WEBVTT Kind: captions Language: en 00:00:05.109 --> 00:00:07.209 Good morning, everyone. 00:00:07.209 --> 00:00:11.050 Welcome to the Earthquake Science Center seminar. 00:00:11.050 --> 00:00:13.120 Just a couple of quick notes. 00:00:13.120 --> 00:00:15.240 Take a note. Next week we’re not going to 00:00:15.240 --> 00:00:18.430 actually have a seminar due to the annual SCEC meeting. 00:00:18.430 --> 00:00:23.020 We’ll reconvene the following week on Wednesday, September 20th. 00:00:23.020 --> 00:00:26.460 We’ll have Dave Lockner give a talk. 00:00:27.600 --> 00:00:30.939 Today it’s my pleasure to introduce Tom Brocher. 00:00:30.939 --> 00:00:37.070 Tom actually just told me he’s been with the USGS for 32 years now. 00:00:37.070 --> 00:00:41.370 Just joined the Pacific Northwest Hazards team about 2-1/2 years ago. 00:00:41.370 --> 00:00:46.170 And prior to that, he was the Earthquake Science Center director. 00:00:46.170 --> 00:00:48.640 So looking forward to a good talk from you today, Tom. 00:00:48.640 --> 00:00:52.290 I’ll let you take it away. - Okay. Thank you very much. 00:00:52.290 --> 00:00:55.800 Thank you all for coming today. If I’m a little distracted 00:00:55.800 --> 00:00:58.360 today, I apologize. My daughter-in-law was just 00:00:58.360 --> 00:01:03.440 induced with my first grandchild, and so I may get a text anytime. 00:01:03.440 --> 00:01:08.480 So if I’m distracted, that might be the explanation. 00:01:08.480 --> 00:01:14.460 What I would like to do today is talk about a magnitude 6-1/2 to 7 00:01:14.460 --> 00:01:19.380 earthquake in central Washington state that happened in 1872. 00:01:19.390 --> 00:01:22.070 So it’s not very well-known even amongst 00:01:22.070 --> 00:01:24.500 the people that live in Washington. 00:01:24.500 --> 00:01:29.479 And also relate that to a seismicity cluster that we see 00:01:29.479 --> 00:01:35.729 in the same area that’s very unusual. And I’m going to try to convince you 00:01:35.729 --> 00:01:39.659 that the evidence suggests that the two are related. 00:01:39.659 --> 00:01:47.229 And I’d like to acknowledge my co- authors, Rick Blakely and Brian Sherrod. 00:01:47.229 --> 00:01:50.590 And also Jim Dewey, who has helped me a great deal 00:01:50.590 --> 00:01:56.540 in learning how to interpret seismic intensity data. 00:01:56.540 --> 00:02:00.369 As well as Margaret Hopper and her colleagues, who are now retired 00:02:00.369 --> 00:02:06.180 or deceased, in our Golden center, who developed much of the information that 00:02:06.180 --> 00:02:12.760 I’ll be summarizing today, as well as the reviewers of a couple of journal papers. 00:02:12.760 --> 00:02:18.510 So the 1872 earthquake, we think now – we now think 00:02:18.510 --> 00:02:23.070 occurred in central Washington. And the red dot shows the proposed 00:02:23.070 --> 00:02:28.950 epicenter, and the large circle shows the felt – the approximate felt area. 00:02:28.950 --> 00:02:32.860 So it was very widely felt in British Columbia, Montana, 00:02:32.860 --> 00:02:38.200 Idaho, Oregon, and all of the state of Washington. 00:02:38.200 --> 00:02:43.440 Although what’s made it very difficult and challenging to work with as a 00:02:43.450 --> 00:02:50.370 net earthquake is that Washington in particular was very sparsely populated. 00:02:50.370 --> 00:02:54.790 It only had a population of about 18,000. 00:02:54.790 --> 00:02:56.980 And at that time, it wasn’t even a state yet. 00:02:56.980 --> 00:03:01.940 It was still a territory, and it didn’t become a state until 1889. 00:03:01.950 --> 00:03:06.790 So there are very few people that resided in the area 00:03:06.790 --> 00:03:10.340 of where we think the epicenter was at the time. 00:03:12.880 --> 00:03:16.660 To give you some local context about the earthquake, 00:03:16.670 --> 00:03:20.410 this is a map of the Bay Area. You’ll recognize the major faults 00:03:20.410 --> 00:03:26.519 with the probabilities of occurrence of a magnitude 6.7 or greater earthquake 00:03:26.519 --> 00:03:29.660 in the next 30 years from the UCERF report. 00:03:29.660 --> 00:03:35.120 And the red – the big red ellipse shows you the approximate size of 00:03:35.120 --> 00:03:41.600 the uncertainty of the location of the 1872 event, based on the intensity data. 00:03:41.600 --> 00:03:45.220 And this is from a very nice paper that Bill Bakun and others did, 00:03:45.220 --> 00:03:50.239 and I’m very pleased to see that Bill has joined the audience today. 00:03:50.239 --> 00:03:54.330 But this shows you that, if this was the level of uncertainty that we had 00:03:54.330 --> 00:03:59.690 for the 1868 Hayward Fault earthquake, we really wouldn’t know whether 00:03:59.690 --> 00:04:02.269 that event had happened on the Hayward Fault or the 00:04:02.269 --> 00:04:06.799 San Andreas Fault or the – or the Calaveras Fault or whatever. 00:04:06.799 --> 00:04:14.799 So the uncertainty in the location on the 1872 earthquake has really been 00:04:14.799 --> 00:04:20.209 a problem in terms of reducing the level of uncertainty 00:04:20.209 --> 00:04:23.080 in hazard assessment in Washington. 00:04:24.110 --> 00:04:28.740 So here’s a view of the seismicity in Washington state 00:04:28.740 --> 00:04:33.830 of crustal earthquakes for about the past 40 years. 00:04:33.830 --> 00:04:35.449 And you can see that, west of the Cascades, 00:04:35.449 --> 00:04:40.550 there’s a lot of seismicity in the Puget Sound area, Puget lowland. 00:04:40.550 --> 00:04:47.400 That north of there, there’s a seismic – seismicity lineations along 00:04:47.400 --> 00:04:52.270 Mount St. Helens – seismic lineation in Mount Rainier. 00:04:52.270 --> 00:04:57.499 But east of the Cascade Range, there’s very diffuse seismicity along the 00:04:57.499 --> 00:05:03.249 northern edge of the Yakima Fold Belt. And then we have this very unusual 00:05:03.249 --> 00:05:08.759 cluster of seismicity that we’re calling the Entiat cluster because 00:05:08.759 --> 00:05:14.190 it’s centered on this city – the town of Entiat in Washington. 00:05:14.190 --> 00:05:16.629 And we now think that the 1872 earthquake 00:05:16.629 --> 00:05:20.600 was located basically in the same region. 00:05:22.440 --> 00:05:27.219 One reason we might be interested in these long-lived aftershock sequences 00:05:27.220 --> 00:05:31.980 is that they have their own implications for seismic hazard analysis. 00:05:31.980 --> 00:05:37.220 And this is a figure that you’ve all seen, or many of you have probably seen, 00:05:37.220 --> 00:05:42.539 from Stein and Liu in a 2009 Science paper in which they argue 00:05:42.540 --> 00:05:50.400 that the duration of aftershocks, which is on the left axis, 00:05:50.400 --> 00:05:53.040 is a function of the loading rate of the fault. 00:05:53.059 --> 00:05:58.059 So that faults that load rapidly, say along the San Andreas Fault system, 00:05:58.059 --> 00:06:01.279 have aftershock durations – sequence durations that are 00:06:01.279 --> 00:06:05.849 typically about 10 years or so – maybe a bit more – tens of years. 00:06:05.849 --> 00:06:14.330 As we get to lower – slower loading rates, we see longer aftershock durations. 00:06:14.330 --> 00:06:18.279 And then they argue, when we get loading rates of about 00:06:18.279 --> 00:06:22.929 1/10 of a millimeter per year or so, then we can get aftershock durations 00:06:22.929 --> 00:06:25.610 that are thousands of years long. 00:06:25.610 --> 00:06:30.169 And they use this – this is one of their arguments to argue that 00:06:30.169 --> 00:06:35.309 the seismicity that we’re seeing in New Madrid region today 00:06:35.309 --> 00:06:41.800 are primarily aftershocks of the 1812 events there. 00:06:41.800 --> 00:06:46.460 So it’s important – this is an important curve to resolve this. 00:06:46.460 --> 00:06:51.710 I should point out that this line is actually a derivation of 00:06:51.710 --> 00:06:55.629 an equation that’s given in Dieterich et al. – 00:06:55.629 --> 00:07:01.160 or, Dieterich in JGR in 1999 – or, 1994, sorry. 00:07:02.249 --> 00:07:07.419 And of course, there’s local interest in knowing where the 1872 event was. 00:07:07.419 --> 00:07:10.879 And this is a map that I’m showing of kind of an expanded view of 00:07:10.879 --> 00:07:16.520 central Washington. You can see the Columbia River snaking through here. 00:07:16.520 --> 00:07:19.830 This is a Lake Chelan, which is a natural lake. 00:07:19.830 --> 00:07:24.169 And then these blue balloons show you the locations of some of 00:07:24.169 --> 00:07:31.169 the hydroelectric dams near the epicenter where we think the earthquake was now. 00:07:31.169 --> 00:07:36.719 And the Wells Dam – all of these have probably some issues associated 00:07:36.719 --> 00:07:41.999 with seismic safety, but in particular the Wells Dam located here 00:07:41.999 --> 00:07:48.029 has the greatest concern with it. And again, this kind of big black circle 00:07:48.029 --> 00:07:55.389 shows you the area identified by Bakun et al. as the most likely location 00:07:55.389 --> 00:08:00.020 of the 1872 earthquake based on seismic intensity data. 00:08:00.020 --> 00:08:06.650 And they estimated a magnitude range of 6-1/2 to 7 for that event. 00:08:06.650 --> 00:08:11.379 So having a good location for that event is very critical 00:08:11.379 --> 00:08:15.069 for the retrofit of some of these major infrastructure. 00:08:15.069 --> 00:08:19.110 This figure just shows you some – gives you some pictures of 00:08:19.110 --> 00:08:25.159 what these dams look like. And again, all of these dams were really 00:08:25.159 --> 00:08:30.569 built before we knew where the 1872 event was, and we really didn’t have a 00:08:30.569 --> 00:08:36.220 good appreciation of the hazard posed by that earthquake to these facilities. 00:08:37.740 --> 00:08:43.279 So the talk I’m going to present today is organized – about half of the talk, 00:08:43.279 --> 00:08:49.110 I’ll be presenting the evidence that suggests that the 1872 event 00:08:49.110 --> 00:08:53.440 occurred near Entiat, Washington, in central Washington. 00:08:53.440 --> 00:08:58.530 And these include things like the seismic intensity data that Bill Bakun looked at. 00:08:58.530 --> 00:09:02.029 A fault scarp that has been recently discovered 00:09:02.029 --> 00:09:04.520 and discussed by Brian Sherrod. 00:09:04.520 --> 00:09:09.120 Persistence of aftershocks after the 1872 event. 