WEBVTT Kind: captions Language: en-US 00:00:04.250 --> 00:00:06.210 Okay, everyone. We’re going to get started. 00:00:06.210 --> 00:00:10.260 Thanks for coming to this week’s Earthquake Science Seminar. 00:00:10.260 --> 00:00:14.840 Next week, there will be no seminar because of the SCEC meeting. 00:00:14.840 --> 00:00:18.680 And we’ll resume our normally scheduled seminars the week after – 00:00:18.680 --> 00:00:22.440 September 18th – with Valerie Sahakian. 00:00:22.440 --> 00:00:24.439 And with that, I’ll hand it over to 00:00:24.440 --> 00:00:28.520 Nick Beeler to introduce today’s speaker, Arjun Kohli. 00:00:30.520 --> 00:00:32.440 - Okay, so we’re really lucky to have Arjun here. 00:00:32.449 --> 00:00:39.540 So he is a – I don’t know – he’s kind of a brilliant rock mechanics kind of guy. 00:00:39.540 --> 00:00:43.400 And he started out as – here. He grew up here. 00:00:43.400 --> 00:00:49.570 But he went to undergraduate at Brown and did his senior thesis with 00:00:49.570 --> 00:00:53.870 Greg Hirth and various other rock mechanics people – Dave Goldsby, 00:00:53.870 --> 00:00:57.840 Terry Tullis – doing high-speed friction. So flash weakening of 00:00:57.840 --> 00:01:02.960 serpentinite. And that was eventually published in the JGR. 00:01:02.960 --> 00:01:08.580 Then he came back out here to Stanford, and we managed to waylay him briefly. 00:01:08.580 --> 00:01:18.070 He did a project on creeping faults in David’s lab using oscillating stress 00:01:18.070 --> 00:01:23.490 to measure essentially the elastic and inelastic rheology of the material 00:01:23.490 --> 00:01:28.509 with an in situ, in pressure vessel, displacement sensor. 00:01:28.509 --> 00:01:33.720 Just absolutely phenomenal work that he hasn’t published yet. [laughter] 00:01:33.720 --> 00:01:36.960 And then he went back – of course, he was a student at Stanford. 00:01:36.960 --> 00:01:39.259 He got his Ph.D. there. He worked with Mark Zoback. 00:01:39.259 --> 00:01:43.750 Ran the lab there. Then he did a postdoc at SLAC doing 00:01:43.750 --> 00:01:48.950 basically imaging – so in situ high- energy imaging of deforming rocks. 00:01:48.950 --> 00:01:52.880 He built pressure vessels. And now he’s back at Stanford as a – as a 00:01:52.880 --> 00:01:57.790 researcher and still in rock mechanics. And he’s doing earthquake work again. 00:01:57.790 --> 00:02:01.440 And so, when he was at Stanford, he also worked with Jessica Warren, 00:02:01.440 --> 00:02:05.640 who is a deep – I guess a deep rock deformation person, 00:02:05.640 --> 00:02:09.860 as opposed to myself, who is a very shallow rock deformation person. 00:02:09.860 --> 00:02:15.160 And so this is work that comes, I guess, from that seed. 00:02:15.160 --> 00:02:20.800 And Jessica is no longer at Stanford, but Arjun’s back on the case. 00:02:26.700 --> 00:02:31.980 - Okay. Thanks, Nick, for that lovely introduction. 00:02:32.920 --> 00:02:37.060 So I’m going to be talking today about oceanic transform faults, 00:02:37.069 --> 00:02:41.129 and in particular, looking at deep seawater circulation. 00:02:41.129 --> 00:02:43.910 And you might be thinking, sort of, what does it have to 00:02:43.910 --> 00:02:48.129 do with earthquakes? Of course, this is the earthquake seminar. 00:02:48.129 --> 00:02:50.230 And really, what we’re thinking about here is, you know, 00:02:50.230 --> 00:02:54.870 our ocean basins are these huge bathtubs filled with seawater. 00:02:54.870 --> 00:02:57.930 And we know from looking at bathymetry and mid-ocean ridges 00:02:57.930 --> 00:03:01.190 that there are cracks in the bottom of the bathtub. 00:03:01.190 --> 00:03:03.769 And those cracks are, in fact, faults that are moving. 00:03:03.769 --> 00:03:08.159 And so sort of a central question, which came about in this research, 00:03:08.159 --> 00:03:12.030 a bit indirectly is, you know, how far does seawater percolate 00:03:12.030 --> 00:03:16.850 through those faults into the crust, and maybe into the mantle? 00:03:16.850 --> 00:03:19.980 And so we’ll take a look at that today. 