WEBVTT Kind: captions Language: en-US 00:00:00.480 --> 00:00:01.880 [inaudible background conversations] 00:00:01.880 --> 00:00:05.880 Okay. Hi. We’re going to get started here. 00:00:05.880 --> 00:00:09.060 Before I introduce Gary, just want to let you know about 00:00:09.060 --> 00:00:11.670 what’s upcoming in seminars. 00:00:11.670 --> 00:00:15.860 Next week, we’re going to have a preview of some AGU talks. 00:00:15.860 --> 00:00:21.320 And then, the week after that, AGU week, we will not be having a seminar. 00:00:21.320 --> 00:00:27.400 So for today’s seminar, it’s my pleasure to introduce our own Gary Fuis. 00:00:27.400 --> 00:00:31.430 I’ve known Gary since I came to the Survey 15 years ago. 00:00:31.430 --> 00:00:35.430 We actually turn out to have pretty similar academic trajectories, both going 00:00:35.430 --> 00:00:39.580 to Cornell and then deciding that we couldn’t handle the weather back east. 00:00:39.580 --> 00:00:44.650 [laughs] And then coming out to Caltech to do Ph.D.s. 00:00:44.650 --> 00:00:46.970 Gary was telling me he actually the first NAGT intern 00:00:46.970 --> 00:00:51.960 at the USGS doing astrogeology in Arizona. 00:00:51.960 --> 00:00:56.460 But then fortunately turned his attentions to seismology. 00:00:56.460 --> 00:01:02.520 And he’s made his career here at USGS doing active source experiments, 00:01:02.520 --> 00:01:07.060 mainly in southern California and Alaska, to study structure, 00:01:07.060 --> 00:01:10.650 including deep structure, and also tectonics, including trying to 00:01:10.650 --> 00:01:14.850 understand the structures of faults. And his most recent big project was 00:01:14.850 --> 00:01:19.330 the Salton Sea Imaging Project looking at the structure of the 00:01:19.330 --> 00:01:23.640 southernmost San Andreas. And that had a lot of interesting results, 00:01:23.640 --> 00:01:27.620 that the southern San Andreas is not, you know, sort of a vertical 00:01:27.620 --> 00:01:30.690 planar feature, but is actually sort of a dipping curved feature. 00:01:30.690 --> 00:01:33.010 And that got him interested thinking about places 00:01:33.010 --> 00:01:36.710 where he’d seen similar kind of nonplanar strike-slip faults. 00:01:36.710 --> 00:01:41.420 Got him thinking back about the Loma Prieta earthquake from 1989. 00:01:41.420 --> 00:01:44.080 And that’s what he’s going to be talking to us about today. 00:01:44.080 --> 00:01:46.340 Take it away, Gary. 00:01:48.000 --> 00:01:54.320 [Silence] 00:01:54.320 --> 00:01:56.740 - Hello? Does it work? 00:01:56.740 --> 00:01:58.900 - Good. - Okay, good. 00:02:00.360 --> 00:02:04.000 Well, thanks, Jeanne, for that kind introduction. 00:02:04.000 --> 00:02:07.120 [chuckles] Us old guys, they can’t just say, 00:02:07.130 --> 00:02:10.090 he needs no introduction anymore, because he does. [laughter] 00:02:10.090 --> 00:02:13.590 There’s too many youngsters in the crowd. 00:02:13.590 --> 00:02:17.890 So this is the title and the authors of an 00:02:17.890 --> 00:02:22.680 Open-File report that came out this last year. 00:02:22.680 --> 00:02:27.659 It was put together by Edward Zhang – that’s how you pronounce it, 00:02:27.659 --> 00:02:30.500 Rufus Catchings, Dan Scheirer, Mark Goldman, 00:02:30.500 --> 00:02:35.140 and Klaus Bauer in Potsdam, and myself. 00:02:35.140 --> 00:02:40.200 And it’s an effort that’s been going on for a couple years now. 00:02:41.580 --> 00:02:46.120 I don’t know how many of you actually look at Open-File reports or read them. 00:02:46.120 --> 00:02:48.000 [laughter] 00:02:48.010 --> 00:02:52.519 So I thought I’d give a talk in order to call attention to 00:02:52.519 --> 00:02:59.139 some new results for an earthquake that the Survey spent a hell of a lot 00:02:59.139 --> 00:03:05.129 of time and energy on 20 years ago. There were five professional papers 00:03:05.129 --> 00:03:10.579 that came out and countless journal articles on this earthquake. 00:03:10.579 --> 00:03:15.540 And it wasn’t very well understood at the time, I don’t think. 00:03:15.540 --> 00:03:18.760 So we’re going to take another crack at it. 00:03:19.600 --> 00:03:26.460 This all started over two years ago when I was at Berkeley giving a seminar and – 00:03:26.469 --> 00:03:32.849 on our Salton Seismic Imaging Project. And we’d found an interesting – 00:03:32.849 --> 00:03:36.709 an unusual and unexpected structure for the – deep structure – subsurface 00:03:36.709 --> 00:03:42.380 structure for the San Andreas Fault. It wasn’t vertical or planar. 00:03:42.380 --> 00:03:48.040 And in my Berkeley talk, I mentioned at the end that, well, maybe we ought to 00:03:48.040 --> 00:03:52.760 take another look at Loma Prieta. There was some similarities there. 00:03:52.840 --> 00:03:54.420 And Roland Burgmann was in the audience. 00:03:54.420 --> 00:03:58.260 He was one of the original workers on this earthquake. 00:03:58.260 --> 00:04:01.959 And he very enthusiastically said, yeah, go for it. 00:04:01.959 --> 00:04:08.099 We had collected seismic – active source seismic across the epicenter in 1991, 00:04:08.099 --> 00:04:13.379 Rufus and I and others. And he says, yeah, go ahead and look at that data. 00:04:13.379 --> 00:04:21.150 And then Barbara Romanowicz sent over a student intern, Edward Zhang, 00:04:21.150 --> 00:04:24.670 to see if we needed any help. And the answer was, yes, we did need help. 00:04:24.670 --> 00:04:32.311 Because Rufus was too busy, and I was retired, so Edward was just the ticket. 00:04:32.311 --> 00:04:37.090 And he saw this project through from the beginning. 00:04:37.090 --> 00:04:40.360 And the early results were no results. 00:04:40.360 --> 00:04:43.520 And then he went back and looked at it again and did it again. 00:04:43.520 --> 00:04:45.340 And then he came up with some 00:04:45.340 --> 00:04:50.980 intriguing results that are worth looking at. 00:04:51.920 --> 00:04:57.500 Okay, so when myself and others were working on Loma Prieta 00:04:57.500 --> 00:05:04.330 in the years following 1989, the prevailing interpretation 00:05:04.330 --> 00:05:09.840 was that the rupture was not on the San Andreas Fault. 00:05:09.840 --> 00:05:14.140 And the reason was because it wasn’t vertical – the rupture wasn’t vertical, 00:05:14.140 --> 00:05:18.320 and it did not project to the surface at the San Andreas Fault. 00:05:19.220 --> 00:05:25.640 Okay, so Walter Mooney and I – I had my doubts about this secondary 00:05:25.650 --> 00:05:29.150 fault interpretation at the time, but Walter Mooney and I dutifully 00:05:29.150 --> 00:05:35.240 drew a vertical fault at Loma Prieta with a fault right next to it – 00:05:35.240 --> 00:05:38.400 dipping fault right next to it that was a Loma Prieta rupture 00:05:38.400 --> 00:05:41.840 in our chapter in the Bob Wallace professional paper 00:05:41.840 --> 00:05:44.420 on the San Andreas Fault. 00:05:45.980 --> 00:05:49.960 Okay, so 20 years later, we did the Salton Seismic Imaging Project, 00:05:49.980 --> 00:05:53.580 and we discovered that the San Andreas is actually steep – 00:05:53.580 --> 00:05:57.360 dips steeply northeast in the Coachella Valley. 00:05:57.360 --> 00:05:58.990 And then bends over about 6 kilometers – 00:05:58.990 --> 00:06:03.940 so moderate depth – 55 to 60 degrees. 00:06:05.160 --> 00:06:12.150 So anyhow, as – I asked myself, where have I seen this kind of thing, 00:06:12.150 --> 00:06:17.900 or suspected it, before? And the answer was Loma Prieta. So take a look at it. 00:06:18.730 --> 00:06:24.060 This is a summary of our interpretations at Loma Prieta. 00:06:24.070 --> 00:06:31.380 This is a seismic – I’m just going to – we’ll see this later, so not much detail. 00:06:31.380 --> 00:06:35.660 Seismic velocity model with fault interpretations – 00:06:35.660 --> 00:06:39.380 two fault interpretations superimposed. 00:06:39.900 --> 00:06:46.120 The – let’s see, are we on over here? Hang on just a second. 00:06:48.200 --> 00:06:49.840 Yep, okay. 00:06:53.080 --> 00:06:57.380 Hope I didn’t turn it off. Anyhow, there are two fault 00:06:57.390 --> 00:07:03.540 interpretations – white and gray – superimposed on this image here. 00:07:03.540 --> 00:07:11.070 The white interpretation is for steep-dip reflection detection and migration. 00:07:11.070 --> 00:07:17.370 And the gray is for double-difference earthquakes that were developed – okay, 00:07:17.370 --> 00:07:21.340 so there were two tools developed since all the research on Loma Prieta. 00:07:21.340 --> 00:07:27.790 One was steep-dip reflection detection and migration, 00:07:27.790 --> 00:07:30.050 shown here with the little black lines. 00:07:30.050 --> 00:07:35.290 And double-difference earthquake relocation, shown by the gray dots. 00:07:35.290 --> 00:07:42.230 And these were available for all of California beginning in about 2008. 