WEBVTT Kind: captions Language: en-US 00:00:01.740 --> 00:00:06.640 Okay. So, a few notes. The history of the lightning talks. 00:00:06.640 --> 00:00:09.691 About five years ago, Jack Boatwright started this tradition. 00:00:09.691 --> 00:00:13.490 And so the themes of the lightning talks were, the first year, lightning. 00:00:13.490 --> 00:00:17.520 Followed by scientists, mad scientists, and superheroes. 00:00:17.520 --> 00:00:20.947 Followed by Jack’s choice of nothing but cute baby goats. 00:00:20.947 --> 00:00:25.712 And then, last year, in memory of Jack, we did baseball. 00:00:25.712 --> 00:00:28.650 The theme of this year’s lightning talks was chosen by Sara McBride. 00:00:28.650 --> 00:00:30.039 It will be marine mammals. 00:00:30.039 --> 00:00:34.898 I think they ended up being all seals except for one set of dolphins. 00:00:34.898 --> 00:00:39.630 And then I believe – was it Jessie has already requested pandas for next year? 00:00:39.630 --> 00:00:43.477 But, if people have requests for the year thereafter, please send them to me. 00:00:43.477 --> 00:00:46.290 I always need suggestions. 00:00:46.290 --> 00:00:49.167 Okay. So, lightning talks. Every talk will be one slide, 00:00:49.167 --> 00:00:52.594 three minutes long. This is your microphone. 00:00:52.594 --> 00:00:54.950 Please be sure that it is up here where your mouth is 00:00:54.950 --> 00:00:57.850 when you come to give your talk. 00:00:57.850 --> 00:01:01.900 So the names of the presenters are in the program. 00:01:01.900 --> 00:01:05.220 They will present in alphabetical order so people who are giving lightning talks 00:01:05.220 --> 00:01:08.180 can see where they are in the lineup. And you’re going to pay attention 00:01:08.180 --> 00:01:11.659 to that because it is very important that you know when you’re supposed 00:01:11.659 --> 00:01:14.675 to be here because your talk might start without you. 00:01:15.625 --> 00:01:18.353 And the way the talks go is something like this. 00:01:18.353 --> 00:01:23.539 First, we have a title slide that gives the name of the person and who they are. 00:01:23.539 --> 00:01:25.329 Then they give a talk. 00:01:25.329 --> 00:01:28.499 And, at the bottom of their talk will be a slider that goes from left to right. 00:01:28.499 --> 00:01:30.310 It will take three minutes to go from left to right. 00:01:30.310 --> 00:01:34.640 In fact, I got this template from Ken Hudnut. Thank you very much. 00:01:34.640 --> 00:01:39.789 And then, when the slider gets to the full end of the slide, in three minutes, 00:01:39.789 --> 00:01:44.013 your slide goes away. This will be your hint to leave. 00:01:44.013 --> 00:01:45.937 [laughter] 00:01:45.937 --> 00:01:48.549 Then the next speaker gets announced, at which point, 00:01:48.549 --> 00:01:50.590 that speaker should already be running up here. 00:01:50.590 --> 00:01:53.530 Because, again, their talk may start without them. 00:01:53.530 --> 00:01:56.739 Okay. 00:01:56.739 --> 00:01:58.460 Are we ready? 00:01:59.882 --> 00:02:02.159 Here we go. 00:02:04.230 --> 00:02:06.412 [laughter] 00:02:06.437 --> 00:02:10.190 - Okay. I’m going to talk about strainmeter or strain data. 00:02:10.190 --> 00:02:14.440 And I wanted to point out, on the heels of that last excellent session, that, 00:02:14.440 --> 00:02:19.310 for the Ridgecrest sequence, there were strainmeters well within 00:02:19.310 --> 00:02:26.530 rupture length of the sequence, as you can see on the map on the top left. 00:02:26.530 --> 00:02:29.880 And you can see the waveforms immediately adjacent to those that – 00:02:29.880 --> 00:02:36.450 at least this one station, B921, these instruments are linearly 00:02:36.450 --> 00:02:42.501 sensitive to all frequencies from static to the highest frequency 00:02:42.501 --> 00:02:45.691 you can sample without aliasing. 00:02:45.691 --> 00:02:48.960 And that’s reflected, I think, in the waveforms there. 00:02:48.960 --> 00:02:54.320 And there’s a lot of complexity in there, which brings us to the lower part of 00:02:54.320 --> 00:03:04.040 that slide, which is a snapshot of the wavefield that Evan Hirakawa did 00:03:04.040 --> 00:03:07.330 at a couple different times. And he used these strainmeter data 00:03:07.330 --> 00:03:14.750 as part of the constraint on what the kinematics of this rupture were. 00:03:14.750 --> 00:03:19.430 And then, to bring it to the Bay Area, I just wanted to show – 00:03:19.430 --> 00:03:24.699 just keep notice of the map on the top – well, up there. 00:03:24.699 --> 00:03:29.230 All those little dots, those are strainmeters that are actively running 00:03:29.230 --> 00:03:32.890 at the moment in the Bay Area. So you can see that we have really 00:03:32.890 --> 00:03:39.240 good spatial coverage in terms of capturing potentially a big event 00:03:39.240 --> 00:03:43.580 on the Hayward Fault or the Calaveras or whatever you want to look at. 00:03:43.580 --> 00:03:49.723 And so I’m – can we stop early? - Sure. 00:03:49.723 --> 00:03:51.961 - Okay. [laughter] 00:03:52.096 --> 00:03:54.015 - [inaudible] - Yeah. 00:03:54.799 --> 00:03:57.340 [laughter] 00:03:57.365 --> 00:04:01.367 [applause] 00:04:04.748 --> 00:04:08.361 - Hello. I’m Phil Beilin. I have a GIS background – 00:04:08.361 --> 00:04:10.980 oh, I have a GIS background. And I started here at the USGS 00:04:10.980 --> 00:04:14.860 in landslides, and I’ve used GIS throughout my career. 00:04:14.860 --> 00:04:17.628 I’ve moved into public safety over my career. 00:04:17.628 --> 00:04:21.539 I’ve worked on data sharing, especially for public safety. 00:04:21.539 --> 00:04:26.610 And this slide in my presentation to you is meant to encourage the scientists 00:04:26.610 --> 00:04:31.659 and engineers of you to reach out to – they’re not really consortiums, but there 00:04:31.659 --> 00:04:35.159 are groups of decision-makers for public safety, such as the fire chiefs, 00:04:35.159 --> 00:04:41.389 the police chiefs, the EOC managers, in which they rarely get a – I’ll suggest 00:04:41.389 --> 00:04:45.829 to you – a presentation focused for them to understand, in regards to 00:04:45.829 --> 00:04:50.190 earthquakes, aftershocks, liquefaction, landslides that might be created 00:04:50.190 --> 00:04:52.977 by earthquakes, what does it mean to them? 00:04:52.977 --> 00:04:58.379 If you had to take all the science and distill it down to, in terms of the actions 00:04:58.379 --> 00:05:02.533 that they need to be thinking about when they’re deploying their people, 00:05:02.533 --> 00:05:06.349 that they don’t have enough information about – in a language 00:05:06.349 --> 00:05:08.990 that they have understood. And I’m saying this because, 00:05:08.990 --> 00:05:13.800 in Bay Area, we have tried to put on workshops, participate in state- and 00:05:13.800 --> 00:05:18.069 national-level earthquake exercises, bringing together the fire chiefs and 00:05:18.069 --> 00:05:21.160 police chiefs, the EOC managers, and get them familiar with the 00:05:21.160 --> 00:05:24.988 kinds of data that are going to be created during an earthquake 00:05:24.988 --> 00:05:29.069 and that are going to be derived from earthquake damage. 00:05:29.069 --> 00:05:32.729 And then, how do they use it for where they’re responding to people. 00:05:32.729 --> 00:05:36.220 They really don’t understand well the idea that there’s going to be aftershocks 00:05:36.220 --> 00:05:39.849 and what that might mean to them. They really would love to hear from 00:05:39.849 --> 00:05:44.490 geologists and engineers that, hey, just pay attention to, if you’re going to 00:05:44.490 --> 00:05:50.240 be staging people, resources, don’t do that under a building 00:05:50.240 --> 00:05:54.159 that’s seismically unsafe. And, in fact, you know what buildings 00:05:54.159 --> 00:06:00.779 in your city might represent those pre-1980 buildings that you should 00:06:00.779 --> 00:06:04.460 probably know which ones of those are unreinforced masonry and which aren’t. 00:06:04.460 --> 00:06:08.017 Which one has hazardous materials? I’ll suggest to you, in the East Bay, 00:06:08.017 --> 00:06:14.206 along San Pablo Boulevard, we have a lot of 1900, 1907, 1913 brick, 00:06:14.206 --> 00:06:17.433 unreinforced masonry garages. And what do garages have? 00:06:17.433 --> 00:06:19.310 Of course, petrochemical kind of things. 