WEBVTT Kind: captions Language: en 00:00:00.179 --> 00:00:07.819 … pleasure to introduce Professor Shri Krishna Singh from UNAM, Mexico. 00:00:07.819 --> 00:00:13.670 I met Shri in 1974 in the field. We were measuring seismic data 00:00:13.670 --> 00:00:20.610 in central Mexico and getting the geometry of the slab beneath the – 00:00:20.610 --> 00:00:25.780 from Acapulco inland. And the experiment was successful, 00:00:25.780 --> 00:00:35.360 and we’ve remained friends for the succeeding 44 years – a long time. 00:00:35.360 --> 00:00:41.440 Shri grew up in northern India, not far from Varanasi. 00:00:41.450 --> 00:00:45.830 Varanasi is the – one of the – the holy city where you often see people bathing. 00:00:45.830 --> 00:00:53.480 He went to the Indian School of Mines, which is kind of like the MIT of India. 00:00:53.480 --> 00:00:58.410 And from the School of Mines, he took a difficult exam and 00:00:58.410 --> 00:01:00.630 did rather well on it and got a full scholarship to 00:01:00.630 --> 00:01:05.019 go to Columbia University in New York City, where he did his Ph.D. 00:01:05.019 --> 00:01:12.400 His classmate was Bob Jachens. Jachens is someone in the – 00:01:12.400 --> 00:01:15.520 you know, GUMP – the geophysical unit in Menlo Park. 00:01:15.520 --> 00:01:18.150 So there’s some connections even to – 00:01:18.150 --> 00:01:21.880 many connections to people here in Menlo Park. 00:01:21.880 --> 00:01:26.580 In 1971, he took an assistant professorship in UNAM and 00:01:26.580 --> 00:01:29.759 rose rapidly to become a full professor. 00:01:29.759 --> 00:01:34.770 And later in his career, he even became what’s called university professor. 00:01:34.770 --> 00:01:40.080 That is an honor which is given to people with outstanding achievement in science. 00:01:40.080 --> 00:01:43.700 He’s now continuing his work. He tells me he goes to work every day. 00:01:43.700 --> 00:01:47.090 He’s what’s called an emeritus professor there. 00:01:47.090 --> 00:01:49.810 And this earthquake that he’s – we’re going to hear about 00:01:49.810 --> 00:01:54.030 is really interesting because it’s not a mega – you know, 00:01:54.030 --> 00:01:57.920 there’s a subduction zone in Mexico, but this is not a megathrust event. 00:01:57.920 --> 00:02:03.899 As the title says, it’s a intraslab event. But I think I better let you tell the 00:02:03.899 --> 00:02:07.119 whole story rather than getting into what’s an intraslab event. [laughs] 00:02:07.119 --> 00:02:08.119 You’re the speaker. 00:02:08.120 --> 00:02:11.980 So thank you for coming, Shri, and we’re delighted to have you. 00:02:13.900 --> 00:02:29.040 [Silence] 00:02:29.040 --> 00:02:31.360 - Is it on? I think so. 00:02:31.379 --> 00:02:35.900 Thanks, Walter. I’m not sure the experiment was all that successful. 00:02:35.900 --> 00:02:40.909 I remember we tried years to look at the data. The data was terrible. 00:02:40.909 --> 00:02:46.470 This is reflection provided from Pinotepa Nacional to a lake in the interior 00:02:46.470 --> 00:02:52.779 of Mexico. And I think we read – we read the phases a hundred times. 00:02:52.779 --> 00:02:55.029 We couldn’t really figure it out. But anyway, good thing that 00:02:55.029 --> 00:02:58.019 he thinks it was a great success. [laughter] 00:02:58.020 --> 00:03:02.200 - Shri, Walter’s experiments are all a success. [laughter] 00:03:02.200 --> 00:03:03.660 - Thanks, Jack. 00:03:03.660 --> 00:03:09.859 - Okay. Well, I’ll talk about this intraslab earthquake. 00:03:09.859 --> 00:03:16.439 As you know, in September of last year, we had two extraordinary intraslab 00:03:16.439 --> 00:03:20.959 earthquakes – the earthquakes that occur in the subducted plate. 00:03:20.959 --> 00:03:25.059 And one of them, of course – the one on 8th of September 00:03:25.059 --> 00:03:29.790 was largest to ever occur in Mexico, include thrust earthquakes. 00:03:29.790 --> 00:03:35.200 The next to this earthquake would be the Jalisco earthquake of 1932. 00:03:35.200 --> 00:03:36.900 So this is largest earthquake. 00:03:36.900 --> 00:03:41.620 It was something like 700 kilometers from Mexico City – very far away. 00:03:41.620 --> 00:03:45.700 Didn’t really cause any damage in the city, but was felt. 00:03:45.700 --> 00:03:48.939 It was early in the morning. An alert sounded. 00:03:48.940 --> 00:03:52.320 And there was – in fact, talking about alerts, the day before, there was 00:03:52.320 --> 00:03:56.920 a false alert. An alert had sounded. Nothing had happened. 00:03:56.930 --> 00:04:00.189 When this sounded, people didn’t really go out. I didn’t go out, anyway. 00:04:00.189 --> 00:04:02.959 Didn’t even feel the earthquake. Anyway. 00:04:02.959 --> 00:04:09.309 So within, what, about 10 days or so later, there was this earthquake, 00:04:09.309 --> 00:04:11.680 which was much closer to Mexico City. 00:04:11.680 --> 00:04:18.379 In fact, you’ll see that this is the closest intraslab earthquake – small or large – 00:04:18.380 --> 00:04:22.660 that has ever been recorded in Mexico City. 00:04:22.660 --> 00:04:25.320 And it was certainly the most damaging 00:04:25.330 --> 00:04:28.800 intraslab earthquake in the history of Mexico. 00:04:28.800 --> 00:04:36.370 So I’ll be talking about the smaller, closer to Mexico City, 00:04:36.370 --> 00:04:39.920 earthquake of 19th of September. 00:04:40.740 --> 00:04:47.980 As was mentioned, this was only 114 kilometers from Mexico City. 00:04:47.990 --> 00:04:51.449 And the depth was, like, about 57 kilometers. 00:04:51.449 --> 00:04:55.629 Okay, well, actually, it turns out that, you know, we had 40 days 00:04:55.629 --> 00:05:03.389 of disasters in Mexico. We had two large earthquakes. 00:05:03.389 --> 00:05:05.830 We had three hurricanes passing by. 00:05:05.830 --> 00:05:13.220 We had two tropical storm and thousands of aftershocks. 00:05:13.220 --> 00:05:19.000 So it was a very agitated period. And, of course, the rescue effort 00:05:19.000 --> 00:05:22.659 in Mexico City were also hampered, not only in Mexico City, 00:05:22.659 --> 00:05:26.370 but also in Oaxaca and Chiapas for the other earthquake because 00:05:26.370 --> 00:05:32.440 we have rains and all that problem. So it was a test, as this title says – 00:05:32.440 --> 00:05:36.760 México puesto a prueba – was put to test. 00:05:36.760 --> 00:05:42.560 Okay, the effect of the earthquake in Mexico City. 00:05:42.560 --> 00:05:47.870 Something like 230 people died. 44 buildings collapsed. 00:05:47.870 --> 00:05:51.169 600 of these were severely damaged. 00:05:51.169 --> 00:05:58.320 And apparently 15,000 additional buildings were – suffered some damage. 00:05:59.080 --> 00:06:02.800 And it was the most – second-most destructive earthquake in the history 00:06:02.800 --> 00:06:08.100 of Mexico. The first one was 1985 earthquake, as we know. 00:06:09.860 --> 00:06:13.140 Next only to that, this earthquake. 00:06:14.020 --> 00:06:18.780 So, actually, you know, the city is still recuperating from this earthquake. 00:06:18.780 --> 00:06:21.729 Doesn’t really – not everything has come back to normal. 00:06:21.729 --> 00:06:27.740 There are protests and problems and problems of rebuilding and so on. 00:06:28.850 --> 00:06:31.280 The damage was not only limited to Mexico City. 00:06:31.280 --> 00:06:38.340 This is map of damage, which a colleague of mine has compiled. 00:06:38.340 --> 00:06:41.879 This is Mexico City, which looks like heart. 00:06:41.879 --> 00:06:44.610 And see the damage concentration in Mexico City. 00:06:44.610 --> 00:06:50.400 But surrounding the epicentral area, especially to the northwest, 00:06:50.400 --> 00:06:56.039 there’s a huge – large distribution of damaged buildings. 00:06:56.039 --> 00:07:05.389 So it was fairly disastrous for – here’s a list of number of structures 00:07:05.389 --> 00:07:10.900 which are damaged – number of them. So Morelos, this date, had some – 00:07:10.900 --> 00:07:14.940 more than Mexico City. And it was also in some other states. 00:07:14.940 --> 00:07:19.759 Ah, this is our ShakeMap. We have – we prepare – we don’t 00:07:19.759 --> 00:07:23.439 call it ShakeMap because that’s, I think, copyright of USGS. 00:07:23.440 --> 00:07:27.560 But these are based on our own attenuation relations and so on. 00:07:27.560 --> 00:07:36.569 Maybe I’ll show you the true ground motion contours later. 00:07:36.569 --> 00:07:43.650 Okay. So immediately, you know, questions were asked – not only 00:07:43.650 --> 00:07:49.060 asked by laymen, but also scientists themselves – researchers. 00:07:49.060 --> 00:07:51.979 Well, was the earthquake unexpected in some sense? 00:07:51.979 --> 00:07:57.800 Meaning, was the magnitude expected? Was it too large? 00:07:57.800 --> 00:08:01.560 Was the distance to Mexico City unusual? 00:08:02.320 --> 00:08:06.120 The other set of questions involved – were the ground motions – why this 00:08:06.120 --> 00:08:11.860 damage? Because it was sort of – we were taken all by somewhat surprise. 00:08:11.860 --> 00:08:15.400 And I think it is vital to answer these questions. 00:08:15.400 --> 00:08:18.920 Were the ground motions especially anomalous? 