00:07:42.230 --> 00:07:47.640 So pretty late in the game for the early research. 00:07:47.640 --> 00:07:53.080 And Waldhauser and Schaff were the guys that put together the catalog. 00:07:53.920 --> 00:07:56.780 So that’s all I’m going to say for the moment. 00:07:56.780 --> 00:08:00.889 Oh, one other thing. We still have this nagging data gap 00:08:00.889 --> 00:08:02.730 between about 5 kilometers and 00:08:02.730 --> 00:08:08.180 10 kilometers that dogged us back in the early days of research 00:08:08.180 --> 00:08:15.200 on Loma Prieta that prevented us from confidently drawing a connection 00:08:15.210 --> 00:08:20.260 between these steep faults in the upper crust – uppermost crust and the rupture, 00:08:20.260 --> 00:08:26.900 which goes from 10 down to 18 kilometers’ depth. 00:08:31.060 --> 00:08:36.540 Okay, so we collected seismic refraction data in 1991. 00:08:36.550 --> 00:08:38.960 Rufus headed up the project. 00:08:38.960 --> 00:08:46.339 This is one of the profiles that he and I shot – I helped, and lots of other people. 00:08:46.339 --> 00:08:48.930 And it ran from Aptos over to Calero Reservoir. 00:08:48.930 --> 00:08:54.740 Let’s look at the blow-up over here – Aptos to Calero Reservoir. 00:08:54.740 --> 00:08:59.840 Had four shot points. I don’t know if you can see the little red dots there. 00:08:59.840 --> 00:09:05.879 And the station spacing was about 150 meters. 00:09:05.880 --> 00:09:07.920 So those were the parameters. 00:09:09.580 --> 00:09:12.680 It was 25 kilometers long. 00:09:12.680 --> 00:09:20.589 And, in contrast, in SSIP in 2011, we shot 40-kilometer-long 00:09:20.589 --> 00:09:27.259 fault-crossing transects. They were very densely shot – 16 shots. 00:09:27.260 --> 00:09:33.240 And the station spacing was 50 to 100 meters, as opposed to 150 meters here. 00:09:34.340 --> 00:09:40.940 Now, to use Klaus Bauer’s steep-dip reflection detection and migration 00:09:40.949 --> 00:09:44.630 method, you’d ideally like to have at least four data points for wavelength, 00:09:44.630 --> 00:09:51.290 and wavelength is about 300 meters. And so 150 meters really didn’t cut it. 00:09:51.290 --> 00:09:54.350 80 meters is what we needed, but we said, ah, we’ll give it 00:09:54.350 --> 00:09:57.420 a shot anyhow to see if there’s anything there. 00:09:58.680 --> 00:10:00.120 Okay. 00:10:00.120 --> 00:10:07.820 So Rufus Open-Filed this velocity model along that profile in 2004. 00:10:07.820 --> 00:10:17.520 And so what it shows here – the four shot points – the yellow stars. 00:10:17.520 --> 00:10:22.580 Various faults – the Zayante Fault. San Andreas is in this little valley here. 00:10:22.580 --> 00:10:27.600 And Sargent Fault is more or less on Loma Prieta, which is the high point. 00:10:27.600 --> 00:10:32.240 And Berrocal Fault is over here. The velocities range from 00:10:32.240 --> 00:10:37.700 2 kilometers per second in blue to 6 kilometers per second in red. 00:10:38.400 --> 00:10:41.360 But I want to call your attention to a low-velocity zone – 00:10:41.370 --> 00:10:47.740 this little fat finger that kind of sticks down about 3 kilometers deep 00:10:47.740 --> 00:10:53.220 from the surface from under Loma Prieta. 00:10:53.960 --> 00:10:58.060 It’s flanked by, over here, a high-velocity zone, 00:10:58.060 --> 00:11:02.220 which underlies the San Andreas itself. The San Andreas is here. 00:11:02.220 --> 00:11:09.009 And this little high-velocity prong actually sticks up and to the northeast – 00:11:09.009 --> 00:11:10.559 to the northeast of the San Andreas Fault. 00:11:10.560 --> 00:11:15.420 There’s a – kind of a mirror image high-velocity zone over here. 00:11:15.420 --> 00:11:20.440 Well, in order to migrate reflections, we needed a bigger model. 00:11:20.440 --> 00:11:25.140 And this is the one we used. Different color scale. 00:11:25.140 --> 00:11:28.629 Here we go from violet, which is 2 kilometers per second, 00:11:28.629 --> 00:11:36.249 to red, which is 7 kilometers per second, with a big – oh, and a big step here 00:11:36.249 --> 00:11:41.500 under about the San Andreas Fault, which is in this little valley here. 00:11:42.500 --> 00:11:45.740 Here, the low-velocity zone is all in green. 00:11:45.740 --> 00:11:49.850 It shows up – this is the feature here. And this is this little high-velocity 00:11:49.850 --> 00:11:54.180 prong that sticks off to the northeast of the San Andreas. 00:11:55.100 --> 00:12:02.540 Okay, these were the earthquakes in red that were used in the 00:12:02.540 --> 00:12:07.040 Loma Prieta research back in the – 20 years ago – catalog locations. 00:12:07.040 --> 00:12:11.100 And in gray are the new double-difference earthquake 00:12:11.100 --> 00:12:17.280 locations from the Waldhauser and Schaff catalog. 00:12:17.280 --> 00:12:22.040 And just to anticipate our interpretation, we made this simple-minded 00:12:22.040 --> 00:12:29.459 interpretation that this cluster, which was pretty diffuse 20 years ago, 00:12:29.459 --> 00:12:35.170 now has at least three sub-clusters in it. So we matched those up with the 00:12:35.170 --> 00:12:39.760 three faults up here – San Andreas, Sargent, and Berrocal. 00:12:39.760 --> 00:12:44.620 There may be other interpretations, but that was the most straightforward one. 00:12:44.620 --> 00:12:49.339 And then here’s the rupture. And then here’s this no-man’s-land. 00:12:49.339 --> 00:12:52.800 We still don’t have any good earthquake alignments in this region 00:12:52.800 --> 00:12:56.790 between 5 and 10 kilometers. So we can’t really confidently 00:12:56.790 --> 00:13:01.060 connect the steep faults in the upper crust with the rupture. 00:13:02.380 --> 00:13:06.680 Okay, now I’ll go through the data processing required for this 00:13:06.680 --> 00:13:11.879 new tool called steep reflection detection and migration. 00:13:11.880 --> 00:13:17.700 These are Fourier transforms of all of our signals from all stations in the 00:13:17.700 --> 00:13:21.780 profile and the average down here. And you can see that, really, 00:13:21.780 --> 00:13:29.180 our explosions generate frequencies between 10 and 20 hertz, by and large. 00:13:29.189 --> 00:13:31.689 So that’s where we’re going to band-pass things. 00:13:31.689 --> 00:13:37.800 And this is the shot gathered from shot point 8, south of the San Andreas Fault. 00:13:37.800 --> 00:13:43.220 And this is when we top-mute it – we take away the first arrivals and a few 00:13:43.230 --> 00:13:50.189 wiggles past so we can begin to see the faint reflections in the record section. 00:13:50.189 --> 00:13:56.140 And then we bottom-mute it to get rid of S waves and surface waves. 00:13:56.140 --> 00:14:03.480 And then we band-pass it between 10 and 20 hertz, AGC it, 00:14:03.480 --> 00:14:09.600 and then spectrally whiten it to bring up all amplitudes in our 00:14:09.600 --> 00:14:16.980 pass band so that we can see reflections at all frequencies in our pass band. 00:14:16.980 --> 00:14:21.960 And this is the result for that shot point – shot point 8. 00:14:21.960 --> 00:14:27.200 I have three panels here. The top three are the time domain, 00:14:27.200 --> 00:14:32.260 and the bottom one is the depth domain. This is the processed data. 00:14:32.260 --> 00:14:36.910 These are the computer picks of steep-dip reflections. 00:14:36.910 --> 00:14:42.440 And these are some reflections of interest that I’ll talk about later. 00:14:43.200 --> 00:14:46.559 So let’s see how this method works. 00:14:48.240 --> 00:14:53.520 Okay. Here’s some synthetic data in the upper left, if I get the mouse 00:14:53.529 --> 00:14:59.699 to move here – with the source is the star, and it’s labeled S. 00:14:59.699 --> 00:15:04.319 And a receiver, labeled R, is the triangle here. 00:15:04.320 --> 00:15:08.320 And here we have the P wave – this thing. 00:15:08.320 --> 00:15:11.069 The S wave. 00:15:11.069 --> 00:15:14.949 A normal move-out reflection from this boundary 00:15:14.949 --> 00:15:18.290 in the model – right here – this horizontal boundary. 00:15:18.290 --> 00:15:22.410 And a reverse move-out reflection – a very strong one – from this 00:15:22.410 --> 00:15:27.200 very narrow but very deep low-velocity zone that’s meant to mimic a fault. 00:15:27.200 --> 00:15:30.489 It’s a little white stripe that you see here. 00:15:30.489 --> 00:15:38.420 So we go down this trace at the receiver, look at all the times, and we look 00:15:38.420 --> 00:15:43.300 in all directions to look for correlations – phase correlations. 00:15:43.300 --> 00:15:49.160 And here at 5.4 seconds, we find a nice phase correlation. 00:15:49.160 --> 00:15:53.100 In fact, its perfect. The semblance is 1. 00:15:53.600 --> 00:15:58.379 And so we pick that phase, draw a line, and then we trace a ray backwards 00:15:58.379 --> 00:16:04.320 from the receiver to the isochrone here, which is the locus of all times of 00:16:04.320 --> 00:16:12.660 equal reflection from source S to receiver R, and draw a tangent to the – 00:16:12.660 --> 00:16:17.060 to that isochrone, and that’s the migrated reflection. 