00:06:19.310 --> 00:06:22.169 And they’re listed in hazardous material sites. 00:06:22.169 --> 00:06:26.990 But having somebody bring that information together and explain to that 00:06:26.990 --> 00:06:31.559 is part of what I wanted to address. I also wanted to mention that the 00:06:31.559 --> 00:06:35.520 National Guard has exercises regularly, and they would love to have 00:06:35.520 --> 00:06:40.039 participation we have from both utilities as well as some of the scientists to 00:06:40.039 --> 00:06:43.319 sit in their JOC – their Joint Operations Center – and see 00:06:43.319 --> 00:06:47.939 how they use your information. What questions they ask – they’re trying 00:06:47.939 --> 00:06:53.080 to make decisions about deploying resources and where to deploy them, 00:06:53.080 --> 00:06:57.800 and they need to learn more from you folks about what it is that you’re 00:06:57.800 --> 00:07:01.539 going to be creating for them and how they should use it for their 00:07:01.539 --> 00:07:04.310 purposes and for decision-making and supporting the state of California. 00:07:04.310 --> 00:07:07.612 Because that’s what the National Guard does is support the government. 00:07:07.612 --> 00:07:09.367 Period. Thank you. 00:07:09.367 --> 00:07:12.800 [applause] 00:07:16.891 --> 00:07:22.999 - So this busy slide is really best done in animation, but I’ll walk you – 00:07:22.999 --> 00:07:27.550 so this – so the entire text here on the left just makes the point 00:07:27.550 --> 00:07:33.139 that observational seismology is essentially exclusively, 00:07:33.139 --> 00:07:37.379 or almost exclusively, done in the – in the far-field. 00:07:37.379 --> 00:07:41.300 Where all the detailed information of what is happening within 00:07:41.300 --> 00:07:44.800 rupture zone is lost. It’s never been good quality. 00:07:44.800 --> 00:07:49.629 So the situation in earthquake physics is similar to where chemists 00:07:49.629 --> 00:07:54.264 were before scientists started looking inside the atom. 00:07:54.264 --> 00:07:57.980 We are doing far-field seismology. Chemists were doing – in the 18th 00:07:57.980 --> 00:08:01.221 and 19th centuries, they were doing far-field chemistry. 00:08:01.221 --> 00:08:04.460 Essentially just based on the concept of atoms. 00:08:04.460 --> 00:08:09.445 Of course, once people looked inside the atoms, there were – that led to 00:08:09.445 --> 00:08:14.369 development of a whole new theory – quantum mechanics – that led to far 00:08:14.369 --> 00:08:18.739 greater predictability and quantitative framework for chemistry. 00:08:18.739 --> 00:08:24.106 Similarly, I think that when we replace the far-field concept of just fault or 00:08:24.106 --> 00:08:29.809 surfaces with fault zones and volumes, and we actually get detailed information 00:08:29.809 --> 00:08:33.813 of what is happening inside these volumes during dynamic rupture, 00:08:33.813 --> 00:08:41.096 we will get more quantitative and far more predictive theory. 00:08:41.096 --> 00:08:47.161 So, to do that, I think time has come to actually instrument the entire plate 00:08:47.161 --> 00:08:52.920 boundary with dense fault zone arrays. This concept is illustrated for southern 00:08:52.920 --> 00:08:57.190 California, but it applies globally. Bay Area, but actually all faults. 00:08:57.190 --> 00:09:00.555 So there’s certain type of information that has never been recorded. 00:09:00.555 --> 00:09:04.560 As I said, we never actually observe dynamic rupture 00:09:04.560 --> 00:09:09.209 in the near field within the rupture itself. 00:09:09.209 --> 00:09:14.970 Now, the faults in southern California and in the Bay Area are all overdue. 00:09:14.970 --> 00:09:18.959 So this is a very good time – this is something that we should be sort of 00:09:18.959 --> 00:09:22.539 doing – thinking into the future. This is a good time to start 00:09:22.539 --> 00:09:27.509 instrumenting the entire hazardous plate boundary faults. 00:09:27.509 --> 00:09:30.570 And the basic instrumentation – this is all dense near-field. 00:09:30.570 --> 00:09:36.585 Each line here is an array of seismometers and high-rate GPS 00:09:36.585 --> 00:09:41.800 stations and cameras also that are just sitting here with densities, 00:09:41.800 --> 00:09:48.183 are bracketing main potential ruptures, 100 meter on each side, and 350 meters 00:09:48.183 --> 00:09:52.860 on each side, and then 500 meters. So there aren’t too many instruments 00:09:52.860 --> 00:09:59.440 here. This is a – sort of a concept. But this – and they’re spaced 00:09:59.440 --> 00:10:02.120 about tens of kilometers apart. 00:10:02.120 --> 00:10:04.949 If we capture – if we instrument now the plate boundary, we should 00:10:04.949 --> 00:10:09.423 be able to capture, for the first time ever, dynamic rupture 00:10:09.423 --> 00:10:14.990 within the rupture zone itself. In the meantime, we will get 00:10:14.990 --> 00:10:18.430 tons of information about fault zones. 00:10:18.430 --> 00:10:22.333 So the instrument are not just waiting there. Thank you. 00:10:22.333 --> 00:10:25.771 [applause] 00:10:28.407 --> 00:10:33.172 [silence] 00:10:33.172 --> 00:10:37.649 - Okay. This is not really a science talk, but I just wanted to present a new 00:10:37.649 --> 00:10:40.298 capability we have in the Earthquake Science Center. 00:10:40.298 --> 00:10:43.860 We have three-component nodal seismographs. 00:10:43.860 --> 00:10:48.509 And, those of you who are not familiar with them, they are basically small 00:10:48.509 --> 00:10:54.720 self-contained units, three-component. They have a GPS sensor on board 00:10:54.720 --> 00:10:58.513 and a battery. So you can put these things out for 35, 40 days, 00:10:58.513 --> 00:11:03.670 let them record, and the frequency range is pretty good on it. 00:11:03.670 --> 00:11:07.456 You can go down to about 4 to 5 seconds all the way up to 1,000 hertz. 00:11:07.456 --> 00:11:10.100 So you can use them for a lot of different types of seismic 00:11:10.100 --> 00:11:13.500 investigations. And I think, for northern California, this is 00:11:13.500 --> 00:11:19.809 going to be really – a really nice tool that we can do for a lot of different 00:11:19.809 --> 00:11:22.990 investigations that we couldn’t previously do. 00:11:22.990 --> 00:11:26.767 We currently have 268 of them. We hope to have about 300 00:11:26.767 --> 00:11:32.439 by the summer, and then we can put out the large arrays. 00:11:32.439 --> 00:11:37.230 And these are some examples of what we’ve done so far with them. 00:11:37.230 --> 00:11:42.130 We’ve done some site characterization studies in southern California. 00:11:42.130 --> 00:11:46.836 In the upper right, it shows the Ridgecrest sequence, where, 00:11:46.836 --> 00:11:53.269 in just a day, we were able to cover the whole sequence – a 50-by-60-kilometer 00:11:53.269 --> 00:12:00.069 array put out over the major seismicity and recorded that for a couple months. 00:12:00.069 --> 00:12:04.399 We can also – we also did it – you can’t see very well across the 00:12:04.399 --> 00:12:08.470 step-over on the Garlock Fault, where we put out an even denser array. 00:12:08.470 --> 00:12:12.061 So you can do many different types of studies – earthquake monitoring. 00:12:12.061 --> 00:12:17.709 You can do the site characterization. Reflection, refraction, ambient noise. 00:12:17.709 --> 00:12:20.759 Tomography – just a lot of different things you can do with these that 00:12:20.759 --> 00:12:24.600 we couldn’t previously do because you can put out such a large number 00:12:24.600 --> 00:12:29.570 of instruments. Where, typically, you might put out, 00:12:29.570 --> 00:12:33.696 you know, one or two broadband instruments in a day, with these, 00:12:33.696 --> 00:12:37.209 you can put out hundreds. So you can really blanket the area 00:12:37.209 --> 00:12:39.160 and see what you’re looking for. 00:12:39.160 --> 00:12:44.889 So these instruments will be available to any of the scientists 00:12:44.889 --> 00:12:48.459 in earthquake hazards or anybody collaborating with them. 00:12:48.459 --> 00:12:52.569 So if anybody wants to use them, they’ll be there. Thank you. 00:12:52.569 --> 00:12:56.567 [applause] 00:12:59.860 --> 00:13:05.423 [silence] 00:13:05.423 --> 00:13:08.730 - Good afternoon. Good afternoon, everyone. 