00:03:19.980 --> 00:03:24.660 And so, initially, our understanding of oceanic transform faults really came 00:03:24.660 --> 00:03:28.459 from looking at earthquakes that have been located teleseismically. 00:03:28.459 --> 00:03:32.209 And looking at sort of global studies of earthquake occurrence. 00:03:32.209 --> 00:03:37.130 And, since transform faults have a variety of different slip rates 00:03:37.130 --> 00:03:42.370 due to different spreading rates around the global transform system, 00:03:42.370 --> 00:03:46.140 it’s very useful to look at these things just as a function of temperature 00:03:46.140 --> 00:03:49.590 instead of depth. Because, of course, temperature as a function of depth 00:03:49.590 --> 00:03:54.290 varies pretty strongly as you move around different mid-ocean ridges. 00:03:54.290 --> 00:03:58.700 So here’s some early work from the ‘80s, and later the 2000s, 00:03:58.700 --> 00:04:02.349 in which earthquakes were located at depth and then 00:04:02.349 --> 00:04:05.900 compared to numerical thermal models of the fault. 00:04:05.900 --> 00:04:11.989 And what you can see approximately is this 600-degree isotherm right here 00:04:11.989 --> 00:04:16.440 becomes sort of a lower limit, or a lower bound, 00:04:16.440 --> 00:04:18.520 for where seismicity was seen. 00:04:18.530 --> 00:04:21.030 And this was sort of kind of the state of the science 00:04:21.030 --> 00:04:26.250 for 20, 30 years until quite recently. 00:04:26.250 --> 00:04:30.650 And so recently, there have been ocean bottom seismic deployments. 00:04:30.650 --> 00:04:34.500 So seismometers on the ocean floor that have been deployed at 00:04:34.500 --> 00:04:38.780 various transform faults. And this is some work by Emily Roland 00:04:38.780 --> 00:04:43.380 and Jeff McGuire and others who, in 2007, 2008, went out and actually 00:04:43.380 --> 00:04:48.820 deployed a series of seismometers on this transform on the East Pacific Rise. 00:04:48.820 --> 00:04:51.250 This is the Gofar Transform Fault. 00:04:51.250 --> 00:04:55.570 And what they found was really quite interesting. 00:04:55.570 --> 00:05:00.320 Instead of the earthquakes being limited to about 600 C, you can see in this – 00:05:00.320 --> 00:05:03.970 in this thermal model down here in which the earthquakes are in red 00:05:03.970 --> 00:05:09.630 and yellow, that they in fact extend quite deep on the fault in terms of 00:05:09.630 --> 00:05:14.820 the temperature structure. Maybe even greater than 1,000 degrees C. 00:05:14.820 --> 00:05:19.100 And so this was a really quite surprising result. 00:05:19.100 --> 00:05:24.560 And it sort of shook up our thinking a little bit because we know that faults are 00:05:24.560 --> 00:05:30.510 often conduits for fluid flow here [audio static] certainly in continental settings. 00:05:30.510 --> 00:05:34.120 And so the question is that [audio back to normal], you know, if we have 00:05:34.120 --> 00:05:37.919 earthquakes going all the way down to 1,000 degrees, is there also 00:05:37.919 --> 00:05:42.020 seawater percolating through a network of fractures and cracks 00:05:42.020 --> 00:05:47.389 down to those depths, and maybe impacting the behavior of the fault? 00:05:47.389 --> 00:05:50.490 And this really sort of becomes a question that kind of 00:05:50.490 --> 00:05:55.449 feeds back on itself. On the right is just a depiction of sort of 00:05:55.449 --> 00:06:02.139 the strength, depth, and then the thermal conductivity profile of those faults. 00:06:02.139 --> 00:06:07.240 And the region in blue right there is sort of the extent of brittle failure 00:06:07.240 --> 00:06:09.240 on the fault. So you would think 00:06:09.240 --> 00:06:13.640 that corresponds to, you know, the depth extent of earthquakes. 00:06:13.640 --> 00:06:18.320 And the blue curve on the right, again, is the thermal conductivity, which, in 00:06:18.330 --> 00:06:24.639 the shallow crust, is increased due to the effect of hydrothermal circulation. 