00:16:17.060 --> 00:16:25.560 Well, the requirements for this method are, first, reverse move-out 00:16:25.560 --> 00:16:29.200 reflections only. That’s the ones that dip away from the source. 00:16:29.209 --> 00:16:31.529 The source is right here in the middle. 00:16:31.529 --> 00:16:35.629 And so the ones that dip away from it are the ones we’re after. 00:16:35.629 --> 00:16:41.709 The second is the semblance has to be between 0.6 and 1. 1 is the maximum. 00:16:41.709 --> 00:16:45.990 And the third is that, in the depth domain, the dips have to be greater 00:16:45.990 --> 00:16:52.120 than 30 degrees. So those are rigid criteria in the algorithm. 00:16:52.120 --> 00:16:55.440 So let’s look at the data again – the top panel. 00:16:55.449 --> 00:17:00.749 This is the processed data. And if you look at it, you can see by eye 00:17:00.749 --> 00:17:07.400 some phases that you would likely pick if you were picking things by eye. 00:17:07.400 --> 00:17:09.970 And then you can see what the computer picked. 00:17:09.970 --> 00:17:14.380 And sometimes they’re the same phases, but sometimes not. 00:17:14.780 --> 00:17:21.540 Sometimes by eye, you can’t estimate what the real semblance is of a phase. 00:17:21.540 --> 00:17:26.940 And so the computer kicks it out if it’s not 0.6 or greater. 00:17:26.940 --> 00:17:29.080 And sometimes you can’t even necessarily 00:17:29.080 --> 00:17:31.820 see some of these phases the computer picks. 00:17:31.820 --> 00:17:36.480 So it is possible to do this all by eye, and I’ve done that – not for Loma Prieta, 00:17:36.480 --> 00:17:40.820 but for imaging LARSE II out in southern California, 00:17:40.820 --> 00:17:44.370 where I picked stuff by eye, and then Klaus picked it by computer, 00:17:44.370 --> 00:17:47.220 and we compared them, and the results were fairly similar. 00:17:47.220 --> 00:17:52.180 I’m not saying that would be true here, but if you do it by eye, then you’re 00:17:52.190 --> 00:17:58.220 subject to the criticism that, well, this is all subjective, and so we didn’t do that. 00:17:59.460 --> 00:18:04.150 And then here are two reflections of interest – B and C. 00:18:04.150 --> 00:18:08.690 And now we’ll go on to the next shot point over here – 00:18:08.690 --> 00:18:17.630 shot point 3 north of the Berrocal Fault. And same thing – processed data, 00:18:17.630 --> 00:18:22.860 computer picks of reverse move-out reflections. 00:18:22.860 --> 00:18:28.250 Special reflections we want to look at – A, D, and E. 00:18:28.250 --> 00:18:33.520 And they fall in these boxes here – A, D, and E. 00:18:33.840 --> 00:18:40.900 And this are all reflections from all shot points plotted on our velocity model. 00:18:40.910 --> 00:18:46.000 Let’s look at boxes A, B, C, D, and E. 00:18:46.000 --> 00:18:49.980 A are reflections that come up, and when last seen, they’re curving 00:18:49.980 --> 00:18:54.250 over toward this little valley here where the San Andreas lives. 00:18:54.250 --> 00:18:59.280 Okay, B are reflections under the Sargent Fault. 00:18:59.280 --> 00:19:06.610 C is kind of a broader zone that seems to be associated with Berrocal Fault. 00:19:06.610 --> 00:19:12.540 D are some reflections that may or may not be associated with the Zayante Fault. 00:19:12.540 --> 00:19:19.660 And E are some weak reflections that seem to crosscut any attempt to draw a 00:19:19.660 --> 00:19:24.200 connection between steep faults in the upper crust and the rupture down here. 00:19:24.200 --> 00:19:28.251 So they’re problematic, and we’ve interpreted them as noise, but they 00:19:28.251 --> 00:19:31.750 may not be. They may just throw this whole thing out. 00:19:31.750 --> 00:19:36.650 But it’s hard to [chuckles] – hard to know exactly what those phases are. 00:19:36.650 --> 00:19:45.310 So now let’s go ahead and look at – well, actually, let me – this curvature 00:19:45.310 --> 00:19:49.980 in the upper couple kilometers here is very curious and peculiar. 00:19:49.980 --> 00:19:53.460 And, you know, we’ve said, what the hell? 00:19:53.460 --> 00:19:59.390 It does, however, bend around this high-velocity prong under the 00:19:59.390 --> 00:20:04.480 San Andreas, from down here at the bottom of the low-velocity zone, 00:20:04.480 --> 00:20:08.200 around the top, and when last seen, it’s headed toward the surface. 00:20:08.200 --> 00:20:14.480 The real reason we drew a curved upper San Andreas Fault there is 00:20:14.480 --> 00:20:19.560 from magnetic data, which was done in 2004 by Jachens and Griscom. 00:20:19.560 --> 00:20:23.460 Here’s our profile with the shot points, LP. 00:20:23.460 --> 00:20:27.290 And we’ll look at this profile here, E-E-prime, which is 5 kilometers 00:20:27.290 --> 00:20:30.440 southeast of our seismic profile. 00:20:30.440 --> 00:20:34.990 And then we’ll look at another profile down here at Watsonville – S-S-prime. 00:20:34.990 --> 00:20:40.740 So this is E-E-prime. And here’s the data up here – 00:20:40.740 --> 00:20:43.960 the gridded data are the – are the little circles. 00:20:43.960 --> 00:20:47.820 And the model predictions are the – is the black line. 00:20:47.820 --> 00:20:53.220 And the bodies down here are labeled with magnetic susceptibilities. 00:20:54.060 --> 00:20:56.220 But the thing that caught my eye was this. 00:20:56.220 --> 00:20:58.560 I mean, what the heck’s going on here? 00:20:58.560 --> 00:21:03.290 The upper part of the San Andreas curves off toward the southwest. 00:21:03.290 --> 00:21:05.880 And the control for that is up here. 00:21:05.880 --> 00:21:11.160 It’s not perfectly matched, but you can get a sense of what’s going on. 00:21:11.170 --> 00:21:17.650 The Berrocal Fault is very well-matched. Berrocal Fault is this boundary here. 00:21:17.650 --> 00:21:21.820 And then off in San Jose, the Silver Creek Fault is also very well-matched, 00:21:21.820 --> 00:21:27.490 the southwest-dipping fault here. The Zayante Fault is modeled as dipping 00:21:27.490 --> 00:21:33.960 southwest here, but the match in the data don’t – are not too convincing. 00:21:33.960 --> 00:21:40.920 Now let’s jump down to Watsonville. And those are very well-constrained. 00:21:40.930 --> 00:21:45.720 Here we have the Logan Gabbro, which is a very magnetic body 00:21:45.720 --> 00:21:51.360 and very nicely defines the surface of the San Andreas Fault. 00:21:51.360 --> 00:21:56.220 You can see the data up here really well-constrained that. 00:21:56.220 --> 00:22:00.300 And so this fault goes down to about 8 or 9 kilometers. 00:22:00.300 --> 00:22:04.610 And the aftershocks that we had at the time all cluster around 00:22:04.610 --> 00:22:09.490 this surface, and they continue going down to depth. 00:22:09.490 --> 00:22:14.490 So here is one place where it looks like the Loma Prieta rupture, 00:22:14.490 --> 00:22:18.710 as indicated by the aftershocks that you see here, was on the 00:22:18.710 --> 00:22:22.190 San Andreas Fault proper, and it goes right to the surface where 00:22:22.190 --> 00:22:24.590 the trace of the San Andreas is located. 00:22:24.590 --> 00:22:27.830 That doesn’t necessarily mean that we’re still on – that the rupture was 00:22:27.830 --> 00:22:33.520 still on the San Andreas Fault at Loma Prieta. But it’s kind of suggestive. 00:22:34.620 --> 00:22:40.840 Okay, so this is our – once again, our interpretation of faults 00:22:40.840 --> 00:22:47.870 through these three groups of reflections up here – steep-dip reflections. 00:22:47.870 --> 00:22:52.020 Everything that’s dashed is – has an uncertain interpretation, 00:22:52.020 --> 00:22:56.620 including the Zayante Fault here. We have it dipping northeast, and 00:22:56.620 --> 00:23:01.610 magnetic has it dipping southwest. It’s the boundary of the La Honda Basin, 00:23:01.610 --> 00:23:06.620 and so we’re not going to interpret anything here. 00:23:06.620 --> 00:23:11.420 And once again, we’re dashed through this area from 5 to 10 kilometers. 00:23:11.420 --> 00:23:13.820 We just don’t have any data. 00:23:13.820 --> 00:23:18.660 Again, not even reflections except for this puzzling feature here. 00:23:18.660 --> 00:23:23.230 Okay. And then, just to remind you again, here’s the double-difference 00:23:23.230 --> 00:23:28.890 earthquake, three sub-clusters up here, and once again, double-difference 00:23:28.890 --> 00:23:34.400 didn’t pull out any more earthquakes in this maddening gap here. 00:23:34.400 --> 00:23:40.500 And then this is the superposition of both those interpretations. 00:23:40.510 --> 00:23:46.740 They’re arguably similar. What you can say is that we 00:23:46.740 --> 00:23:50.140 don’t know the absolute location of the double-difference 00:23:50.140 --> 00:23:54.140 earthquake clusters here. They can move one way or the other. 