00:09:09.120 --> 00:09:12.490 And then the concentration of ground failures and water effects 00:09:12.490 --> 00:09:16.709 that resulted from the earthquake. And then, perhaps less compelling, 00:09:16.709 --> 00:09:21.529 but sort of interesting at least to me, is the reported directions of 00:09:21.529 --> 00:09:24.400 horizontal shaking agree with an Entiat epicenter, 00:09:24.400 --> 00:09:28.760 but it’s not as compelling as these other evidence. 00:09:28.760 --> 00:09:31.560 The second half of the talk, I’ll be talking about the aftershocks 00:09:31.560 --> 00:09:37.589 themselves and describing what we know about the modern seismicity 00:09:37.589 --> 00:09:41.560 cluster at Entiat and the fact that they’re very shallow earthquakes. 00:09:41.560 --> 00:09:44.830 They have mainly oblique thrust mechanisms. 00:09:44.830 --> 00:09:49.170 They have a typical kind of a garden variety b-value of 1. 00:09:49.170 --> 00:09:55.940 They have a constant seismicity rate and a low seismic moment rate. 00:09:55.940 --> 00:09:59.930 And then we’ll talk – I’ll talk about aftershock forecast modeling 00:09:59.930 --> 00:10:04.269 for three time periods of aftershock – or earthquakes 00:10:04.269 --> 00:10:07.940 in the Entiat area following the 1872 event. 00:10:07.940 --> 00:10:10.850 So the first is the immediate aftershocks 00:10:10.850 --> 00:10:14.320 in the first seven hours or so after the earthquake. 00:10:14.320 --> 00:10:20.910 And then we’ll look at earthquakes that were felt between 1900 and 1976. 00:10:20.910 --> 00:10:27.120 And then we’ll look at the instrumental catalog between 1976 and 2006. 00:10:27.120 --> 00:10:32.980 And the only reason we’re not looking at the earthquakes that happened right after 00:10:32.980 --> 00:10:37.800 the earthquake and before 1900 is that there were really none to talk about – 00:10:37.800 --> 00:10:44.779 at least none – no felt earthquakes, really, were reported in that interval, 00:10:44.779 --> 00:10:49.509 which I think reflects the low – very low density of population 00:10:49.509 --> 00:10:54.149 in that area of Washington until about 1900. 00:10:54.149 --> 00:10:57.800 So where did the 1872 event happen? 00:10:57.800 --> 00:11:03.160 This is kind of a laundry list of some of the major papers on the event. 00:11:03.160 --> 00:11:06.899 And the first one that was really published was by Milne, 00:11:06.899 --> 00:11:09.860 who only looked at felt reports in British Columbia. 00:11:09.860 --> 00:11:12.379 And he got a – he got an earthquake location 00:11:12.379 --> 00:11:15.850 in British Columbia, and I’ll show you that. 00:11:15.850 --> 00:11:21.779 Coombs et al. is a report by a – is a consultant report who is very thorough 00:11:21.779 --> 00:11:31.879 and extensive, but basically unpublished. So it hasn’t contributed to the literature. 00:11:31.879 --> 00:11:37.879 Malone and Bor published a paper in BSSA in 1979, which looked at 00:11:37.879 --> 00:11:44.000 most of the intensity data, but not all. And I’ll discuss their findings. 00:11:44.000 --> 00:11:49.301 Bill Bakun and his – and his group updated the location and the magnitude 00:11:49.301 --> 00:11:55.649 of the 1872 event using what I’m calling modal seismic intensities. 00:11:55.649 --> 00:11:58.240 And I’ll explain what those are. 00:11:58.240 --> 00:12:02.800 And Bill’s paper, I think, was really, really good and really gave us a – 00:12:02.800 --> 00:12:07.329 the first good idea of where the earthquake was and its magnitude. 00:12:07.329 --> 00:12:12.230 Brian – as I mentioned before, Brian Sherrod, a couple years ago now, 00:12:12.230 --> 00:12:18.279 gave a poster at the AGU meeting, which he talked about identifying 00:12:18.279 --> 00:12:24.240 a scarp on Lidar data that he has done some trenching on and looked at 00:12:24.240 --> 00:12:32.180 some of the trees that were killed – probably killed by that earthquake or 00:12:32.180 --> 00:12:35.000 grew after the earthquake that give us some 00:12:35.000 --> 00:12:38.380 age dating for the – for this fault scarp. 00:12:38.380 --> 00:12:45.260 And then I’ll be summarizing a paper – two papers by myself and others that 00:12:45.260 --> 00:12:50.280 summarize the evidence for the location of the 1872 earthquake, 00:12:50.280 --> 00:12:55.400 and the first of which is really based on a paper that was put together 00:12:55.410 --> 00:12:59.579 by Hopper and her colleagues in Golden but never published. 00:12:59.580 --> 00:13:05.680 It was submitted as a – to BSSA, but was never published. 00:13:05.680 --> 00:13:09.439 But it’s been referred to in the literature since, so I think – I felt it was 00:13:09.439 --> 00:13:15.589 important to try to document those – that information in the literature. 00:13:15.589 --> 00:13:19.379 And the second paper talks about the relationship 00:13:19.380 --> 00:13:24.380 about the 1872 event and the Entiat seismicity cluster. 00:13:25.220 --> 00:13:30.730 So this figure is from a paper by Malone and Bor in ’79, 00:13:30.730 --> 00:13:37.149 and it shows three place – this is – the A-asterisk is the 00:13:37.149 --> 00:13:40.779 location identified by Milne for the earthquake. 00:13:40.779 --> 00:13:44.470 And these two are two other locations that were 00:13:44.470 --> 00:13:47.519 also investigated by Malone and Bor. 00:13:47.519 --> 00:13:52.269 And although they had a site very close to where we think – now think 00:13:52.269 --> 00:13:57.279 the earthquake was, this location was disfavored by them, 00:13:57.279 --> 00:14:06.110 and they preferred a solution closer to the solution identified by Milne. 00:14:06.110 --> 00:14:11.880 And there’s a couple points to note about the Malone work. 00:14:11.880 --> 00:14:16.820 First, they assumed a fairly large magnitude of 7.4. 00:14:16.839 --> 00:14:22.410 And they also assumed a epicenter that’s 40 to 60 kilometers deep. 00:14:22.410 --> 00:14:26.160 And that was, I think, based on the experience that we’ve had 00:14:26.160 --> 00:14:29.720 of in-slab earthquakes and underneath Puget lowland 00:14:29.720 --> 00:14:34.420 in 1949 and 1965 of deeper events. 00:14:34.430 --> 00:14:38.949 And so they were thinking that this was likely a deeper event. 00:14:38.949 --> 00:14:43.620 I think the other reason they came up with preferring this location 00:14:43.620 --> 00:14:47.850 was that they did not use the complete set of intensity observations. 00:14:47.850 --> 00:14:52.480 They used a subset, and I think that pulled that 00:14:52.480 --> 00:14:55.759 epicenter away from where we think it is now. 00:14:55.759 --> 00:15:03.579 The circle shows – again, that’s the area that – defined by Bakun et al. for their – 00:15:03.579 --> 00:15:07.510 for their location, and the red dot is where I would 00:15:07.510 --> 00:15:12.860 prefer to – or, where I would suggest the earthquake occurred. 00:15:13.980 --> 00:15:17.360 So what are these modal seismic intensity data? 00:15:17.360 --> 00:15:21.069 These, again, were intensity data that were developed by Hopper et al. 00:15:21.069 --> 00:15:24.550 in this unpublished manuscript. 00:15:24.550 --> 00:15:29.110 These were actually published by Bill Bakun and his group, 00:15:29.110 --> 00:15:34.550 but they didn’t really discuss very much how they were developed. 00:15:34.550 --> 00:15:38.439 And, well, I’ve color-coded them – 00:15:38.439 --> 00:15:41.639 these observations by the Did You Feel It scale. 00:15:41.639 --> 00:15:46.480 So you can see the high intensities here and the lower intensities here. 00:15:46.480 --> 00:15:49.749 Probably very hard to see it, but there’s a – there’s a large 00:15:49.749 --> 00:15:55.079 radius here of – which, more or less, goes around the felt intensities 00:15:55.079 --> 00:15:59.100 of about a 650-kilometer radius. 00:15:59.100 --> 00:16:02.869 And there’s a shorter – a smaller-radius circle about – 00:16:02.869 --> 00:16:08.879 of about 270-kilometer radius that includes all of these square points, 00:16:08.879 --> 00:16:13.850 which are places where we have observations of some type of 00:16:13.850 --> 00:16:20.819 ground failure – a landslide or liquefaction or water effects, but for 00:16:20.820 --> 00:16:28.460 which Hopper and her group did not – did not assign seismic intensities. 00:16:29.120 --> 00:16:34.720 So let’s dive into what these modal seismic intensities are a bit more. 00:16:34.720 --> 00:16:42.970 So what Hopper and her co-workers did is that they used 10 different 00:16:42.970 --> 00:16:51.089 estimates of the intensities to each newspaper report at a community. 00:16:51.089 --> 00:16:55.699 So five of these were from the five USGS authors of that 00:16:55.699 --> 00:17:00.779 unpublished Hopper paper, with one of the authors making a second estimate. 00:17:00.779 --> 00:17:05.410 So that – we have a total of six assignments from USGS. 00:17:05.410 --> 00:17:10.539 And then they used four assignments by consultants, which were – 00:17:10.540 --> 00:17:15.020 are in the gray literature, but not really formally published. 00:17:15.840 --> 00:17:19.990 They didn’t assign any intensities to observations that were solely of 00:17:19.990 --> 00:17:24.449 ground failure and water effects because these effects can happen 00:17:24.449 --> 00:17:29.309 at intensities lower than you would assign an intensity to 00:17:29.309 --> 00:17:37.360 if you had some other estimate of building damage or human reaction. 00:17:38.940 --> 00:17:41.640 So what they did with these – once they had the 10 estimates, 00:17:41.640 --> 00:17:45.950 they assigned the mode at the 10 estimates as the intensity, 00:17:45.950 --> 00:17:49.770 but if four of the estimates were higher than the mode, 00:17:49.770 --> 00:17:53.860 then they chose the next intensity higher above the mode. 00:17:54.880 --> 00:18:01.780 So let’s discuss these for just a moment. So this methodology is assuming 00:18:01.780 --> 00:18:07.480 that all of the intensity assignments are of equal value and independent. 00:18:07.480 --> 00:18:14.870 But the Coombs et al. report was commissioned after these three other 00:18:14.870 --> 00:18:22.670 consulting reports provided pretty widely disagreeing intensity estimates. 00:18:22.670 --> 00:18:27.950 So Coombs were brought in as kind of an uber-consultant to look at these 00:18:27.950 --> 00:18:32.669 reports and look at the evidence and derive their own estimates. 00:18:32.669 --> 00:18:38.140 So it’s probably not fair to say that it’s independent of the other – at least 00:18:38.140 --> 00:18:45.060 these three other consulting reports. And probably not equal to them, either. 