00:08:19.540 --> 00:08:24.960 Was the damage pattern in Mexico City, which immediately became clear that 00:08:24.960 --> 00:08:28.840 it was different from 1985 earthquake – was it expected? 00:08:29.640 --> 00:08:32.640 Was it – something else was happening? 00:08:33.600 --> 00:08:38.360 Did the ground motions exceed the building code specifications? 00:08:38.360 --> 00:08:43.260 Building code was revised after the 1985 earthquakes, so it’s useful 00:08:43.260 --> 00:08:48.390 to know whether that code is doing all right or not. 00:08:48.390 --> 00:08:52.350 Finally, what is the exceedance rate of such ground motions in Mexico City? 00:08:52.350 --> 00:09:01.180 So we’d like to go through these – attempt to answer these questions 00:09:01.180 --> 00:09:11.339 that were – that actually are vital to the reconstruction of the city and 00:09:11.339 --> 00:09:16.089 understanding the seismic hazard in the city for future earthquakes. 00:09:16.089 --> 00:09:25.310 So, before I go through each of these issues, first a very brief 00:09:25.310 --> 00:09:31.519 outline of things which are basic to answer these questions. 00:09:31.519 --> 00:09:35.660 One of them – I should first point out that, you know, 00:09:35.660 --> 00:09:42.440 this is not the only historical – or only intraslab earthquake which has 00:09:42.440 --> 00:09:48.010 caused damaged to Mexican altiplano. There are several earthquakes – large 00:09:48.010 --> 00:09:51.780 earthquakes which are damaging, beginning in 1858, that 00:09:51.780 --> 00:09:55.660 we have some description of. The magnitude is not known. 00:09:55.660 --> 00:10:02.340 There was this earthquake in 1931 which caused severe damage to Oaxaca. 00:10:02.340 --> 00:10:07.340 Then there was an earthquake In ’64 – 7.3. ’73 and so on. 00:10:07.340 --> 00:10:15.700 So, as you can see, these are earthquakes not very far from Mexico City. 00:10:15.700 --> 00:10:20.800 If we go to Chiapas and so on, the list would increase, of course. 00:10:23.990 --> 00:10:29.660 And it’s interesting. The one in 1931, which was magnitude 7.8, 00:10:29.670 --> 00:10:32.740 and I think this was the first non-faulting earthquake 00:10:32.740 --> 00:10:41.120 that we studied in some detail, this destroyed the city of Oaxaca. 00:10:41.120 --> 00:10:49.260 And there’s an excellent documentary by Eisenstein – famous movie maker 00:10:49.260 --> 00:10:54.720 from Russia, who was in Mexico making a film called ¡Que Viva México! 00:10:54.720 --> 00:10:59.000 The film was never finished, but the documentary is beautiful. 00:10:59.000 --> 00:11:03.920 Ten minutes long or so if you wish to see this – very enlightening. 00:11:03.920 --> 00:11:11.320 In fact, it’s really – as you can expect from such a distinguished director, 00:11:11.340 --> 00:11:15.660 the movie is wonderful. So you might want to see it. 00:11:15.660 --> 00:11:19.240 Ah, there’s some – so there was a paper that was 00:11:19.240 --> 00:11:22.000 written on this earthquake – the non-faulting earthquake. 00:11:22.000 --> 00:11:26.300 Because we thought before that it was actually usual thrust earthquake. 00:11:26.300 --> 00:11:32.460 And this was published just before the Michoacán earthquake. 00:11:32.460 --> 00:11:36.560 And we were accused by the press that we knew that an earthquake 00:11:36.560 --> 00:11:40.840 was going to occur in this gap, but we didn’t advise the city and so on. 00:11:40.840 --> 00:11:46.250 So we were traitors and so on. [laughs] 00:11:46.250 --> 00:11:50.720 Anyway. That section – historical. 00:11:50.720 --> 00:11:57.440 Coming back to this earthquake, as I mentioned, the issues that 00:11:57.440 --> 00:12:03.050 became clear that need answer, for them to be answered, 00:12:03.050 --> 00:12:05.880 some things – I must mention some things first. 00:12:05.880 --> 00:12:12.630 One is that we have an accelerograph in UNAM working since 1964. 00:12:12.630 --> 00:12:15.630 So 54 years of recordings of both 00:12:15.630 --> 00:12:19.200 intraplate as well as interplate earthquakes. 00:12:19.200 --> 00:12:22.720 And that’s a mine of information that we can compare – we can see what goes 00:12:22.730 --> 00:12:27.589 on – what sort of wavefield are involved in each of these kinds of earthquakes. 00:12:27.589 --> 00:12:32.310 So that’s a very vital source of information. 00:12:32.310 --> 00:12:37.230 UNAM is here. This is a map of Mexico City. 00:12:37.230 --> 00:12:45.399 And you can see the periods – natural or dominant period of the soil. 00:12:45.399 --> 00:12:49.281 These are, like, between 1 and 2 seconds. 00:12:49.281 --> 00:12:55.480 And you can see some part of the lake, the dominant period reaches 5 seconds. 00:12:56.280 --> 00:13:01.240 And this is, of course, very important as we go further in the talk. 00:13:01.250 --> 00:13:04.790 So this – as I mentioned, this is very precious information. 00:13:04.790 --> 00:13:11.540 In fact, if you look at this table – this is – this is a table of intraslab 00:13:11.540 --> 00:13:19.960 earthquakes prepared in 2013. These were the 10 top intraslab 00:13:19.960 --> 00:13:28.900 earthquakes recorded in UNAM and with decreasing PGA values. 00:13:28.900 --> 00:13:31.740 The new earthquake is here. 00:13:32.460 --> 00:13:39.180 It is – it had 57 centimeters per second squared – far above 00:13:39.190 --> 00:13:46.930 anything else that we record before. So this is now the first in the list. 00:13:46.930 --> 00:13:52.790 But even when you see all earthquakes, including interplate and intraplate 00:13:52.790 --> 00:13:59.050 earthquakes – and these are the 20, I think, largest – this earthquake now 00:13:59.050 --> 00:14:03.930 is 57, and it becomes, again, the top here. 00:14:03.930 --> 00:14:12.260 Twice the PGA at UNAM plus two times greater than 1985 earthquake. 00:14:13.600 --> 00:14:20.740 Well, but I must emphasize also that, although the PGA was two times 00:14:20.740 --> 00:14:26.980 smaller than the ’85, the damage was very much larger during ’85 – 00:14:26.980 --> 00:14:30.740 maybe 10,000 people lost their lives, and many buildings were collapsed. 00:14:30.740 --> 00:14:36.480 So just the PGA, of course, is not a very meaningful measure, as you would see. 00:14:37.300 --> 00:14:45.560 Okay, also this data tells that actually Mexico City – the exceedance rate – 00:14:45.579 --> 00:14:55.230 exceedance rates of PGA – of interplate and intraplate is about the same. 00:14:55.230 --> 00:14:58.970 And this was somewhat surprising to us too. 00:14:58.970 --> 00:15:05.720 The damage is not the same, but the exceedance rates of the PGA of 00:15:05.720 --> 00:15:11.020 intra- and interplate is about the same. So this is also useful information. 00:15:11.020 --> 00:15:13.800 I will come back to it later on. 00:15:13.819 --> 00:15:20.079 Okay, so we have – on one hand is, you know, beautiful – really not 00:15:20.080 --> 00:15:24.300 very beautiful – the older records are very poor, as you can understand. 00:15:24.300 --> 00:15:28.420 ‘64 we had this – you know, I don’t know which kind of – strange kind of 00:15:28.420 --> 00:15:35.120 accelerographs on films and so on. So the quality of the data has improved. 00:15:37.260 --> 00:15:43.400 So we have that set of records and there’s another vital piece of – 00:15:43.400 --> 00:15:45.640 thing that we should remember is the amplification 00:15:45.649 --> 00:15:50.379 of seismic waves in the Valley of Mexico. 00:15:50.380 --> 00:15:54.820 Which is actually the main cause for all the damage and so on. 00:15:56.800 --> 00:16:01.920 This was known that the valley amplifies seismic waves. 00:16:01.920 --> 00:16:06.480 There are some papers in 1957 – 00:16:06.480 --> 00:16:10.100 rough calculations by Emilio Rosenblueth and so on. 00:16:10.100 --> 00:16:20.040 But the earthquake of ’85, which was better recorded, made it very clear that – 00:16:20.040 --> 00:16:22.170 what was happening in the Valley of Mexico. 00:16:22.170 --> 00:16:26.930 All these stations reading from here – CUP – this is the UNAM station – 00:16:26.930 --> 00:16:31.100 all the way until here, they’re all in the Valley of Mexico very close 00:16:31.100 --> 00:16:34.310 to one another, yet you can see the ground motion difference. 00:16:34.310 --> 00:16:38.320 And you can see the amplification of the seismic waves coming in. 00:16:38.320 --> 00:16:43.510 So this was of course very clear. I mean, it was clear before. 00:16:43.510 --> 00:16:47.180 Now we had some quantitative information. 00:16:47.180 --> 00:16:54.370 But since then – well, anyway, also I think I should mention that, 00:16:54.370 --> 00:16:59.800 for geotechnical purposes, the valley is divided into three zones. 00:16:59.800 --> 00:17:04.150 One is the Hill Zone, which consists of lava flows and volcanic tuffs and so on. 00:17:04.150 --> 00:17:07.260 The velocity is, like, 750 meters per second. 00:17:07.260 --> 00:17:11.620 Then there’s the Lake-bed Zone, which is 30 to 80 meters thick, usually. 00:17:11.630 --> 00:17:14.680 Highly compressible, high water-content clay. 00:17:14.680 --> 00:17:19.220 The velocities are, I think, 50 to 100 meters per second. 