00:23:54.140 --> 00:23:57.480 We also don’t know exactly where these reflections are. 00:23:57.480 --> 00:24:00.570 So these interpretations may be coincident, 00:24:00.570 --> 00:24:04.330 or they may be even further apart. We don’t know. 00:24:04.330 --> 00:24:13.270 It would be nice to redo this with 3D – with hypocenters obtained by 00:24:13.270 --> 00:24:19.020 3D modeling as initial hypocenters for the double-difference modeling. 00:24:19.020 --> 00:24:24.320 These are taken from Dave Oppenheimer’s 1D series of 00:24:24.320 --> 00:24:29.650 models up and down the San Andreas. And the initial hypocenters 00:24:29.650 --> 00:24:37.240 are from those 1D models. But it’d be interesting to do 3D and see, 00:24:37.240 --> 00:24:39.270 if you take those as initial epicenters, 00:24:39.270 --> 00:24:42.360 how it moves things. Down at Parkfield, it moved it a few 00:24:42.360 --> 00:24:47.730 hundred meters, and it moved it to the northeast when it was done down there. 00:24:47.730 --> 00:24:51.170 Okay, so now, where did all this come from? 00:24:51.170 --> 00:24:56.150 Down in the Salton Trough. The Salton Sea is blue. 00:24:56.150 --> 00:25:00.160 The Coachella Valley is the flat area here. 00:25:00.160 --> 00:25:04.860 These are – this is the Peninsular Ranges. San Andreas Fault is in orange. 00:25:04.860 --> 00:25:09.280 And interpreted flower structure faults are in yellow. 00:25:09.280 --> 00:25:15.640 The shot points are these little tiny red and larger red dots. 00:25:15.640 --> 00:25:18.800 So we shot from the Peninsular Ranges across the valley, 00:25:18.800 --> 00:25:21.050 through the Mecca Hills – this is Mecca Hills – Mike Rymer 00:25:21.050 --> 00:25:24.340 spent a lot of his career mapping this stuff. 00:25:24.340 --> 00:25:29.440 And then all the way out to – this is I-10, which goes through Indio. 00:25:29.440 --> 00:25:34.400 And so we have two models for that profile. 00:25:34.400 --> 00:25:38.410 The seismic model at the top and a potential field model, 00:25:38.410 --> 00:25:42.540 done by Vicky Langenheim, down here on the bottom. 00:25:42.540 --> 00:25:45.601 And she actually published hers – or, showed it at AGU before 00:25:45.601 --> 00:25:50.000 we were done with ours, so they were done more or less independently. 00:25:50.000 --> 00:25:51.880 Well, they were done independently. 00:25:51.880 --> 00:25:58.730 And what I’ve done is – I’ll blow this – show a blow-up here in a second, 00:25:58.730 --> 00:26:03.980 but I’ve taken interpretive faults in gray up here from our seismic model 00:26:03.980 --> 00:26:07.010 and superimposed them on Vicky’s model down here 00:26:07.010 --> 00:26:08.810 to show you how similar they are. 00:26:08.810 --> 00:26:12.360 She has a two-part dip to the San Andreas just like we do. 00:26:12.360 --> 00:26:15.840 Okay, let’s look at the seismic model. 00:26:16.860 --> 00:26:20.820 Okay, so steep-dip reflections are shown here a little differently 00:26:20.830 --> 00:26:22.590 from what we did at Loma Prieta. 00:26:22.590 --> 00:26:26.300 We’re not showing individual reflections, but we’re contouring 00:26:26.300 --> 00:26:30.380 reflection density, and the highest density is red. 00:26:30.380 --> 00:26:38.610 So you can see this package of steep-dip reflections here, here, here, here. 00:26:38.610 --> 00:26:43.490 And then they change dip at about 6 kilometers, and then we lose 00:26:43.490 --> 00:26:47.160 any image of the San Andreas Fault all the way to the surface, except for 00:26:47.160 --> 00:26:52.500 these possible flowers up here. We also image – the PCF here 00:26:52.500 --> 00:26:57.500 is the Painted Canyon Fault, which looks like a flower structure. 00:26:57.500 --> 00:27:05.000 And the Platform Fault over here has a bit of an image – a reflection image – 00:27:05.010 --> 00:27:09.220 and may or may not correlate with this group of hypocenters, which are shown 00:27:09.220 --> 00:27:11.570 in orange and are actually southeast – 00:27:11.570 --> 00:27:14.190 from 4 to 6 kilometers southeast of our line. 00:27:14.190 --> 00:27:19.710 The black hypocenters are plus or minus 2 kilometers from our line. 00:27:19.710 --> 00:27:24.510 And you can see that they – actually, both groups define 00:27:24.510 --> 00:27:30.500 a fairly nicely northeast-dipping structure here. 00:27:30.500 --> 00:27:33.550 And many of the mechanisms are actually strike-slip, which gives 00:27:33.550 --> 00:27:40.010 you confidence that, well, yeah, we’re looking at the San Andreas. 00:27:40.010 --> 00:27:49.240 So – okay, so we’ll take these blue – these gray lines, color them blue, 00:27:49.240 --> 00:27:54.260 and reverse them, and we’ll superimpose them on our inferred 00:27:54.270 --> 00:27:59.010 Loma Prieta structure here, anchoring them at the San Andreas. 00:27:59.010 --> 00:28:03.200 And you can see a similarity. Steep faults in the upper crust 00:28:03.200 --> 00:28:10.600 connecting – or, bending over to a moderately dipping fault at depth. 00:28:12.060 --> 00:28:17.940 Now, a little bit of geology here. I hope this isn’t too confusing, but in 00:28:17.940 --> 00:28:23.960 both Loma Prieta and Mecca Hills, North America – North American 00:28:23.960 --> 00:28:26.000 rocks are subduction complex rocks. 00:28:26.000 --> 00:28:31.060 And, in both places, the Pacific Plate rocks are batholithic rocks. 00:28:31.840 --> 00:28:37.100 In the Coachella Valley, the North American rocks are in the hanging wall. 00:28:37.100 --> 00:28:43.440 And at Loma Prieta, the Pacific Plate rocks are in the hanging wall. 00:28:43.440 --> 00:28:46.611 So they’re slightly different that way. But the flower structure seems to be 00:28:46.620 --> 00:28:51.760 mostly confined to North American rocks in both places. 00:28:51.760 --> 00:28:54.360 So, for all that’s worth. 00:28:55.680 --> 00:29:05.680 Okay. So just want to summarize what I’ve tried to show here so far is that the 00:29:05.680 --> 00:29:14.680 San Andreas Fault can be non-vertical and can have a non-planar dip. 00:29:14.680 --> 00:29:17.640 And here we see it in these two places. 00:29:17.640 --> 00:29:23.180 Now, where else in California do we see this kind of thing? 00:29:24.200 --> 00:29:30.500 Well, there are 11 places where we have fault-crossing profiles that contain 00:29:30.510 --> 00:29:40.420 either seismic or potential field data. And of those, only three – there’s only 00:29:40.420 --> 00:29:43.480 three where the San Andreas is vertical. And it appears to be vertical, as far as 00:29:43.480 --> 00:29:47.900 we can tell, all the way down to the Moho and even into the lithosphere. 00:29:47.900 --> 00:29:55.020 On the other eight fault-crossing profiles, the San Andreas dips moderately. 00:29:55.020 --> 00:29:58.770 And every place where we’ve looked at it in detail – these moderately dipping 00:29:58.770 --> 00:30:03.490 San Andreas Faults – it comes up to the upper crust and then bends over 00:30:03.490 --> 00:30:10.500 to a steep fault at that juncture. Now, you have to beware 00:30:10.500 --> 00:30:14.020 of generalizations. It may be that the fourth one of 00:30:14.030 --> 00:30:19.310 these dipping San Andreas locations, if we get enough detail, 00:30:19.310 --> 00:30:23.660 will show something different from this non-planar structure. 00:30:23.660 --> 00:30:27.140 But anyhow, this is – this is what we – what we have at the moment. 00:30:27.150 --> 00:30:32.190 So let me speculate about why the dipping of the San Andreas 00:30:32.190 --> 00:30:38.020 might flip over to a steep fault in this upper crustal layer. 00:30:38.120 --> 00:30:42.660 And here’s the idea. 00:30:44.420 --> 00:30:48.900 A dipping San Andreas comes up to a layer where the rheology changes. 00:30:48.900 --> 00:30:52.240 The earthquakes stop, or start petering out, 00:30:52.240 --> 00:30:56.640 about 6 kilometers – they go up to 2 kilometers at Loma Prieta. 00:30:57.620 --> 00:30:59.820 And so there’s something different about this upper layer. 00:30:59.820 --> 00:31:04.680 It’s not – it’s probably nonelastic, at least at some time scales. 00:31:04.680 --> 00:31:10.960 And, whereas, earthquakes occur down where you can store and 00:31:10.960 --> 00:31:15.240 release strain in an elastic fashion. So there’s something – some nonlinear 00:31:15.240 --> 00:31:21.430 elements added in this layer above 5 or 6 kilometers. 00:31:21.430 --> 00:31:26.860 So when a dipping fault hits that, it may be easier – take less energy 00:31:26.860 --> 00:31:33.680 for it just to go straight to the surface through this layer with 00:31:33.680 --> 00:31:39.940 different rheology, rather than continuing on as a longer fault 00:31:39.940 --> 00:31:43.240 through this viscous layer up to the surface. 00:31:43.240 --> 00:31:49.800 It’d be less energy to move the blocks past a shorter fault than a longer fault. 