and our understanding of the physics of the process and the uncertainties in our ability to predict shaking intensity in order to do the best job we can and then make sure that we're delivering the product that the users need. That's very important, Richard. I will hardly agree with that. Thanks a lot. So Eric, there's a question for you. There were some questions associated with the proxy damage maps. And one was more specific and related to the source of the significant proxy damage on the east side of the slide, I think that you showed, and then related to the proxy damage maps themselves and how they may be able to help emergency managers and responders in that case. So I think generally talking a little bit more about the proxy damage maps and what they can and can't do. Yes, we think that the damage in the Searles Lake and Trona area was due to a liquefaction and lateral spreading of material. It's not a direct fault rupture. It's a secondary non-tectonic deformation, but clearly doing significant damage to, I think there's quite a few people that are no longer able to live in their houses in Trona. The damage proxy maps, we have been producing these for earthquakes and for other types of disasters including hurricanes and floods. So we do get these two emergency responders as soon as we can. We have to wait for the satellite to come back over the area and collect the second satellite image. And that's one of the reasons why we're working on getting access to more radar satellite systems. And there's some commercial companies these days that are working on launching large constellations of satellites to possibly get data very quickly after earthquakes. Great. Thank you very much, Eric. That's helpful. So actually, this question could go to both of you, but I think, Richard, it came up in context of your presentation, and that's just simply clarifying for folks the difference between the magnitude and the Richter scale and the intensity and the Mercale scales, if you could. Sure. Yeah, this is actually a real challenge I think we face as a community is that there is a huge amount of confusion by people using early warning about the difference between magnitude and intensity. So magnitude, for a given earthquake, there is one magnitude. It was a magnitude 7.1 earthquake. And so that magnitude describes the size of the earthquake. Intensity, there are many values of intensity. Intensity is about how strong the shaking is at a given location. And of course, the shaking intensity is strongest closest to the epicenter of the earthquake. So when you write out the epicenter, that's where you have the strongest shaking and there will be the highest intensity for that earthquake. And as you get further and further away from the earthquake, the intensity decreases. And so when it comes to thinking about the effect of an earthquake on you as an individual, on me as an individual, what matters to me is the intensity of shaking at my location. And that is a function of both the magnitude and the distance it was from that particular location. There's actually a third piece. The third most important factor in intensity is the site conditions. What the soils, whether you're on soft soils in a deep basin or on hard rock. So for example, in the Los Angeles basin, there are shaking intensity increases compared to the shaking intensity in the adjacent mountains just because of an amplification effect. So the intensity is what's happening, the shaking at your location and the magnitude is just a single number that describes the size of the earthquake. Great. And Eric, I'm actually going to ask you to follow up on that if you could, because there's some relationship obviously between the behavior of the materials and the deformation you see. So could you add on to what Richard shared there, please, with regard to your work? Yes. When we make the satellite measurements, we are only able to see the overall deformation, which is related to the size of the earthquake source. We then have to use some type of model to estimate what the intensity is going to be at a given location. But we do also have this second type of measurement, the damage proxy map, which actually measures how much the ground surface changed in between, before the earthquake and after the earthquake. So in that case, if the shaking is intense enough, then it causes a disruption of the ground surface that we can detect with the radar. Thank you very much, Eric. I'm going to follow up on it with a slightly different track for you. And obviously, Richard, feel free to add on to this as well. So for Eric, did the stress from the first ridge crest earthquake increase the probability of the second large earthquake or also speaking more towards the interaction between those two different earthquakes? What was the relationship there? Do we know that? Do we have enough information? We're actively studying that subject. We haven't come to a final conclusion, but there's no question that the two earthquakes are closely related. They're very close together in both space and time. The details we're still trying to work out. And one of the issues is being able to separate the ground deformation that was due to the 6.4 from the much larger 6.7.1 and understand what happened in between the two earthquakes. Yeah, if I can add to that, I mean, it's really interesting. If you watch an animation of the time history of the events, these are two, you know, clearly two earthquakes that were on two perpendicular fault planes, but clearly were related to one another. The first event, the smaller event, the magnitude 6.4, it's very difficult to describe this without pointing at a figure, but it has a strike that points to the northeast. But then if you look at the pattern of aftershocks, you can kind of see how it actually starts to see a cluster of aftershocks that is sort of close to the location of what then became the magnitude 7.1 epicentre as well. So the point being simply that there is clearly a very interesting and complex relationship between the earthquakes as you go from one to the next, which of course is all determined by this transfer of stress that you're asking about Elizabeth. And, you know, so this is a complex rupture. It wasn't just sequential earthquakes on the same fault plane. It was a complex rupture. And this isn't the first time we've seen this. We've now seen several of these complex fault ruptures where the seismicity seemed to be jumping from one fault plane to another fault plane. I mean, it just shows us, you know, as we see events like this, that these processes are much more complicated than we really, I think, have imagined in the past. And what's different now, again, it comes back to sort of one of the points I wanted to make, that we have this fantastic observational infrastructure today that is the reason that we can see all of this complexity that we were not aware of in the past. I think I'm going to add that there was a similar sequence of earthquakes in 1987 called the Superstition Hills where a smaller earthquake happened on one fault and then a larger earthquake happened