00:18:46.130 --> 00:18:50.760 But one of the good things that – one of the strong points of the 00:18:50.770 --> 00:18:56.000 Hopper assignments is that, by not assigning intensities 00:18:56.000 --> 00:19:00.490 to those places that only had observations of ground failures or 00:19:00.490 --> 00:19:06.070 water effects, that really reduced the scatter in the intensity assignments. 00:19:06.070 --> 00:19:11.880 And so the intensity assignments that were derived by Hopper et al. 00:19:11.880 --> 00:19:16.279 and used by Bakun in his paper are really not very different from 00:19:16.279 --> 00:19:21.800 those reported by Coombs et al. And Bill noted that in his paper as well. 00:19:21.800 --> 00:19:29.529 So I think whether – you might have some disagreement with the way 00:19:29.529 --> 00:19:34.299 Hopper et al. did things, but I think, in the end of the day, their answer 00:19:34.300 --> 00:19:38.080 is not very different from the one that was derived by Coombs. 00:19:39.340 --> 00:19:43.480 So here’s a blow-up view of these intensity data. 00:19:43.480 --> 00:19:46.460 Maybe it’s a little bit easier to see them now. 00:19:46.460 --> 00:19:51.520 And this is the intensity – well, this is the region 00:19:51.520 --> 00:19:56.049 of 95% confidence assigned by Bill Bakun’s group. 00:19:56.049 --> 00:20:00.080 And here is their – the triangle is their intensity center. 00:20:00.080 --> 00:20:03.059 And you can see the highest intensities are close, 00:20:03.060 --> 00:20:08.600 or proximal, to this – to this region, which makes sense. 00:20:11.409 --> 00:20:14.560 So the other piece of evidence – another piece of evidence 00:20:14.560 --> 00:20:17.940 is this recently identified Lidar scarp. 00:20:17.940 --> 00:20:23.700 And I’ll show you a kind of a map that puts this scarp into better regional 00:20:23.700 --> 00:20:34.470 context, but this is a – this is a map of the relief of the Lidar elevation data. 00:20:34.470 --> 00:20:38.419 And I’ve sort of highlighted the scarp, but actually the scarp is 00:20:38.419 --> 00:20:41.799 just to the south of my highlighting. And so if you look, you can see 00:20:41.799 --> 00:20:46.570 a dark line that’s just south of the red line, which is the scarp. 00:20:46.570 --> 00:20:52.190 And in this – proximal – this scale bar is 2 kilometers, and very proximal to this 00:20:52.190 --> 00:20:58.610 fault scarp, or this scarp, we can see at least five landslides that occurred. 00:20:58.610 --> 00:21:03.710 And two of these are fairly large, and one of them – at least one of them 00:21:03.710 --> 00:21:08.769 near the headwall of the scarp, the sagging produced a little pond. 00:21:08.769 --> 00:21:16.250 And in 1884, people observed that there were dead trees in the pond. 00:21:16.250 --> 00:21:22.340 And it’s suggested that the pond formed before 1884 00:21:22.340 --> 00:21:26.380 and that the trees died before then. 00:21:26.380 --> 00:21:30.730 The trees growing on these parts of the landslide have been 00:21:30.730 --> 00:21:35.720 dated with a tree ring analysis and dated to about 1885, 00:21:35.720 --> 00:21:42.240 so about 12 years or so after the 1872 event. 00:21:42.240 --> 00:21:44.490 And then there were – Brian and his co-workers – 00:21:44.490 --> 00:21:48.320 and I should mention that this is Brian’s work, and hopefully Brian – 00:21:48.320 --> 00:21:53.460 he’s working on writing these results up, and he’ll come down and give us a talk 00:21:53.460 --> 00:21:57.190 when he’s finished, but he’s – they dug a couple of trenches across 00:21:57.190 --> 00:22:02.540 the scrap in this location. And I’ll talk about the results from that as well. 00:22:02.540 --> 00:22:05.720 So I talked a little bit about the tree ring dating. 00:22:05.720 --> 00:22:11.890 The radiocarbon dating of the colluvium in the trenches suggests an earthquake 00:22:11.890 --> 00:22:18.840 between 1840 and 1890, but we know there wasn’t a bigger event after 1872, 00:22:18.840 --> 00:22:27.480 so it’s really – this constrains the age of the scarp to about 1840 and 1872 or so. 00:22:27.480 --> 00:22:34.300 The scarp reflects offset on a northwest-dipping plane. 00:22:34.309 --> 00:22:36.970 And the scarp height is about 1 or 2 meters high. 00:22:36.970 --> 00:22:42.779 So if you use Wells and Coopersmith’s relationship, you would – you would 00:22:42.779 --> 00:22:49.269 estimate a magnitude of something between 6-1/2 up to 6.7, which is 00:22:49.269 --> 00:22:55.330 at the – it’s within the range estimated by Bakun for the earthquake. 00:22:55.330 --> 00:22:59.830 However, the length of the fault scarp is a little bit short for a 00:22:59.830 --> 00:23:05.320 magnitude earthquake of that – and in fact, the 6-kilometer-long scarp 00:23:05.320 --> 00:23:10.180 would only be consistent with about a magnitude 5.8 event. 00:23:11.410 --> 00:23:16.960 But important for our – for our work here is that the – 00:23:16.960 --> 00:23:20.650 one of the trenches identified at least two events. 00:23:20.650 --> 00:23:23.909 And these two events are both post-Mazama ash, 00:23:23.909 --> 00:23:28.600 so they’re roughly within the past 8,000 years or so. 00:23:28.600 --> 00:23:32.690 So we seem to be getting events on this structure every few – 00:23:32.690 --> 00:23:35.580 or every several thousand years. 00:23:37.160 --> 00:23:43.640 Another piece of evidence supporting an Entiat location for the earthquake 00:23:43.640 --> 00:23:49.600 is that – how long the aftershocks from that earthquake were felt. 00:23:49.600 --> 00:23:54.480 And another piece of information from that aftershock information is that, 00:23:54.480 --> 00:24:00.000 because so many aftershocks were felt, I think we can be fairly clear – 00:24:00.000 --> 00:24:04.630 confident that this must be a crustal earthquake and not a deeper earthquake. 00:24:04.630 --> 00:24:08.260 And this was noted by Bill Bakun and his group as well that the 00:24:08.260 --> 00:24:12.309 number of aftershocks is consistent with a crustal earthquake. 00:24:12.309 --> 00:24:18.059 But immediately after the main shock, large aftershocks were felt for 00:24:18.059 --> 00:24:23.230 about the first hour over much of the felt area of the main shock. 00:24:23.230 --> 00:24:27.230 But within the next day or so, that area really shrunk. 00:24:27.230 --> 00:24:29.860 And within a few months, we’re looking at this area – 00:24:29.860 --> 00:24:37.960 and after a year, we only see aftershocks in this area of the most likely 00:24:37.960 --> 00:24:43.019 location of the earthquake based on the intensity data shown in here. 00:24:43.020 --> 00:24:49.480 So that, again, supports an Entiat location for the 1872 event. 00:24:50.600 --> 00:24:56.840 And a third piece of evidence – or, a fourth piece of evidence, rather, 00:24:56.840 --> 00:25:01.580 is that the – many of the observed ground failures 00:25:01.590 --> 00:25:05.100 and water effects were concentrated in the Entiat area. 00:25:05.100 --> 00:25:10.110 So this is the Columbia River snaking through the map here. 00:25:10.110 --> 00:25:13.840 Scale here is about a 20-kilometer scale bar. 00:25:13.840 --> 00:25:17.610 Here’s the proposed epicenter location. 00:25:17.610 --> 00:25:22.289 This is the location of one of the intensity VIII observations here. 00:25:22.289 --> 00:25:26.539 The other one down here in Wenatchee. Here is the location of the 00:25:26.540 --> 00:25:30.750 Spencer Canyon Scarp that I showed you on the Lidar image. 00:25:31.980 --> 00:25:35.300 Here’s the location of – it’s called the Ribbon Cliff failure 00:25:35.300 --> 00:25:41.680 where much of the cliff sloughed off and fell into the Columbia River. 00:25:41.680 --> 00:25:47.640 And here’s the location of an observation of fountaining that was 00:25:47.640 --> 00:25:52.600 up to 20 or 30 feet high after the earthquake. 00:25:53.420 --> 00:26:00.510 The rectangle serves two purposes. It kind of outlines the area that would 00:26:00.510 --> 00:26:05.130 produce a magnitude 6.8 earthquake. That’s kind of the area 00:26:05.130 --> 00:26:09.100 that you’re looking at. And the northwest dip of the 00:26:09.100 --> 00:26:13.000 Spencer Canyon Scarp suggests that many of these features were 00:26:13.000 --> 00:26:18.160 likely to have been in the hanging wall block of the thrust fault. 00:26:18.880 --> 00:26:23.160 But the area of the rectangle also is the area that sort of 00:26:23.169 --> 00:26:29.280 outlines the distribution – the surface area of the seismicity cluster. 00:26:29.280 --> 00:26:35.480 So it shows that – it helps illustrate that a lot of these features – the scarp, 00:26:35.480 --> 00:26:42.400 a lot of the prominent ground failures, the largest intensities, are co-located 00:26:42.420 --> 00:26:47.920 with this Entiat seismicity cluster. And I’ll show that in a different slide as well. 00:26:49.080 --> 00:26:52.210 So here’s a view of the Ribbon Cliff landslide. 00:26:52.210 --> 00:27:00.100 And this was published – a nice study of the landslide was published in BSSA. 00:27:00.100 --> 00:27:04.480 But you can see it still looks fairly fresh, and you can even see parts of the 00:27:04.480 --> 00:27:10.090 landslide sticking up above the surface of Columbia River. 00:27:10.090 --> 00:27:14.380 And there were a few people living in the area at the time, 00:27:14.380 --> 00:27:20.510 and they reported that the cliff fell into the river and that the river was actually 00:27:20.510 --> 00:27:26.820 dammed by the landslide, and people downstream of the landslide reported 00:27:26.820 --> 00:27:32.940 that the flow of the Columbia River actually stopped for a few hours. 00:27:35.090 --> 00:27:38.620 So that’s all, I think, fairly compelling evidence 00:27:38.620 --> 00:27:43.620 for an Entiat location for the epicenter. Here is some evidence that I think is 00:27:43.620 --> 00:27:49.059 interesting, but it’s not as compelling and not as restrictive as this other evidence. 00:27:49.060 --> 00:27:54.760 And that’s the directions that horizontal shaking were reported. 00:27:54.760 --> 00:27:56.320 And so here’s a map view. 00:27:56.330 --> 00:28:00.139 Here’s where we think the epicenter was in this location. 00:28:00.139 --> 00:28:04.730 And all these little bars show you the locations and the orientation 00:28:04.730 --> 00:28:10.889 that horizontal shaking was reported from the – from the main shock. 