00:31:49.809 --> 00:31:53.559 This could be tested numerically, and anybody interested in doing this, 00:31:53.559 --> 00:31:56.970 I’d be all ears. Anybody interested – anybody have any 00:31:56.970 --> 00:32:06.080 other explanations for this phenomenon, I’d also be all ears to hear about. 00:32:08.040 --> 00:32:12.060 So anyhow, you know, I think that’s – I’m going to 00:32:12.070 --> 00:32:15.871 stop at this point and take questions, so … 00:32:15.880 --> 00:32:22.260 [Applause] 00:32:25.280 --> 00:32:27.300 - Thanks, Gary, for a really interesting … 00:32:27.300 --> 00:32:28.500 - Okay. [chuckles] 00:32:28.500 --> 00:32:30.330 - … ideas there. Do we have audience questions? 00:32:30.330 --> 00:32:33.300 - How did you come up with the extended velocity model 00:32:33.300 --> 00:32:36.590 down to 20 kilometers? - You did, Mark. You tell us. 00:32:36.590 --> 00:32:41.140 - No, I mean, but which one – which version did you use of the … 00:32:41.140 --> 00:32:44.610 - [chuckles] You’ve got all the versions on your computer. 00:32:44.610 --> 00:32:49.280 I don’t know. We modeled it after a profile that Rufus had 00:32:49.280 --> 00:32:52.960 collected 50 kilometers north across the peninsula, and … 00:32:52.960 --> 00:32:55.560 - Oh, okay, so you based it on a larger scale. 00:32:55.560 --> 00:33:01.720 - Yeah, a larger-scale model. And that step in the – in the – 00:33:01.720 --> 00:33:07.980 the big step in the velocity model there actually matches a step that 00:33:07.980 --> 00:33:11.050 Donna Eberhart- Phillips and Andy Michael had in their 00:33:11.050 --> 00:33:20.790 tomographic model of the peninsula. So that step is not something we – 00:33:20.790 --> 00:33:23.100 that we alone came up with. - Right. That’s what I was 00:33:23.100 --> 00:33:25.680 wondering, where … - Yeah. Right. 00:33:29.140 --> 00:33:32.800 - Gary, I thought that was an interesting idea about the rheology. 00:33:32.809 --> 00:33:35.200 And, as you were speaking, I was wondering about the 00:33:35.200 --> 00:33:38.140 thermal structure in these areas, and I was kind of looking around 00:33:38.140 --> 00:33:40.370 to see if Colin Williams was in the room, but … 00:33:40.370 --> 00:33:42.170 - Well, Wayne Thatcher is right behind you. 00:33:42.170 --> 00:33:45.110 - Wayne. Okay. Well, the – you know, the thermal structure, I think, would be 00:33:45.110 --> 00:33:49.200 fairly well-known in both these areas. And would we expect, for instance, 00:33:49.200 --> 00:33:52.890 relief on the brittle-ductile transition because of the contrast in thermal 00:33:52.890 --> 00:33:54.460 conductivity between the Pacific Plate and the … 00:33:54.460 --> 00:33:56.900 - This is not – this is not the brittle-ductile transition. 00:33:56.900 --> 00:33:59.700 That’s way down. - Right. But I would think that 00:33:59.700 --> 00:34:04.520 there might be – well, you know, the brittle-ductile transition is typically 00:34:04.520 --> 00:34:07.242 10 to 15 kilometers in continental crust. - Yeah. 00:34:07.242 --> 00:34:10.560 - And I was thinking it could certainly be shallower in the Salton Sea. 00:34:10.560 --> 00:34:12.869 - Well … - And possibly here it’s a 00:34:12.869 --> 00:34:15.060 fairly high heat flow area. Anyway, that was my … 00:34:15.060 --> 00:34:16.560 - Yeah. It’s definitely – definitely high heat flow. 00:34:16.560 --> 00:34:21.880 But we have earthquakes from about 6 down to 10 kilometers there. 00:34:21.880 --> 00:34:28.070 So it’s elastic, at least, down that far before you get to more ductile. 00:34:32.480 --> 00:34:34.280 - Hi, Gary. Nice talk. 00:34:34.280 --> 00:34:39.000 I had – I didn’t fully follow your explanation why you would expect 00:34:39.000 --> 00:34:42.830 that you would have the shallower- dipping section in deeper. 00:34:42.830 --> 00:34:46.050 I understood the explanation for the transition, but I didn’t understand 00:34:46.050 --> 00:34:51.070 why it would be shallower – why you would have the shallow dip at depth. 00:34:51.070 --> 00:34:56.940 And, to sort of follow on to that, would we – would you expect to see 00:34:56.940 --> 00:35:01.720 this sort of behavior elsewhere? - Okay, so, in the 11 places where 00:35:01.720 --> 00:35:05.760 we’ve crossed the San Andreas in central and southern California, 00:35:05.760 --> 00:35:09.490 we’ve seen vertical or steep dips only in three places – 00:35:09.490 --> 00:35:15.560 the two LARSE profiles and Parkfield are the three steep places. 00:35:15.560 --> 00:35:18.970 All the rest of them dip moderately everywhere we’ve looked. 00:35:18.970 --> 00:35:24.320 - By elsewhere, I meant other faults. - Oh. Other than San Andreas Fault? 00:35:24.320 --> 00:35:27.300 - Right. - No. We haven’t – we see … 00:35:27.300 --> 00:35:30.720 - But would you expect to see this sort of behavior? 00:35:32.160 --> 00:35:36.640 - [chuckles] That’s a good question. 00:35:36.640 --> 00:35:42.120 It’d be fun to model numerically what happens to a fault 00:35:42.120 --> 00:35:44.510 when it hits a changing rheology. 00:35:44.510 --> 00:35:49.800 And then after that, we can say a little bit more about what to expect. 00:35:52.380 --> 00:36:01.540 I forget his first name – Ma showed, in many cases, normal fault strike-slip 00:36:01.540 --> 00:36:06.000 thrust faults, as the fault approaches the free service, which is another 00:36:06.000 --> 00:36:11.460 parameter that I didn’t really bring into it, you get flowering. 00:36:12.200 --> 00:36:17.080 And so some of these other faults are actually flower structures 00:36:17.080 --> 00:36:21.620 which have somewhat different origins from the plate boundary. 00:36:24.440 --> 00:36:31.140 But they’re there because, among other things, as you approach the free surface, 00:36:31.140 --> 00:36:37.730 you get this diffusion of deformation outward from the trace. 00:36:37.730 --> 00:36:42.000 I’m kind of beating around the bush there, but [chuckles] … 00:36:42.000 --> 00:36:45.540 - Hey, Gary. Great talk. I have two questions. 00:36:45.540 --> 00:36:49.810 The first one is, wouldn’t you expect that this mechanical system – 00:36:49.810 --> 00:36:57.330 the San Andreas Fault – would behave in a similar way everywhere? 00:36:57.330 --> 00:37:01.840 But what you’re showing us here is, in southern California, it’s dipping to the 00:37:01.840 --> 00:37:06.560 east, and it’s just the exact opposite … - [chuckles] Okay. All right. 00:37:06.560 --> 00:37:09.180 - … in the Bay Area. It’s dipping to the west. 00:37:09.180 --> 00:37:11.860 - Yeah, it is. - So why doesn’t it behave 00:37:11.860 --> 00:37:16.080 in a consistent way? - Well, good question. 00:37:16.080 --> 00:37:19.800 So this shows southern California. And we’ve published a paper back in 00:37:19.800 --> 00:37:23.880 2012 showing that the San Andreas is actually propeller-shaped down there. 00:37:23.880 --> 00:37:29.220 It dips off to the northeast in the Coachella Valley, and especially 00:37:29.220 --> 00:37:34.000 in the San Bernardino Mountains. It goes to vertical, or very steep, 00:37:34.000 --> 00:37:38.800 in Cajon Pass, through the western Mojave – okay. 00:37:40.620 --> 00:37:43.660 It dips off to the northeast around this bend. 00:37:44.580 --> 00:37:49.280 It’s actually dipping about 37 degrees northeast at this point. 00:37:49.290 --> 00:37:54.660 And then goes to vertical – the LARSE profile came through here, here. 00:37:54.660 --> 00:37:58.540 And then flips around the other direction and dips off under Santa Barbara 00:37:58.540 --> 00:38:06.040 as you go around this bend here, and that’s this ribbon shape to it. 00:38:06.040 --> 00:38:09.840 So it has to do with bends in southern California. 00:38:11.700 --> 00:38:17.470 And for Loma Prieta, I would have to go back and think about that 00:38:17.470 --> 00:38:20.770 a little bit more as to whether there’s a bend. 00:38:20.770 --> 00:38:26.010 Ellsworth and – Dietz and Ellsworth proposed a bend as the reason 00:38:26.010 --> 00:38:31.980 for a southwestward dip of the Loma Prieta rupture. 00:38:31.980 --> 00:38:35.640 But I was never actually able to find it looking on a map. 00:38:35.640 --> 00:38:41.820 - Well, I wondering, could you go back to the example from southern California? 00:38:41.820 --> 00:38:47.540 The fault seems to dip away from the high-velocity – the more competent 00:38:47.540 --> 00:38:50.980 higher-velocity, higher-density rock. Yeah, here it’s dipping 00:38:50.980 --> 00:38:54.920 away from the red. The red is high-velocity. 