00:13:08.730 --> 00:13:10.250 Thank you for being here. 00:13:10.250 --> 00:13:14.120 So the main focus of my research is investigating the usefulness 00:13:14.120 --> 00:13:17.959 of incorporating strain data in earthquake early warning systems 00:13:17.959 --> 00:13:21.360 in general and ShakeAlert system in particular. 00:13:21.360 --> 00:13:24.830 In this work, we used thousands of earthquake records from borehole 00:13:24.830 --> 00:13:28.630 strainmeters deployed along the West Coast of the United States 00:13:28.630 --> 00:13:33.572 to find a relationship – empirical relationship between raw, 00:13:33.572 --> 00:13:38.220 uncalibrated peak dynamic strains and earthquake moment magnitudes, 00:13:38.220 --> 00:13:40.690 which is shown here on your left. 00:13:40.690 --> 00:13:44.220 As a proof of concept to test this relationship, 00:13:44.220 --> 00:13:48.800 we used strain data from the 2019 Ridgecrest sequence. 00:13:48.800 --> 00:13:54.029 The figure on your left shows how magnitudes estimated from two 00:13:54.029 --> 00:14:00.389 very close stations to the Ridgecrest main shock – this is magnitude 7.1 00:14:00.389 --> 00:14:05.129 shown here as blue and green lines, how they compare with the actual 00:14:05.129 --> 00:14:10.488 performance of ShakeAlert system shown here as a black line. 00:14:10.488 --> 00:14:15.510 And what would have been the system’s response with perfect data, 00:14:15.510 --> 00:14:19.870 with no data transmission latency, internal or external. 00:14:19.870 --> 00:14:24.750 And I would like to thank Jen Andrews for providing this data to us. 00:14:24.750 --> 00:14:28.410 So, whether we consider data transmission latency or not, 00:14:28.410 --> 00:14:31.590 looking at the magnitude estimates from the strainmeters and their 00:14:31.590 --> 00:14:36.939 close proximity to the – to the fault, we can see the strain data has the 00:14:36.939 --> 00:14:42.149 potential to complement seismometer data for earthquake early warning. 00:14:42.149 --> 00:14:46.980 In this case, I assume it might even help – it might have helped ShakeAlert 00:14:46.980 --> 00:14:51.507 system reach the four-station alerting threshold faster. 00:14:51.507 --> 00:14:55.959 The other work that we did in this regard – we also utilized strain data 00:14:55.959 --> 00:14:59.379 from several recent earthquakes – I’m only showing one here – 00:14:59.379 --> 00:15:05.700 in California to find the associated peak ground velocities. 00:15:05.700 --> 00:15:11.000 So this is for the main shock again. And the strain-derived PGVs are shown 00:15:11.000 --> 00:15:15.069 as the red squares, and you can see they’re almost indistinguishable from the 00:15:15.069 --> 00:15:20.170 seismic-derived ones, which are shown as the – as gray circles here for the same 00:15:20.170 --> 00:15:24.579 distance from the rupture. And putting all this together, we can – 00:15:24.579 --> 00:15:28.670 it can mean that data from borehole strainmeters that are already there – 00:15:28.670 --> 00:15:32.600 they’re already deployed – can be complementary to seismometer 00:15:32.600 --> 00:15:35.949 data in estimating important parameters like earthquake 00:15:35.949 --> 00:15:42.089 ground motions and magnitudes. And also maybe help avoid triggering 00:15:42.089 --> 00:15:46.050 of false events by spurious noise by – with strainmeters providing 00:15:46.050 --> 00:15:48.389 verification that there is actual shaking happening 00:15:48.389 --> 00:15:51.290 and not just noise bursts on seismometers. 00:15:51.290 --> 00:15:56.189 But what I’m really excited about in the future is actually the great 00:15:56.189 --> 00:16:00.319 potential for scalability and offshore monitoring when we consider the fiber 00:16:00.319 --> 00:16:05.180 optic strain-sensing technology for this. And that’s actually the main focus 00:16:05.180 --> 00:16:09.290 of my current research. And, with that, I thank you very much for your attention 00:16:09.290 --> 00:16:11.687 and invite you to check out my poster. 00:16:11.687 --> 00:16:15.733 [applause] 00:16:18.467 --> 00:16:21.939 [silence] 00:16:21.939 --> 00:16:25.970 - Okay. So I have an update on a lightning talk I gave last year. 00:16:25.970 --> 00:16:29.319 And I also have a poster next-door. So this is about a year on, and the 00:16:29.319 --> 00:16:33.800 story is that we went on a hunt for relationships between the Cascadia 00:16:33.800 --> 00:16:37.630 subduction zone and the San Andreas. Ten years ago, we published a paper 00:16:37.630 --> 00:16:41.860 that showed that a number of events in both fault systems had radiocarbon 00:16:41.860 --> 00:16:45.680 overlap, which made an interesting story, but not quite a smoking gun. 00:16:45.680 --> 00:16:50.930 But we went on a hunt for the smoking gun, and we realized at some point, 00:16:50.930 --> 00:16:54.759 in the upper panel, that we could – we’re capable or recording historic 00:16:54.759 --> 00:16:57.810 earthquakes on both fault systems, which helped a lot. 00:16:57.810 --> 00:17:04.900 So, on the Cascadia side, we have events in 1980 and 1992 that were 00:17:04.900 --> 00:17:08.150 precisely dated with bomb carbon, and that’s in the upper panel. 00:17:08.150 --> 00:17:11.120 And so that helped us realize that we had more resolution than 00:17:11.120 --> 00:17:15.380 we thought in the Cascadia record. And so – and it also told us that, 00:17:15.380 --> 00:17:20.689 if we can record historical events of that size – and we also found the 00:17:20.689 --> 00:17:25.130 1906 earthquake in a – in a nice age model of – in Cascadia 00:17:25.130 --> 00:17:29.260 in Trinidad Canyon, 100 kilometers north of the San Andreas. 00:17:29.260 --> 00:17:32.330 So that told us, if we can record San Andreas earthquakes on the 00:17:32.330 --> 00:17:35.340 Cascadia side, we should also be seeing Cascadia earthquakes 00:17:35.340 --> 00:17:38.940 on the San Andreas side at a similar range in Noyo Canyon, which is 00:17:38.940 --> 00:17:43.240 100 kilometers to the south. And so what happens is, though, 00:17:43.240 --> 00:17:47.950 when you look in the Noyo Canyon cores at the – at around 1700, 00:17:47.950 --> 00:17:51.179 when we have an earthquake on both fault systems – north coast 00:17:51.179 --> 00:17:55.085 San Andreas and Cascadia – we don’t see two separate events. 00:17:55.085 --> 00:17:58.850 What we see is one event. And it’s that doublet in the middle there 00:17:58.850 --> 00:18:04.549 where we see a stacking of two separate elements of what may 00:18:04.549 --> 00:18:08.610 be actually a single event. And, for several hundred years later 00:18:08.610 --> 00:18:11.460 and several hundred years earlier, there’s nothing else in there. 00:18:11.460 --> 00:18:14.950 And we should see both events. So what I’m proposing is that we’re 00:18:14.950 --> 00:18:18.320 seeing actually Cascadia and San Andreas stacked together 00:18:18.320 --> 00:18:21.740 in single turbidites with very little time between them. 00:18:21.740 --> 00:18:25.809 And there’s seven other examples in the late Holocene like that. 00:18:25.809 --> 00:18:28.700 We look in the time range, where we see – expect to see 00:18:28.700 --> 00:18:31.730 San Andreas and Cascadia represented, and we don’t see that. 00:18:31.730 --> 00:18:36.650 We see stacking of these odd doublets at those times. 00:18:36.650 --> 00:18:41.750 And, for comparison, in 1906, just a single San Andreas event is on 00:18:41.750 --> 00:18:44.490 the left, and that’s what that looks like. So we have eight of those doublets like 00:18:44.490 --> 00:18:49.830 in the middle, and the idea is that Cascadia may be partially synchronized 00:18:49.830 --> 00:18:53.920 and triggering San Andreas earthquakes eight times in the last 3,000 years. 00:18:53.920 --> 00:18:57.409 But essentially not at all in the prior 7,000 years. 00:18:57.409 --> 00:19:04.498 So partial synchronization, maybe a la Chris Scholz’s 2010 paper 00:19:04.498 --> 00:19:08.567 talking about fault synchronization. So there’s the short version. 00:19:08.567 --> 00:19:13.833 Come by the poster, and I’ll – and try to talk me out of it. [laughter] 00:19:13.833 --> 00:19:18.000 [applause] 00:19:20.781 --> 00:19:27.932 [silence] 00:19:27.932 --> 00:19:31.000 - Hi. I’m Ruth, and I’m going to talk about a project that we’ve been working 00:19:31.000 --> 00:19:35.200 on for a little while. Co-authors in the room, please raise your hand. 00:19:35.200 --> 00:19:37.444 Yay. All right. Thank you for coming. 