00:28:10.889 --> 00:28:13.010 So we have 23 different reports. 00:28:13.010 --> 00:28:18.730 They have a fairly uniform azimuthal distribution along this 00:28:18.730 --> 00:28:21.740 proposed epicenter. And you can see that some of them 00:28:21.740 --> 00:28:26.370 look like they’re fairly radial from the epicenter. 00:28:26.370 --> 00:28:30.840 I’ve just drawn in little linear ray paths. 00:28:32.120 --> 00:28:33.870 But some of them look like they’re transverse. 00:28:33.870 --> 00:28:41.450 So we’re going to interpret these as either Rayleigh waves – 00:28:41.450 --> 00:28:45.500 radio Rayleigh waves or transverse Love waves. 00:28:45.500 --> 00:28:50.509 And we calculated the misfit we would get from assuming – 00:28:50.509 --> 00:28:55.220 that those assumptions and the assumption of linear ray paths. 00:28:55.220 --> 00:29:03.880 Now, I should also point out that the directions of shaking are pretty much 00:29:03.880 --> 00:29:09.620 restricted to north-south, east-west, north-south, northeast-southwest, 00:29:09.620 --> 00:29:16.940 or northwest-southeast. So about 45-degree increments. 00:29:18.060 --> 00:29:24.620 But for the record, and just – if you – if you calculate the misfits, you do get 00:29:24.620 --> 00:29:32.450 the smallest misfit at Entiat if you – under this – under these assumptions. 00:29:32.450 --> 00:29:41.289 Now, this whole area shown in this kind of broad eclipse – ellipse has misfits 00:29:41.289 --> 00:29:46.350 of between about 20 and 24 degrees. And that doesn’t sound very impressive, 00:29:46.350 --> 00:29:51.539 but if you go outside of it, you get misfits of 30 or 31 degrees. 00:29:51.539 --> 00:29:56.149 So there is a difference between the misfits that you get in this kind of – 00:29:56.149 --> 00:29:58.960 in this kind of region versus outside of this region. 00:29:58.960 --> 00:30:03.929 And it’s kind of – the region is elongated to the north, and we 00:30:03.929 --> 00:30:09.720 appear to have more control in the east and west direction of the epicenter. 00:30:09.720 --> 00:30:16.240 So that’s kind of evidence that’s consistent with an Entiat location for the 00:30:16.240 --> 00:30:23.420 epicenter, but I wouldn’t argue strongly that it constrains the epicenter closely. 00:30:24.440 --> 00:30:30.990 Now I want to move into – hopefully convinced you that the 1872 event was 00:30:30.990 --> 00:30:37.240 likely in the Entiat area, and now I want to try to answer the question is, could 00:30:37.240 --> 00:30:43.600 the ongoing Entiat seismicity cluster represent aftershocks of that earthquake? 00:30:44.680 --> 00:30:49.280 And again, here’s our map of the crustal seismicity of Washington. 00:30:49.280 --> 00:30:52.580 Here is our Entiat location for the earthquake. 00:30:52.580 --> 00:30:55.950 And here’s the seismicity cluster that we’ll be looking at. 00:30:55.950 --> 00:31:00.940 And we’ll expand that view in this slide. 00:31:00.940 --> 00:31:03.830 So here’s the seismicity – Entiat seismicity cluster. 00:31:03.830 --> 00:31:10.260 Here’s the town of Entiat. Here’s the Columbia River, Lake Chelan. 00:31:10.260 --> 00:31:12.779 So here’s the cluster. The focal mechanisms – 00:31:12.779 --> 00:31:17.250 again, they’re mainly thrust fault mechanisms or oblique thrust. 00:31:17.250 --> 00:31:20.559 There’s a few strike-slip events in there. 00:31:20.559 --> 00:31:26.260 The clusters trend – seem to trend sort of to the northeast. 00:31:26.260 --> 00:31:32.070 And it’s subparallel to this lineation in the aeromagnetic data. 00:31:32.070 --> 00:31:36.009 Here’s the location of the Spencer Canyon Scarp 00:31:36.009 --> 00:31:38.429 that was identified from the Lidar. 00:31:38.429 --> 00:31:44.309 Here’s the Ribbon Canyon landslide, which appeared to dam the Columbia 00:31:44.309 --> 00:31:49.519 River, so – and the highest intensities are over here and just down here. 00:31:49.519 --> 00:31:54.519 So again, I’ve drawn a rectangle around the seismicity cluster. 00:31:54.519 --> 00:31:57.210 You can quibble, I’m sure, with how I’ve drawn it. 00:31:57.210 --> 00:32:01.970 I’ve tried to capture some of the seismicity in the Entiat area that was 00:32:01.970 --> 00:32:08.399 reported before we had a good regional network in the location up in here. 00:32:08.399 --> 00:32:12.039 So you could probably make this a little bit smaller if you’d like, 00:32:12.039 --> 00:32:16.269 but again, the area of the cluster seems to be compatible 00:32:16.269 --> 00:32:20.880 with a magnitude – moment magnitude 6.8 earthquake. 00:32:22.309 --> 00:32:26.100 The depth distribution of these aftershocks is very – 00:32:26.100 --> 00:32:31.789 they have kind of a broad peak between 3 and 8 kilometers or so. 00:32:31.789 --> 00:32:37.320 There’s a very sharp spike that’s shallow that we don’t think is real 00:32:37.320 --> 00:32:41.860 because the errors in these depths are several kilometers, 00:32:41.860 --> 00:32:45.190 whereas the errors in these depths are just a kilometer or two. 00:32:45.190 --> 00:32:49.480 So we think this is a spurious peak, and that the cluster is really limited to 00:32:49.480 --> 00:32:55.080 sort of the 3- to 8-kilometer depth range, and it depends on which – a little bit 00:32:55.080 --> 00:33:01.720 what – which earthquakes you look at. Cluster 1 is these earthquakes that are 00:33:01.730 --> 00:33:06.309 sort of along the magnetic anomaly, and cluster 2 is maybe – sub-cluster 2 00:33:06.309 --> 00:33:13.450 is in this region. But the bottom line is that these are very shallow earthquakes. 00:33:13.450 --> 00:33:17.880 If you – if you look at the cumulative number of these earthquakes, 00:33:17.880 --> 00:33:24.400 you see that the instrument record for the past 40 years seems to be complete 00:33:24.400 --> 00:33:31.710 down to about a magnitude 1-1/2 or so. The b-value is very ordinary of 1. 00:33:31.710 --> 00:33:35.110 And we’ve done a – just regressing this curve where we’re doing a 00:33:35.110 --> 00:33:39.549 maximum likelihood method, we get 1 in both cases. 00:33:39.549 --> 00:33:44.060 We really have no events greater than a magnitude 3.7 00:33:44.060 --> 00:33:49.160 in 40 years of observations, which is interesting. 00:33:49.920 --> 00:33:53.279 This – so I showed you the – kind of the magnitude-time 00:33:53.279 --> 00:33:56.309 distribution of the earthquakes in the cluster. 00:33:56.309 --> 00:34:01.440 If we look at their cumulative number, we see a very linear rate of seismicity 00:34:01.440 --> 00:34:05.360 for the past 40 years. We don’t see any decrease. 00:34:05.360 --> 00:34:08.580 And we also see a more or less linear increase 00:34:08.580 --> 00:34:12.710 in the moment rate that reflects this. 00:34:12.710 --> 00:34:16.970 And I’ll show you an expanded view of the moment rate here. 00:34:16.970 --> 00:34:22.850 If you do a kind of a Brune-style inversion of this moment rate that you 00:34:22.850 --> 00:34:28.820 infer from this curve for a fault plane, the dimensions that would produce 00:34:28.820 --> 00:34:35.990 the 1872 event, you would get a slip rate of about 0.004 millimeters per year. 00:34:35.990 --> 00:34:41.140 So it’s a very, very low slip rate. And that reflects, again, 00:34:41.140 --> 00:34:45.610 the absence of seismicity that’s greater than 4. 00:34:45.610 --> 00:34:51.600 Just very low rates of – or no – large events. 00:34:52.640 --> 00:34:58.800 So now that’s a little description of what the aftershock cluster looks like. 00:34:58.800 --> 00:35:03.960 And now I want to look at whether these earthquakes and other earthquakes 00:35:03.970 --> 00:35:09.530 might be aftershocks of the 1872 event. So, again, we’re going to look at 00:35:09.530 --> 00:35:13.260 the events in the first seven hours or so after the earthquake. 00:35:13.260 --> 00:35:17.140 We’re going to look at felt events from 1900 to 1976. 00:35:17.140 --> 00:35:24.150 And then, the last 40 years or so of instrumentally recorded events. 00:35:24.150 --> 00:35:27.690 And we’re going to use Omori’s law for this, as implemented by 00:35:27.690 --> 00:35:33.270 Reasenberg and Jones. We’re going to use the median 00:35:33.270 --> 00:35:37.640 onshore subduction zone region for the aftershock parameters, 00:35:37.640 --> 00:35:41.840 and this is from a paper by Morgan Page and her co-authors. 00:35:41.840 --> 00:35:46.080 And we’re going to use a region that’s appropriate for Washington state. 00:35:46.080 --> 00:35:50.960 And for those people who are really into aftershock forecast modeling, 00:35:50.960 --> 00:35:57.530 these are the parameters that we used. And I’ll just note that the b-value 00:35:57.530 --> 00:36:02.110 is 1 for this region, and that’s exactly what we see for the cluster. 00:36:02.110 --> 00:36:07.540 And the p-value is very close to 0.8, which is a very typical value. 00:36:09.280 --> 00:36:13.200 So the – what we’re going to use for the immediate aftershocks 00:36:13.200 --> 00:36:20.130 of the 1872 main shock is an eyewitness report from Wenatchee, which is about 00:36:20.130 --> 00:36:26.060 20 kilometers south of the Lidar scarp that I showed you. 00:36:28.040 --> 00:36:34.060 The eyewitness was awoken by the earthquake. He was in his house. 00:36:34.070 --> 00:36:40.250 And he reported that the shocks, which lasted until 5:00 Sunday 00:36:40.250 --> 00:36:44.980 morning – and the earthquake was around 10:00 p.m. or so – 00:36:44.980 --> 00:36:51.370 so about – for seven hours, there were 64 – he felt 64 shocks, 00:36:51.370 --> 00:36:54.260 and eight of these were very severe. 00:36:56.260 --> 00:37:02.240 So for the felt earthquakes between 18 – I’m sorry, 00:37:02.240 --> 00:37:08.310 1900 and 1976 – I pulled these together from various sources, 00:37:08.310 --> 00:37:14.900 including the NOAA intensity database and other public sources. 00:37:14.900 --> 00:37:20.990 And I plotted these as a function of the Modified Mercalli Intensity 00:37:20.990 --> 00:37:26.070 of the felt report – the Maximum Modified Mercalli Intensity 00:37:26.070 --> 00:37:31.360 as a function of year, so from 1900 to 1980. 00:37:31.360 --> 00:37:37.200 And you can see from this plot as well as this cumulative plot that this 00:37:37.200 --> 00:37:42.370 does not look like a complete catalog. We have long intervals of time 00:37:42.370 --> 00:37:45.950 where we’re not seeing any events, or very few events. 