on a nearly perpendicular fault. So this is something that seems to happen in California that's maybe not always happening in other parts of the world, but something we have to be ready for. In fact, one of my colleagues at the USGS who went immediately to Ridgecrest after the 6.4 had told the city of Ridgecrest that there was a significant chance of a second larger earthquake before the 7.1 hit. So they were at least partially warned. I'm going to follow up. There's a question a lot of people have been asking, not just here, but on the news. And maybe both of you can respond to this. And that's with regard to the number of aftershocks. And some folks indicate I saw the number 35,000, maybe aftershocks. And of course there are different magnitudes. Some people have said, is this unusual to have this number of aftershocks? Would both of you care to respond to that? Eric, do you want to go first? Well, one of the reasons that we are actually seeing more aftershocks is that the seismic networks have gotten so much more sensitive than they used to be. So we're able to locate many more aftershocks at lower magnitudes than they were able to locate in the past. So even the same earthquake sequence today has more aftershocks just because we can see the smaller ones. And there's always many more smaller ones than bigger ones. Yeah, I have to say the reason I was happy for Eric to go first is I don't have a good answer for you, Elizabeth. It's not clear to me whether we are seeing an unusual number of aftershocks. Of course, you know, the norm. So the norm that we would expect is for one magnitude seven that we would have one magnitude six aftershock, 10 magnitude five, et cetera, et cetera. And it's not clear to me that we've necessarily deviated that much from that history. But I'm not an expert on this and I haven't been watching closely, so I'll stop talking. Great, that's super. Thanks. I'm seeing that we're nearing the end of the hour. I'm going to ask each of you probably a short question, a short remaining question, and then we'll do a wrap up. And so this one to you, Eric, and this was a more specific question, wondering if you looked, or did JPL look for deformation south of the garlic fault? Yes, we've analyzed the satellite images for the whole region. We did see what appears to be some triggered slip on a small section of the garlic fault by perhaps a centimeter or so. And I believe the field geologists have confirmed that there was a little amount of triggered slip on the garlic fault itself. But we haven't seen any significant deformation south of the garlic, at least so far. Thank you. And Richard, then final question to you. And it's actually, we bundled two questions together. So the question began, how would the system have worked? The shakler system have worked. If the epicenter had been closer to LA, and in combination with that, why is getting a warning even of 30 seconds or less an important thing? Sure. So yeah, so first of all, the warning time. So the warning time is a function of how far you are from the earthquake. So that's why I kind of highlighted at the beginning that the city of LA is about 200 kilometers from the epicenters of these earthquakes. And so that's why you got the 48 seconds worth of warning. If the earthquake was closer, then you would have less warning time. There's no question about that. As an example, that magnitude 4.3 earthquake that I mentioned in the Bay Area a few days ago, that was just east of the Bay Area, close to Concord. And so that, of course, it's only a magnitude 4.3, but the shakler, of course, detected it and sent out a warning. And in fact, my colleagues at Berkeley got about four seconds of warning. And so that's an earthquake that was essentially on the margins of the Bay Area to the east of the Bay Area. Still Berkeley got about four seconds of warning. The city of San Francisco would get perhaps an additional two seconds of warning over that. So that's a nice example that, yes, even when the earthquakes are close by, of course, you get less warning time. But you still get a few seconds of warning. It's still enough time perhaps to get under a desk. If not, it's enough time to brace yourself. And this is one of the things that have come out of social science studies in Japan, that one of the real values of early warning in Japan is just being able to brace yourself and being ready for that shaking. Now, I've forgotten the second half of the question. Sorry, can you remind me? No, that's all right. How would the system have worked if the epicenter had been closer to LA? Yeah, so the system, I mean, it would just mean there was less warning time. That was the only difference, is that you would have less warning time. Okay, thanks for that. I'm going to give each of you 10 or 15 seconds to give a final takeaway to the audience. And then I want to do a quick wrap up and provide some instructions about the audio recording from today's webinar. So, Richard, how about you first and then Eric, your final takeaway for our audience here? Well, my favorite fact about earthquakes in California remains true. And that is that we still haven't had an earthquake in California that wasn't a surprise. And I mean, I think that's a really kind of shocking thing to realize is that all of the major earthquakes that we've had have not been on major known faults. That was true of the Napa earthquake, the magnitude 6 earthquake up in the Bay Area a few years ago, and that's true of this earthquake. And I think it's good for us, it's humbling, and I think it's important for us to remember that, that we haven't had a surprising event and so we obviously have to remain vigilant. Eric? Yes, I agree that the faults that have ruptured in earthquakes in the last 100 years in California except for the Parkfield earthquake of 2014, which was predicted but was only magnitude 6 and didn't cause any damage have all been on faults that weren't mapped in vans. So, it's one of the reasons we go out and collect data of various types both in the field and with satellites to better understand where deformation is occurring over long time intervals and where earthquakes are more likely to happen in the long run. Okay, so we're at the end of our time here. I wanted to thank Richard and Eric for their excellent presentations and willingness to take your questions and to the members of the audience who have great questions. We apologize that we weren't able to get to all of those. So, the last slide is up here for those of you in the audience. If you have any questions about our committee on seismology and geodynamics or board on our sciences and resources you can feel free to reach out to us here on the staff. You have our emails there. I think the most important element that you're interested in is that we'll be sharing the presentations and audio recordings from today's webinar that will be posted within 7 to 10 days. And you should please watch your email for announcements about when that is posted and also for future webinars and events. Thanks again so much Richard and Eric and thanks to our audience. Goodbye, everyone.