00:19:37.444 --> 00:19:41.580 All right. So this paper – the manuscript is going to get submitted 00:19:41.580 --> 00:19:46.200 to internal review this week. And this is looking at dynamic rupture. 00:19:46.200 --> 00:19:50.740 So physically soft consistent models of large earthquakes on some 00:19:50.740 --> 00:19:54.110 Bay Area faults. These are faults that have the highest – one of the highest 00:19:54.110 --> 00:19:57.470 probabilities in this region. And what we do is we have a dynamic 00:19:57.470 --> 00:20:00.820 rupture model that goes all the way from just opposite Gilroy on the 00:20:00.820 --> 00:20:03.430 Calaveras Fault all the way through the Hayward Fault. 00:20:03.430 --> 00:20:07.490 And to the Rodgers Creek Fault. It ends just south of Santa Rosa. 00:20:07.490 --> 00:20:11.299 And, in this complicated fault geometry – 3D fault geometry, 00:20:11.299 --> 00:20:15.200 we also include a branch over to the northern Calaveras Fault. 00:20:15.200 --> 00:20:18.470 All of these – excuse me. All of these faults do creep, 00:20:18.470 --> 00:20:23.260 so that makes it kind of interesting. And all these faults also have geologic 00:20:23.260 --> 00:20:26.669 or historical evidence of large earthquakes on them, 00:20:26.669 --> 00:20:30.270 which makes them scary. So we want to put together a model. 00:20:30.270 --> 00:20:33.821 On the right side is some of the information that we’re including in this, 00:20:33.821 --> 00:20:38.059 and this includes what the rocks – the 3D rock structure looks like. 00:20:38.059 --> 00:20:44.490 And this is from the 3D GUMP model. We have fault friction information from 00:20:44.490 --> 00:20:47.990 our Earthquake Science Center Rock Mechanics Lab, based on 00:20:47.990 --> 00:20:51.380 surface outcrop measurements. We have creep information from 00:20:51.380 --> 00:20:56.060 Gareth Funning and Estelle Chaussard’s slip rate models for these faults. 00:20:56.060 --> 00:20:59.470 And then we put it all together into our favorite dynamic rupture code, which, 00:20:59.470 --> 00:21:02.549 in our case, is Michael Barall’s FaultMod code. 00:21:02.549 --> 00:21:07.960 And then we get simulations. The bottom three – Hayward nucleation 00:21:07.960 --> 00:21:12.110 on the left, Rodgers Creeks nucleation in the middle, and then northern 00:21:12.110 --> 00:21:15.580 Calaveras nucleation on the right. And it’s showing the ground – 00:21:15.580 --> 00:21:19.820 3D ground velocity, so the 3D ground shaking of the Earth’s surface for 00:21:19.820 --> 00:21:24.380 each of these scenarios – for three scenarios that we looked at. 00:21:24.380 --> 00:21:27.480 And then, just under that, the bottom final slip is showing 00:21:27.480 --> 00:21:31.780 what the fault slip would look like on the fault surfaces for each of those 00:21:31.780 --> 00:21:37.460 scenarios. So these are a subset of the many possibilities that we ran. 00:21:37.460 --> 00:21:40.500 All of our stuff does show that you can get large earthquakes. 00:21:40.500 --> 00:21:44.430 Sometimes the creep – the effect of the creep does stop these events. 00:21:44.430 --> 00:21:46.640 Sometimes it does not. 00:21:46.640 --> 00:21:50.033 So manuscript going into internal review this week. 00:21:50.033 --> 00:21:53.120 And if you have any questions, please come ask me or any of 00:21:53.120 --> 00:21:56.152 my enthusiastic co-authors. Thank you. 00:21:56.152 --> 00:22:00.133 [applause] 00:22:03.215 --> 00:22:08.762 [silence] 00:22:08.787 --> 00:22:10.161 - So cute. 00:22:10.867 --> 00:22:11.817 Ooh. 00:22:13.229 --> 00:22:15.567 Okay. Hi, everyone. My name is Alex Hatem. 00:22:15.567 --> 00:22:18.791 I’m a Mendenhall in Golden – the USGS office there. 00:22:18.791 --> 00:22:24.429 And I’m here to beat our drum of trying to collect more geologic data 00:22:24.429 --> 00:22:27.300 to update the National Seismic Hazard Model. 00:22:27.300 --> 00:22:30.390 The national model will be updated in 2023. 00:22:30.390 --> 00:22:33.160 And we’re currently collecting geologic input data 00:22:33.160 --> 00:22:36.010 that will be used in the model. 00:22:36.010 --> 00:22:43.130 Our group is trying to collate about – well, many years of geologic data that’s 00:22:43.130 --> 00:22:47.010 not been included in the model. California is in the best shape in the 00:22:47.010 --> 00:22:50.799 country, so congratulations to you. But you can come by my poster and 00:22:50.799 --> 00:22:53.730 see, actually, how little data exists in California. 00:22:53.730 --> 00:22:59.899 So I’m hoping that people here can help me add to our database. 00:22:59.899 --> 00:23:03.339 This is just a basic timeline for our progress. 00:23:03.339 --> 00:23:07.549 We’re in the process of reaching out to regional coordinators that know 00:23:07.549 --> 00:23:12.280 of new studies that have been done. And we’re currently compiling what 00:23:12.280 --> 00:23:16.793 we have for the rest of the country as well as from UCERF in the state 00:23:16.793 --> 00:23:21.230 of California. And then, we’ll be accepting these new contributions 00:23:21.230 --> 00:23:24.400 until the end of May. And then, over the summer, 00:23:24.400 --> 00:23:29.687 we’ll put together this new database that encompasses all 50 states, 00:23:29.687 --> 00:23:32.182 which will be the first time that that’s been done. 00:23:32.182 --> 00:23:39.720 And we’re hoping to get this submitted as a Open-File report or something 00:23:39.720 --> 00:23:42.590 with a DOI that’ll be included in the national model. 00:23:42.590 --> 00:23:46.990 So the data that we’re looking for are geologic slip rates, paleoearthquake 00:23:46.990 --> 00:23:50.470 recurrence, slip per event, and any updates to fault geometries. 00:23:50.470 --> 00:23:56.110 So really basic geology that I know we all have and love. 00:23:56.110 --> 00:24:00.400 So I want to know what you know, so please come talk to me at my poster, 00:24:00.400 --> 00:24:03.121 and you can contact me. My email is up there, and we 00:24:03.121 --> 00:24:06.333 can talk more another time. Thank you. 00:24:06.333 --> 00:24:10.200 [applause] 00:24:13.187 --> 00:24:18.574 [silence] 00:24:18.599 --> 00:24:22.340 - So I hope I made the point well with my talk earlier about the importance of 00:24:22.340 --> 00:24:28.330 going to the field quickly to capture data that are perishable in nature. 00:24:28.330 --> 00:24:33.255 And the Ridgecrest sequence provided us with a great example of that. 00:24:33.255 --> 00:24:37.917 I know that probably not all of you would agree, and Chris said earlier, 00:24:37.917 --> 00:24:41.570 he used to go rushing out to the field, but now he stays at the computer. 00:24:41.570 --> 00:24:46.650 I know that, with seismic networks and open availability of data, that that 00:24:46.650 --> 00:24:53.100 is the case in many senses nowadays, but I also think that being ready to go 00:24:53.100 --> 00:24:57.040 quickly, either by ground-based vehicle or by helicopter, to do 00:24:57.040 --> 00:25:01.400 reconnaissance can really pay off. And perhaps Ridgecrest was just 00:25:01.400 --> 00:25:05.630 a special case. So perhaps, in general, maybe we don’t really need that. 00:25:05.630 --> 00:25:07.740 Maybe we’ll get lucky with the right satellite images, 00:25:07.740 --> 00:25:11.240 or somebody else will do some airborne imagery that we can use. 00:25:11.240 --> 00:25:16.228 But, no, I think what I feel strongly about, and what I would like to 00:25:16.228 --> 00:25:19.929 encourage and try to help with passing along the information I have 00:25:19.929 --> 00:25:25.039 on this is that it is important to be ready to go as quickly as possible. 00:25:25.039 --> 00:25:29.990 And it won’t be every time that those of us that, in general, are ready, 00:25:29.990 --> 00:25:33.370 are able because, you know, maybe we’re away on a trip, or maybe we’re 00:25:33.370 --> 00:25:36.590 sick or have something else that’s happening in our lives. 00:25:36.590 --> 00:25:40.080 And so we need some depth to this group of people 00:25:40.080 --> 00:25:41.700 that are really ready to go. 00:25:41.700 --> 00:25:48.982 Okay, so that’s my main point. And I wanted to offer to have meetings. 