00:06:24.639 --> 00:06:28.910 And one thing you can see in either of these two scenarios – what they 00:06:28.910 --> 00:06:34.980 have in common is that the depth of hydrothermal cooling is really the depth 00:06:34.980 --> 00:06:38.910 of the brittle-ductile transition, where you have frictional slip 00:06:38.910 --> 00:06:44.100 in the shallow crust and then some ductile flow law due to creep 00:06:44.100 --> 00:06:47.300 at high pressures and temperatures. And the intersection of those 00:06:47.300 --> 00:06:52.110 really determines where these authors sort of ended the 00:06:52.110 --> 00:06:54.389 depth of hydrothermal circulation. 00:06:54.389 --> 00:07:01.170 So there’s really sort of a mismatch here that we maybe want to look at. 00:07:01.170 --> 00:07:05.630 And one way to do that is to look at olivine data from the lab. 00:07:05.630 --> 00:07:09.660 And so olivine, of course, is the mineral that really defines the strength 00:07:09.660 --> 00:07:13.120 and properties of the oceanic crust and the mantle. 00:07:13.120 --> 00:07:16.500 And what we really want to think of when we’re thinking of 00:07:16.500 --> 00:07:21.220 slip on faults is two rough surfaces just sliding together. 00:07:21.220 --> 00:07:25.300 And the area of contact on those rough surfaces won’t necessarily be 00:07:25.310 --> 00:07:29.120 the nominal area of the two surfaces, but it’ll actually be made of these 00:07:29.120 --> 00:07:34.440 small asperity contacts, in which stress is actually concentrated. 00:07:34.440 --> 00:07:39.099 And Dieterich and Kilgore did some nice work earlier describing what 00:07:39.099 --> 00:07:44.630 sort of this real area of contact – what I’m calling A-r, actually is. 00:07:44.630 --> 00:07:49.910 And you can think of it as actually the ratio of that real area of contact to the 00:07:49.910 --> 00:07:55.460 nominal area is related to the applied normal stress compared to the 00:07:55.460 --> 00:08:00.370 yield stress – how much it actually takes to – how much stress it actually 00:08:00.370 --> 00:08:06.520 takes to cause these contacts to flow rather than slip frictionally. 00:08:06.520 --> 00:08:10.050 And so Margaret Boettcher did some interesting work in which 00:08:10.050 --> 00:08:13.419 she deformed olivine at various pressures and temperatures and 00:08:13.419 --> 00:08:18.280 actually looked at the relationship of this asperity stress to temperature. 00:08:18.280 --> 00:08:20.169 And what she found was quite interesting. 00:08:20.169 --> 00:08:26.250 So, at fairly high strain rates in the lab – this is 10 to the minus 4 per second 00:08:26.250 --> 00:08:30.530 to 10 to the minus 7 – she found that the yield strength of these asperities 00:08:30.530 --> 00:08:36.160 was somewhere in the neighborhood of about 1,200 to 800 MPa. 00:08:36.160 --> 00:08:42.120 And so this really corresponded to a temperature of 1,000 degrees C. 00:08:42.120 --> 00:08:47.770 So this says that, in the lab, at these high strain rates, at 1,000 degrees C, 00:08:47.770 --> 00:08:53.550 you can actually push the rock – in this case olivine – from frictional slip 00:08:53.550 --> 00:08:57.580 to a more ductile to a more plastic deformation response. 00:08:57.580 --> 00:09:01.290 And so, in order to extrapolate that constraint to the Earth, she just 00:09:01.290 --> 00:09:05.990 looked at strain rates that are more representative of intraplate oceanic 00:09:05.990 --> 00:09:10.820 settings and found that, okay, that point turns out to be 00:09:10.820 --> 00:09:15.350 somewhere around 600 degrees C. So, again, there’s this indication 00:09:15.350 --> 00:09:19.700 that the brittle-ductile transition for olivine, in this case, 00:09:19.700 --> 00:09:22.620 is somewhere around that 600 degree value. 00:09:22.620 --> 00:09:27.310 But, again, you know, we see these earthquakes that extend 00:09:27.310 --> 00:09:29.120 much deeper in model thermal structures. 00:09:29.120 --> 00:09:33.520 So there’s a bit of a conflict here to resolve. 