00:17:19.220 --> 00:17:25.560 I think Usain Bolt runs faster than this. [laughs] Maybe. 00:17:25.560 --> 00:17:27.210 So extremely low velocity. 00:17:27.210 --> 00:17:32.340 You can see this contrasts – is the cause of great amplification. 00:17:32.340 --> 00:17:40.640 Then there’s the Transition Zone between the Hill and Lake-bed Zone. 00:17:41.560 --> 00:17:46.420 And you see that, this time, this area suffered huge damage. 00:17:46.420 --> 00:17:55.130 Okay, so apart from – well, this is UNAM station and 00:17:55.130 --> 00:17:57.490 a few stations that were operating in ’85. 00:17:57.490 --> 00:18:03.580 Now we have a network which is almost, like, 80 or 100 accelerographs in the city. 00:18:04.560 --> 00:18:09.160 These are – not all of them – all of the stations, but some stations. 00:18:09.160 --> 00:18:18.300 And, on top of these – this map, you see the dominant period of the sites. 00:18:18.300 --> 00:18:22.580 This is 1 second, then there’s 2-second period, and all that. 00:18:25.140 --> 00:18:32.860 So, as you can imagine, we have transfer functions at these sites 00:18:32.860 --> 00:18:39.810 with respect to reference hard site, which is the UNAM CU station. 00:18:39.810 --> 00:18:43.550 And this is a very valuable piece of information. 00:18:43.550 --> 00:18:48.320 Because if we have the transfer functions for all these stations, 00:18:48.320 --> 00:18:51.920 and since we have records since 1964, we can simulate what the 00:18:51.920 --> 00:18:58.280 ground motions were during all those earthquakes in the past. 00:18:58.280 --> 00:19:05.440 Not only that, this permits us to immediately compute – I hate to say 00:19:05.440 --> 00:19:12.380 ShakeMap because I’m [chuckles] maybe accused of violating copyrights, maybe. 00:19:12.380 --> 00:19:18.280 We can make our predictions of ground motion without any problem. 00:19:18.280 --> 00:19:23.320 This is what we do. We have the recording here. 00:19:23.320 --> 00:19:27.080 And once we have the recording here, we can – we compute Fourier 00:19:27.080 --> 00:19:33.330 acceleration spectra, multiplied by the transfer functions here. 00:19:33.330 --> 00:19:35.950 So we have spectra there, and we go through random 00:19:35.950 --> 00:19:39.590 vibration theory, immediately produce all kinds of maps. 00:19:39.590 --> 00:19:46.800 This is the expected ground motion at different periods, PGA, whatever. 00:19:46.800 --> 00:19:52.020 Okay. Fine. But we need to demonstrate that actually this 00:19:52.020 --> 00:19:56.670 transfer function is stable. That is actually – it doesn’t change. 00:19:56.670 --> 00:19:59.900 An intraplate earthquake, an interplate earthquake – you guys 00:19:59.900 --> 00:20:06.250 do same transfer functions, mostly. That’ll help, otherwise it’ll be a mess. 00:20:06.250 --> 00:20:14.130 And here is a station. This is recording since 1985. 00:20:14.130 --> 00:20:16.760 And this is the transfer function of this station 00:20:16.760 --> 00:20:22.240 with respect to our reference station, which is at UNAM campus. 00:20:22.240 --> 00:20:30.380 Interplate earthquakes, stacked spectral ratio. 00:20:30.380 --> 00:20:34.330 And that’s the smoothed spectral ratio of interplate earthquakes. 00:20:34.330 --> 00:20:37.180 These are intraslab earthquakes. 00:20:37.880 --> 00:20:40.000 They are – you can see they’re relatively the same. 00:20:40.000 --> 00:20:43.340 This is the smoothed version of this. 00:20:43.340 --> 00:20:51.250 Here I’m plotting the median – or mean – I don’t know, median value 00:20:51.250 --> 00:20:57.930 of the amplification and the observed spectral ratio during 00:20:57.930 --> 00:21:04.190 the earthquakes of 1985, 2017. You can see, more or less same. 00:21:04.190 --> 00:21:08.760 So you can – actually, it should work in the sense that, if you have – 00:21:08.760 --> 00:21:13.460 as I was saying, if you have the recording from UNAM, you can 00:21:13.460 --> 00:21:18.590 compute the spectra at all the other sites, and you can compute, then, 00:21:18.590 --> 00:21:22.160 ground motion without actually looking at the data itself 00:21:22.160 --> 00:21:25.530 with each of these individual stations [inaudible]. 00:21:25.530 --> 00:21:32.720 Which is – helps. We can get this thing within a minute or so. 00:21:34.110 --> 00:21:45.740 So here is the first look at the damage. And the modeled response spectra. 00:21:45.740 --> 00:21:50.140 This is 1985 earthquake, for which we had very little data, remember? 00:21:50.140 --> 00:21:56.960 We had few stations. So just been done through transfer functions. 00:21:56.970 --> 00:22:01.380 2017 – here also – what I’m showing you is – has been done 00:22:01.380 --> 00:22:05.160 from the transfer functions. These are not real data, 00:22:05.160 --> 00:22:11.540 but this is the expected – is this double-L? I don’t know. 00:22:11.540 --> 00:22:16.740 Anyway, so this is at 1 second. You can see what is being predicted here is, 00:22:16.740 --> 00:22:24.510 at 1 second, response spectra – Sa – is very intense during this new earthquake. 00:22:24.510 --> 00:22:27.020 Which is intraslab earthquake, which is what you’d expect. 00:22:27.020 --> 00:22:28.560 We’ll go through it again. 00:22:28.560 --> 00:22:35.290 But if you go to 2 seconds – 1985 earthquake is far more 00:22:35.290 --> 00:22:41.680 intense than 2017 earthquake, right? 00:22:43.760 --> 00:22:50.040 And the damage has been superimposed on these maps. 00:22:50.050 --> 00:22:54.650 The damage, you see – well, you see the contours – but I don’t know 00:22:54.650 --> 00:22:58.430 that you can see, but this damage – but I’ll come back to it later on. 00:22:58.430 --> 00:23:03.211 This damage is concentrated also in the – in the – within the contour 00:23:03.211 --> 00:23:07.840 of 2 and 3 seconds, while this one is 1- and 2-second contour. 00:23:07.840 --> 00:23:17.360 So, in fact, within 2 minutes, you can get immediately first cut on what to expect. 00:23:17.360 --> 00:23:19.730 And this says that, well, the damage should be different. 00:23:19.730 --> 00:23:25.470 In fact, this was the first observation of people who went to see the damage, 00:23:25.470 --> 00:23:31.400 that the damage patterns were extremely different between the two earthquakes. 00:23:31.400 --> 00:23:39.560 Okay, so now, with this introduction – and how much time did I waste? 00:23:39.560 --> 00:23:41.600 [laughs] 00:23:41.600 --> 00:23:44.100 Something like 30 minutes? Good. 00:23:44.100 --> 00:23:51.090 So we now go to – try to answer some of the questions that were raised. 00:23:51.090 --> 00:23:54.730 So the first thing, was the earthquake too close? 00:23:54.730 --> 00:23:59.440 Should we have expected an earthquake like this at that distance? 00:24:00.260 --> 00:24:05.120 And let’s try to see. The answer is actually fairly straightforward. 00:24:05.120 --> 00:24:08.720 Okay. So this is – these are focal mechanisms. 00:24:08.720 --> 00:24:13.120 You can see all these intraplate earthquakes, normal faulting, occurring. 00:24:13.120 --> 00:24:16.740 Mexico City is here. 00:24:18.220 --> 00:24:23.320 These contours are supposed to be the depth – representing 00:24:23.320 --> 00:24:28.160 the depth of the subducted – the top of the subducted plate. 00:24:28.160 --> 00:24:32.490 You see it remains flat. Here is 40 kilometers, and 60 is here. 00:24:32.490 --> 00:24:39.460 It’s essentially flat for a long distance, then it plunges below. 00:24:39.460 --> 00:24:43.410 Volcanoes are – two – couple of volcanoes are marked here – 00:24:43.410 --> 00:24:50.500 Toluca and Popocatépetl. You see it – okay. 00:24:50.500 --> 00:24:55.760 So there’s some relevance, the reason why I’m mentioning these volcanoes. 00:24:55.760 --> 00:25:04.340 Okay, so if we look at it, there are at least four earthquakes in recent years 00:25:04.340 --> 00:25:13.280 between a magnitude 5.6 to, say, a 6.5 within 148 to 194 kilometers. 00:25:13.280 --> 00:25:16.240 These are these events somewhere. 00:25:18.420 --> 00:25:23.180 So – but you had – you know, these are relatively moderate earthquakes. 00:25:23.180 --> 00:25:28.130 But there are three earthquakes also – magnitude 6.9 to about 7.3 – 00:25:28.130 --> 00:25:32.100 within a distance range of 184 to 225. 00:25:32.100 --> 00:25:41.460 So it doesn’t seem that the distance was too close. 00:25:41.460 --> 00:25:45.420 I mean, the truth is that it is actually the closest earthquake. 00:25:45.420 --> 00:25:48.620 Here is this number one. Number one because that’s 00:25:48.620 --> 00:25:56.080 the list of accelerations – the top event. It is the closest. 00:25:56.080 --> 00:26:01.180 It is – this one – the one is closest to Mexico City than any other earthquake. 00:26:02.420 --> 00:26:05.920 But not necessarily really surprisingly close. 00:26:05.920 --> 00:26:10.780 I mean, we – before the earthquake, I’d have said, well, the closest distance 00:26:10.780 --> 00:26:14.570 to Mexico City would be defined by this green line. 00:26:14.570 --> 00:26:16.360 Now we have to push it back somewhere. 