00:38:54.920 --> 00:38:57.140 - I’m not sure what I’m doing here. What the heck? 00:38:57.140 --> 00:38:59.040 [laughter] - Well, any cross-section 00:38:59.050 --> 00:39:02.590 you show – there you go. So here, it’s – the fault dips 00:39:02.590 --> 00:39:05.210 away from the red. The red is high-velocity, so that 00:39:05.210 --> 00:39:11.500 maybe is a gabbro – a stronger layer – a stronger body, and it dips away from it. 00:39:11.500 --> 00:39:15.550 And then likewise, in the Coachella example, it dips eastward because 00:39:15.550 --> 00:39:19.300 it’s dipping away from – yeah, see, it’s kind of 00:39:19.300 --> 00:39:23.130 a darker yellow or getting towards the brown. 00:39:23.130 --> 00:39:29.490 So maybe it’s a lithological control rather than a geometrical. 00:39:29.490 --> 00:39:33.270 Maybe the rheology is controlling it. - Well, that’s a good point. 00:39:33.270 --> 00:39:39.700 It’s certainly the geometry because it dips differently going around a curvature 00:39:39.700 --> 00:39:45.220 to the – to the west than it does going around Bakersfield – 00:39:45.220 --> 00:39:49.700 curvature to the north. So that clearly is a factor. 00:39:49.700 --> 00:39:54.880 Hadn’t thought too much about this. By the way, the Peninsular Ranges’ 00:39:54.880 --> 00:39:58.430 batholithic rocks extend all the way under the Coachella Valley 00:39:58.430 --> 00:40:02.570 to the San Andreas Fault here. And so you’re right in pointing that out. 00:40:02.570 --> 00:40:06.580 Thanks for pointing that out. This basement right here 00:40:06.580 --> 00:40:09.670 is granitic batholithic rocks. 00:40:09.670 --> 00:40:14.640 It’s not the same basement we see in Imperial Valley, where the basement 00:40:14.640 --> 00:40:17.900 is metasedimentary Colorado River sediments, as you and I … 00:40:17.900 --> 00:40:20.720 - Right. And the last thing I’ll say – 00:40:20.720 --> 00:40:24.060 and I promise to shut up – is that the reason you have to work 00:40:24.060 --> 00:40:29.500 so hard to get your results is that you don’t have enough seismic stations. 00:40:29.500 --> 00:40:33.620 You know, if you had 1,000 stations – like, 1,000 nodes, the data would 00:40:33.620 --> 00:40:37.280 just show you exactly what you’re teasing out of the data. 00:40:37.280 --> 00:40:40.330 I mean, you’re really working hard now because – what did you 00:40:40.330 --> 00:40:45.410 say the station spacing was? - Loma Prieta is average 150 meters. 00:40:45.410 --> 00:40:49.250 - Yeah. If you’re doing it at 25 meters, you would just see these reflectors 00:40:49.250 --> 00:40:51.820 in the – in the data without a lot of work. 00:40:51.820 --> 00:40:53.820 People are giving talks now that show that. 00:40:53.820 --> 00:40:55.859 You know this as well as I do. - Yeah. Yeah. 00:40:55.859 --> 00:41:01.040 These are long profiles, so you’d need a lot of stations. And – but you’re right. 00:41:01.040 --> 00:41:06.540 I mean, there’s no question that more data would help resolve that. 00:41:08.320 --> 00:41:14.760 - It’s actually – I can go first. Hi, back here. Don’t be surprised that I 00:41:14.760 --> 00:41:17.489 have some opinions about these things. - Good. 00:41:17.489 --> 00:41:21.310 - So at Loma Prieta, first of all, there’s definitely a bend. 00:41:21.310 --> 00:41:25.170 If you look at the focal mechanism south of Loma Prieta rupture, 00:41:25.170 --> 00:41:31.310 the strike-slip mechanisms and then into the rupture have different strikes. 00:41:31.310 --> 00:41:37.160 And that’s seen with the 3D relocations and 3D ray tracing that Donna and I did. 00:41:38.240 --> 00:41:41.520 And of course, the Loma Prieta rupture itself, some of the – you know, 00:41:41.520 --> 00:41:45.310 we originally had it as a single – it’s almost funny that the first motion 00:41:45.310 --> 00:41:49.580 representative of the hypocenter shows an oblique mechanism. 00:41:49.580 --> 00:41:53.359 But probably some of the more finite source models suggest that it’s actually 00:41:53.359 --> 00:41:57.800 closer to more of a strike-slip – first strike-slip on the southern part 00:41:57.800 --> 00:42:01.450 of the rupture and more of a dip-slip on the northern part of the rupture. 00:42:01.450 --> 00:42:04.540 So, as you have this bend, and then you have the dip-slip, 00:42:04.540 --> 00:42:07.470 it makes sense to start supporting – that you have to start having that 00:42:07.470 --> 00:42:10.460 on a dipping surface. Because there’s no point in having 00:42:10.460 --> 00:42:14.320 dip-slip motion on a vertical surface, tectonically. 00:42:14.320 --> 00:42:17.240 It doesn’t – it doesn’t shorten or extend the crust. 00:42:17.240 --> 00:42:19.230 - Yeah. - But similarly, it also 00:42:19.230 --> 00:42:23.270 makes no energetic sense. It’s sort of a violation of Hamilton’s 00:42:23.270 --> 00:42:28.040 principle of – you want to minimize your transfer of, you know, 00:42:28.040 --> 00:42:31.860 potential to kinetic energy – back to mechanics. 00:42:31.860 --> 00:42:36.160 And if you have a dipping pure strike-slip fault, you have so much extra 00:42:36.160 --> 00:42:41.700 surface area on that fault that it shouldn’t be sustained over a long period of time 00:42:41.700 --> 00:42:45.980 because you’re just – you’re just dealing with a lot more energy in the system. 00:42:45.980 --> 00:42:49.760 So, like at Parkfield, we do see a very vertical fault, and we see the strike – 00:42:49.760 --> 00:42:54.730 we see the dip-slips off on a separate fault at Coalinga, you know, 00:42:54.730 --> 00:42:57.360 to have the cross-fault compression. - Right. Right. 00:42:57.360 --> 00:43:00.890 - So, I mean, I think the – you know, if we start looking at Loma Prieta 00:43:00.890 --> 00:43:04.900 in detail – and maybe – so I don’t like worrying about fault names, 00:43:04.900 --> 00:43:07.640 but there’s a lot of complexity in there because there is a bend. 00:43:07.640 --> 00:43:11.460 And there is this dip-slip motion in the main shock. 00:43:11.460 --> 00:43:13.930 When you start doing this in southern California, I think that then raises 00:43:13.930 --> 00:43:18.170 the question, do we expect these to be pure strike-slip earthquakes 00:43:18.170 --> 00:43:23.640 on these non-vertical surfaces? And if so, what is maintaining, 00:43:23.640 --> 00:43:26.810 you know, the ability of that to happen in a – you know, probably – 00:43:26.810 --> 00:43:28.420 what’s probably a highly fractured crust? 00:43:28.420 --> 00:43:32.480 Why isn’t it evolving towards a vertical fault? 00:43:33.060 --> 00:43:36.020 Or are we going to be very surprised by the motion on the fault when we actually 00:43:36.020 --> 00:43:38.800 have an earthquake down there where we actually know what happened? 00:43:38.800 --> 00:43:45.080 - Well, um – thanks, Andy. As far as Loma Prieta goes, 00:43:45.080 --> 00:43:48.680 I put a ruler down on a map – trace of the San Andreas Fault, 00:43:48.690 --> 00:43:51.540 and I couldn’t find a curve. And maybe Bob McLaughlin, 00:43:51.540 --> 00:43:54.300 who I thought I saw here … - [inaudible] [laughs] 00:43:54.300 --> 00:43:59.700 - … could show me that – where that curve is. 00:43:59.700 --> 00:44:02.980 - You’re talking about the [inaudible] restraining bend? 00:44:02.980 --> 00:44:05.720 - Restraining bend, yes. - Yeah. 00:44:08.040 --> 00:44:12.260 Well, I’m afraid I don’t see that much of a restraining bend myself, either. 00:44:12.260 --> 00:44:17.279 You’ll have to ask Roland Burgmann, who puts a lot of stock in that. 00:44:17.279 --> 00:44:21.020 - Okay. - But I guess I did have a comment, 00:44:21.020 --> 00:44:26.869 and speaking mainly from the point of view of the long-term 00:44:26.869 --> 00:44:30.720 framework geology, and as a field geologist, 00:44:30.720 --> 00:44:38.080 that I’m not sure that this kind of geometry should be unexpected, 00:44:38.080 --> 00:44:46.880 in that the San Andreas is actually a boundary that is superimposed on – 00:44:46.880 --> 00:44:52.280 or replaces a much older boundary involved in subduction. 00:44:52.280 --> 00:44:59.440 And looking at the modeling, for instance, of Jachens and Griscom 00:44:59.440 --> 00:45:05.880 from the magnetic and gravity, the crust over Loma Prieta, at least, 00:45:05.880 --> 00:45:10.800 which is what I speak more from, is quite complicated. 00:45:10.800 --> 00:45:19.880 And there’s a lot of variation – large blocks of different density and – 00:45:19.880 --> 00:45:25.190 which – whose boundaries have been offset by younger faults 00:45:25.190 --> 00:45:31.510 of the San Andreas system. But we see elsewhere that a lot of these 00:45:31.510 --> 00:45:37.010 older boundaries are being reactivated by the younger strike-slip faults. 00:45:37.010 --> 00:45:45.430 So I think that perhaps some really good insights into the answers 00:45:45.430 --> 00:45:52.750 that you’re seeking would be to consider the deeper crustal structure 00:45:52.750 --> 00:45:58.840 involved in these different boundaries between these various blocks. 00:45:58.840 --> 00:46:04.020 - Yeah. Thanks. I’m glad you couldn’t find it, either. [laughs] 00:46:04.020 --> 00:46:06.620 Anyhow, I’ll look for it. Okay, and thanks for bringing 00:46:06.620 --> 00:46:13.180 that up – the deeper tectonics. Let me bore you with one more slide. 