00:09:33.520 --> 00:09:38.480 And the way we sort of want to do that is by actually looking at rocks 00:09:38.480 --> 00:09:42.200 that come from these transform faults. And so this is just sort of 00:09:42.200 --> 00:09:46.029 a global picture. And I’m going to zoom in on 00:09:46.029 --> 00:09:50.450 maybe six of these transforms from [volume too low to hear] ridges. 00:09:50.450 --> 00:09:56.480 So the first ones we’ll look at are samples from this very slow 00:09:56.480 --> 00:10:03.740 southwest [audio static] [inaudible] … 00:10:03.740 --> 00:10:05.680 … both of these are [volume very low] 00:10:05.680 --> 00:10:11.860 ultra-slow-slipping transforms that are on the order of [inaudible] … 00:10:13.480 --> 00:10:14.760 Sorry. 00:10:14.760 --> 00:10:22.940 [Silence] 00:10:23.540 --> 00:10:25.060 Testing. 00:10:25.060 --> 00:10:27.340 - Can you put it on the top? 00:10:34.600 --> 00:10:37.020 - Okay. - I think it might just be [inaudible]. 00:10:37.020 --> 00:10:39.200 - Oh, no worries. 00:10:40.820 --> 00:10:45.060 So just looking at this ultra-slow- slipping Southwest Indian Ridge, 00:10:45.079 --> 00:10:49.769 there are two transforms here. And, from both of these transforms, 00:10:49.769 --> 00:10:53.510 highly deformed rocks have been recovered – what we’re calling 00:10:53.510 --> 00:10:56.620 peridotite mylonites. So these are olivine mylonites 00:10:56.620 --> 00:10:59.740 that have been picked up from the surface – or, rocks that have 00:10:59.740 --> 00:11:05.120 been deformed ductilely at high pressures and temperatures 00:11:05.120 --> 00:11:07.580 on these faults. 00:11:07.580 --> 00:11:10.900 If we look at the Mid-Atlantic Ridge, there’s a couple interesting faults here. 00:11:10.900 --> 00:11:17.260 Vema is a quite large transform fault from which mylonites 00:11:17.260 --> 00:11:19.950 have again been recovered by dredging. 00:11:19.950 --> 00:11:24.240 And there’s also actually an island exposure at St. Peter St. Paul’s Rocks, 00:11:24.240 --> 00:11:28.800 in which a piece of deformed transform fault actually sticks out of the ocean’s 00:11:28.800 --> 00:11:35.000 surface onto an island that’s, you know, hundreds of meters in diameter. 00:11:35.000 --> 00:11:39.269 And actually, mylonites have also been recovered from there. 00:11:39.269 --> 00:11:44.411 And the last class are these extremely fast-slipping transform faults on the 00:11:44.420 --> 00:11:49.780 order of hundreds of millimeters per year from the East Pacific Rise. 00:11:49.780 --> 00:11:54.480 And mylonites have again been recovered from Garrett Transform Fault. 00:11:54.490 --> 00:11:58.280 And actually, Gofar Transform Fault, which we looked at earlier, that had 00:11:58.280 --> 00:12:02.870 the OBS deployment, there’s actually a cruise planned for this November 00:12:02.870 --> 00:12:06.810 to actually dredge that fault to recover rocks from there. 00:12:06.810 --> 00:12:11.850 So you can see that, you know, across the wide variety of slip rates, we can 00:12:11.850 --> 00:12:17.209 recover deformed rocks from these transforms at the bottom of the ocean. 00:12:17.209 --> 00:12:22.920 And so, for this talk, I’ll specifically focus on this Shaka Transform Fault, 00:12:22.920 --> 00:12:29.950 which is, again, right off the coast of Africa on the Southwest Indian Ridge. 00:12:29.950 --> 00:12:34.470 This is just a bathymetry map of the fault plotted here with 00:12:34.470 --> 00:12:38.510 historical earthquakes dating back to 1950. 00:12:38.510 --> 00:12:42.589 As you can see from the rose diagram on the right, most of the seismicity 00:12:42.589 --> 00:12:48.160 here is right-lateral transform faults as well as some normal faults, 00:12:48.160 --> 00:12:51.600 which are on the ridge boundaries. 00:12:52.200 --> 00:12:57.160 And the star symbols and triangle symbols denote the location of dredges 00:12:57.160 --> 00:13:02.130 in which basically a metal basket has been dragged across the ocean floor 00:13:02.130 --> 00:13:06.900 to collect whatever rocks are sitting in the transform valley. 00:13:06.900 --> 00:13:10.930 So we’ll spend a bit of time taking a look at these rocks and figuring out 00:13:10.930 --> 00:13:14.660 what they might tell us about the interactions with seawater 00:13:14.660 --> 00:13:17.440 and the deformation properties of the fault. 