00:26:16.360 --> 00:26:23.030 Now, of course, a related question is, can these be – what are the chances 00:26:23.030 --> 00:26:26.490 that there are other earthquakes that might get closer to Mexico City? 00:26:26.490 --> 00:26:32.600 And we come back to it later on and, as usual, that’s the real problem 00:26:32.600 --> 00:26:35.380 in these things, right? Nobody knows. 00:26:35.380 --> 00:26:41.260 Okay. In fact, we thought that magnitude 7 was a likely possibility, 00:26:41.260 --> 00:26:47.070 and we used, a few years earlier, this earthquake – recording of this 00:26:47.070 --> 00:26:50.980 earthquake to synthesize ground motion for a magnitude 7 earthquake in 00:26:50.980 --> 00:26:54.030 Mexico City to see what would it look like and 00:26:54.030 --> 00:26:58.620 what the consequences of such an earthquake might be. 00:26:59.740 --> 00:27:01.740 Oh, do I have it? 00:27:02.680 --> 00:27:04.080 Yeah, okay. 00:27:04.090 --> 00:27:08.790 So the answer of the first one, so well, Mw and R actually 00:27:08.790 --> 00:27:14.200 weren’t all that unexpected. Yeah. 00:27:14.200 --> 00:27:20.280 We didn’t really have one, but considering what information 00:27:20.280 --> 00:27:25.460 we had, it was not unlikely. Now we come to ground motions. 00:27:25.460 --> 00:27:32.600 Were they anomalous in – at the university? 00:27:32.600 --> 00:27:35.490 And that’s somewhat easier to answer at the university 00:27:35.490 --> 00:27:38.400 because we have this continuous recording. 00:27:38.400 --> 00:27:41.380 Other places, very few earthquakes have been recorded – 00:27:41.380 --> 00:27:45.060 well, the small ones have been recorded, so it’s difficult to answer. 00:27:45.060 --> 00:27:50.220 But since – as I mentioned, that since everything is related to the university 00:27:50.220 --> 00:27:57.100 record, you can infer from this – for other sites in the Valley of Mexico. 00:27:57.100 --> 00:28:00.900 Okay. So let’s try to do that. 00:28:02.820 --> 00:28:09.860 Okay. As I was mentioning that, you know, we had synthesized 00:28:09.860 --> 00:28:13.380 ground motion using this earthquake as an empirical Green's function. 00:28:13.380 --> 00:28:16.300 Let’s see the results from that old simulation. 00:28:16.300 --> 00:28:19.760 It was a very simple exercise. 00:28:19.770 --> 00:28:27.700 So the event was 5.8 at a distance of 150 kilometers. It occurred in 2000. 00:28:27.700 --> 00:28:31.130 And here is the empirical Green’s function, and these are – 00:28:31.130 --> 00:28:39.360 this is our predicted line for – as a function of magnitude – 6.5, 7, 7.5. 00:28:39.360 --> 00:28:43.910 This is the velocity and the displacement predicted 00:28:43.910 --> 00:28:47.020 from this simulation – very simple simulation. 00:28:47.020 --> 00:28:49.080 And the observed values are not very far. 00:28:49.080 --> 00:28:54.010 The observed value for this earthquake is here and here and here. 00:28:54.010 --> 00:28:59.220 Sorry about the quality of the figure. That’s the best I can produce these days. 00:29:00.640 --> 00:29:06.630 So it seems that, if looking at this, things were all right. 00:29:06.630 --> 00:29:11.670 I mean, it wasn’t very anomalous in terms of 00:29:11.670 --> 00:29:15.760 ground motions for a magnitude 7.1 at that distance. 00:29:15.760 --> 00:29:20.560 Here are some simulations. This is the Green’s function – 00:29:20.560 --> 00:29:30.540 6.5, 7, and, in this simulation – this is one sample – A-max – 00:29:30.540 --> 00:29:34.000 PGA looks much larger on the accelerogram. 00:29:34.000 --> 00:29:37.540 Here are the velocities and displacements. 00:29:37.550 --> 00:29:45.670 So not especially energetic according to this summation. 00:29:45.670 --> 00:29:53.970 But if you look at the spectra of many other earthquakes at UNAM, 00:29:53.970 --> 00:30:00.510 this is the latest earthquake of 2017. Goes far higher – above. 00:30:00.510 --> 00:30:06.470 These are magnitude 5.6 and other earthquakes. 00:30:06.470 --> 00:30:10.160 We noticed that there’s one which is really – all of the magnitude 00:30:10.160 --> 00:30:14.660 of this earthquake is only 5.9. It’s well above the other ones. 00:30:16.000 --> 00:30:21.020 And it remains above – when we analyze these spectra, the same distance at which 00:30:21.030 --> 00:30:27.320 this thing was recorded, by correcting for Q and geometrical dispersion. 00:30:27.320 --> 00:30:32.860 So this black line is here. This is larger earthquake – 6.5. 00:30:32.860 --> 00:30:37.160 So you’d expect it to be higher. But this is somewhat high. 00:30:37.160 --> 00:30:39.850 If you use that as an empirical Green’s function, 00:30:39.850 --> 00:30:44.660 of course we get much higher values than were observed. 00:30:44.660 --> 00:30:51.030 But this is the – sort of the uncertainties that these things involve. 00:30:51.030 --> 00:30:58.060 [inaudible]. I don’t want to show it because it takes too much time. 00:30:58.060 --> 00:31:01.390 The reason, of course, is that that earthquake had a very strong 00:31:01.390 --> 00:31:03.720 directivity towards Mexico City. 00:31:03.720 --> 00:31:10.500 See the P pulse as we go towards Mexico City. It’s much shorter. 00:31:10.500 --> 00:31:18.040 So then, as you focus towards Mexico City, we use that seed to 00:31:18.040 --> 00:31:22.810 synthesize without correcting for directivity, obviously get large number. 00:31:22.810 --> 00:31:30.850 Okay, so that says, well, depends. But let’s see some other way to 00:31:30.850 --> 00:31:35.510 look at this problem, whether these motions in UNAM 00:31:35.510 --> 00:31:40.160 were normal or were to be expected. 00:31:40.160 --> 00:31:46.270 Okay. We come back again to this recorded spectra at UNAM. 00:31:46.270 --> 00:31:48.710 This is the new earthquake. 00:31:48.710 --> 00:31:54.100 These are all other intraplate earthquake of magnitude 6.9 to 7.3. 00:31:54.100 --> 00:31:55.990 The distance of these earthquakes, somewhat greater. 00:31:55.990 --> 00:32:00.520 So, of course, the amplitude – the Fourier solution spectra is lower. 00:32:00.520 --> 00:32:02.120 No big surprise. 00:32:02.130 --> 00:32:07.810 But we normalize it again to the same distance, it still remains high. 00:32:07.810 --> 00:32:12.510 In band of frequency which is critical to Mexico City. 00:32:12.510 --> 00:32:15.120 So with respect to the other – these other earthquakes, 00:32:15.120 --> 00:32:22.720 seems that the earthquake was – well, fairly more – a lot more energetic 00:32:22.720 --> 00:32:26.590 than we would have inferred from this normalization. 00:32:26.590 --> 00:32:31.690 Let’s look at the response spectra at the university. 00:32:31.690 --> 00:32:35.160 The predicted response spectra are these blue lines, 00:32:35.160 --> 00:32:39.890 and the observed ones are here. And we again see, in the critical 00:32:39.890 --> 00:32:43.500 frequency range to Mexico City, the earthquake seems to be 00:32:43.500 --> 00:32:49.800 more towards the higher side – towards the [inaudible]. 00:32:49.800 --> 00:32:53.840 If you look at PGA as a function of distance, and PGV, 00:32:53.840 --> 00:33:00.530 we again see that PGV is somewhat higher than 00:33:00.530 --> 00:33:05.070 predicted by the ground motion prediction equations. 00:33:05.070 --> 00:33:08.510 If we look at the teleseismic source spectra, you don’t see much. 00:33:08.510 --> 00:33:13.950 There’s a little blip here. I don’t know whether one would 00:33:13.950 --> 00:33:19.110 think that, well, the source was very energetic or more – but anyway. 00:33:19.110 --> 00:33:26.920 The local data indicates that, yes, there was enriched variation often 00:33:26.920 --> 00:33:34.950 in this 0.4 to 1 hertz band, which is the critical band for Mexico City. 00:33:34.950 --> 00:33:39.840 Okay, so the answer is, well, probably the earthquake 00:33:39.840 --> 00:33:43.700 was more energetic than we’d have expected. 00:33:43.700 --> 00:33:48.260 Okay. Now we come to the issue of why the damage patterns 00:33:48.260 --> 00:33:58.220 were different in 2017 as compared to 1985 in the city. 00:33:58.220 --> 00:34:04.100 We actually talked about it, but let’s look at it more carefully. 00:34:04.100 --> 00:34:16.520 First, since – you know, we showed that the spectral ratios are nearly invariant. 00:34:18.440 --> 00:34:21.420 What decides what will happen is the input ground motion 00:34:21.420 --> 00:34:24.720 to this station at the university. 00:34:24.720 --> 00:34:28.740 And you can – as you can imagine, all the earthquakes which are 00:34:28.740 --> 00:34:34.340 intraslab are closer. And we know that they involve – 00:34:34.340 --> 00:34:38.940 they’re deeper, they’re closer, and they involve – 00:34:38.940 --> 00:34:45.380 and we have computed these things many times – higher stress drop. 00:34:47.700 --> 00:34:51.200 Well, not really true. Stress drop, I think – 00:34:51.200 --> 00:34:56.280 I don’t know who measures them, whether they can trust it, 00:34:56.280 --> 00:34:58.849 but these are the stress drops that you need to explain 00:34:58.849 --> 00:35:04.359 the observed spectra without too many corrections for the site. 00:35:04.360 --> 00:35:09.560 Okay, so, as a consequence of these things, deeper earthquake, closer sources, 00:35:09.560 --> 00:35:17.680 and higher stress drop, the input ground motion to the station in Mexico City 00:35:17.690 --> 00:35:22.539 and to Mexico – the Valley of Mexico – is enriched at high frequencies. 