WEBVTT Kind: captions Language: en-US 00:00:03.040 --> 00:00:08.056 Good afternoon, and thank you for inviting me to present in this session. 00:00:08.080 --> 00:00:12.320 My presentation today is a bit fragmented to the extent that I will 00:00:12.320 --> 00:00:15.120 give a brief overview of where we stand with our update of 00:00:15.120 --> 00:00:18.240 the real-time double-difference system – a project that has been 00:00:18.240 --> 00:00:22.320 funded by NEHRP. And then, in the second half, 00:00:22.320 --> 00:00:28.183 I will focus on the recent activity on the Maacama Fault near Willits. 00:00:29.040 --> 00:00:34.240 We have recently completed a new update of the double-difference base 00:00:34.240 --> 00:00:40.536 catalog for northern California for events between 1984 and 2019. 00:00:40.560 --> 00:00:45.680 The catalog includes precise locations for close to 900,000 earthquakes. 00:00:45.680 --> 00:00:50.671 So these locations are based on analysis of 4 billion phase delay times 00:00:50.695 --> 00:00:55.336 formed from arrival time picks and about 5 billion delay times 00:00:55.360 --> 00:00:59.258 measured from cross-correlation of P and S waves. 00:01:00.400 --> 00:01:04.720 An initial look at the results indicate similar or better resolution 00:01:04.720 --> 00:01:11.440 compared to the previous catalogs. The figure on the right side show – 00:01:11.440 --> 00:01:14.880 for the region where the San Andreas Fault and Calaveras Fault intersect, 00:01:14.880 --> 00:01:18.560 a comparison between the double- difference location and the network 00:01:18.560 --> 00:01:23.816 locations for both the new, at the top, and the old catalog at the bottom. 00:01:23.840 --> 00:01:27.560 Essentially, what we see is an increase in event density 00:01:27.560 --> 00:01:32.936 as the earthquakes since 2011 reactivate pre-existing structures. 00:01:32.960 --> 00:01:37.440 The complete catalog is shown in the middle as blue dots – 00:01:37.440 --> 00:01:44.216 events between 1984 and 2019 – and then, in red, as the real-time 00:01:44.240 --> 00:01:48.856 locations that are continuously being computed. 00:01:48.880 --> 00:01:53.760 You can see the active areas in the Geysers and [inaudible] popping up, 00:01:53.760 --> 00:01:57.840 and then the red arrow points to the Maacama Fault that I will 00:01:57.840 --> 00:02:01.672 show in a little bit. 00:02:03.360 --> 00:02:07.440 Just a reminder that the high-precision double-difference locations are being 00:02:07.440 --> 00:02:12.456 computed and made available in near real time for over a decade now. 00:02:12.480 --> 00:02:16.960 So the way the real-time solutions are computed is a little bit different 00:02:16.960 --> 00:02:21.176 from the way we compute base catalogs. 00:02:21.200 --> 00:02:24.480 So the new base catalog offers the opportunity to evaluate the 00:02:24.480 --> 00:02:30.000 performance of the real-time system. We looked at 250,000 real-time 00:02:30.000 --> 00:02:33.520 locations and compared them to the corresponding locations the new – 00:02:33.520 --> 00:02:36.480 in the new base catalog. You can see on the figure 00:02:36.480 --> 00:02:41.120 on the left that the differences are typically less than 100 meter, 00:02:41.120 --> 00:02:44.376 especially along well-monitored faults. 00:02:44.400 --> 00:02:47.920 Once we update the real-time system with the new base catalog, 00:02:47.920 --> 00:02:51.096 then the differences will be even smaller. 00:02:51.120 --> 00:02:55.920 Differences between the real-time locations and the NCSN network 00:02:55.920 --> 00:03:00.424 locations are quite larger in most regions. 00:03:03.840 --> 00:03:08.240 Moving to the second part of the talk, I am showing on the left a figure 00:03:08.240 --> 00:03:11.360 with the new base catalog for the Maacama Fault in blue 00:03:11.360 --> 00:03:14.720 and the real-time double-difference locations since the beginning 00:03:14.720 --> 00:03:20.859 of 2020, which includes the Willits sequence in red. 00:03:22.400 --> 00:03:26.400 Since it – since I will be using real-time double-difference locations 00:03:26.400 --> 00:03:29.520 in all upcoming slides, I just wanted to make sure that we are looking at 00:03:29.520 --> 00:03:38.000 a high-resolution picture of this area. So I am showing on the right side 00:03:38.000 --> 00:03:41.920 of the slide the difference between the real-time double-difference 00:03:41.920 --> 00:03:46.456 locations and the new base catalog locations just like in the previous slide. 