00:46:16.180 --> 00:46:17.560 This one. 00:46:17.570 --> 00:46:21.980 Okay, there’s a good reason, probably, for these two bends 00:46:21.980 --> 00:46:25.100 in the San Andreas in southern California. 00:46:25.100 --> 00:46:30.790 If you imagine a locomotive with a cow catcher coming northward – 00:46:30.790 --> 00:46:35.560 the Peninsular Ranges – and pushing under the 00:46:35.560 --> 00:46:40.780 plate boundary there, then you would get a dip off to the northeast. 00:46:40.780 --> 00:46:46.240 The Sierra Nevada and its foothill rocks are coming southward. 00:46:46.240 --> 00:46:53.220 Another cow catcher pushing southward underneath the plate boundary there. 00:46:53.221 --> 00:46:56.670 And there you have the propeller. 00:46:56.670 --> 00:47:06.280 So this is a very deep explanation – or, guess at what’s going on. 00:47:06.280 --> 00:47:09.780 And we do see the Pacific Plate downwelling under the 00:47:09.780 --> 00:47:13.369 San Bernardino Mountains. I mean, it goes 30 kilometers 00:47:13.369 --> 00:47:17.660 past the trace of the San Andreas up toward Barstow, when last seen, 00:47:17.660 --> 00:47:20.590 by the time it gets to the Moho. 00:47:20.590 --> 00:47:25.650 So it’s really downwelling under southern California. 00:47:25.650 --> 00:47:30.560 It’s not a subduction, but it’s downwelling, I guess. 00:47:30.560 --> 00:47:32.660 So any other questions? 00:47:33.400 --> 00:47:37.400 - That was really a provocative talk, in the best sense of the word. 00:47:37.400 --> 00:47:44.460 The issue of what’s happening in the shallower crust – say, if we consider 00:47:44.460 --> 00:47:48.260 the Loma Prieta case, up to, I don’t know, down to 00:47:48.260 --> 00:47:53.990 6 kilometers or something like that. And then what’s happening between 00:47:53.990 --> 00:48:02.780 11 and 17, which is roughly the depth range for the Loma Prieta earthquake. 00:48:02.780 --> 00:48:05.070 The issue that you raised of, how are they connected … 00:48:05.070 --> 00:48:08.980 - Yep. - … seems to be an open question. 00:48:08.980 --> 00:48:10.800 - Yes. Still is. 00:48:10.800 --> 00:48:19.030 - Yeah. And in 1906, as you know, there was lateral offsets of about 00:48:19.030 --> 00:48:24.970 2 meters on the San Andreas Fault – you know, roughly near your profile. 00:48:24.970 --> 00:48:32.280 So the mechanism of the – of deformation has changed as well. 00:48:32.280 --> 00:48:38.660 - Well, I’m not sure, Wayne. It’s just that the – my guess is that 00:48:38.660 --> 00:48:43.200 displacement was large enough in that – in 1906 to actually make it up 00:48:43.200 --> 00:48:48.020 through the – up into the upper crust into the surface. 00:48:48.020 --> 00:48:52.171 Deformation in this Loma Prieta earthquake wasn’t maybe 00:48:52.180 --> 00:48:57.700 quite big enough to do that. That’s a hypothesis. 00:48:57.700 --> 00:49:03.660 - But there’s no vertical offset on the 1906 trace at the surface. 00:49:03.660 --> 00:49:05.920 It was pure strike-slip. 00:49:05.920 --> 00:49:10.790 - Well, and Loma Prieta – Loma Prieta actually went down. 00:49:10.790 --> 00:49:17.790 So there was – the hanging wall went up, and the mountain itself went down. 00:49:17.790 --> 00:49:23.040 - Right. - And so I don’t know how well-known 00:49:23.040 --> 00:49:28.660 elevation changes were in 1906. I mean, I agree the rupture probably 00:49:28.660 --> 00:49:32.260 didn’t show a lot of vertical, but … - No. It was strike-slip. 00:49:32.260 --> 00:49:38.869 And I don’t think there’s any data bearing on vertical displacement 00:49:38.869 --> 00:49:42.120 of Loma Prieta Mountain. - Yeah. 00:49:43.460 --> 00:49:46.380 - One more question. In talking to the public … 00:49:46.390 --> 00:49:47.390 - You’ve had yours. [laughter] 00:49:47.390 --> 00:49:48.390 - He said he wasn’t going to talk anymore. 00:49:48.390 --> 00:49:52.020 - I said I wasn’t going to talk. In talking to the public, 00:49:52.020 --> 00:49:56.500 we sometimes tell people that, oh, yeah, the Loma Prieta earthquake 00:49:56.500 --> 00:49:58.930 didn’t occur on the San Andreas Fault. 00:49:58.930 --> 00:50:03.340 It occurred on, you know, an adjacent fault. 00:50:03.340 --> 00:50:04.800 Would you agree with that? 00:50:04.800 --> 00:50:08.560 Or would you say that it occurred on the San Andreas Fault? 00:50:08.570 --> 00:50:12.260 - Okay. I would say it most likely occurred on the San Andreas Fault. 00:50:12.260 --> 00:50:14.870 We can’t prove it. 00:50:14.870 --> 00:50:17.560 As you know, in geophysics and physics, there’s no proof, just math. 00:50:17.560 --> 00:50:18.810 - That’s what you’re showing. - Huh? 00:50:18.810 --> 00:50:21.760 - You’re showing it on the San Andreas. - It’s not – that’s not proof. 00:50:21.760 --> 00:50:25.213 [laughter] - You don’t believe your own result. 00:50:25.220 --> 00:50:28.660 - I believe my – we have three data sets showing this. 00:50:28.660 --> 00:50:36.920 And so, with one data set, I wouldn’t believe it with just our reflections. 00:50:36.920 --> 00:50:40.120 Double-difference earthquakes, I wouldn’t believe it. 00:50:40.120 --> 00:50:42.400 Potential field, maybe. Maybe not. 00:50:42.410 --> 00:50:47.870 But when you get all three of them coinciding to show a structure that 00:50:47.870 --> 00:50:53.960 we’re showing here, I would say the evidence is in favor of that being the San 00:50:53.960 --> 00:50:59.180 Andreas – that shape that you’re – and white lines there being the San Andreas. 00:50:59.180 --> 00:51:04.369 It’s – however, it is an alternate hypothesis to the assumption – 00:51:04.369 --> 00:51:08.210 largely an assumption that was made 20 years ago that a strike-slip 00:51:08.210 --> 00:51:14.560 fault has to be vertical and planar. So it’s an alternate model to that model. 00:51:15.300 --> 00:51:23.000 - So I want to issue a caveat about earthquake – small earthquake locations. 00:51:23.000 --> 00:51:26.560 So a little bit further south, there’s a paper by Catherine Dorbath, 00:51:26.560 --> 00:51:30.070 Falk Amelung, and I where we looked at the San Andreas at the 00:51:30.070 --> 00:51:34.240 confluence of the Calaveras Fault. And there, the San Andreas – 00:51:34.240 --> 00:51:37.630 the mechanisms looks like garden-variety strike-slip, 00:51:37.630 --> 00:51:44.290 yet the fault dips about 5 – if I recall correctly, 5 degrees to the west. 00:51:44.290 --> 00:51:46.550 So, in that paper … - 85? 00:51:46.550 --> 00:51:48.860 - 85. - I’m sorry, 85. Yes. 00:51:48.860 --> 00:51:50.460 - Yeah. - It’s 5 degrees off vertical. 00:51:50.460 --> 00:51:55.460 And Catherine – and, in this paper, they did a 3D relocation. 00:51:55.460 --> 00:51:57.880 - Mm-hmm. - And the dip persisted. 00:51:57.880 --> 00:52:01.680 And we said, well, must be real. Except that, when you look at the 00:52:01.680 --> 00:52:07.580 first motion mechanisms, you see strong evidence of lateral refraction. 00:52:07.580 --> 00:52:12.700 And the point is that the 3D model does not explain the travel times. 00:52:12.700 --> 00:52:15.250 The model – the 3D approach just isn’t sufficient. And … 00:52:15.250 --> 00:52:17.579 - The 3D earthquake relocation. - Yes. 00:52:17.579 --> 00:52:21.099 - Yeah, okay. - And, if that’s the case, then you’re 00:52:21.099 --> 00:52:25.590 pushing unmodeled travel times into locations, and maybe the dip 00:52:25.590 --> 00:52:28.730 is an artifact. So I’m not saying what we’re 00:52:28.730 --> 00:52:34.150 seeing here is that, but I just wanted to caution you that 3D locations 00:52:34.150 --> 00:52:36.880 may not be any better than the 1D locations. 00:52:36.880 --> 00:52:40.100 - Then the 1D? Well, these are all 1Ds. 00:52:40.110 --> 00:52:44.480 I mean, they’re your – they used your model. 00:52:44.480 --> 00:52:47.720 So I was just speculating that maybe 3D would be better. 00:52:47.720 --> 00:52:49.090 Maybe it’d be worse. [laughs] 00:52:49.090 --> 00:52:53.100 - And my point is, though, that artificial dips can be introduced 00:52:53.100 --> 00:52:57.020 by laterally varying velocity structure. So be aware of that. 00:52:57.020 --> 00:52:58.020 - Yep. 00:52:59.180 --> 00:53:05.300 But when they did 3D at Parkfield, they did get those earthquakes onto the fault, 00:53:05.300 --> 00:53:08.820 as determined by drilling. So it sort of worked there. 00:53:08.820 --> 00:53:10.160 - [whispering] That guy's had his hand up for a long time. 00:53:10.160 --> 00:53:11.240 [inaudible] - [whispering] Of course. 00:53:13.320 --> 00:53:18.940 - Yeah, Gary? You hinted at part of my question. 00:53:18.940 --> 00:53:25.780 Given that your model would simplify things by putting Loma Prieta onto the 00:53:25.780 --> 00:53:35.640 San Andreas, and if it’s San Andreas activity that is making the Coast Ranges, 00:53:35.640 --> 00:53:38.790 you can’t make the Coast Ranges with this earthquake. 