00:35:22.539 --> 00:35:25.700 And you can see what the consequence of that would be. 00:35:25.700 --> 00:35:28.840 Let’s see that in some detail. 00:35:30.000 --> 00:35:35.040 Ah, first – so let’s say – this you would infer without doing any exercise. 00:35:35.050 --> 00:35:40.450 Well, you’d expect the intraplate – intraslab earthquakes to be energetic 00:35:40.450 --> 00:35:44.140 at high frequencies, and these ones – 00:35:44.140 --> 00:35:48.329 and these are recorded ground motions at the university. 00:35:48.329 --> 00:35:52.109 These are interplate earthquakes, and these are intraslab earthquakes – 00:35:52.109 --> 00:35:54.349 magnitude 7 and above. 00:35:54.349 --> 00:35:59.299 And you can see that these spectra are peaked, and they are shifted 00:35:59.299 --> 00:36:04.789 towards longer period – more than 2 seconds, generally. 00:36:04.789 --> 00:36:11.700 While the spectra of the intraslab earthquakes are more flat. 00:36:11.700 --> 00:36:16.220 They are more enriched at high frequencies as compared to these. 00:36:16.220 --> 00:36:18.480 And let’s try to compare it. 00:36:20.860 --> 00:36:23.940 This comparison, I think, illustrates the point quite well. 00:36:23.940 --> 00:36:27.940 This is the 1985 earthquake at the university – north-south component. 00:36:27.940 --> 00:36:32.720 And the same – at the same site, this is the 2017 earthquake. 00:36:32.720 --> 00:36:38.079 Look at the ground motion – as I was mentioning, the peak values were about 00:36:38.079 --> 00:36:43.789 really half – this one, which is here. These are band-pass filter accelerograms. 00:36:43.789 --> 00:36:48.720 If you filter it between 2.5 and 8 hertz, there’s nothing left. 00:36:48.720 --> 00:36:50.760 So all the energy is at longer period. 00:36:50.769 --> 00:36:55.440 While this, when you filter it, still is very energetic, right? 00:36:55.440 --> 00:36:59.730 So this thing is full of high frequency. In fact, look at the Fourier spectra. 00:36:59.730 --> 00:37:03.000 The continuous line here is 2017. 00:37:03.000 --> 00:37:08.390 And the dashed ones are for the 1985 earthquake. 00:37:08.390 --> 00:37:14.160 This earthquake, at the university, is more energetic, say, 00:37:14.160 --> 00:37:20.740 at all frequencies above this. While, at lower frequencies, 00:37:20.740 --> 00:37:25.620 later periods, the 1985 earthquake is more energetic. 00:37:25.620 --> 00:37:28.480 You can see what the consequences of this would be. 00:37:28.480 --> 00:37:33.840 This difference is the whole cause of the difference 00:37:33.840 --> 00:37:37.369 in the pattern that you see damaged in Mexico City. 00:37:37.369 --> 00:37:41.700 Look at the response spectra at the university during the two earthquakes. 00:37:41.700 --> 00:37:46.410 This is for the recent intraslab earthquake, and this is ’85. 00:37:46.410 --> 00:37:55.789 You can see that, at periods smaller than – I don’t know, 1.5, 1.6 seconds, 00:37:55.789 --> 00:37:58.890 the intraslab earthquake is much more energetic. 00:37:58.890 --> 00:38:03.690 At longer periods, the 1985 earthquake was. 00:38:03.690 --> 00:38:09.240 Good. Once – well, we can see the same thing when you do the 00:38:09.240 --> 00:38:13.140 same exercise at SCT, which is in the Lake-bed Zone. 00:38:13.140 --> 00:38:19.220 The recordings of the 1985 earthquake now are much more higher at that 00:38:19.220 --> 00:38:21.610 Lake-bed because of the amplification. 00:38:21.610 --> 00:38:29.809 Then, during 2017, and filtered things, and same thing that we had inferred 00:38:29.809 --> 00:38:37.740 from spectral ratios and the data at UNAM is, of course, true. 00:38:37.740 --> 00:38:45.779 Okay. Well, so let’s see. This is what you would have expected. 00:38:45.779 --> 00:38:47.869 Because it is enriched at high frequencies, 00:38:47.869 --> 00:38:54.630 the damages would be at sites with period around 1 second or so. 00:38:54.630 --> 00:38:58.580 So let’s see. This is 2017. Now it’s computed. 00:38:58.580 --> 00:39:03.380 It’s not the estimated by random vibration theory, 00:39:03.380 --> 00:39:04.940 but it’s the computed one. 00:39:04.940 --> 00:39:13.380 It is 1-second spectral values. This is 2-second spectral values. 00:39:13.380 --> 00:39:17.980 In 1985, of course, it’s inferred. So here, as you can see, 00:39:17.980 --> 00:39:24.220 2-second is very much illuminated. Here, 1-second is more illuminated. 00:39:24.220 --> 00:39:29.060 And if you plot the damage – damaged buildings versus period 00:39:29.060 --> 00:39:34.030 of the site, the new earthquake – you know, it’s fairly broadband, 00:39:34.030 --> 00:39:43.300 as the spectra was – the Fourier spectra at UNAM – fairly broadband, 00:39:43.300 --> 00:39:48.270 but centered around, say, 1, 1.4 second site period. 00:39:48.270 --> 00:39:53.579 While this one – 1985 earthquake – was peaked at – the damage is 00:39:53.579 --> 00:39:58.559 peaked at 2 seconds, as you would have expected, right? 00:39:58.559 --> 00:40:02.380 So there’s really no mystery in that sense. 00:40:02.380 --> 00:40:05.059 And you can see the damage. 00:40:05.060 --> 00:40:13.380 This is – it just falls where the 2-second spectral values are very high. 00:40:15.520 --> 00:40:18.460 Okay. There is still a mystery. 00:40:20.900 --> 00:40:24.860 [Silence] 00:40:25.620 --> 00:40:28.980 Well – well, this is the thing I mentioned, right? 00:40:28.980 --> 00:40:33.140 If you look at the estimated Sa, this seemed to better correlate 00:40:33.140 --> 00:40:38.029 with the damage than the real data. There’s subtle discrepancy between 00:40:38.029 --> 00:40:43.420 what we were getting from the real data and from random vibration theory. 00:40:43.420 --> 00:40:45.880 I think there’s some problem with interpolation and so on. 00:40:45.880 --> 00:40:48.120 We’ll have to figure that out. 00:40:48.120 --> 00:40:53.080 Okay. So there – the mystery is really not a mystery. 00:40:53.089 --> 00:40:57.230 In fact, we have, even before, contemplated this possibility and 00:40:57.230 --> 00:41:02.369 had mentioned in several papers that this would happen during 00:41:02.369 --> 00:41:06.019 intraslab earthquakes, that the damage would be more like 00:41:06.019 --> 00:41:10.309 where the periods are, like, what, 1 second rather than 2 seconds. 00:41:10.309 --> 00:41:14.960 The other question is, how often such ground motions 00:41:14.960 --> 00:41:20.279 are exceeded in the valley. It’s critical again. 00:41:20.279 --> 00:41:25.930 And, again, we take recourse to our data at the university. 00:41:25.930 --> 00:41:29.109 These are the exceedance rates for the 54 years of data 00:41:29.109 --> 00:41:31.400 that we have – da da da da da da. 00:41:31.400 --> 00:41:33.940 The last data point is here – new earthquake. 00:41:33.940 --> 00:41:37.200 Now that, of course, becomes difficult in the sense 00:41:37.200 --> 00:41:42.319 that we know that this occurred in the last 54 years. 00:41:42.319 --> 00:41:47.380 But we also probably know that, in the last 117 years, 00:41:47.380 --> 00:41:52.420 no such earthquake had – so maybe this data should be here, right? 00:41:52.420 --> 00:41:59.480 And the recurrence period is at least of that order – 117 years or so. 00:41:59.480 --> 00:42:07.020 The interesting thing here is this PSHA estimation, which is based on – 00:42:07.020 --> 00:42:11.609 it’s very interesting that these two observed values are roughly 00:42:11.609 --> 00:42:19.660 following the probability hazard – seismic hazard estimation values. 00:42:19.660 --> 00:42:24.440 And I think there are very few places in the world where you can really test this. 00:42:24.440 --> 00:42:29.259 You know – as you know, the exercise that is done to compute 00:42:29.260 --> 00:42:33.640 PSHA, you have to have a catalog. You have a b value. 00:42:33.640 --> 00:42:36.400 You have attenuation logs and so forth and so on. 00:42:36.400 --> 00:42:40.130 And so, at any given point, you use all this information to predict 00:42:40.130 --> 00:42:44.230 what the motion would be, right? Here we have observed motions. 00:42:44.230 --> 00:42:49.710 And this has to be checked out with probabilistic hazard estimation. 00:42:49.710 --> 00:42:54.509 And, in this case, luckily, it does turn out to be true. 00:42:54.509 --> 00:43:03.240 This black line here is our design value for intraslab earthquakes in Mexico City. 00:43:03.240 --> 00:43:08.760 And they correspond to something like 300 years of return period. 00:43:08.760 --> 00:43:13.900 So I would say that says ground motions probably exceed every 150 years or so. 00:43:13.900 --> 00:43:15.680 Maybe. 