00:25:48.982 --> 00:25:52.190 Let’s start out biweekly. Tuesdays at 10:00. 00:25:52.190 --> 00:25:55.837 And we can try using Microsoft Teams, but we’ll probably need to use Skype 00:25:55.837 --> 00:26:00.910 or maybe something else. But what I’d like to do is, with CGS and 00:26:00.910 --> 00:26:06.340 USGS people and academic colleagues, just get together and talk about things 00:26:06.340 --> 00:26:10.510 like – in the upper right there, it emphasizes teamwork. 00:26:10.510 --> 00:26:14.630 And, you know, the steps that you need to be sure that you’re ready 00:26:14.630 --> 00:26:18.250 to complete in a timely manner when the time comes. 00:26:18.250 --> 00:26:22.950 And so this just shows all of the gear that I keep in my two bags – a backpack 00:26:22.950 --> 00:26:26.407 and a flight bag. And that’s really all you need in order to 00:26:26.407 --> 00:26:30.140 go to the field quickly. It’s all listed there. 00:26:30.140 --> 00:26:33.179 That’s what I bring and what I would encourage others to bring. 00:26:33.179 --> 00:26:38.460 So if you currently would like to participate in rapid-response field work, 00:26:38.460 --> 00:26:43.223 this can be kind of a list of things that would be useful. 00:26:43.223 --> 00:26:48.650 And basically shows, when you gear up, you bring your pack, and everything 00:26:48.650 --> 00:26:52.970 else is on you or left behind. And so know what’s in your 00:26:52.970 --> 00:26:57.630 go bags and be ready to go in a hurry. Because maybe next time, it’ll be you 00:26:57.630 --> 00:27:01.380 that gets some really important data that otherwise we would never 00:27:01.380 --> 00:27:03.970 know what happened. And so, with Ridgecrest, 00:27:03.970 --> 00:27:08.429 between the 6.4 and the 7.1, this approach of going quickly 00:27:08.429 --> 00:27:12.480 to the field really paid off. Let’s hope that, by sharing information 00:27:12.480 --> 00:27:17.567 about how to do this, we can all be better prepared in the future. Thanks. 00:27:17.567 --> 00:27:22.633 [applause] 00:27:25.735 --> 00:27:33.961 [silence] 00:27:33.961 --> 00:27:38.780 - Hello. So my slide shows an interesting correlation I found last fall. 00:27:38.780 --> 00:27:43.730 But first some background. The SAFOD deep drilling project across 00:27:43.730 --> 00:27:46.760 the central creeping section of the San Andreas Fault proved, 00:27:46.760 --> 00:27:51.690 for the first time, that serpentinite and its magnesium-rich alteration 00:27:51.690 --> 00:27:58.090 products will promote fault creep. So I – since SAFOD, which I worked 00:27:58.090 --> 00:28:03.150 on a lot, I used what I learned there, started looking at creeping fault 00:28:03.150 --> 00:28:06.750 sections and segments in northern California to see if they can 00:28:06.750 --> 00:28:10.169 be explained in the same way. So the first stop was the 00:28:10.169 --> 00:28:13.960 Bartlett Springs Fault – the creeping segment around Lake Pillsbury – 00:28:13.960 --> 00:28:19.220 and that, indeed, seems to be caused by serpentinite there as well. 00:28:19.220 --> 00:28:23.809 So what I – this past year, I’ve started looking at the Rodgers Creek Fault, 00:28:23.809 --> 00:28:28.490 and the map in the center here is taken from a recent paper by 00:28:28.490 --> 00:28:33.750 Shakibay Senobari and Funning. They developed a new technique, 00:28:33.750 --> 00:28:38.889 which they used to kind of improve the identification of repeating earthquakes, 00:28:38.889 --> 00:28:46.480 which are considered to be indicators of fault creep at depth in the fault zone. 00:28:46.480 --> 00:28:50.730 And so the northern half of the Rodgers Creek Fault, shown here, is known from 00:28:50.730 --> 00:28:55.820 a variety of data sources to be creeping. And they, indeed, found a large number 00:28:55.820 --> 00:29:00.210 of repeating earthquakes with various degrees of confidence 00:29:00.210 --> 00:29:03.919 in the northern half, extending down to a – this little outlier 00:29:03.919 --> 00:29:08.029 a few kilometers southeast of Santa Rosa there. 00:29:08.029 --> 00:29:12.029 So there are – geologic maps of the area show three places where serpentinite 00:29:12.029 --> 00:29:17.130 outcrops within the Rodgers Creek Fault – and there are photographs 00:29:17.130 --> 00:29:20.601 of them on the left side and the upper right over here. 00:29:20.601 --> 00:29:23.570 And, when I went to put my AGU talk together last fall, 00:29:23.570 --> 00:29:28.090 I wanted to use this map and plot their locations on them. 00:29:28.090 --> 00:29:31.980 And it turned out that the three serpentinite outcrops all ended up lying 00:29:31.980 --> 00:29:35.330 directly above some of the repeating earthquakes that they found. 00:29:35.330 --> 00:29:38.721 More or less – they’re shown in the green circles here, more or less 00:29:38.721 --> 00:29:45.458 bracketing the two end members – extremes of the creeping section here. 00:29:45.458 --> 00:29:49.529 But there’s also one other point as well. 00:29:49.529 --> 00:29:54.299 A few years ago, Belle Philibosian told me about a sliver of serpentinite 00:29:54.299 --> 00:29:58.364 in the northern extension of the West Napa Fault. 00:29:58.364 --> 00:30:03.010 And it’s a few kilometers to the northwest of the area that ruptured 00:30:03.010 --> 00:30:06.640 in the 2014 earthquake. So it’s well beyond that. 00:30:06.640 --> 00:30:11.040 I went and sampled it. That’s the photograph in the lower right. 00:30:11.040 --> 00:30:15.029 Not much was known about it, but – so I sampled it but didn’t 00:30:15.029 --> 00:30:19.429 do anything about it, but I went – looked at it, and also plots 00:30:19.429 --> 00:30:22.059 directly above some of the repeating earthquakes. 00:30:22.059 --> 00:30:26.309 So the question is, is this possible correlation a coincidence, or is there 00:30:26.309 --> 00:30:29.950 some sort of genetic relationship between the serpentinite at the surface 00:30:29.950 --> 00:30:33.380 and the repeating earthquakes at depth? 00:30:33.380 --> 00:30:35.945 And that’s what I’m going to pursue next. Thank you. 00:30:35.945 --> 00:30:39.533 [applause] 00:30:41.801 --> 00:30:44.808 [silence] 00:30:44.833 --> 00:30:48.649 - Hi. I’m Carol Ostergren. I’m with the National Geospatial Program. 00:30:48.649 --> 00:30:52.210 And I wanted to give everyone an awareness of the Lidar 00:30:52.210 --> 00:30:56.880 that the NGP has ready for your GIS. 00:30:56.880 --> 00:31:02.500 The darker colors shown in the map – the dark blue and the dark green 00:31:02.500 --> 00:31:08.206 are data that’s already available through national map servers. 00:31:08.206 --> 00:31:13.910 The shades of blue in this map show Quality Level 1 data. 00:31:13.910 --> 00:31:16.630 That’s eight points per square meter. 00:31:16.630 --> 00:31:20.270 The green shades show Quality Level 2 data. 00:31:20.270 --> 00:31:23.336 That’s two points per square meter. 00:31:23.336 --> 00:31:28.669 Much of this data that you see across this swath of northern California 00:31:28.669 --> 00:31:33.750 and the area down around Ventura and Santa Barbara was funded 00:31:33.750 --> 00:31:39.370 through 2018 supplemental funds that we received. 00:31:39.370 --> 00:31:44.519 The areas up around Plumas and the Camp Fire area are 00:31:44.519 --> 00:31:50.309 now posted and available for download. 00:31:50.309 --> 00:31:56.419 So one of the things that we’re watching this upcoming year is a new 00:31:56.419 --> 00:32:01.059 line item in the governor’s proposed budget for $80 million. 00:32:01.059 --> 00:32:06.620 And this would complete Quality Level 1 for the entire state 00:32:06.620 --> 00:32:14.750 in partnership with the 3DEP program and with local partners. 00:32:14.750 --> 00:32:18.700 So there’s a couple of URLs for looking for data, but I would really encourage 00:32:18.700 --> 00:32:24.440 you to get a hold of me, and I can help you get the data that you need. 00:32:24.440 --> 00:32:26.009 Thanks. 00:32:26.009 --> 00:32:31.000 [applause] 00:32:33.567 --> 00:32:38.233 [silence] 00:32:38.233 --> 00:32:40.631 - Hi, there. 00:32:40.631 --> 00:32:43.659 My name is Jessie Pearl. I’m a new Mendenhall postdoctoral 00:32:43.659 --> 00:32:46.700 research fellow up in the Seattle field office. 