00:53:38.790 --> 00:53:42.680 Because, as you said, Loma Prieta – the highest thing around went down. 00:53:42.680 --> 00:53:45.320 - Yep. - So what else is going on? 00:53:46.640 --> 00:53:48.640 - That’s a good question. 00:53:50.640 --> 00:53:56.140 You’d have to ask some of the GPS-ers in the crowd. 00:53:57.060 --> 00:54:04.760 I don’t know if Jim Savage is still looking at that trend or not. 00:54:05.660 --> 00:54:14.460 - [inaudible] showed that there was a problem, geodetics seemed to indicate 00:54:14.460 --> 00:54:20.580 that the fault was offset from [inaudible]. - Oh, that’s right, yeah. Right. 00:54:21.860 --> 00:54:24.880 - I showed have warned him about that. 00:54:28.920 --> 00:54:32.840 - Hey. Gary, I’m going to take another shot on a different topic. 00:54:32.859 --> 00:54:37.330 You listed a litany of geophysical fields that you’ve looked at 00:54:37.330 --> 00:54:41.020 to come up with this model. But the one you actually left out is Vs. 00:54:41.020 --> 00:54:43.080 - The which? - Shear wave velocity. 00:54:43.080 --> 00:54:46.300 - Oh, shear – yeah. Yeah. - So, actually – so, in Donna’s and 00:54:46.300 --> 00:54:50.650 my velocity model, the thing we actually pinned some evidence – and there’s, 00:54:50.650 --> 00:54:54.550 like, very sparse – one or two earthquakes right under the San Andreas 00:54:54.550 --> 00:54:56.830 Fault, was we did actually see a Vp/Vs change. 00:54:56.830 --> 00:55:02.780 We didn’t directly invert for Vs for reasons of resolution, but in Vp/Vs, 00:55:02.780 --> 00:55:06.089 there is a hint that there is actually a structure. 00:55:06.089 --> 00:55:12.440 And a lot of people didn’t interpret a single high-velocity P wave body 00:55:12.440 --> 00:55:17.530 going from the Loma Prieta main shock across the San Andreas 00:55:17.530 --> 00:55:20.200 and over to some of the other faults. 00:55:20.200 --> 00:55:26.310 But that body doesn’t seem to have constant Vp/Vs ratio in it. 00:55:26.310 --> 00:55:28.070 - Yeah. - And I think there’s always a thing that, 00:55:28.070 --> 00:55:34.610 you know, each geophysical field sort of gives us one view of contrasts, right? 00:55:34.610 --> 00:55:38.300 Where the density contrasts are. Where the magnetic material is. 00:55:38.300 --> 00:55:42.600 And so, you know, there’s still other things out there in terms of 00:55:42.600 --> 00:55:45.740 possibly a vertical San Andreas. The other thing is that also I think 00:55:45.740 --> 00:55:49.950 I agree with Dave’s comment about earthquake locations. 00:55:49.950 --> 00:55:53.690 Station corrections in 1D velocity models are extreme powerful at 00:55:53.690 --> 00:55:58.020 correcting a lot of what goes on in the 3D structure. 00:55:59.560 --> 00:56:02.800 Largely, what we wind up doing, actually, in the 3D velocity models 00:56:02.810 --> 00:56:06.310 is pushing the station corrections into structure. 00:56:06.310 --> 00:56:10.140 And we actually – in, I think, the Loma Prieta paper, we had an objection that we 00:56:10.140 --> 00:56:15.360 couldn’t justify our 3D velocity model based on an F-test of residual reduction. 00:56:15.360 --> 00:56:18.200 And so we said, well, we think there’s an a priori basis for believing in 00:56:18.200 --> 00:56:23.440 3D changes in the Earth, that the Earth is not – you know, has 3D structure. 00:56:23.440 --> 00:56:25.610 But I – you know, I think they get a little better. 00:56:25.610 --> 00:56:28.250 I wouldn’t say they’re not better at all. And they definitely can do better 00:56:28.250 --> 00:56:31.440 at getting the focal mechanisms in terms of getting the 00:56:31.440 --> 00:56:35.020 lateral refractions with a good enough model. 00:56:35.020 --> 00:56:39.190 But, yeah, they’re not a magic thing that just changes the – even with the 3D, 00:56:39.190 --> 00:56:44.060 or even with the 1D, there’s still some slop and tradeoffs in the locations. 00:56:44.060 --> 00:56:47.930 - Yeah. Great point. Okay, so one of the bits of evidence 00:56:47.930 --> 00:56:53.800 for a vertical San Andreas Fault was this cluster of earthquakes. 00:56:53.800 --> 00:56:57.540 Unfortunately, it occurs north of the trace of the San Andreas – 00:56:57.540 --> 00:57:02.869 a couple kilometers. Actually, more or less under the Sargent Fault. 00:57:02.869 --> 00:57:05.510 And if you look at mechanisms in there, they’re up and down. 00:57:05.510 --> 00:57:08.839 They’re not strike-slip. That was the one piece of evidence 00:57:08.839 --> 00:57:16.530 that – for this allegedly vertical San Andreas under Loma Prieta. 00:57:16.530 --> 00:57:20.500 And, like you say, laterally, it changes. 00:57:20.500 --> 00:57:25.820 And you don’t see this on – at least too many adjacent profiles. 00:57:25.820 --> 00:57:32.720 So it’s hard to use it as evidence for a – for a vertical fault. Do you … 00:57:32.720 --> 00:57:37.690 - That’s not what we used, but yeah. - Okay. 00:57:38.440 --> 00:57:41.180 - Gary. Over here. - Yeah. 00:57:41.190 --> 00:57:46.160 - I think there was a report of the Santa Cruz Harbor entrance marker going – 00:57:46.160 --> 00:57:49.310 you know, rising a meter after Loma Prieta. 00:57:49.310 --> 00:57:52.500 Are you aware of that one? Or does that fit in this? 00:57:52.500 --> 00:57:55.339 - Not that particular fact. - Pardon? 00:57:55.340 --> 00:57:59.820 - I’m not aware of that particular fact. - I think it came from Latitude 38. 00:57:59.820 --> 00:58:01.720 [laughter] 00:58:01.720 --> 00:58:07.600 The other – you’ve done a similar display for around the Parkfield – 00:58:07.600 --> 00:58:13.380 or, you know, SAFOD site? I mean, this looks kind of like just 00:58:13.380 --> 00:58:16.240 a giant scale of a flower structure, isn’t it? 00:58:16.240 --> 00:58:22.020 - Well, yeah, that’s what I’m calling it – a flower-like structure. 00:58:22.680 --> 00:58:29.780 And – but – so you’re talking about surface deformation, and I don’t – we 00:58:29.780 --> 00:58:35.300 really haven’t done a whole lot – I mean, I haven’t really looked much into that. 00:58:36.960 --> 00:58:44.500 Rowland Tabor – Rowland – Tabor – Roland Burgmann did have a model 00:58:44.510 --> 00:58:52.109 based on surface deformation right there. Okay. 00:58:52.109 --> 00:58:57.580 The ones on this side, these two are Dietz and Ellsworth, 00:58:57.580 --> 00:59:01.770 where they interpreted two possible configurations for the San Andreas. 00:59:01.770 --> 00:59:06.450 Here goes the San Andreas Fault up here. It cuts off. Sorry. 00:59:06.450 --> 00:59:09.440 Here goes the Loma Prieta rupture here, 00:59:09.440 --> 00:59:12.640 and it cuts off a vertical San Andreas Fault. 00:59:12.640 --> 00:59:22.650 Another interpretation is that it is the – it is the San Andreas Fault with a change 00:59:22.650 --> 00:59:27.800 in depth like we’re proposing here. And here’s that cluster of earthquakes 00:59:27.800 --> 00:59:33.320 with an up-and-down motion that may have been evidence of a 00:59:33.320 --> 00:59:37.180 deeper vertical San Andreas, but unfortunately, it’s displaced 00:59:37.180 --> 00:59:44.570 by about 2 kilometers to the north. Roland Burgmann saw these 00:59:44.570 --> 00:59:49.840 displacements in his model of the geodetics. 00:59:49.840 --> 00:59:52.660 He saw displacement on the Berrocal Fault 00:59:52.660 --> 00:59:55.100 and then displacement on the rupture. 00:59:55.100 --> 00:59:59.600 But he never connected the two because that’s what he saw. 00:59:59.600 --> 01:00:08.380 So, yeah, I mean, deformation is an important element in all this. 01:00:08.980 --> 01:00:10.260 Yep? 01:00:10.340 --> 01:00:14.940 - I was just going to say that Vicky Langenheim has a paper 01:00:14.940 --> 01:00:25.070 in Geology, which she’s argued that the seismicity under the Berrocal Fault 01:00:25.070 --> 01:00:31.010 in off – in the – beneath the Santa Clara Valley is actually triggered slip. 01:00:31.010 --> 01:00:36.550 Certainly, there’s never really been documented surface rupture 01:00:36.550 --> 01:00:43.810 associated with that seismicity out there, so it’s considered to be something 01:00:43.810 --> 01:00:49.980 that was triggered by the shaking associated with the main rupture. 01:00:54.420 --> 01:00:57.080 - All right. Well, thank you, again, Gary, for a 01:00:57.089 --> 01:00:59.350 really interesting and thought-provoking talk. 01:00:59.350 --> 01:01:02.800 And thank you for the audience for the really interesting discussion. 01:01:02.800 --> 01:01:05.280 Join me in giving Gary another round of applause. 01:01:05.280 --> 01:01:11.440 [Applause] 01:01:11.440 --> 01:01:15.380 [inaudible background conversations] 01:01:15.380 --> 01:01:17.360 Yep. People definitely remembered Loma Prieta. 01:01:17.400 --> 01:01:18.620 - They certainly did. 01:01:18.800 --> 01:01:24.860 [Silence]