00:37:45.950 --> 00:37:50.140 And so I think it’s safe to say that this is not a complete catalog, 00:37:50.140 --> 00:37:56.070 but hopefully these events that are intensity V or even intensity VI, 00:37:56.070 --> 00:38:04.090 which we looked at an empirical magnitude versus MMI relations. 00:38:04.090 --> 00:38:08.640 And these would correspond to magnitude 4 or maybe these 00:38:08.640 --> 00:38:13.860 intensities VI would correspond to magnitude 5 earthquakes. 00:38:14.410 --> 00:38:21.140 So hopefully those earthquakes – those stronger earthquakes are more complete. 00:38:21.140 --> 00:38:26.450 And then, of course, we have the recorded seismicity for the past 40 years. 00:38:26.450 --> 00:38:28.820 This is a slide that I’ve shown you before. 00:38:28.820 --> 00:38:34.430 And, again, we’re going to focus on the magnitude 2 and greater earthquakes 00:38:34.430 --> 00:38:40.200 because that’s where the instrumental catalog seems to be complete. 00:38:40.200 --> 00:38:45.620 So this is a summary of the aftershock modeling, and there’s three figures – 00:38:45.620 --> 00:38:51.710 one for the immediate aftershocks, one for the felt events 1900 to 1976, 00:38:51.710 --> 00:38:56.920 and then for the past 40 years. And each one of these diagrams is the same. 00:38:56.920 --> 00:39:01.850 They show the expected number of aftershocks on the vertical axis as a 00:39:01.850 --> 00:39:08.320 function of the magnitude of the 1872 event, which is assumed. 00:39:08.320 --> 00:39:13.060 And then these curves show you the predicted number of aftershocks 00:39:13.060 --> 00:39:17.040 that you would get for any magnitude range. 00:39:17.940 --> 00:39:24.120 So, and the gray – this gray bar shows you the magnitude range that Bill Bakun 00:39:24.120 --> 00:39:31.120 and his group developed for the 1872 event between 6-1/2 and 7. 00:39:31.120 --> 00:39:34.940 And this darker gray box is the estimate of the magnitude that we 00:39:34.940 --> 00:39:42.950 derived from the height of the scarp identified from the Lidar information. 00:39:42.950 --> 00:39:50.200 So this is the – this is the observations from Wenatchee of 64 events, so that this 00:39:50.200 --> 00:39:55.300 bar here is 64 events and eight stronger events. 00:39:56.000 --> 00:40:01.440 And so this curve is for intensity II felt at Wenatchee would correspond to 00:40:01.450 --> 00:40:06.620 magnitude 3 earthquakes. So within the uncertainties, we could – 00:40:06.620 --> 00:40:10.760 we could certainly explain the observations of the numbers of 00:40:10.760 --> 00:40:16.300 aftershocks with the magnitude range that Bill Bakun and his group proposed. 00:40:16.300 --> 00:40:19.490 And if the stronger earthquakes felt in Wenatchee correspond to 00:40:19.490 --> 00:40:24.760 magnitude 4 and larger earthquakes, we see some overlap there as well. 00:40:24.760 --> 00:40:29.750 So we could conclude that the immediate aftershock data seem to 00:40:29.750 --> 00:40:35.240 be consistent with the magnitude 6-1/2 to 7 earthquake. 00:40:35.240 --> 00:40:40.150 Now, looking at the felt – this next region – this next time interval 00:40:40.150 --> 00:40:48.440 between 1900 and 1976, focusing on the stronger of these felt earthquakes – 00:40:48.440 --> 00:40:54.240 the ones that were either at intensity V or VI – so the – 00:40:54.240 --> 00:40:58.070 again, these correspond to magnitude 4 or maybe 5 earthquakes. 00:40:58.070 --> 00:41:02.640 We can see the predicted number of those earthquakes, 00:41:02.640 --> 00:41:10.730 that also seems to match the magnitude range defined by the intensity data. 00:41:10.730 --> 00:41:15.120 And if we look at the – at recorded seismicity for the past 40 years, 00:41:15.120 --> 00:41:18.670 and now, again – now we can just look at all events, say, 00:41:18.670 --> 00:41:22.960 larger than magnitude 2 or magnitude 2-1/2 or 3, we can see 00:41:22.960 --> 00:41:29.200 the events that are magnitude 3 or 3-1/2, they kind of butt against the 00:41:29.200 --> 00:41:36.030 lower region of the magnitude range preferred by Bakun et al. 00:41:36.030 --> 00:41:40.460 The magnitude 2 and greater earthquakes are a little bit deficient. 00:41:40.460 --> 00:41:45.270 We don’t see as many as we might expect to see for that magnitude range. 00:41:45.270 --> 00:41:49.410 And the magnitude 3-1/2 earthquakes are also a little deficient. 00:41:49.410 --> 00:41:57.470 But overall, I think the bottom line is that, to answer the question, can these 00:41:57.470 --> 00:42:03.140 earthquakes represent aftershocks of the 1872 earthquake, I think – I don’t 00:42:03.140 --> 00:42:07.420 think we can rule out that possibility with the information that we have. 00:42:07.420 --> 00:42:11.420 Now, it is possible – and I don’t want to argue that some of these 00:42:11.420 --> 00:42:14.990 earthquakes don’t represent some ongoing deformation, 00:42:14.990 --> 00:42:20.220 but that’s not really required by the aftershock forecast modeling. 00:42:21.420 --> 00:42:24.660 So this was – I should say that this was not the conclusion 00:42:24.660 --> 00:42:30.480 I expected to find when I started this. I expected this not to be the case. 00:42:30.480 --> 00:42:33.980 So then I started to wonder, well, where do we see – 00:42:33.980 --> 00:42:41.840 are there other examples of – good examples of aftershocks sequences 00:42:41.840 --> 00:42:46.320 that have lasted 100 years or more? And there is at least one. 00:42:46.320 --> 00:42:51.140 And this is from an earthquake in Japan that was actually used by 00:42:51.140 --> 00:42:56.710 Omori back in his day to help establish Omori’s law. 00:42:56.710 --> 00:43:02.320 But Utsu published a paper over 20 years ago now, in which he 00:43:02.320 --> 00:43:07.740 showed that that aftershock sequence has persisted for over 100 years. 00:43:08.740 --> 00:43:12.740 So that’s a – that’s a nice example. And then, Dieterich, you know, 00:43:12.750 --> 00:43:20.120 in his JGR paper from 1994, developed theory that suggests that the relationship 00:43:20.120 --> 00:43:24.740 between the recurrence intervals of the main shock and the duration 00:43:24.740 --> 00:43:29.680 of the aftershock sequences that they produce is a linear one. 00:43:29.680 --> 00:43:35.200 And they have a proportion – a constant of proportionality that 00:43:35.200 --> 00:43:39.880 suggests that the recurrence intervals are between – 00:43:39.890 --> 00:43:45.290 and these are some earthquake sequences that Jim plotted mainly 00:43:45.290 --> 00:43:50.750 for California, Alaska, and, I think, Hawaii. 00:43:50.750 --> 00:43:57.460 But the constant proportionality is such that the recurrence interval 00:43:57.460 --> 00:44:00.370 between main shocks is somewhere between 00:44:00.370 --> 00:44:06.200 10 and 40 times longer than the duration of the aftershocks. 00:44:07.170 --> 00:44:10.440 And this is a result of rate-state theory, and it’s – 00:44:10.440 --> 00:44:18.780 and it’s also – happens for low stress drop for main shocks. 00:44:18.780 --> 00:44:23.300 But here – so the red horizontal bar – so this is the Entiat – 00:44:23.310 --> 00:44:28.090 this would be the Entiat cluster at 145 years. 00:44:28.090 --> 00:44:31.900 So that intersects Jim’s curves in this location. 00:44:31.900 --> 00:44:36.580 And if we go down, Jim’s series suggests that 00:44:36.580 --> 00:44:39.410 we would have main shock recurrence intervals of 00:44:39.410 --> 00:44:45.260 thousands of years – maybe one to two – 1,000 to 5,000 years. 00:44:45.260 --> 00:44:48.450 And that’s consistent with the observation that there had been 00:44:48.450 --> 00:44:54.230 two post-Mazama events in the trench that Brian Sherrod 00:44:54.230 --> 00:44:58.680 and his workers – co-workers have identified. 00:44:58.680 --> 00:45:08.290 These are – again, they are mainly higher slip rates, but this point is the 00:45:08.290 --> 00:45:13.430 example from Japan that I showed on the last – and it should be updated now. 00:45:13.430 --> 00:45:15.940 It should be moved a little bit higher. 00:45:15.940 --> 00:45:19.420 But I don’t know – I haven’t had a chance to go back to see 00:45:19.420 --> 00:45:24.540 whether that sequence is still going on, if the evidence suggests that. 00:45:26.040 --> 00:45:30.760 So then I thought, well, if this is a long-lived aftershock sequence, why are 00:45:30.760 --> 00:45:35.600 these events – why are we seeing this one, and why are they so uncommon? 00:45:35.600 --> 00:45:41.370 Because as I noted before, the recurrence intervals are 40 – 10 to 40 times longer 00:45:41.370 --> 00:45:47.140 than the aftershock durations. That suggests that more than 90% of the 00:45:47.140 --> 00:45:53.740 time, the seismicity along the fault zone is going to be at background level. 00:45:53.750 --> 00:46:00.870 So perhaps that’s the explanation that, most faults having long recurrence intervals, it’s 00:46:00.870 --> 00:46:06.060 unlikely that we’ll see the aftershock sequence because it will have decayed away. 00:46:06.960 --> 00:46:12.560 It’s also very challenging for us to get good observations of long-lived 00:46:12.560 --> 00:46:17.000 aftershock sequences because – because these are long-lived, or long-length, 00:46:17.000 --> 00:46:23.000 they live – you know, they’re active for two or three or four or more generations. 00:46:23.000 --> 00:46:30.640 So it’s a very challenging issue to monitor such long behavior. 00:46:30.640 --> 00:46:38.080 And the other factor is that many of these long-lived sequences tend to be far 00:46:38.090 --> 00:46:42.100 from plate boundaries, which tend to – tend to be not as well-instrumented. 00:46:42.100 --> 00:46:46.500 So they’re not going to be as easy to define. 00:46:48.700 --> 00:46:53.110 So the other question is, okay, so if the – if the Entiat cluster 00:46:53.110 --> 00:46:57.110 earthquakes are not aftershocks, well, what are they? 00:46:57.110 --> 00:47:00.520 And the obvious answer is, well, they’re just background seismicity 00:47:00.520 --> 00:47:07.200 along that fault system produced by ongoing deformation. 00:47:08.220 --> 00:47:10.160 Well, let’s look at that for a minute. 00:47:10.170 --> 00:47:15.250 So as I showed before, if you – if you do a moment rate inversion 00:47:15.250 --> 00:47:20.120 for the slip rate, you get a very low number for the slip rate. 00:47:21.030 --> 00:47:26.020 If you look at the local GPS data, and this is from a very nice fairly recent 00:47:26.030 --> 00:47:32.360 paper by McCaffrey and others, if you look at the local GPS data, 00:47:32.360 --> 00:47:38.010 you get an estimate in this region of 2 to 3 nanostrains, which suggests 00:47:38.010 --> 00:47:44.690 a deformation rate of maybe – of a few millimeters per year 00:47:44.690 --> 00:47:48.620 over a baseline of about 100 kilometers’ length. 