00:32:46.700 --> 00:32:50.445 So I am happy that I have a chance today to talk about one of the tools that 00:32:50.445 --> 00:32:55.820 I’m using up in the Washington area but also in coastal Oregon and northern 00:32:55.820 --> 00:33:01.828 California to precisely date earthquake events, so – using dendrochronology. 00:33:01.828 --> 00:33:04.809 So, back in the 1990s, dendrochronology was famously 00:33:04.809 --> 00:33:08.880 used to the date the 1700 megathrust earthquake event with the idea that 00:33:08.880 --> 00:33:13.020 coastal forests are subsiding after the – during the earthquake. 00:33:13.020 --> 00:33:16.700 The saltwater, or even just the trauma itself, will kill those forests. 00:33:16.700 --> 00:33:19.789 And we can go back decades or centuries later and look at actually 00:33:19.789 --> 00:33:22.630 the physical characteristics of those rings in the tree. 00:33:22.630 --> 00:33:26.539 The narrow and wide ring patterns act as a barcode. 00:33:26.539 --> 00:33:33.029 If we can match that barcode with a long tree ring record from that area 00:33:33.029 --> 00:33:37.390 that’s pretty much dictated by climate regionally, we can match exactly that 00:33:37.390 --> 00:33:41.910 barcode and get a precise year, if not a season, on when that earthquake event 00:33:41.910 --> 00:33:45.590 happened, when that ghost forest died. So this is a well-known technique, 00:33:45.590 --> 00:33:49.360 and it’s a really wonderful and accurate technique, but it’s limited 00:33:49.360 --> 00:33:52.460 by how long these reference chronologies are. 00:33:52.460 --> 00:33:57.179 So how long we can make and push back these living tree chronologies, 00:33:57.179 --> 00:34:01.360 or even extending it back farther using timbers or log cabins. 00:34:01.360 --> 00:34:04.870 So I’m proposing that we turn to the chemical characteristics of the wood. 00:34:04.870 --> 00:34:07.990 So we’re looking not only at flooding signals in the wood, but we’re also 00:34:07.990 --> 00:34:11.129 looking at the radiocarbon within each tree ring. 00:34:11.129 --> 00:34:15.629 So, similar to the bomb spike we think of in the 1950s, ’60s, and ’70s, 00:34:15.629 --> 00:34:19.490 we see natural radiocarbon spikes. And, once we detect those within the 00:34:19.490 --> 00:34:23.480 rings, we can precisely date some of these floating ghost forests, which will 00:34:23.480 --> 00:34:27.399 allow us to more accurately date the penultimate earthquake and even other 00:34:27.399 --> 00:34:32.379 earthquakes over the past 2,000 years, wherever we have those forests. 00:34:32.379 --> 00:34:35.690 So another component to this research that I’m actually really excited about 00:34:35.690 --> 00:34:42.050 here in northern California specifically is we can take trees that die in landslide 00:34:42.050 --> 00:34:46.129 events – so there’s a piece of wood up there right above the graph of 00:34:46.129 --> 00:34:51.649 the radiocarbon spike in 774 that’s from a landslide event at Red Lassic 00:34:51.649 --> 00:34:55.770 Peak in northern California. So, using this wood, I can not only 00:34:55.770 --> 00:34:58.900 precisely date when this earthquake occurred – or, when – sorry, when this 00:34:58.900 --> 00:35:03.079 landslide occurred, but we can also think about what the 00:35:03.079 --> 00:35:06.099 climate effect might be on the landslide versus seismic. 00:35:06.099 --> 00:35:10.380 So there’s plenty of hydroclimate- sensitive tree rings chronologies 00:35:10.380 --> 00:35:14.030 in northern California. We can think about the years preceding 00:35:14.030 --> 00:35:18.650 this landslide were very wet or very dry and start to detangle the seismic 00:35:18.650 --> 00:35:21.310 versus climate effects and what might be driving some of these 00:35:21.310 --> 00:35:26.140 large earthquake events in the early part of the common era. 00:35:26.140 --> 00:35:29.290 So I’m excited to talk with you all about some possible avenues 00:35:29.290 --> 00:35:31.890 to explore in northern California using these techniques, 00:35:31.890 --> 00:35:35.500 and I’ll be milling about. Thanks very much. 00:35:35.500 --> 00:35:39.500 [applause] 00:35:42.689 --> 00:35:52.438 [silence] 00:35:53.983 --> 00:35:56.867 [laughter] 00:35:58.680 --> 00:36:00.661 - Okay. I’ll just hold it. 00:36:00.661 --> 00:36:06.110 So the – returning to Ridgecrest, the 2019 Ridgecrest earthquakes 00:36:06.110 --> 00:36:10.250 have occasionally been characterized as surprising because only about 00:36:10.250 --> 00:36:16.848 a third of the surface ruptures occurred on faults that were included in the 00:36:16.848 --> 00:36:20.810 Quaternary Fault and Fold Database, shown in the figure on the left. 00:36:20.810 --> 00:36:25.600 The black lines are the 2019 rupture, and the bold orange lines are the 00:36:25.600 --> 00:36:29.702 parts of those that were in Q-Faults. 00:36:29.702 --> 00:36:33.375 Along with my co-authors, Jessica Jobe, Colin Chupik, 00:36:33.375 --> 00:36:38.010 and a number of others, we took on the project of attempting 00:36:38.010 --> 00:36:41.859 to map out near-tectonic features that existed in the landscape 00:36:41.859 --> 00:36:47.250 prior to the 2019 events. And, for that project, we used the 00:36:47.250 --> 00:36:51.970 2-meter VM that was released on OpenTopography as well as 00:36:51.970 --> 00:36:56.490 satellite and aerial imagery and recent – a compilation 00:36:56.490 --> 00:37:00.819 of recent field observations that were made sort of as a byproduct 00:37:00.819 --> 00:37:04.730 of mapping the 2019 ruptures themselves. 00:37:04.730 --> 00:37:10.180 So our compiled mapping, based on that, what we find is that 00:37:10.180 --> 00:37:15.180 pre-existing neotectonic features did exist in the landscape distributed 00:37:15.180 --> 00:37:21.460 throughout basically the entire length of the 2019 ruptures. 00:37:21.460 --> 00:37:26.980 And that – and, including, in fact, most of the secondary traces as well as 00:37:26.980 --> 00:37:31.780 the primary traces. So what’s shown in the figure on the right – again, 00:37:31.780 --> 00:37:37.140 the black lines are the 2019 ruptures, but they’re colored in red where there 00:37:37.140 --> 00:37:42.780 were identifiable features pre-existing before 2019. 00:37:42.780 --> 00:37:46.859 And the yellow lines there are actually additional faults that did not have 00:37:46.859 --> 00:37:50.670 rupture reported in 2019 that’s sort of illustrating this broad 00:37:50.670 --> 00:37:54.810 orthogonal fault network in that area. 00:37:54.810 --> 00:37:59.140 So, based on our mapping, we estimate that between 50 and 70% 00:37:59.140 --> 00:38:04.069 of the 2019 ruptures could have been identified as active faults if this 00:38:04.069 --> 00:38:09.640 kind of a detailed study had been done prior to the events. 00:38:09.640 --> 00:38:14.609 And this figure pretty much represents the punchline of our study. 00:38:14.609 --> 00:38:19.036 But, for many more intriguing details, please come to the poster, 00:38:19.036 --> 00:38:21.391 and we can talk about it. Thank you. 00:38:21.391 --> 00:38:24.567 [applause] 00:38:26.967 --> 00:38:42.440 [silence] 00:38:42.440 --> 00:38:45.200 [laughter] 00:38:45.200 --> 00:38:55.055 [silence] 00:38:55.055 --> 00:38:56.829 - Hi, everybody. I’m Arben Pitarka. 00:38:56.829 --> 00:39:00.437 I work at Lawrence Livermore National Lab and Rob Graves 00:39:00.437 --> 00:39:05.349 and Arthur Rodgers are co-authors of this study. 00:39:05.349 --> 00:39:10.944 This is just a snapshot of a long collaboration. 00:39:10.944 --> 00:39:14.359 And we have a poster, actually, that will show probably more details, 00:39:14.359 --> 00:39:20.790 but what we wanted to do is, we wanted to see how well the rupture generator 00:39:20.790 --> 00:39:27.329 that we have recently refined works in predicting strong ground motion, 00:39:27.329 --> 00:39:30.947 starting with the earthquakes that are already happening, and they 00:39:30.947 --> 00:39:35.700 have really good recording data. In particular, our model includes 00:39:35.700 --> 00:39:41.349 stochastic and deterministic features that are constrained based on 00:39:41.349 --> 00:39:46.050 rupture dynamics but also recorded earthquakes. 