00:13:18.760 --> 00:13:23.519 So, on the left here is what we consider a peridotite protolith. 00:13:23.519 --> 00:13:29.610 It’s a fairly coarse-grained rock, 1- to 10- millimeter grain size, 00:13:29.610 --> 00:13:35.380 made up of mostly olivine and also clinopyroxene and orthopyroxene. 00:13:35.380 --> 00:13:38.670 On the right is the type of deformed rocks that we actually recover 00:13:38.670 --> 00:13:43.100 from transform faults. And so we’ll call these peridotite mylonites. 00:13:43.100 --> 00:13:46.460 And they’re really made of an extremely fine-grain matrix, 00:13:46.460 --> 00:13:51.510 maybe 1 to 10 microns in grain size, sitting amidst a range of porphyroclasts, 00:13:51.510 --> 00:13:54.720 which are basically flowing inside that matrix. 00:13:54.720 --> 00:13:58.810 And so Jessica Warren and Greg Hirth first looked at these rocks back in 00:13:58.810 --> 00:14:03.420 2006 and actually found some very interesting components 00:14:03.420 --> 00:14:07.290 of their deformation behavior. And we’ll zoom in a little bit 00:14:07.290 --> 00:14:11.300 on that as I revisited these rocks for this study. 00:14:12.720 --> 00:14:17.700 So these are three examples of mylonites that we find from 00:14:17.709 --> 00:14:21.430 the Shaka Transform Fault. The first on the left is what 00:14:21.430 --> 00:14:25.089 we’ll call an ultra-mylonite. It’s the most deformed, 00:14:25.089 --> 00:14:28.610 has the most strain. The grain size is extremely small – 00:14:28.610 --> 00:14:31.689 almost down to 1 micron in size. 00:14:31.689 --> 00:14:36.640 And it basically has this anastomosing structure of shear zones 00:14:36.640 --> 00:14:40.320 amidst these larger porphyroclasts. 00:14:40.329 --> 00:14:43.279 In the middle is something we’ll just call a mylonite. 00:14:43.279 --> 00:14:47.510 It has, you know, a moderate amount of grain size reduction. 00:14:47.510 --> 00:14:51.240 Also has some porphyroclasts. And interestingly, it has this interesting 00:14:51.240 --> 00:14:58.050 behavior where we see orthopyroxene deformed in sort of these domino-style 00:14:58.050 --> 00:15:02.880 normal faults undergoing cataclastic flow while the olivine 00:15:02.880 --> 00:15:05.860 around it actually flows ductilely. 00:15:05.860 --> 00:15:09.510 And that’s something we’ll revisit in a minute. 00:15:09.510 --> 00:15:13.940 And then finally, we see these things that we called the altered mylonites. 00:15:13.940 --> 00:15:18.339 And what these are are much less grain size reduction and a lot more 00:15:18.340 --> 00:15:21.480 fracturing and serpentinization. 00:15:22.160 --> 00:15:26.110 And so the initial ductile microstructure has been significantly 00:15:26.110 --> 00:15:34.540 overprinted by both fracturing and also interaction with water. 00:15:34.540 --> 00:15:40.070 And so the focus on this will be to look at the structure of these rocks to, 00:15:40.070 --> 00:15:43.880 one, pull out information about their deformation behavior, and two, 00:15:43.880 --> 00:15:48.949 figure out what the nature of their interaction with fluids are. 00:15:48.949 --> 00:15:52.300 And so the first evidence for hydration of these rocks actually 00:15:52.300 --> 00:15:57.040 comes from looking inside a individual large grain. 00:15:57.040 --> 00:16:06.110 So, on the left is a view of shear bands, basically, in this mylonite sample. 00:16:06.110 --> 00:16:10.209 And what we’re looking at is a large relic grain that sort of got trapped in 00:16:10.209 --> 00:16:13.740 the middle of the shear bands. It didn’t reduce with the rest of the 00:16:13.740 --> 00:16:20.980 grain size and actually maintains these healed fractures that have little beads 00:16:20.980 --> 00:16:26.010 of fluid inclusions throughout them. And these are subparallel planes that 00:16:26.010 --> 00:16:29.900 actually go through the grain, basically subvertically. 