00:47:50.520 --> 00:47:55.620 That estimate is consistent with the low deformation rate that we see, 00:47:55.620 --> 00:48:03.280 or don’t see, rather, in InSAR data in the Entiat area that basically looked at 00:48:03.280 --> 00:48:08.760 10 years of interferograms – about 90 interferograms – and didn’t see 00:48:08.760 --> 00:48:16.680 any discernable deformation. So those estimates are compatible. 00:48:18.060 --> 00:48:22.980 But there are other faults in the Pacific Northwest that have low 00:48:22.980 --> 00:48:29.410 slip rates, and we don’t see anywhere near this level of seismicity along those. 00:48:29.410 --> 00:48:34.400 So it does beg the question of why we would see so much 00:48:34.400 --> 00:48:38.420 seismicity along this fault system in this location, 00:48:38.420 --> 00:48:43.520 and we don’t see that in other places in the Pacific Northwest. 00:48:44.930 --> 00:48:49.700 So this gets back to this slide I showed in the beginning of my talk, 00:48:49.710 --> 00:48:55.850 and this is from the paper by Stein and Liu in their Science article. 00:48:55.850 --> 00:49:02.920 So the kind of elongated ellipse here is our kind of guesstimate 00:49:02.920 --> 00:49:07.470 of the deformation rates in the Entiat area, and it’s probably 00:49:07.470 --> 00:49:12.700 in the few tenths of a millimeter per year at the most. 00:49:13.900 --> 00:49:19.020 So that – according to this curve, that observation would suggest 00:49:19.020 --> 00:49:24.740 that we will see aftershocks in the Entiat area for thousands of years. 00:49:26.560 --> 00:49:33.160 If we saw earthquakes for thousands of years following main shocks, 00:49:33.160 --> 00:49:37.440 I might think that we would see more examples of long-lived aftershock 00:49:37.440 --> 00:49:42.860 sequences than we seem to see. But that’s my guess. 00:49:44.320 --> 00:49:48.520 One of the things we’re doing to more closely monitor 00:49:48.520 --> 00:49:52.800 the Entiat cluster – and again, this is a close-up view of the cluster. 00:49:52.800 --> 00:49:55.860 This is a 10-kilometer scale bar here. 00:49:55.860 --> 00:50:03.300 The unfilled triangles show you the permanent seismic network 00:50:03.300 --> 00:50:09.600 in the Entiat area, and it shows that the cluster is very nicely bounded 00:50:09.600 --> 00:50:15.580 by the existing stations, but there are no stations directly over the cluster. 00:50:15.580 --> 00:50:22.010 And given that – the fact that so many of these events are shallow, 00:50:22.010 --> 00:50:26.250 we wanted to put in more stations directly over the cluster, 00:50:26.250 --> 00:50:30.010 and the Washington Department of Natural Resources deployed 00:50:30.010 --> 00:50:35.700 about 12 RefTeks, which are these filled triangles, about a year ago, 00:50:35.700 --> 00:50:40.800 to get better depth control and focal mechanisms. 00:50:40.800 --> 00:50:46.470 And to date, the biggest event we’ve recorded is this magnitude 3.3 00:50:46.470 --> 00:50:50.910 about six weeks ago, which doesn’t sound like much for Californians, 00:50:50.910 --> 00:50:54.220 but we only get one of these every few years. 00:50:54.220 --> 00:50:57.080 So we were quite fortunate to record that. 00:50:58.300 --> 00:51:02.160 Other lines of evidence that are suggested by this work is that I think 00:51:02.170 --> 00:51:06.240 we’re all looking forward to Brian Sherrod’s paper on his work 00:51:06.240 --> 00:51:13.730 on the Lidar scarp in Spencer Canyon is – he hasn’t gotten all the 00:51:13.730 --> 00:51:18.420 age dating yet, but when that happens, it should be very interesting to look at 00:51:18.420 --> 00:51:23.170 the recurrence intervals of the events on the – on the scarp – 00:51:23.170 --> 00:51:28.610 on the fault and compare them with some of these other predictions 00:51:28.610 --> 00:51:33.060 for long-lived aftershock sequences. 00:51:35.340 --> 00:51:39.840 Lake Chelan is a natural lake, as I think I pointed out before, 00:51:39.840 --> 00:51:43.160 and it’s been there for a long time. And so it would be possible to 00:51:43.160 --> 00:51:48.650 look for turbidites or other deposits in the lake that might have been 00:51:48.650 --> 00:51:52.920 created by previous earthquakes in the region. 00:51:52.920 --> 00:51:57.090 And I think that might be a very promising line of work. 00:51:57.090 --> 00:51:58.650 I think we can do a little bit more to 00:51:58.650 --> 00:52:03.980 refine estimates of the deformation rates in the region. 00:52:04.850 --> 00:52:10.840 We definitely need to look at the seismicity that we’ve recorded on this 00:52:10.840 --> 00:52:15.860 temporary network in the Entiat area. Unfortunately, we haven’t been able to 00:52:15.860 --> 00:52:21.030 telemeter that in real time, and so we’ll have to do that post facto. 00:52:21.030 --> 00:52:24.160 And I think it might be also very interesting to look at 00:52:24.160 --> 00:52:29.020 three-component seismograms that have been recorded by earthquakes 00:52:29.020 --> 00:52:35.070 in the Entiat area at sites where we have some of our – 00:52:35.070 --> 00:52:41.070 that felt the 1872 event to see whether they can give us any clarity 00:52:41.070 --> 00:52:45.840 on the horizontal directions of shaking that were reported. 00:52:47.420 --> 00:52:51.120 So, to summarize, I’ve tried to argue and present the evidence 00:52:51.130 --> 00:52:59.070 that an Entiat-area epicenter for the 1872 earthquake best matches 00:52:59.070 --> 00:53:04.520 the intensity data and the aftershock data and the fault – the location of the 00:53:04.520 --> 00:53:09.280 fault that produced surface offset close by in Spencer Canyon. 00:53:10.900 --> 00:53:19.620 I’ve shown, I hope, that the magnitude 6.5 to 7 earthquake in 1872 is 00:53:19.630 --> 00:53:25.800 consistent with the observed scarp heights and the modal intensity data. 00:53:25.800 --> 00:53:28.720 Although I didn’t show that. Bill Bakun showed that. 00:53:30.240 --> 00:53:34.140 The aftershock forecast modeling that I presented here indicates that 00:53:34.150 --> 00:53:41.440 the seismicity since 1872 is consistent with an earthquake 00:53:41.440 --> 00:53:47.020 in that location and that magnitude range inferred from the intensity data. 00:53:47.790 --> 00:53:55.660 The inferred 140-year duration of the aftershocks in the Entiat area 00:53:55.660 --> 00:54:02.670 would be consistent with the theory for aftershock durations that suggest that 00:54:02.670 --> 00:54:07.530 the main shock recurrence intervals should be several – a few to several 00:54:07.530 --> 00:54:15.310 thousand years in this region, which seemed to be pointed to by the 00:54:15.310 --> 00:54:21.580 available information on the number of events for the past 8,000 years or so. 00:54:21.580 --> 00:54:27.200 But, you know, these long-lived aftershock sequences are unusual. 00:54:27.200 --> 00:54:32.760 And we really – the theory suggests we really shouldn’t expect to see that many. 00:54:32.770 --> 00:54:39.340 So the ones that we do see, I think, warrant as much study 00:54:39.340 --> 00:54:42.940 as we can devote to them. So thank you very much. 00:54:42.940 --> 00:54:49.120 [ Applause ] 00:54:50.460 --> 00:54:54.000 [ Silence ] 00:54:54.760 --> 00:54:58.980 - Thank you very much, Tom. Any questions for Tom? 00:55:00.560 --> 00:55:05.180 [ Silence ] 00:55:05.880 --> 00:55:08.140 - Thanks, Tom. Nice talk, and I’m especially happy 00:55:08.140 --> 00:55:12.760 to see that there’s more data being collected to bear on these questions. 00:55:13.520 --> 00:55:15.320 So I was curious. You talked a lot 00:55:15.330 --> 00:55:18.200 about aftershock duration. - Right. 00:55:18.200 --> 00:55:23.170 - But I don’t think you defined how aftershock duration is defined. 00:55:23.170 --> 00:55:24.860 And the other question I had is, 00:55:24.860 --> 00:55:29.440 how is it measured in practice from seismicity data? 00:55:29.440 --> 00:55:39.260 - Okay, so to go back to this figure, I think you would say the aftershock – 00:55:39.260 --> 00:55:43.460 you know, you have to think of some region that you’re looking at. 00:55:43.460 --> 00:55:45.750 And you count the number of earthquakes in that region. 00:55:45.750 --> 00:55:53.540 And as long as this curve continues to decay, you could argue that the sequence 00:55:53.540 --> 00:55:58.280 is continuing. But when – at some point, this curve will flatten out. 00:55:59.160 --> 00:56:05.650 And you’ll see about the same number of events year to year. 00:56:05.650 --> 00:56:10.660 So at that point, you would say the aftershock sequence has ended. 00:56:12.000 --> 00:56:17.180 - Okay. But – so then it’s getting to the background rate. 00:56:17.180 --> 00:56:19.300 - Okay. - And I think it does get tricky because 00:56:19.300 --> 00:56:22.860 you said we don’t see a lot of these, you know, long-lived aftershock sequences. 00:56:22.860 --> 00:56:27.820 But if you take this sort of to its illogical conclusion, you know, 00:56:27.820 --> 00:56:31.140 maybe all the background seismicity we see is a super-position of many 00:56:31.140 --> 00:56:35.320 long-lived aftershock sequences. So I think it gets a little bit tricky to 00:56:35.330 --> 00:56:43.160 disentangle all of these effects, but … - I think I see your issue. 00:56:43.160 --> 00:56:49.710 I don’t know – in practice, you know, I – you know, I think you’re right. 00:56:49.710 --> 00:56:53.430 I think we have a challenge in all these long-lived aftershock sequences. 00:56:53.430 --> 00:57:01.190 We really don’t have an idea of what, you know, an event more than 100 00:57:01.190 --> 00:57:05.370 years ago, of what the background rate of seismicity was, more than likely. 00:57:05.370 --> 00:57:08.300 I mean, here in Japan, we might. 00:57:08.300 --> 00:57:10.850 There might – there might be good enough records. 00:57:10.850 --> 00:57:13.850 So perhaps that should be done here is that, you know, you should 00:57:13.850 --> 00:57:19.420 look ahead of the earthquake to see what the background rate is, and you look – 00:57:19.420 --> 00:57:24.930 you know, you watch for that and to see whether your rate 00:57:24.930 --> 00:57:27.210 behind the earthquake matches. 