00:39:46.050 --> 00:39:51.190 And one of the features that we’re more interested is how – the unknown 00:39:51.190 --> 00:39:54.829 for future earthquakes, which is the large slip patches will affect 00:39:54.829 --> 00:39:58.720 the ground motion. So we did some sensitivity analysis, 00:39:58.720 --> 00:40:04.770 and we noticed that, no matter how you put those patches into your slip 00:40:04.770 --> 00:40:10.069 in general, you may fit the ground motion, on average, well. 00:40:10.069 --> 00:40:14.720 But, when you look at the particular locations and time histories like they 00:40:14.720 --> 00:40:21.270 are in the left – on the right panel, you can see that the performance of 00:40:21.270 --> 00:40:26.480 each model kind of deteriorate, or it’s better than the others, and some 00:40:26.480 --> 00:40:33.099 of them fit the data very well, some not. So what we notice is that, depending on 00:40:33.099 --> 00:40:40.329 the depth of these patches, you may introduce or not really high frequency 00:40:40.329 --> 00:40:44.471 into your rupture or – which is reflected in higher ground motion. 00:40:44.471 --> 00:40:50.010 And, as Julian mentioned in this case, we need to take care of the fact that the 00:40:50.010 --> 00:40:54.240 stress drop in our rupture model should be depth-dependent, and also the 00:40:54.240 --> 00:40:58.660 rupture kinematics should reflect that. So that means that, for slip patches 00:40:58.660 --> 00:41:03.280 that reach the free surface, which is one of those cases here, you may expect 00:41:03.280 --> 00:41:06.780 really good behavior, and for the ones that have really deep patches, 00:41:06.780 --> 00:41:11.890 then you kind of overestimate the effect of the high frequency generated 00:41:11.890 --> 00:41:15.359 by the rupture. And then you have overestimate of ground motion. 00:41:15.359 --> 00:41:19.339 So, for more details, you can come and discuss with us at the poster. 00:41:19.339 --> 00:41:22.475 And I’ll be – I’ll be glad to show you more details 00:41:22.475 --> 00:41:25.133 about this exciting research. Thank you. 00:41:25.133 --> 00:41:29.233 [applause] 00:41:33.635 --> 00:41:37.373 [laughter] 00:41:39.809 --> 00:41:43.700 - All right. I’m going to talk about some work my colleagues and I are 00:41:43.700 --> 00:41:48.390 doing at Livermore to model large Hayward Fault simulations with 00:41:48.390 --> 00:41:52.007 high-performance computing. We’ve been working on simulations 00:41:52.007 --> 00:41:55.010 of a magnitude 7 earthquake on the Hayward Fault for the last few years, 00:41:55.010 --> 00:42:00.080 shown in the map view there in a domain centered on the Hayward Fault. 00:42:00.080 --> 00:42:03.500 And you can see the PGV map in the background. 00:42:03.500 --> 00:42:08.164 We used the code SW4, which is an anelastic wave propagation code. 00:42:08.164 --> 00:42:12.853 It’s recently been enhanced to run efficiently on GPU systems. 00:42:12.853 --> 00:42:17.885 And what used to be a challenging calculation to 5 hertz is now a routine 00:42:17.885 --> 00:42:22.360 calculation. We’ve actually run 50 of those now in another study. 00:42:22.360 --> 00:42:28.510 We got a opportunity to run on the Sierra supercomputer before it went 00:42:28.510 --> 00:42:32.220 behind the fence of Livermore. And we ran a few calculations resolved 00:42:32.220 --> 00:42:36.307 to 10 hertz, which I believe are some of the largest calculations that have ever 00:42:36.307 --> 00:42:41.260 been done for an earthquake in California. And we get good agreement with 00:42:41.260 --> 00:42:44.800 ground motion intensity measures – or ground motion models shown by 00:42:44.800 --> 00:42:49.200 the epsilon values in the lower left there from 100 hertz or 00:42:49.200 --> 00:42:53.120 0.01 seconds up to 10-second periods in PGA and PGV. 00:42:53.120 --> 00:42:57.050 That’s for a network of over 2,000 sites at the surface. 00:42:57.050 --> 00:43:01.550 But there are two important and related shortcomings to this linear elastic 00:43:01.550 --> 00:43:05.170 modeling, or linear anelastic modeling, that have to do with representing 00:43:05.170 --> 00:43:08.390 the geotechnical layer. Firstly, we assume a minimum shear 00:43:08.390 --> 00:43:12.410 wave speed of 500 meters per second. And we also know that, when you hit 00:43:12.410 --> 00:43:17.040 rocks with large accelerations, they do not behave as linear elastic solids. 00:43:17.040 --> 00:43:22.277 We use – we’ve developed a method to correct for this based on a new ground 00:43:22.277 --> 00:43:25.530 motion model from Jeff Bayless and Norm Abrahamson based on Fourier 00:43:25.530 --> 00:43:31.140 amplitude spectrum that allows us to form a transfer function that essentially 00:43:31.140 --> 00:43:34.400 we can have the numerator be the Fourier amplitude spectrum for 00:43:34.400 --> 00:43:38.680 a nonlinear response for the in situ geotechnical properties. 00:43:38.680 --> 00:43:41.970 The denominator is what we calculate, and that’s the before the Fourier 00:43:41.970 --> 00:43:46.258 spectrum of 500 meter per second for linear only. 00:43:46.258 --> 00:43:49.609 The black seismograms there are what comes out of SW4. 00:43:49.609 --> 00:43:51.460 The green are after applying the transfer function. 00:43:51.460 --> 00:43:55.750 The transfer function is shown in the middle plot there. 00:43:55.750 --> 00:43:59.900 The dashed line is just the purely elastic transfer function. 00:43:59.900 --> 00:44:03.809 That’s going from low – from 500 meter per second to the in situ for – this is 00:44:03.809 --> 00:44:07.280 a case in Berkeley at 100 meters per second. 00:44:07.280 --> 00:44:09.360 The black line includes nonlinear response, 00:44:09.360 --> 00:44:12.010 which beats down those high frequencies. 00:44:12.010 --> 00:44:16.490 And the response spectra are shown in the lower right there. 00:44:16.490 --> 00:44:20.260 And geotechnical engineers will recognize that that shape is diagnostic 00:44:20.260 --> 00:44:25.290 of a weak soil site, in this case Berkeley, at 100 meters per second. 00:44:25.290 --> 00:44:29.990 So, when we perform this, we think we’re getting more accurate motions 00:44:29.990 --> 00:44:36.720 and overcoming the shortcomings in linear elastic modeling as a intermediate 00:44:36.720 --> 00:44:41.807 step to doing fully nonlinear modeling. And I’ll stop there. Thanks. 00:44:41.807 --> 00:44:46.200 [applause] 00:44:49.100 --> 00:44:54.155 [silence] 00:44:54.155 --> 00:44:55.930 - Hi, everyone. I’m Valerie Sahakian, 00:44:55.930 --> 00:44:58.910 and I’m presenting on some work on an earthquake hazards program 00:44:58.910 --> 00:45:01.334 project that I’ve been doing with students Elias King and 00:45:01.334 --> 00:45:04.740 Alexis Klimasewski. The goal is to estimate site parameters 00:45:04.740 --> 00:45:07.920 in the Bay Area to use for ground motion estimation, and in particular, 00:45:07.920 --> 00:45:10.849 we’re starting with kappa. Kappa is the slope of the 00:45:10.849 --> 00:45:13.601 high-frequency fall-off on a site spectra that you can see in the 00:45:13.601 --> 00:45:17.520 cartoon over there. So a high kappa means higher attenuation. 00:45:17.520 --> 00:45:20.310 In order to compute this, we compile a database of records like the 00:45:20.310 --> 00:45:23.859 seismogram on the top left. We get the spectra of each record, 00:45:23.859 --> 00:45:26.772 and this represents contributions from the event and the site. 00:45:26.772 --> 00:45:30.380 And then we put this into a large spectral decomposition to get event 00:45:30.380 --> 00:45:33.990 and site spectra for every earthquake and station in the data set. 00:45:33.990 --> 00:45:37.510 We then constrain that instead of using a reference site with a deviation 00:45:37.510 --> 00:45:40.720 from a Brune spectrum. And then, using the site spectra 00:45:40.720 --> 00:45:43.440 that comes out of that, we fit the equation on the bottom 00:45:43.440 --> 00:45:46.261 to it in order to get kappa out of it. 00:45:47.142 --> 00:45:49.910 We’re doing this in the Bay Area starting with just broadband stations. 00:45:49.910 --> 00:45:52.810 So the map on the bottom left shows our data set. 00:45:52.810 --> 00:45:55.102 The circles are earthquakes. The triangles, which are 00:45:55.102 --> 00:45:58.790 a little hard to see, are the stations. We have about 30 stations, 00:45:58.790 --> 00:46:02.540 3,800 earthquakes. And it’s about 20,000 recordings. 