00:16:29.900 --> 00:16:35.269 And they contained actually trapped fluids that were in fractures that formed 00:16:35.269 --> 00:16:39.610 and then healed very rapidly, holding the fluids inside. 00:16:39.610 --> 00:16:43.579 Here’s another example from a larger porphyroclast, this time 00:16:43.579 --> 00:16:47.029 from the altered mylonite. And here you can really see that, 00:16:47.029 --> 00:16:51.490 in this grain that’s, you know, heavily fractured, that there are these 00:16:51.490 --> 00:16:58.740 sort of curtains of these fluid inclusions, which we really interpret as fractures 00:16:58.740 --> 00:17:02.970 that formed, were filled with fluids, and then healed extremely rapidly 00:17:02.970 --> 00:17:09.600 to trap those fluids as actually individual bubbles inside the grain. 00:17:09.600 --> 00:17:15.560 And this is something that’s been seen in continental settings for quite a while. 00:17:15.560 --> 00:17:20.700 On the left is some work by Tom Mitchell and others looking at 00:17:20.700 --> 00:17:25.740 a transform fault zone in Chile, and they’re actually looking at 00:17:25.740 --> 00:17:29.880 some quartz mylonites in which they see very similar behavior. 00:17:29.880 --> 00:17:38.240 So on the top is a view of rock microstructure in which the slip on the 00:17:38.240 --> 00:17:43.150 transform fault is in this orientation. And they actually see this large pattern 00:17:43.150 --> 00:17:48.620 of fluid inclusions opening up at a high angle to the fault over there. 00:17:48.620 --> 00:17:53.860 And they sort of related the formation of these fluid inclusions to actually 00:17:53.860 --> 00:17:56.880 rupture propagation on the transform fault. 00:17:56.880 --> 00:18:00.920 So on the right here is a numerical model from Di Toro and others. 00:18:00.920 --> 00:18:05.160 And they’re basically looking at the dominant stress fields, both the 00:18:05.160 --> 00:18:09.559 direction and the magnitude – tension versus compression – 00:18:09.559 --> 00:18:13.750 of the stress fields due to a propagating rupture on a strike-slip fault. 00:18:13.750 --> 00:18:17.820 And what they found, interestingly, is that, near the tip, there’s actually 00:18:17.820 --> 00:18:24.210 this zone of quite large tensional forces, which are oriented almost normal 00:18:24.210 --> 00:18:28.299 to the strike of the fault. And so this is a potential mechanism 00:18:28.299 --> 00:18:32.510 by which, during earthquake slip, you can get these tensional stresses 00:18:32.510 --> 00:18:37.070 that possibly open up mode-I cracks and allow fluids to percolate and 00:18:37.070 --> 00:18:41.200 then eventually heal into these fluid inclusions. 00:18:41.200 --> 00:18:45.980 And this is sort of particularly relevant to what we’re doing here. 00:18:45.980 --> 00:18:49.870 Because, although these are samples that are dredged from the ocean floor, 00:18:49.870 --> 00:18:55.080 you know, we have no geographic way to relate them to any fault zone 00:18:55.080 --> 00:18:58.059 structures on the ocean floor or anything. 00:18:58.059 --> 00:19:03.530 We can sort of make, you know, a broad assumption that the foliation 00:19:03.530 --> 00:19:08.809 here, which is horizontal in each of these end section views, is actually – 00:19:08.809 --> 00:19:11.940 represents sort of a map view of the transform fault. 00:19:11.940 --> 00:19:17.070 So, if you look on the image on the left, if we were to sort of look down on it 00:19:17.070 --> 00:19:22.570 from the – from the Earth’s surface perspective, that would sort of be the 00:19:22.570 --> 00:19:26.120 foliation that’s parallel to the strike of the transform fault. 00:19:26.120 --> 00:19:31.990 So, if we view the fluid inclusion orientations in that reference frame, 00:19:31.990 --> 00:19:37.350 these fluid inclusions really, we find, are almost normal to the foliation 00:19:37.350 --> 00:19:42.610 in each case. So we think it’s potentially likely that this mechanism 00:19:42.610 --> 00:19:46.640 of earthquake rupture propagation actually popped open these tensional 00:19:46.640 --> 00:19:52.990 cracks, allowed fluids to flow in, where they later healed at higher temperature, 00:19:52.990 --> 00:19:56.480 eventually forming these fluid inclusion planes. So that’s sort of 00:19:56.480 --> 00:20:01.262 the first evidence for hydration. And we already see that there is some 00:20:01.262 --> 00:20:06.140 evidence that hydration is connected to both brittle and ductile processes. 