00:03:46.480 --> 00:03:50.080 So, again, these differences are – differences are typically smaller 00:03:50.080 --> 00:03:55.336 than 100 meter in the area we are interested in. 00:03:57.120 --> 00:04:04.160 So those – the next couple of slides show seismicity map and – map view 00:04:04.160 --> 00:04:12.088 and cross-sections color-coded by time for the sequence near Willits. 00:04:13.040 --> 00:04:17.760 So you can see that the first sequence actually started in 2019 – 00:04:17.760 --> 00:04:24.296 in April 2019, shown in blue. 00:04:24.320 --> 00:04:33.123 And the second sequence started in August of 2020, shown mostly in red. 00:04:34.880 --> 00:04:39.920 This is obviously a very interesting sequence, which I have to admit, 00:04:39.920 --> 00:04:43.945 I just started looking at in preparation for this talk. 00:04:45.200 --> 00:04:50.320 The first sequence started with a magnitude 4.3 at about 8 kilometers’ 00:04:50.320 --> 00:04:55.336 depth and then went straight up through the crust to reach the surface. 00:04:55.360 --> 00:04:59.360 And the second sequence started with a magnitude 4 at about 00:04:59.360 --> 00:05:04.320 7 kilometers’ depth and, similar to the 2019 sequence, 00:05:04.320 --> 00:05:08.720 went also straight to the surface. The 2020 sequence included the 00:05:08.720 --> 00:05:13.736 second magnitude 4 event at about 8 kilometers’ depth. 00:05:13.760 --> 00:05:18.240 The seismicity in the time period between the two sequences – 00:05:18.240 --> 00:05:23.920 mostly that green color – mostly occurred in the 00:05:23.920 --> 00:05:29.150 6- to 8-kilometer depth range with no significant large events. 00:05:32.480 --> 00:05:38.000 This slide is the same as the previous slide except that I show a close-up of 00:05:38.000 --> 00:05:44.804 the Willits sequence, again with the 2019 in blue and the 2020 in red. 00:05:45.920 --> 00:05:50.560 In a fault – in a fault-perpendicular cross-section, as shown in the 00:05:50.560 --> 00:05:54.400 middle panel, we see that the cluster does not image a narrow fault 00:05:54.400 --> 00:05:59.523 but rather a near-vertical tube, or pipe, of seismicity. 00:06:01.920 --> 00:06:08.000 This slide shows the seismicity until the start of the second sequence. 00:06:08.000 --> 00:06:12.160 So the second sequence is not imaged here. 00:06:12.160 --> 00:06:15.440 The interesting thing to notice is that the location of the second 00:06:15.440 --> 00:06:19.736 sequence was already active before the first sequence. 00:06:19.760 --> 00:06:23.680 Even included some larger events at about 5 kilometers’ depth. 00:06:23.680 --> 00:06:26.856 So those events are shown in gray. 00:06:26.880 --> 00:06:29.280 There also seems to be a seismogenic connection 00:06:29.280 --> 00:06:33.840 between the two sequences at the base of the brittle crust. 00:06:36.240 --> 00:06:42.616 This slide shows the temporal evolution of the second sequence in 2020. 00:06:42.640 --> 00:06:48.000 It started at the base of the – of the seismogenic zone 00:06:48.000 --> 00:06:52.080 just like the 2019 sequence. The 2020 sequence shows 00:06:52.080 --> 00:06:56.320 a pipe similar to the 2019 sequence that has more events with 00:06:56.320 --> 00:06:59.167 a somewhat more diffused distribution. 00:07:01.120 --> 00:07:04.160 Again, it is interesting to note that the 2020 sequence started 00:07:04.160 --> 00:07:08.960 close to the nucleation location of the 2019 sequence and then veered off 00:07:08.960 --> 00:07:11.680 to the northwest, suggesting that they are somehow triggered 00:07:11.680 --> 00:07:14.376 by a common driving force. 