00:43:19.040 --> 00:43:23.460 So we’re coming to – let’s see. How am I doing with time? 00:43:23.460 --> 00:43:26.880 Still 15 minutes, right? Good. 00:43:27.840 --> 00:43:32.780 I can bore you some more. Okay. So and the next question, of course, 00:43:32.790 --> 00:43:37.509 is how close these earthquakes can be. I mean, as I was mentioning, although 00:43:37.509 --> 00:43:43.759 100-kilometer depth contour is here, more or less where the volcanoes are, 00:43:43.760 --> 00:43:50.400 there is no seismicity beyond this green line. For some reason, it ceases. 00:43:50.400 --> 00:43:56.900 And it doesn’t – it’s not that it ceases – this seismicity ceases in a 00:43:56.900 --> 00:44:02.300 gradual fashion or something. These are magnitude 5, 5-1/2, and greater. 00:44:02.300 --> 00:44:05.040 And then suddenly it just vanishes. 00:44:06.560 --> 00:44:12.700 For, what – I’m sure – experts in petrology and so on – those who 00:44:12.710 --> 00:44:15.530 work with this kind of thing would have some explanation. 00:44:15.530 --> 00:44:20.579 But it’s not very clear to me. Looking at these things very carefully, 00:44:20.579 --> 00:44:22.869 we have recently identified a couple of earthquakes 00:44:22.869 --> 00:44:28.009 which are actually near the volcano. But these are magnitude 3.3. 00:44:28.009 --> 00:44:31.650 This earthquake, as I was mentioning, had it not occurred, we’d have said 00:44:31.650 --> 00:44:35.980 that this line would be going like this. Now we have to perturb it. 00:44:35.980 --> 00:44:42.960 Could it be that one closer to Mexico City occurred [inaudible] magnitude? 00:44:42.960 --> 00:44:45.510 So we’ll leave that question because I think 00:44:45.510 --> 00:44:49.600 the answer not blowing in the wind. 00:44:50.380 --> 00:44:57.640 Okay. So we come back to the issue of, well, how did everything perform? 00:44:57.650 --> 00:45:01.859 What about the building code? Well, was it exceeded or not? 00:45:01.860 --> 00:45:13.860 Here is some example of the data – response spectra – the data and the code. 00:45:13.860 --> 00:45:18.940 The code, as you know, it’s almost site-specific now in Mexico City. 00:45:18.940 --> 00:45:22.490 For each site, you can have your own code, depending on the period 00:45:22.490 --> 00:45:27.560 and the amplification and so on. So you see, these dashed lines are – 00:45:27.560 --> 00:45:33.530 scale is the same – has different amplitude because of the site conditions. 00:45:33.530 --> 00:45:36.450 And we see that – this is just some samples, right? 00:45:36.450 --> 00:45:45.980 If you look at 80 stations, few stations were actually – were – I think four 00:45:45.980 --> 00:45:50.569 stations where the response spectra observed exceeded the building code. 00:45:50.569 --> 00:45:53.780 Like, for example, one component here. 00:45:53.780 --> 00:45:57.450 Maybe here. So there are some three or four sites only. 00:45:57.450 --> 00:46:03.299 So the code actually – after the earthquake, is not going to be revised. 00:46:03.300 --> 00:46:08.200 Because I think the engineers are quite happy with the code itself. 00:46:09.520 --> 00:46:13.999 So it did quite well. There’s one other item that affects 00:46:14.000 --> 00:46:21.640 ground motions in Mexico City from intraslab earthquakes, which is this. 00:46:21.640 --> 00:46:26.660 Here is, again, university. Here is the volcano. 00:46:26.660 --> 00:46:29.800 Popocatépetl – which is active, as you know. 00:46:29.809 --> 00:46:32.900 If you look at recording at these two stations 00:46:32.900 --> 00:46:35.880 for events which are farther away … 00:46:38.720 --> 00:46:40.980 They come from this side. 00:46:42.640 --> 00:46:44.540 That’s this part. 00:46:45.340 --> 00:46:48.759 Event 8, 9, 10, 11 are coming from this side. 00:46:48.760 --> 00:46:53.620 The wave path crosses Popocatépetl. 00:46:54.480 --> 00:47:00.340 And the top trace here is at university, so the wave path 00:47:00.340 --> 00:47:07.759 to the university crosses volcano. But to this station, Yautepec, it does not. 00:47:07.759 --> 00:47:12.900 So look at these amplitudes. This is crossing the volcano. 00:47:12.900 --> 00:47:16.540 And this part – same earthquakes. 00:47:16.540 --> 00:47:20.970 At this place, much more energetic, as you can see. 00:47:20.970 --> 00:47:25.300 The distances are about the same, so it’s not a problem of the distance. 00:47:27.240 --> 00:47:31.540 But when the earthquakes come from this side – we’ve got 1, 2, 7 – 00:47:31.540 --> 00:47:36.060 if we could find more events. When they come from this side, they 00:47:36.070 --> 00:47:41.890 don’t pass through here, [inaudible], and the amplitudes are about the same. 00:47:41.890 --> 00:47:47.599 So there’s something along the path which attenuates and most likely is 00:47:47.599 --> 00:47:51.170 Popocatépetl, which acts, actually, as shield in some ways 00:47:51.170 --> 00:47:55.390 for strong motions, especially high frequencies. 00:47:55.390 --> 00:48:02.310 This earthquake was, of course, as you see – this earthquake was here. 00:48:02.310 --> 00:48:05.940 Popocatépetl is here. The city is here. 00:48:05.940 --> 00:48:11.400 The path does not pass through the volcano. 00:48:11.400 --> 00:48:17.360 And perhaps because of that, the frequencies were not attenuated, 00:48:17.369 --> 00:48:21.589 as one would have expected, they were not expected – I mean, not attenuated. 00:48:21.589 --> 00:48:27.961 So that’s okay. However, it also follows that, if we had a station above and a 00:48:27.961 --> 00:48:32.859 station at a similar distance below, we would see this difference again. 00:48:32.859 --> 00:48:37.260 So if we look for one – yes, we have it. 00:48:37.260 --> 00:48:41.539 Okay. This is the PGA contour of the whole – this thing. 00:48:41.539 --> 00:48:43.809 Let’s not go through it. 00:48:43.809 --> 00:48:46.970 So here is the epicenter of this earthquake. 00:48:46.970 --> 00:48:49.519 Here is Popo, and we have a station which is called DHIG, 00:48:49.519 --> 00:48:52.779 which is up above. Let’s compare this record with 00:48:52.779 --> 00:48:57.190 a station which is at equal distance, which is Pinotepa Nacional. 00:48:57.190 --> 00:49:03.869 So we – this part, since it passes through the volcano, should be more attenuated 00:49:03.869 --> 00:49:08.289 than the one – wave path 2 – Pinotepa Nacional – 00:49:08.289 --> 00:49:11.769 where we were doing the experiment – successful experiment with 00:49:11.769 --> 00:49:17.589 Bob Meier and Walter Mooney a hundred years ago. 00:49:17.589 --> 00:49:25.710 Okay. So here are the spectra. The trajectory wave path to the coast, 00:49:25.710 --> 00:49:32.730 not passing the volcano, is way above as compared to the one with depth. 00:49:32.730 --> 00:49:36.109 Both sites are very competent. You know, these are limestone sites. 00:49:36.109 --> 00:49:38.039 Local site effects aren’t a problem. 00:49:38.039 --> 00:49:40.859 Most likely it is the effect of the volcano. 00:49:40.860 --> 00:49:49.300 So volcano also plays tricks and so on and – it’s nice to have volcano along the 00:49:49.300 --> 00:49:54.540 path to Mexico City and [inaudible] for this earthquake that didn’t really happen. 00:49:54.540 --> 00:49:57.980 Okay, well, so conclusions. 00:49:59.060 --> 00:50:03.540 Well, it was – conclusions. I don’t know that they’re conclusions, 00:50:03.540 --> 00:50:08.509 but let’s just note that this was the closest intraslab 00:50:08.509 --> 00:50:12.489 earthquake in Mexico City. Distance was 127 kilometers. 00:50:12.489 --> 00:50:15.880 Previous earthquakes that we had seen – 00:50:15.880 --> 00:50:24.220 magnitude 5.8 or so – were at 145, 150, so somewhat closer. 00:50:24.220 --> 00:50:29.630 Magnitude 7 and at a distance of 145 kilometers [inaudible] – 00:50:29.630 --> 00:50:32.579 well, I had computed – thought that it was a likely scenario. 00:50:32.579 --> 00:50:35.880 So, in that sense, was not really unusual. 00:50:35.880 --> 00:50:41.640 This is the most destructive intraslab event in the history of the city. 00:50:42.600 --> 00:50:49.360 Highest PGA, 57 gal, at UNAM, last 54 years. 00:50:49.360 --> 00:50:54.360 Next-highest, 29 gal in 1985. 00:50:54.360 --> 00:50:59.900 Relatively enriched ground motion between 0.4 and 1 hertz. 00:50:59.900 --> 00:51:04.599 And, hence, its consequence with the damage pattern, which differed 00:51:04.599 --> 00:51:10.630 [inaudible] earthquake – was expected – anticipated, in fact. 00:51:10.630 --> 00:51:14.640 And it is clear that this pattern will be repeated during future 00:51:14.640 --> 00:51:18.039 damaging earthquakes. There is no other possibility. 00:51:18.039 --> 00:51:21.789 In other words, intraslab earthquakes would demolish 00:51:21.789 --> 00:51:26.480 buildings less than 10 floors. Interplate earthquake would 00:51:26.480 --> 00:51:29.869 demolish buildings [laughs] which are greater than 10 floors. 00:51:29.869 --> 00:51:33.600 So we get screwed from both sides. 00:51:35.420 --> 00:51:39.560 Ground motion exceeded the current building code at only three sites, 00:51:39.569 --> 00:51:43.239 here says. Maybe there are four. I don’t know. 00:51:43.239 --> 00:51:47.800 There were 44 collapsed buildings, and I am told that they are all – 00:51:47.800 --> 00:51:53.989 were all built before 1987 when the new code became effective. 