00:20:07.190 --> 00:20:10.280 The next evidence we have for hydration, there were actually hydrous 00:20:10.289 --> 00:20:15.020 minerals that are in the deformation structure of these rocks. 00:20:15.020 --> 00:20:20.539 So this is returning back to, again, one of these orthopyroxene porphyroclasts, 00:20:20.540 --> 00:20:22.880 which is actually deforming by cataclasis. 00:20:22.880 --> 00:20:26.660 It’s deforming by this grain-scale faulting process. 00:20:26.669 --> 00:20:30.670 And what’s really interesting is that, at the same time that’s happening, 00:20:30.670 --> 00:20:34.650 there’s actually fine-grain minerals from the matrix being entrained 00:20:34.650 --> 00:20:39.720 in this fault and actually are being deformed sort of within the boundaries 00:20:39.720 --> 00:20:44.540 of that grain. It’s sort of a fault zone within a fault zone. 00:20:45.440 --> 00:20:49.280 And what’s really interesting about that is it’s not only olivine that’s 00:20:49.290 --> 00:20:54.539 being deformed within this. It’s also tremolite, which is a hydrous 00:20:54.540 --> 00:21:00.240 calcium-rich amphibole mineral. And it’s particularly interesting because 00:21:00.240 --> 00:21:05.400 it’s not only fluids being emplaced into the rock post-deformation. 00:21:05.419 --> 00:21:09.610 We can actually see it’s happening right as olivine is deforming ductilely. 00:21:09.610 --> 00:21:16.060 So, again, there’s this suggestion that whatever the fluid process is happening, 00:21:16.060 --> 00:21:19.520 it’s occurring under both brittle and ductile conditions. 00:21:20.620 --> 00:21:25.380 What’s interesting about amphibole also is some work by folks in 00:21:25.390 --> 00:21:29.440 Jessica’s group actually showed that the composition of them really indicates 00:21:29.440 --> 00:21:34.780 that the fluid source is seawater. So, on the left is a plot of chlorine 00:21:34.780 --> 00:21:41.640 versus silica, which, for this mineral system, actually represents sort of 00:21:41.650 --> 00:21:46.130 a temperature-dependent system. So, over to the right here – or, sorry, 00:21:46.130 --> 00:21:50.730 over to the left is relatively low temperature amphiboles. 00:21:50.730 --> 00:21:54.220 And towards the right is relatively high-temperature amphiboles. 00:21:54.220 --> 00:21:58.570 And what they saw is, when they plotted both amphiboles inside those 00:21:58.570 --> 00:22:04.450 grain-scale faults and inside the matrix is that the composition actually is much 00:22:04.450 --> 00:22:09.980 more similar to that of seawater than the depleted mantle. 00:22:09.980 --> 00:22:14.470 So this was already some indication that the fluid source deformed these 00:22:14.470 --> 00:22:17.980 amphiboles that actually, you know, entrained in these rocks 00:22:17.980 --> 00:22:22.100 and was involved in the ductile deformation process, 00:22:22.100 --> 00:22:25.640 was actually coming from the ocean. 00:22:26.610 --> 00:22:31.660 And when they looked at sort of this temperature dependence that was in the 00:22:31.660 --> 00:22:38.340 data, they found basically using this orthopyroxene thermometer that the 00:22:38.340 --> 00:22:41.410 temperature at which these amphiboles formed was somewhere in the 00:22:41.410 --> 00:22:48.440 range of 650 to 850 degrees C. So, again, if we think back to the 00:22:48.440 --> 00:22:52.600 earthquakes we were looking at somewhere in the neighborhood of 600, 00:22:52.610 --> 00:22:56.740 maybe even up to 1,000 degrees C, this is already some indication 00:22:56.740 --> 00:23:02.100 that there’s seawater actually on these transform faults quite deep 00:23:02.100 --> 00:23:05.659 in their temperature structure. And this is something we’ll look at 00:23:05.659 --> 00:23:10.809 a little bit more in terms of rheological constraints, but I think it’s important 00:23:10.809 --> 00:23:14.460 to sort of put these geochemical constraints first. 