00:07:14.400 --> 00:07:18.960 The sequence is still ongoing. And I created a special website 00:07:18.960 --> 00:07:25.200 that tracks the activity near Willits in real time as part of 00:07:25.200 --> 00:07:30.160 the DDRT system operation. You can go to the DDRT website 00:07:30.160 --> 00:07:33.656 and click on the link under “earthquake of special interest,” 00:07:33.680 --> 00:07:39.997 and then you find the same figure that I showed you. 00:07:43.040 --> 00:07:48.720 Now those pipes – those pipe-like structures resemble very much 00:07:48.720 --> 00:07:52.800 pipe structures that we observed in subduction zones in Java, Sumatra, 00:07:52.800 --> 00:07:58.376 and on the Marianas, and which we all interpreted as being fluid-driven. 00:07:58.400 --> 00:08:09.096 So shown on the right lower panels are examples for Java and then Sumatra. 00:08:09.120 --> 00:08:15.360 So, if I make similar plots for the Willits sequences, and plot the timing 00:08:15.360 --> 00:08:20.160 of the events as a function of depth, as shown on the left side of the – 00:08:20.160 --> 00:08:24.320 of the slide, we observe bursts of seismicity that migrate upward 00:08:24.320 --> 00:08:29.200 at high speeds. I had a very quick look at the ascent rates, and they are in the 00:08:29.200 --> 00:08:34.880 kilometers per hour range, so similar to ascent rates we observed for fluids 00:08:34.880 --> 00:08:39.016 that come off the subduction interface after large thrust events. 00:08:39.040 --> 00:08:43.840 So that’s definitely something that we should have a closer look at. 00:08:43.840 --> 00:08:48.160 But these initial results suggest that somehow fluids got released 00:08:48.160 --> 00:08:53.840 by breaking some kind of seal by a larger event. 00:08:57.680 --> 00:09:03.280 So, in summary, we have – we’re close to having 00:09:03.280 --> 00:09:07.576 a new double-difference base catalog for California. 00:09:07.600 --> 00:09:14.291 We need to do some tests and evaluate the location robustness. 00:09:15.440 --> 00:09:18.720 Then preliminary evaluation of the real-time double-difference locations 00:09:18.720 --> 00:09:22.320 that we produced over the last 10 or so years indicate resolution 00:09:22.320 --> 00:09:25.576 better than 100 meter in well-monitored regions. 00:09:25.600 --> 00:09:32.696 So moving forward, that system is only going to become better 00:09:32.720 --> 00:09:35.736 in terms of location precision. 00:09:35.760 --> 00:09:38.480 The Maacama Fault near Willits experienced two earthquake 00:09:38.480 --> 00:09:44.776 sequences in 2019 and 2020. The 2021 is still ongoing. 00:09:44.800 --> 00:09:47.360 I would say the double sequence is still ongoing. 00:09:47.360 --> 00:09:51.816 Preliminary spatiotemporal analysis suggests that they are fluid-driven, 00:09:51.840 --> 00:09:57.028 similar to observations we made in subduction zones. 00:09:58.320 --> 00:10:01.280 There are three repeating event sequences that occurred near 00:10:01.280 --> 00:10:05.920 the recent Willits activity. That’s in the catalog of 00:10:05.920 --> 00:10:09.920 repeating earthquakes we have. But it is not clear yet whether 00:10:09.920 --> 00:10:13.326 any of the new events are repeating events. 00:10:14.240 --> 00:10:16.394 So we will check on that. 00:10:18.400 --> 00:10:28.960 Ongoing work in terms of the real-time operation for the current project is we 00:10:28.960 --> 00:10:34.216 will be able to release a new double- difference base catalog sometime soon. 00:10:34.240 --> 00:10:39.760 We will update and improve the RT-DD system for northern California 00:10:39.760 --> 00:10:42.136 over the next couple of months. 00:10:42.160 --> 00:10:46.536 So they will be at the same location online. 00:10:46.560 --> 00:10:50.560 And then we will be working on implementing the repeating 00:10:50.560 --> 00:10:55.040 earthquake database that we have produced over the several – 00:10:55.040 --> 00:11:01.336 over the last couple of years and implement that into the 00:11:01.360 --> 00:11:06.776 RT-DD system for real-time tracking of recurring events. 00:11:06.800 --> 00:11:12.000 Thank you, and I will see you soon.