00:51:53.989 --> 00:52:00.299 Many newer buildings were damaged, and apparently that was because of 00:52:00.299 --> 00:52:07.039 poor construction – other usual problems with these things. 00:52:07.039 --> 00:52:14.970 PGA of this kind of value, like 60 gals or so, I think probably is a 00:52:14.970 --> 00:52:19.319 return period of about 160 years or so. So this is the situation. 00:52:19.320 --> 00:52:24.600 I thank you. And I – let’s see. Five minutes before. 00:52:24.600 --> 00:52:30.740 [Applause] 00:52:30.740 --> 00:52:34.700 So if you have any questions, ask me. I might be able to even answer. 00:52:34.700 --> 00:52:38.200 May be able to answer some of them. 00:52:40.280 --> 00:52:43.600 - Could you go back to the concluding slide? 00:52:43.609 --> 00:52:47.340 So you – number 4 – highest PGA, 57 gal. 00:52:47.340 --> 00:52:51.990 What was the peak ground velocity? How many centimeters per second? 00:52:51.990 --> 00:52:55.940 - I think it was 8 or so. Let me – there’s values in the table. 00:52:55.940 --> 00:52:59.220 I’ll find it. - It has to be higher than 8. 00:52:59.220 --> 00:53:01.080 - No, I … 00:53:03.020 --> 00:53:06.400 We are in Mexico City, my friend. - [inaudible] 00:53:06.400 --> 00:53:10.120 - It’s a mind-boggling place. 00:53:11.120 --> 00:53:14.480 Okay – 8.9. I’m sorry. 00:53:14.480 --> 00:53:16.799 - That’s not very high. So it’s surprising so much 00:53:16.799 --> 00:53:19.900 damage for – usually damage … - Yeah, but you see – you know, 00:53:19.900 --> 00:53:24.770 what you have to recall is this is on a Hill Zone. 00:53:24.770 --> 00:53:29.609 I mean, 130-40 kilometers, you shouldn’t expect any damage anywhere, right? 00:53:29.609 --> 00:53:34.039 - Right. Okay. - So, in fact, you could almost say that 00:53:34.039 --> 00:53:39.789 never in the history of Mexico City there has been any damage in the Hill Zone. 00:53:39.789 --> 00:53:42.670 Which is where we expect, with reasonable ground conditions, 00:53:42.670 --> 00:53:45.320 at those distances, why should there be any damage? 00:53:45.320 --> 00:53:48.620 So, yeah, the value is correct. Here are the other – 00:53:48.630 --> 00:53:54.359 for other earthquakes, like 3.2. And many of them were badly felt. 00:53:54.359 --> 00:54:01.480 For the – oh, here. For ’85, I don’t have it for some reason here in this table. 00:54:02.589 --> 00:54:10.739 - So the other point is that the – this kind of ground motion 00:54:10.739 --> 00:54:18.589 can be anticipated and only three sites exceeded the code. 00:54:18.589 --> 00:54:23.860 So the buildings that collapsed were pre-building code. 00:54:23.860 --> 00:54:26.200 - Right. - So basically, the problem in 00:54:26.200 --> 00:54:30.860 Mexico City has to do with older buildings that … 00:54:30.860 --> 00:54:33.680 - Yes. - … are vulnerable to ground motions 00:54:33.680 --> 00:54:36.680 that you – that you can anticipate. - Yes. 00:54:36.680 --> 00:54:39.819 - So what to do about that? - Well, I don’t know. 00:54:39.819 --> 00:54:44.400 Maybe – but I think – let me make a correction here. 00:54:44.400 --> 00:54:50.700 Yes, true – collapsed buildings, apparently they were built before 1987. 00:54:50.700 --> 00:54:55.200 But there were many newer buildings which suffered severe damage. 00:54:55.200 --> 00:55:01.720 Okay? And they were newer – they were supposed to have fulfilled the – 00:55:01.720 --> 00:55:04.780 but the thing, they were badly made. I mean, there’s – the problem is, 00:55:04.790 --> 00:55:09.499 the code in Mexico City is very advanced, I think. 00:55:09.499 --> 00:55:12.319 I don’t know about other codes here in California. 00:55:12.319 --> 00:55:15.390 So for that – you know, as I was saying, it is site-specific. 00:55:15.390 --> 00:55:18.460 You want to build at this … - Yeah. 00:55:18.460 --> 00:55:25.220 - You can go and look at response spectra – the building design spectra. 00:55:25.220 --> 00:55:29.809 The problem is that – and so, when actually you make a building, 00:55:29.809 --> 00:55:35.650 you are supposed to have a map – I mean, structural details and 00:55:35.650 --> 00:55:38.920 so on which would say that, yeah, it’ll fulfill the requirements. 00:55:38.920 --> 00:55:42.329 But there is no supervision. They don’t – it’s not [inaudible]. 00:55:42.329 --> 00:55:45.960 - So there’s a problem with enforcement of the building code. 00:55:45.960 --> 00:55:48.950 - Enforcement, corruption, and also ineptitude on the 00:55:48.950 --> 00:55:51.289 part of architects and engineers. 00:55:51.289 --> 00:55:58.880 Because architects like to have corner windows and glasses and so forth. 00:55:58.880 --> 00:56:01.839 And they are – buildings become very vulnerable. 00:56:01.839 --> 00:56:03.619 - Yeah. Okay. - The engineers would know 00:56:03.619 --> 00:56:05.079 more than I would know. - All right. Thanks. 00:56:05.079 --> 00:56:07.769 - Hi, Shri. That was a great talk. 00:56:07.769 --> 00:56:09.900 Make a comment first and then have a question. 00:56:09.900 --> 00:56:12.800 At least in the United States, the goal of the building code is 00:56:12.809 --> 00:56:16.210 not to prevent damage to the building, but to protect life safety. 00:56:16.210 --> 00:56:19.599 So if the building is lost, but no one died, the codes worked. 00:56:19.599 --> 00:56:21.369 - Well, yeah … - So it’s a question 00:56:21.369 --> 00:56:24.059 what the objective is that … - But I think most codes 00:56:24.059 --> 00:56:29.039 actually are designed that way. You’re – they are supposed to degrade 00:56:29.039 --> 00:56:32.049 and so on, but not collapse on you, right? This is true. 00:56:32.049 --> 00:56:36.079 - Right. So if they – if buildings built to the code did not collapse, 00:56:36.079 --> 00:56:40.099 but also did not claim any lives, then perhaps the code was successful. 00:56:40.099 --> 00:56:41.460 I think that’s a possibility. 00:56:41.460 --> 00:56:43.730 - Well, okay, then, good. You’re more optimistic. 00:56:43.730 --> 00:56:47.910 - So a question I had was about the – your very interesting use of the 00:56:47.910 --> 00:56:51.740 empirical Green’s functions to compare the different events. 00:56:51.740 --> 00:56:54.359 And you made a comment about the stress drops. 00:56:54.359 --> 00:56:57.670 And I’m wondering, has anyone looked at – so using empirical 00:56:57.670 --> 00:57:02.039 Green’s functions to get a better handle as to what the source spectra 00:57:02.039 --> 00:57:06.059 are for these damaging intraslab earthquakes? 00:57:06.059 --> 00:57:11.440 You suggested maybe 30 GPa was – MPa was too high as a stress drop. 00:57:11.440 --> 00:57:17.700 - Well, no. Are you saying that the convolving empirical – little earthquakes 00:57:17.700 --> 00:57:21.839 from bigger ones look at the true source spectra and try to figure it out. 00:57:21.839 --> 00:57:27.819 Well, this game is very – while very promising, often doesn’t work. 00:57:27.819 --> 00:57:32.940 Because, you know, periods where you need the information, it is noisy. 00:57:32.940 --> 00:57:39.130 And so it’s – no. I am always willing to – looking for this kind of situation. 00:57:39.130 --> 00:57:42.240 Never successfully have I – except for, you know, 00:57:42.240 --> 00:57:44.339 tiny earthquakes – you know, a couple of them. 00:57:44.339 --> 00:57:49.460 And maybe you could infer where the corner frequency is. 00:57:49.460 --> 00:57:56.460 These estimates are based on – these were done to predict ground motion. 00:57:56.460 --> 00:58:00.520 So we say, all our stations are hard site – nominally. 00:58:00.520 --> 00:58:05.079 What is the value of stress drop that we have to put in a given model 00:58:05.080 --> 00:58:09.280 to bring us to that level, right? 00:58:09.280 --> 00:58:16.520 And we see that interplate ones are 7 bars – 7 megapascals – 70 bars. 00:58:16.520 --> 00:58:18.580 And this one is 30 or so, right? 00:58:18.580 --> 00:58:24.319 So it’s a different – in the valley, stress drop, that you need to jack it up, right? 00:58:24.319 --> 00:58:27.630 But they are real? Of course not. 00:58:27.630 --> 00:58:30.779 I wouldn’t know what the real value is of any earthquake when you’re 00:58:30.779 --> 00:58:35.529 talking about stress drop. It looks like hocus-pocus to me still. 00:58:35.529 --> 00:58:38.839 But you have done – dedicated much of your life to [chuckles] doing this. 00:58:38.840 --> 00:58:42.420 You maybe should defend it. [laughter] 00:58:42.420 --> 00:58:46.980 But at any rate, I mean, for practical purposes, these estimates are wonderful. 00:58:46.980 --> 00:58:51.839 I mean, if I have to predict, on a typical hard site in Mexico, 00:58:51.839 --> 00:58:59.979 where we have stations – because then these values are just wonderful, right? 00:58:59.979 --> 00:59:04.239 And if there is amplification with respect to that typical hard site for 00:59:04.240 --> 00:59:08.800 which we have GMPE, well, you add something more, right? 