00:57:27.210 --> 00:57:29.620 - Okay, thanks. - Yeah. 00:57:33.000 --> 00:57:34.680 - Hey, Tom. - Yeah. 00:57:34.680 --> 00:57:39.060 - So a lot of the GPS data and the focal mechanism data that you’ve 00:57:39.060 --> 00:57:42.760 looked at recently and the long-term paleomagnetic data 00:57:42.760 --> 00:57:46.240 all kind of converge on this idea that the upper plate is rotating, 00:57:46.240 --> 00:57:48.700 you know, clockwise around this rotation pole. 00:57:48.700 --> 00:57:51.700 - Right. - You know, just off to the east here. 00:57:51.700 --> 00:57:57.990 And the location of this earthquake in that kind of conceptual framework – 00:57:57.990 --> 00:58:00.150 this earthquake, you’d predict, you know, 00:58:00.150 --> 00:58:05.040 probably north- or northeast-directed compression. 00:58:05.040 --> 00:58:09.540 And that’s – I think that’s corroborated by the GPS data. 00:58:09.550 --> 00:58:13.960 However, this earthquake and the fault plain and the focal mechanisms 00:58:13.960 --> 00:58:17.350 that are shown here suggest kind of northwest-southeast 00:58:17.350 --> 00:58:22.160 compression, which is kind of orthogonal to what one would expect. 00:58:22.160 --> 00:58:24.300 So I was wondering if you can comment on that and maybe comment 00:58:24.300 --> 00:58:31.579 on kind of the tectonic mechanism for this event in that greater context. 00:58:31.579 --> 00:58:37.560 - Yeah, so Scott’s alluding to some work I’ve done looking at 00:58:37.560 --> 00:58:41.700 fault geometries and focal mechanisms throughout the Pacific Northwest, 00:58:41.700 --> 00:58:50.020 and I found a pole of rotation near Spokane that best matched those data. 00:58:50.020 --> 00:58:55.000 And you’re right. This orientation of, say, the black lines 00:58:55.000 --> 00:59:02.020 that show the aeromagnetic lineation are oblique to that pole. 00:59:02.020 --> 00:59:06.480 So, but the focal mechanisms suggest that as well. 00:59:06.480 --> 00:59:12.340 We’re not seeing pure – generally not seeing – we’re seeing a lot of pure thrust 00:59:12.340 --> 00:59:18.260 fault mechanisms, but we’re also seeing a lot of oblique thrust fault mechanisms. 00:59:18.260 --> 00:59:21.620 So that seems to be compatible. 00:59:23.740 --> 00:59:29.120 [ Silence ] 00:59:29.800 --> 00:59:33.140 - I had a question that sort of went on with the focal mechanisms 00:59:33.140 --> 00:59:37.060 that you’ve shown in this aftershock sequence. 00:59:37.060 --> 00:59:40.070 So I was wondering – so I know a lot of these are fairly small in magnitude, 00:59:40.070 --> 00:59:43.260 and I was wondering if you had resolved some of the focal mechanisms 00:59:43.260 --> 00:59:46.190 for the aftershocks, if you see any evolution in how they 00:59:46.190 --> 00:59:50.200 may change over time. And if there are any implications 00:59:50.200 --> 00:59:57.270 for what the 1872 event – sort of style of rupture – maybe if it was pure thrust 00:59:57.270 --> 01:00:01.740 or if you think there was also an oblique component to the main shock. 01:00:01.740 --> 01:00:05.380 - Oh, you’re asking me to speculate quite a bit. 01:00:07.820 --> 01:00:10.420 I haven’t looked at the time history of the 01:00:10.420 --> 01:00:13.960 focal mechanisms to see if there’s been any change. 01:00:15.280 --> 01:00:18.990 I think it may be more profitable to do that after we’ve had a chance 01:00:18.990 --> 01:00:23.730 to bring in this local temporary network data and get a better sense 01:00:23.730 --> 01:00:30.080 of how good the network focal mechanisms might be. 01:00:32.280 --> 01:00:39.310 If you – if you look at the average of these focal mechanisms, you get – 01:00:39.310 --> 01:00:43.590 you get an average strike of about 230 degrees, 01:00:43.590 --> 01:00:48.450 which is sort of similar to this strike. And you get an average dip of about 01:00:48.450 --> 01:00:54.310 50 degrees, which suggests, you know, maybe primarily thrust faulting, 01:00:54.310 --> 01:01:00.060 but I couldn’t rule out, you know, some component of oblique faulting. 01:01:04.660 --> 01:01:07.920 - Yeah, so you documented the seismicity since 1900. 01:01:07.920 --> 01:01:12.340 So have the seismicity rates been declining, in fact? 01:01:12.340 --> 01:01:15.060 I don’t recall where you showed that. 01:01:17.220 --> 01:01:21.060 - It’s hard to do because we have three different data sets, 01:01:21.060 --> 01:01:23.480 and they don’t really match. 01:01:25.460 --> 01:01:30.990 So I’ve just tried to analyze each one – you know, each different time period. 01:01:30.990 --> 01:01:36.180 So, by the fact that each time frame seems to match – 01:01:36.180 --> 01:01:38.340 well, to go back to your question … 01:01:42.760 --> 01:01:44.240 Sorry. 01:01:45.800 --> 01:01:50.140 [ Silence ] 01:01:50.940 --> 01:01:55.760 If you look at the events right after the event, the numbers are 01:01:55.770 --> 01:02:00.420 more consistent with the high end of the magnitude range. 01:02:00.420 --> 01:02:04.220 If you look at the last 40 years, 01:02:04.220 --> 01:02:07.860 the numbers are only consistent with the low end. 01:02:09.800 --> 01:02:14.060 So, you know, if you could – you could say, well, you know, 01:02:14.070 --> 01:02:20.520 we’re not seeing exactly the same rate. You know, we’re not overlapping 01:02:20.520 --> 01:02:26.010 that well, but we are at least consistent with that range. 01:02:26.010 --> 01:02:29.910 Now – but, you know, we’re dealing with small numbers here. 01:02:29.910 --> 01:02:33.580 You know, we’re talking about 64 earthquakes reported here 01:02:33.580 --> 01:02:39.190 by a person who was involved with trading with Native Americans. 01:02:39.190 --> 01:02:41.300 He was not a – he was not a scientist. 01:02:41.300 --> 01:02:44.760 I don’t know how much education he had. 01:02:45.790 --> 01:02:48.720 So you can quibble with each set of these numbers, 01:02:48.720 --> 01:02:52.580 except perhaps with the past 40 years. 01:02:52.580 --> 01:02:58.620 But, no, I haven’t done a – I haven’t tried to do a single calculation for all 01:02:58.630 --> 01:03:04.650 three different timeframes because I couldn’t figure out how to do it. 01:03:04.650 --> 01:03:10.880 So if you have some suggestions for me, I’d be happy to hear them. 01:03:12.520 --> 01:03:14.100 Bill, you have a question? - Well, I … 01:03:14.100 --> 01:03:17.120 - Oh. - Just more of a comment. 01:03:19.820 --> 01:03:21.790 Thank you very much for a very nice paper. 01:03:21.790 --> 01:03:24.220 It’s always nice to see somebody bring up 01:03:24.220 --> 01:03:26.820 an old paper to say nice things about it. 01:03:26.820 --> 01:03:29.860 [laughter] - Well-deserved. 01:03:29.860 --> 01:03:31.910 - But I want to say a little bit about the history. 01:03:31.910 --> 01:03:36.060 The 1872 earthquake is really an interesting earthquake. 01:03:36.060 --> 01:03:40.490 I got involved after the dust had sort of settled. 01:03:40.490 --> 01:03:46.230 And the background – in the 1970s, the 1872 earthquake was a football 01:03:46.230 --> 01:03:49.210 in the sense that there were two groups that were trying to 01:03:49.210 --> 01:03:53.530 build additional nuclear power plants in that region. 01:03:53.530 --> 01:03:59.090 WPPSS was trying to build nuclear power plants at Hanford. 01:03:59.090 --> 01:04:01.820 Puget Sound Power & Light was trying to build 01:04:01.820 --> 01:04:05.180 nuclear power plants on the Skagit River. 01:04:05.180 --> 01:04:09.840 So they had – the area was swarming with consultants. 01:04:09.840 --> 01:04:13.230 One group of consultants was doing everything they could to boot 01:04:13.230 --> 01:04:15.710 the earthquake down towards Hanford. [laughter] 01:04:15.710 --> 01:04:17.990 The other group was doing everything they could to 01:04:17.990 --> 01:04:24.800 boot the earthquake up to Skagit River, where Steve Malone put it. 01:04:24.800 --> 01:04:32.980 And Ivan Wong told me that the area was swarming with consultants, 01:04:32.990 --> 01:04:36.830 both steaming up and down Lake Chelan collecting data. 01:04:36.830 --> 01:04:41.880 The lawyers got involved, and all of the consultants were 01:04:41.880 --> 01:04:45.160 shut down and said, never say anything about this earthquake. 01:04:45.160 --> 01:04:49.680 The only people that could say anything were the lawyers. 01:04:49.680 --> 01:04:54.250 So then, the bloom came off the nuclear power plant issue, 01:04:54.250 --> 01:04:56.290 and no one was interested in building it. 01:04:56.290 --> 01:05:07.430 So that’s when I got involved, and I started looking at the data, and – 01:05:07.430 --> 01:05:10.140 the data that Margaret Hopper put together. 01:05:10.140 --> 01:05:15.700 And I came across Steve Malone’s paper, and it was an interesting paper 01:05:15.700 --> 01:05:20.380 from what his conclusions were – one of the conclusions was something 01:05:20.380 --> 01:05:24.390 I had never seen in a scientific paper. He said where he thought the 01:05:24.390 --> 01:05:28.900 earthquake was, which was up near Skagit River area, but he – 01:05:28.900 --> 01:05:34.190 very pointed conclusion that the earthquake did not occur 01:05:34.190 --> 01:05:39.200 in the Lake Chelan/Entiat area. That was one of his stated conclusions. 01:05:39.200 --> 01:05:43.880 Now, Steve was, at the time, the president of the 01:05:43.880 --> 01:05:46.420 Seismological Society of America. 01:05:46.420 --> 01:05:51.640 And so this paper I wrote had to be very carefully written. 01:05:51.650 --> 01:05:57.150 And one of the things I’m proud to say is that Steve and I are still friends. 01:05:57.150 --> 01:06:02.060 And that says more about Steve Malone, I think, than it says about my paper, 01:06:02.060 --> 01:06:04.440 and it’s something that – one of the things I wanted to say 01:06:04.440 --> 01:06:08.120 to the younger scientists is that it’s important, 01:06:08.120 --> 01:06:13.160 when someone disagrees with you, not to take it personally. 01:06:13.160 --> 01:06:16.740 They’re not your enemy from then on. 01:06:16.740 --> 01:06:21.450 And they are just someone that has a different opinion. 01:06:21.450 --> 01:06:26.880 And so it’s really important not to take a disagreement personally. 01:06:26.880 --> 01:06:33.720 And it’s a credit to Steve Malone, and that’s the way I think about it. 01:06:33.720 --> 01:06:39.100 But there is a lot more data locked away in these consultants’ 01:06:39.100 --> 01:06:42.660 drawers that was shut down by the lawyers, and you might 01:06:42.660 --> 01:06:46.080 talk to Ivan Wong about … - Okay. 01:06:46.080 --> 01:06:49.040 - …you know, what’s available in Lake Chelan. 