00:46:02.540 --> 00:46:04.930 We do this inversion with a number of different parameters. 00:46:04.930 --> 00:46:08.830 So, starting with signal-to-noise ratio of 3 to 5, different waveform lengths 00:46:08.830 --> 00:46:11.800 from the S wave arrival, as well as number of frequency bins 00:46:11.800 --> 00:46:14.124 that we run the inversion on. 00:46:14.124 --> 00:46:16.630 In the right-hand side, you can see some of the results and also 00:46:16.630 --> 00:46:19.900 some of the future work. So those two middle plots are 00:46:19.900 --> 00:46:22.425 site spectra for all of these different inversion runs for 00:46:22.425 --> 00:46:25.850 station CVS in the North Bay and EMR in the delta. 00:46:25.850 --> 00:46:29.869 You can see the variation in these and also the variability and uncertainty 00:46:29.869 --> 00:46:32.270 that we get out of the model that also goes into uncertainty on 00:46:32.270 --> 00:46:36.590 the kappa values themselves. Another thing to point out is that there 00:46:36.590 --> 00:46:41.309 are regional variations in kappa itself. The top right map shows the stations 00:46:41.309 --> 00:46:44.850 colored by their kappa value. So you can see that stations in the delta 00:46:44.850 --> 00:46:49.622 tend to have higher kappa values on average than other stations elsewhere. 00:46:49.622 --> 00:46:53.030 One of the problems is, if you can see the units on the color bar there, 00:46:53.030 --> 00:46:56.130 the kappa values are a little higher than we would expect. 00:46:56.130 --> 00:46:59.240 And we think that this is because there’s seismic attenuation 00:46:59.240 --> 00:47:01.920 that is not removed by this inversion that we have. 00:47:01.920 --> 00:47:06.050 So one of the next steps is to remove an average Q from the records 00:47:06.050 --> 00:47:09.920 before we invert them. Or to find a path-specific Q from 00:47:09.920 --> 00:47:14.380 a Q model like Eberhart-Phillips in the region and remove that first. 00:47:14.380 --> 00:47:18.320 And then finally, what we would like to have ideally is an interpolated map of 00:47:18.320 --> 00:47:21.591 kappa values and their uncertainties like exists for New Zealand. 00:47:21.591 --> 00:47:24.000 So, in order to do that, we need more data points. 00:47:24.000 --> 00:47:28.619 So one of the next steps is to redo this calculation with a new database that 00:47:28.619 --> 00:47:32.540 also includes records recorded on strong motion stations and then create those 00:47:32.540 --> 00:47:36.792 values to obtain both a kappa map and a kappa uncertainty map. 00:47:36.792 --> 00:47:39.597 I do not have a poster with me, but I have it on my computer, so if 00:47:39.597 --> 00:47:42.932 you’re interested in learning more, please come find me. Thanks. 00:47:42.932 --> 00:47:46.932 [applause] 00:47:50.468 --> 00:47:53.406 [laughter] 00:47:56.049 --> 00:47:59.282 [silence] 00:47:59.282 --> 00:48:02.480 - Hi, everyone. I think I’m the last talk. So, last, but hopefully not least. 00:48:02.480 --> 00:48:05.771 I’m Maureen Walton. I’m a postdoc at the USGS in Santa Cruz. 00:48:05.771 --> 00:48:10.465 I’m here representing our team today to essentially advertise that we have now 00:48:10.465 --> 00:48:16.430 put together a compilation of offshore faults offshore of California. 00:48:16.430 --> 00:48:19.400 So basically, Quaternary faults offshore – a database 00:48:19.400 --> 00:48:21.270 offshore of California. 00:48:21.270 --> 00:48:25.032 And the reason that we started doing this is because there’s lots of updates 00:48:25.032 --> 00:48:27.412 that are needed to the current Quaternary Fault and Fold 00:48:27.412 --> 00:48:33.109 Database offshore. And, in the last decade-plus, we’ve gotten a lot of new 00:48:33.109 --> 00:48:36.579 high-resolution data sets offshore that we’ve been able to do new mapping 00:48:36.579 --> 00:48:42.710 from, both in our group at the USGS but also with external collaborators as well. 00:48:42.710 --> 00:48:45.460 So new data sets include high-resolution bathymetry 00:48:45.460 --> 00:48:48.160 and high-resolution multi-channel seismic. 00:48:48.160 --> 00:48:52.300 And we’ve been working on these data, mapping new faults 00:48:52.300 --> 00:48:56.480 for, yeah, 10-plus years. And lots of new publications 00:48:56.480 --> 00:48:59.103 have come out, and so we decided it was high time to compile 00:48:59.103 --> 00:49:02.422 all of the new mapping into an offshore database. 00:49:02.422 --> 00:49:06.480 And, as you all know, offshore mapping of faults is also important 00:49:06.480 --> 00:49:10.690 for accurate hazard assessments. And so a lot of the – again, all of these 00:49:10.690 --> 00:49:14.829 faults come from published sources. So a lot of the faults in northern and 00:49:14.829 --> 00:49:18.569 central California are from Sam Johnson’s work through the California 00:49:18.569 --> 00:49:21.890 Seafloor Mapping Program. And then a lot of the faults that we have 00:49:21.890 --> 00:49:26.270 in our database offshore of southern California have been published in 00:49:26.270 --> 00:49:29.750 various peer review publications through the USGS West Coast Marine 00:49:29.750 --> 00:49:34.230 Geohazards Project led by Danny Brothers out of my office. 00:49:34.230 --> 00:49:38.410 And we have – just like the Quaternary Fault and Fold Database, we have 00:49:38.410 --> 00:49:41.766 comprehensive attribute information that we’ve kind of optimized for the 00:49:41.766 --> 00:49:45.980 offshore environment, including fault age, slip rate, and slip sense 00:49:45.980 --> 00:49:48.160 where we know it; mapping accuracy; 00:49:48.160 --> 00:49:52.250 and also the reference from where that mapping came from. 00:49:52.250 --> 00:49:56.559 And we have designed – we’ve been working closely with CGS to get this 00:49:56.559 --> 00:50:00.440 database out there. And it’s been designed for easy ingestion by partners 00:50:00.440 --> 00:50:05.329 and coordinated products at CGS, the USGS Quaternary Fault and Fold 00:50:05.329 --> 00:50:09.020 Database, SCEC, National Seismic Hazard Maps, among others. 00:50:09.020 --> 00:50:13.500 And our database will be versionable and capable of accepting published 00:50:13.500 --> 00:50:16.720 input from the community, hopefully, down the road. 00:50:16.720 --> 00:50:22.680 And we are starting with California, but we’re hoping to also add Cascadia – 00:50:22.680 --> 00:50:25.893 offshore Cascadia faults next and then also Alaska. 00:50:25.893 --> 00:50:29.982 And we do have a poster, so please come and see us at our poster. 00:50:29.982 --> 00:50:32.732 And, just so everyone knows where we’re at, this database has been – 00:50:32.732 --> 00:50:34.865 it’s going to be a USGS data release. 00:50:34.865 --> 00:50:38.473 It’s in internal review now, so it should be out within the month, I would say. 00:50:38.473 --> 00:50:41.766 So, yeah, come see us at our poster for more. 00:50:41.766 --> 00:50:45.599 [applause] 00:50:51.686 --> 00:50:56.166 [laughter] 00:50:59.399 --> 00:51:02.747 [silence] 00:51:02.747 --> 00:51:05.532 [laughter] 00:51:05.532 --> 00:51:11.771 [silence] 00:51:11.771 --> 00:51:15.566 - Okay. Thank you, everyone. That was an amazing lightning talk 00:51:15.566 --> 00:51:18.966 session. So all of our speakers – let’s give them all a big hand again. 00:51:18.966 --> 00:51:24.349 [applause] 00:51:24.349 --> 00:51:28.779 So now we have a break to go look at posters. 00:51:28.779 --> 00:51:32.049 A couple of them have been advertised in the lightning talks you just heard, but 00:51:32.049 --> 00:51:35.761 there are many, many more out there. So definitely go check them all out. 00:51:35.761 --> 00:51:37.930 And, if you haven’t checked out the fire truck yet, and it hasn’t 00:51:37.930 --> 00:51:40.270 gone home yet, you should check that out as well. 00:51:40.270 --> 00:51:43.119 And then we’re going to meet back here at 3:15 – is that right – 00:51:43.119 --> 00:51:47.650 for the last session of the day, which is earthquake early warning. 00:51:47.650 --> 00:51:50.632 Okay, see you in a little while. 00:51:52.740 --> 00:51:54.849 [laughter] 00:51:55.613 --> 00:51:58.013 [inaudible background conversations]