00:59:11.450 --> 00:59:14.080 - Thank you for that talk, Krishna. I’m curious what you think is 00:59:14.099 --> 00:59:18.720 going on with Popocatépetl. You got other volcanoes in Mexico. 00:59:18.720 --> 00:59:22.329 I guess they’re not causing that problem. And, at any rate, it seems like 00:59:22.329 --> 00:59:26.240 Popocatépetl is a rather shallow phenomenon that maybe the magma 00:59:26.240 --> 00:59:29.240 reservoir extends as deep as 20 kilometers, but the earthquakes 00:59:29.240 --> 00:59:33.450 are coming from 50 or 60 kilometers. Seems like there – would completely 00:59:33.450 --> 00:59:38.630 bypass and go under the volcano. What do you think’s going on? 00:59:38.630 --> 00:59:43.520 - Well, it could be. The thing is that, when I showed you a 00:59:43.520 --> 00:59:50.080 suite of earthquakes for which the whole thing was deficient – all right, let’s see. 00:59:50.700 --> 00:59:57.220 Well, I mean, wave path, I mean, maybe the entire region is quite hot, you know? 00:59:57.220 --> 01:00:00.099 I mean, it doesn’t have to pass through the magma to – 01:00:00.100 --> 01:00:07.390 for the waves to attenuate at all. In fact – let’s see where this is. 01:00:09.080 --> 01:00:13.180 Actually, you know, this we did in 2002 or something. 01:00:13.190 --> 01:00:15.630 And we were very curious to see whether we could 01:00:15.630 --> 01:00:20.309 sort of map something. Well, we need stations above – 01:00:20.309 --> 01:00:26.950 on the other side of volcano. Well, priority is not trying to figure out 01:00:26.950 --> 01:00:30.670 this kind of thing [inaudible] to put on the coast recorders [inaudible]. 01:00:30.670 --> 01:00:33.970 So we have never managed to do it. 01:00:33.970 --> 01:00:39.320 And it’s one of the priorities where some money is needed – and manpower. 01:00:39.320 --> 01:00:40.920 Where is this? Okay. 01:00:40.920 --> 01:00:44.309 So many of these earthquakes are probably shallow. 01:00:44.309 --> 01:00:45.980 I don’t really remember the depths anymore. 01:00:45.980 --> 01:00:48.480 But for example, this one is not. 01:00:48.480 --> 01:00:52.380 This is an intraplate earthquake of magnitude 6.9. 01:00:52.380 --> 01:00:56.010 It is also 60 kilometers deep. 01:00:56.010 --> 01:01:00.160 So 9 is here – 9 – somewhere here is the epicenter. 01:01:00.160 --> 01:01:05.540 It was close to Tehuacán. It is called Tehuacán earthquake. 01:01:07.200 --> 01:01:09.640 Where, you know, mineral water comes from. 01:01:10.140 --> 01:01:16.380 Okay. So, from here, at that depth, the wave passing here – I mean, 01:01:16.380 --> 01:01:21.920 all kinds of – we’d have to see where the waves that we’re recording are actually – 01:01:21.920 --> 01:01:24.520 what is the wave path? But much of it probably – 01:01:24.520 --> 01:01:27.100 some of it will go below. Others are passing by above. 01:01:27.109 --> 01:01:30.960 Lg waves propagating – not very much excited. 01:01:30.960 --> 01:01:35.549 So I don’t know where the wave path is. And, of course, as I was saying, 01:01:35.549 --> 01:01:38.820 it doesn’t really have to go through the magma chamber. 01:01:38.820 --> 01:01:40.839 Probably the whole region is hot. 01:01:40.840 --> 01:01:43.920 However, when the waves pass Nevado de Toluca, 01:01:43.920 --> 01:01:46.040 we don’t see any attenuation. 01:01:49.680 --> 01:01:54.160 So, for some reason, Nevado is not all that hot, according to this 01:01:54.160 --> 01:01:59.820 observation, right? We should [inaudible] this observation now. 01:02:01.080 --> 01:02:04.660 So, what is your question? - Where’s my microphone? 01:02:05.420 --> 01:02:08.000 - No, but you can scream. [laughs] 01:02:09.800 --> 01:02:12.580 - Actually, I have a different question. 01:02:13.540 --> 01:02:17.859 If we go – and I apologize for asking you to do this – 01:02:17.859 --> 01:02:28.420 to any representation of the spectra from the 2017 event, either a comparison 01:02:28.420 --> 01:02:34.390 with the 1857 or just – just any spectra that you have there, 01:02:34.390 --> 01:02:40.100 it seems to me that that spectra is anomalous at long periods. 01:02:40.100 --> 01:02:48.450 That is to say, it has a perceptible hole at about 3 seconds, and then … 01:02:48.450 --> 01:02:51.660 - The [inaudible]. - And then a peak below that. 01:02:51.660 --> 01:02:55.309 - You mean 2017 spectra? - Yeah, 2017 spectra. 01:02:55.309 --> 01:02:58.220 - Let’s just look at it. Oh. 01:02:58.220 --> 01:03:01.300 - We went right off the edge of the Earth there. 01:03:01.319 --> 01:03:05.180 Who thought it was flat? - The spectra. The spectra. 01:03:05.180 --> 01:03:08.240 Wow, I’m going wrong way. Okay, let’s try … 01:03:09.660 --> 01:03:13.400 But now we are very small – oh, you should see the movie, 01:03:13.400 --> 01:03:17.459 though, of Eisenstein. - I will look it up. 01:03:17.459 --> 01:03:22.150 - Do so. I don’t know what – I’m supposed to do something here. 01:03:22.150 --> 01:03:26.200 I didn’t [inaudible]. - He’s the maker of Potemkin, right? 01:03:26.200 --> 01:03:28.780 - [laughs] Yeah. Yeah, right, right, right. 01:03:30.300 --> 01:03:32.740 No, no, just … - Oh, full screen. 01:03:32.740 --> 01:03:34.240 - Full screen. There. 01:03:38.660 --> 01:03:41.260 Let’s see. Where was it? 01:03:45.000 --> 01:03:51.780 Yeah. Let’s look at the spectra. So, like, maybe I can learn something. 01:03:53.300 --> 01:03:55.460 I doubt. - It’s a little late for – okay, this is fine. 01:03:55.460 --> 01:03:58.240 - You mean this stuff here? - Yes. 01:03:58.240 --> 01:04:00.640 And the … - Let’s see. 01:04:00.640 --> 01:04:02.630 There are many other earthquakes there too, right? 01:04:02.630 --> 01:04:06.080 There is another one which has this bump too, right? 01:04:06.860 --> 01:04:10.240 - Well, the question is, where is that coming from? 01:04:10.240 --> 01:04:15.600 Is that a propagation effect? Or is that a source effect? 01:04:16.440 --> 01:04:19.320 And, I mean, I think that that … - Yeah. But one thing I … 01:04:19.320 --> 01:04:22.340 - That shape underlies your comparison of damage patterns. 01:04:22.340 --> 01:04:27.300 - Right. One thing that I must – I don’t know the answer offhand, but I would 01:04:27.300 --> 01:04:31.020 say the damage is, of course, in this part, right? Not this – this is too long period. 01:04:31.020 --> 01:04:33.589 There is 5 seconds – so when the thing happens, right? 01:04:33.589 --> 01:04:35.640 So, like, 0.2 or something? 01:04:36.400 --> 01:04:40.139 That’s, like, what, 5 seconds and … - That’s 5 seconds. You’re right. 01:04:40.139 --> 01:04:42.430 - The damage is, of course, in this part, right? 01:04:42.430 --> 01:04:46.479 - But the – when you compare the … - Michoacán and this? 01:04:46.480 --> 01:04:50.900 - The Michoacán and this. And please don’t look for another slide. 01:04:50.900 --> 01:04:52.360 [laughter] 01:04:52.360 --> 01:04:56.500 - No. Because I want to know. I’m sure you’re telling something 01:04:56.500 --> 01:04:59.600 which is very wise, and I should learn it. - No, I … 01:04:59.609 --> 01:05:02.030 - Are you kidding? - This is only a question. 01:05:02.030 --> 01:05:05.620 - No, good question. Okay. So you see this bump here, 01:05:05.620 --> 01:05:07.220 you’re talking about? - Yes. 01:05:07.220 --> 01:05:13.200 - There’s a little bump here as well. But what would your explanation be? 01:05:13.209 --> 01:05:18.559 - No, no. But I’m just curious as to – as to where you think that 01:05:18.559 --> 01:05:19.859 that’s coming from. - I really [inaudible] … 01:05:19.859 --> 01:05:22.670 - Or whether you haven’t regarded it … - No, no, not at all. 01:05:22.670 --> 01:05:26.749 - Not at all. Okay. - Because that’s outside our ballgame. 01:05:26.749 --> 01:05:30.479 [laughs] Where the action is. But I think – yeah, 01:05:30.479 --> 01:05:34.660 we should actually – let’s … - Well, the … 01:05:34.660 --> 01:05:37.500 - Let’s look at teleseismic inversion. 01:05:38.920 --> 01:05:41.480 Is there anything there? 01:05:42.260 --> 01:05:45.480 - It was too small for me to see when you showed it. 01:05:45.490 --> 01:05:49.349 - Yeah. Let’s see. Oh, I don’t know where it is. Yeah. 01:05:49.349 --> 01:05:50.349 - Okay. 01:05:50.349 --> 01:05:52.160 - I don’t want to give you a headache. 01:05:52.160 --> 01:05:54.720 - No, no – this is. We can take this up at lunch. 01:05:54.720 --> 01:05:57.360 - Okay. Lunch is … - Without the slides. 01:05:57.360 --> 01:06:03.420 - Lunch is for discussing more science? [laughs] Lunch is lunch. Okay. 01:06:03.420 --> 01:06:08.519 Here it is. Do you see anything – we’re talking … 01:06:08.519 --> 01:06:10.910 This is [inaudible]. Forget it, yeah, yeah. 01:06:10.910 --> 01:06:12.580 We’ll talk about it later. 01:06:13.280 --> 01:06:15.500 So the answer is, I don’t know. 01:06:18.220 --> 01:06:20.839 - We have time for one final question. 01:06:22.240 --> 01:06:24.760 All right. If not, let’s thank him one more time. 01:06:24.760 --> 01:06:25.680 - Thank you. 01:06:25.680 --> 01:06:31.320 [Applause] 01:06:34.660 --> 01:06:36.520 - Mute yourself. - Ah. [laughs] 01:06:36.520 --> 01:06:38.359 - Before you say anything too embarrassing. 01:06:38.359 --> 01:06:40.640 - [laughs] Yeah, right.