01:06:49.040 --> 01:06:51.360 - I appreciate your comment, and I’d like to second it. 01:06:51.360 --> 01:06:55.210 And I also – and in particular about Steve Malone because I know he was 01:06:55.210 --> 01:06:58.160 a reviewer on at least one of my papers, 01:06:58.160 --> 01:07:01.160 and he’s given me some very good suggestions. 01:07:01.740 --> 01:07:04.180 I think we had a question over here. 01:07:06.060 --> 01:07:15.840 [ Silence ] 01:07:16.580 --> 01:07:22.020 - Yeah, nice talk, Tom. It’s a pretty good thing that that 01:07:22.020 --> 01:07:25.990 Lidar image of the scarp showed up. That focuses things nicely. 01:07:25.990 --> 01:07:29.470 Is there anything else in the geomorphology that’s consistent 01:07:29.470 --> 01:07:33.650 with active tectonic deformation there, albeit at a low rate? 01:07:33.650 --> 01:07:37.720 Like, you know, canyon depth changes and things like that 01:07:37.720 --> 01:07:44.360 that might focus the search 01:07:44.360 --> 01:07:47.180 or constrain things a little bit? 01:07:48.100 --> 01:07:51.060 - Ah, that’s not a question that I can address. 01:07:51.060 --> 01:07:54.580 But I don’t know. Scott or Lydia, you have any … 01:07:55.000 --> 01:07:57.360 - There is a – there is a canyon on the east side of the river 01:07:57.360 --> 01:08:03.820 that’s oriented kind of northeast that one could imagine hosts a fault. 01:08:03.820 --> 01:08:07.300 But I think the hill slope processes are fast enough 01:08:07.300 --> 01:08:09.560 where the physical expression is not clear. 01:08:09.560 --> 01:08:13.480 I know Ralph Haugerud has done a lot of mapping on the east side of the river – 01:08:13.480 --> 01:08:17.800 kind of [inaudible] mapping, and he’s mentioned some features 01:08:17.800 --> 01:08:22.120 that have centimeter amounts of [inaudible] across them. 01:08:22.120 --> 01:08:25.340 You got to talk to Ralph, but he could also be here in a couple weeks, but … 01:08:25.340 --> 01:08:29.560 - No. I’m familiar with some of Ralph’s work, and that was behind my question. 01:08:29.560 --> 01:08:32.360 - Yeah. But there’s no obvious scarp, unfortunately. 01:08:32.360 --> 01:08:35.300 And so you could extend the scarp – you could take – you could add 01:08:35.300 --> 01:08:37.780 the width of the river to the – to the length of the scarp 01:08:37.780 --> 01:08:41.060 and assume that the river has obliterated it. 01:08:41.060 --> 01:08:44.000 That might be a fair extrapolation, but no obvious scarp on the 01:08:44.000 --> 01:08:46.300 east side that I – [inaudible] that I’ve seen. 01:08:46.300 --> 01:08:50.600 - For those of you who are listening online, one of the responses 01:08:50.600 --> 01:08:54.800 was that there isn’t an obvious scarp yet identified. 01:08:54.800 --> 01:08:58.540 I don’t know if the area – there’s quite a bit of Lidar, but I 01:08:58.540 --> 01:09:04.560 don’t know if the whole region has as much Lidar as we’d like or need yet. 01:09:04.569 --> 01:09:07.089 But certainly the area near the scarp that’s been identified, 01:09:07.089 --> 01:09:09.559 there’s a lot of Lidar around there. 01:09:09.559 --> 01:09:15.830 So there isn’t any obvious scarp to the – to the southwest. 01:09:15.830 --> 01:09:19.980 Could be a little bit of scarp underneath the river, but we don’t know that. 01:09:22.540 --> 01:09:26.720 - Generating a lot of interesting conversation here, Tom. 01:09:26.720 --> 01:09:29.960 - Yeah, this seems, you know, really relevant given what’s going on 01:09:29.969 --> 01:09:33.219 in the central United States, for example, and Oklahoma 01:09:33.219 --> 01:09:38.400 where strain rates are extremely low. So I wonder if you could return back 01:09:38.400 --> 01:09:43.160 to your interpretation of Dieterich’s data set. 01:09:45.760 --> 01:09:49.670 So how do we – my question is, how do we decide where those 01:09:49.670 --> 01:09:56.019 contours are if it’s an effect of the shear stress, A, the direct effect, 01:09:56.020 --> 01:09:58.940 or the effective normal stress? 01:10:03.160 --> 01:10:06.360 - Could you repeat the question part? - Well, we know – we know that tau, 01:10:06.369 --> 01:10:11.090 the strain rate, is low. How do we go about separating out 01:10:11.090 --> 01:10:17.340 the possible – the possibility that A and sigma are also different? 01:10:18.160 --> 01:10:20.320 - Right. - Does that make sense? 01:10:21.860 --> 01:10:25.660 [ Silence ] 01:10:26.340 --> 01:10:32.040 - I guess we don’t – do we expect the normal stress to change that much? 01:10:33.820 --> 01:10:37.740 - It’s an open question, I guess. - Yeah. 01:10:37.740 --> 01:10:41.920 I guess my first thought it that it’s mainly that the stress – 01:10:41.920 --> 01:10:47.610 the shear stress drop, that’s a variable. But, you know, for Entiat, we – 01:10:47.610 --> 01:10:52.420 you know, we have no idea whether the ratio is 10 or 40. 01:10:52.420 --> 01:10:56.679 If we had a – if we could really constrain the recurrence interval, 01:10:56.679 --> 01:11:00.960 then we could make a better estimate. 01:11:00.960 --> 01:11:05.140 I mean, I would be surprised if this was a low stress drop earthquake. 01:11:05.140 --> 01:11:06.800 I mean, it doesn’t seem like it should be 01:11:06.800 --> 01:11:13.920 a low stress drop environment, but, yeah, don’t know. 01:11:17.760 --> 01:11:20.480 One in the – Tom Hanks, I think. 01:11:22.520 --> 01:11:26.400 [ Silence ] 01:11:27.080 --> 01:11:34.260 - Yeah, I have two things to mention. But while we’re here – so in the 01:11:34.260 --> 01:11:41.570 Basin and Range, which we’ve had a number of historic earthquake ruptures, 01:11:41.570 --> 01:11:46.699 we would expect aftershocks to last then – aftershock sequence 01:11:46.700 --> 01:11:50.980 to last for hundreds of years? For Borah Peak and the Sonoran 01:11:50.980 --> 01:11:59.320 earthquake and Dixie Valley and the one – the 1915 earthquake? 01:12:00.500 --> 01:12:04.160 - That’s what the theory suggests, yeah. 01:12:04.179 --> 01:12:08.169 - That seems to be [chuckles] pretty testable at this point for … 01:12:08.169 --> 01:12:09.980 - Yeah. - … for a number of the Basin 01:12:09.980 --> 01:12:12.159 and Range earthquakes. - Yeah, good point. 01:12:12.160 --> 01:12:18.900 - And the other thing is – gets to the – to the – to the fault scarp. 01:12:19.760 --> 01:12:27.420 So you see a clear break over 6 kilometers, but – 01:12:27.429 --> 01:12:32.639 and nothing has ruptured twice in the last 7,000 years, or 7,700. 01:12:32.640 --> 01:12:40.320 So shouldn’t you have some more identifiable scarp away from that? 01:12:41.180 --> 01:12:43.560 Or do you just have repeating earthquakes 01:12:43.560 --> 01:12:47.580 over the 6-kilometer-long break? 01:12:50.980 --> 01:12:54.040 - Well, there’s two events in one trench on that one scarp. 01:12:54.040 --> 01:12:55.639 - Okay. 01:12:56.860 --> 01:12:59.480 - So, yeah, I mean, it’s a re-rupture … - Yeah. 01:12:59.489 --> 01:13:02.130 - … of that fault. - And, I mean, this may not be 01:13:02.130 --> 01:13:04.650 the best question to you, but for Scott and Lydia that … 01:13:04.650 --> 01:13:06.909 - Yeah. - … who have worked with this thing. 01:13:06.909 --> 01:13:09.860 Do you have any comments on that? 01:13:11.640 --> 01:13:15.320 Because it seems to me, for a magnitude 7 earthquake 01:13:15.321 --> 01:13:21.920 to give you 6 kilometers of fault scarp, particularly when you have so much 01:13:21.920 --> 01:13:27.750 shallow seismic activity, just, you know, across the whole, what, 01:13:27.750 --> 01:13:35.250 30 kilometers or so of the – of the aftershocks, such as Tom can identify, 01:13:35.250 --> 01:13:42.260 that it seems like there should be a longer Holocene fault scarp. 01:13:43.140 --> 01:13:46.480 - So the … - Oh, here. Use a microphone. 01:13:48.340 --> 01:13:56.120 [ Silence ] 01:13:56.860 --> 01:13:58.820 - All right. Is this working now? - Yes. 01:13:58.820 --> 01:14:00.820 - Yeah. - Okay. So at the site of 01:14:00.830 --> 01:14:04.159 the fault trenches, we know that there were two ruptures. 01:14:04.160 --> 01:14:07.900 But that’s one very specific area. And so the overlap for the 01:14:07.900 --> 01:14:12.820 actual ruptures is – we know it exists there, but how long the 01:14:12.820 --> 01:14:17.180 actual scarp from the older earthquake was, we can’t exactly constrain. 01:14:17.180 --> 01:14:20.531 And then, you’re right, of course, that the actual scarp that we see 01:14:20.540 --> 01:14:24.200 at the surface in the Lidar is only about 6 kilometers long, 01:14:24.200 --> 01:14:27.180 give or take maybe a couple more if you add the river. 01:14:28.400 --> 01:14:33.960 But – and that does seem a bit short, but I guess we can’t exactly rule out 01:14:33.960 --> 01:14:38.700 a lot longer of a rupture at depth that just didn’t reach the surface 01:14:38.700 --> 01:14:42.800 or has subsequently been eroded. 01:14:43.960 --> 01:14:50.000 [ Silence ] 01:14:50.780 --> 01:14:56.780 - Yeah, I agree with you, Tom, that the scarp should be longer. 01:14:56.780 --> 01:15:01.920 And for the previous event, if it – if that scarp did exceed the 6 kilometers 01:15:01.920 --> 01:15:06.809 that can see in the Lidar, if it was only – if it was a post-7,000-year-old event, 01:15:06.809 --> 01:15:10.300 we should probably see that in that kind of climate and environment 01:15:10.300 --> 01:15:13.139 in the geomorphology, but we really don’t. 01:15:13.139 --> 01:15:18.329 So I think I – I think I share your surprise that the scarp 01:15:18.329 --> 01:15:21.710 at the surface is as short as it is. Even if you – even if you were to 01:15:21.710 --> 01:15:24.260 add a couple kilometers for the river and a couple kilometers 01:15:24.260 --> 01:15:26.010 for this canyon that I’m talking about, it’s still, 01:15:26.010 --> 01:15:30.130 I think, kind of a little bit shorter than one would expect. 01:15:30.130 --> 01:15:35.710 But my time up in Seattle pre-dates the trenching, so I was never at the site. 01:15:35.710 --> 01:15:38.740 So that’s pretty much all I can comment on. 01:15:44.100 --> 01:15:48.300 - Very good. Any final comments or questions for Tom? 01:15:49.500 --> 01:15:51.460 All right. Then we’ll leave it there. Thanks a lot, Tom. 01:15:51.460 --> 01:15:52.460 - Thank you. 01:15:52.460 --> 01:15:57.380 [ Applause ] 01:16:00.500 --> 01:16:10.220 [ Silence ]