00:23:15.480 --> 00:23:19.230 The last evidence for hydration in these rocks that we have is, of course, 00:23:19.230 --> 00:23:23.610 serpentinization, which is, you know, quite an abundant feature throughout 00:23:23.610 --> 00:23:28.470 the entire oceanic lithosphere and in most ultramafic rocks. 00:23:28.470 --> 00:23:33.350 And so, on the left here is, again, one of – a micrograph of the 00:23:33.350 --> 00:23:37.660 ultra-mylonite sample in which we have, you know, a relatively 00:23:37.660 --> 00:23:40.640 coarse-grained area on the top of the image and a relatively 00:23:40.640 --> 00:23:44.320 fine-grained area on the bottom. And, zooming into the relatively 00:23:44.320 --> 00:23:46.970 coarse-grained area, you can see that it’s sort of 00:23:46.970 --> 00:23:52.060 fingered and entrained by all these serpentine veins. 00:23:52.060 --> 00:23:56.820 What’s interesting about that is they seem to sort of selectively propagate 00:23:56.830 --> 00:24:00.309 through the relatively coarse-grained regions of the rock. 00:24:00.309 --> 00:24:03.260 And this is an idea we’ll return to later when we think a little bit 00:24:03.260 --> 00:24:06.559 about the rheology. But right now, what we really want 00:24:06.559 --> 00:24:13.820 to focus on is what actually caused this massive grain size reduction and 00:24:13.820 --> 00:24:17.340 these interesting textures that we see? You know, what’s sort of the difference 00:24:17.340 --> 00:24:22.680 between these coarse-grained regions up here and these extremely 00:24:22.680 --> 00:24:27.260 fine-grained regions? Or, said differently, you know, 00:24:27.260 --> 00:24:31.860 what’s sort of the driving force for strain localization in these rocks? 00:24:34.200 --> 00:24:38.380 One way to sort of do that is to first constrain the pressure/temperature 00:24:38.380 --> 00:24:41.730 conditions of that strain localization process. 00:24:41.730 --> 00:24:44.750 And so some early work on these mylonites was actually done 00:24:44.750 --> 00:24:48.360 by Jaroslow et al. And what they did was used various 00:24:48.360 --> 00:24:52.730 mineral thermometers on different components of the microstructure 00:24:52.730 --> 00:24:56.890 to actually bound the conditions under which ductile deformation 00:24:56.890 --> 00:24:58.580 was taking place. 00:24:58.580 --> 00:25:03.600 And so they looked at neoblasts, or new small recrystallized grains in the 00:25:03.600 --> 00:25:09.059 mylonites, as well as porphyroclasts in the mylonites, and found this range 00:25:09.059 --> 00:25:14.230 of about 600 to 700 degrees C, which they interpreted as sort of the closure 00:25:14.230 --> 00:25:19.789 temperature for ductile deformation. And so we sort of took that 00:25:19.789 --> 00:25:25.210 and combined it with several mineralogical stability fields 00:25:25.210 --> 00:25:28.840 and the geotherm for the Shaka Transform Fault. 00:25:28.840 --> 00:25:33.120 Here’s the geotherm for the much faster-slipping, hotter Gofar. 00:25:33.130 --> 00:25:36.380 And, just looking in pressure/ temperature space, we can 00:25:36.380 --> 00:25:41.900 look at two main constraints for sort of the conditions of strain localization. 00:25:42.620 --> 00:25:46.960 The upper temperature constraint is given by the fact that we know 00:25:46.970 --> 00:25:50.130 tremolite was present in the ductile microstructure when 00:25:50.130 --> 00:25:52.880 things were deforming. So we’ll assume that the 00:25:52.880 --> 00:25:57.470 high-temperature limit for tremolite is sort of the high point 00:25:57.470 --> 00:26:01.340 at which this strain localization process took place. 00:26:01.340 --> 00:26:06.170 The lower limit, we’ll take as the lower limit of this thermometer, 00:26:06.170 --> 00:26:08.600 which was defined by Jaroslow et al. 00:26:08.600 --> 00:26:14.679 And so this sort of gives this range of maybe 750 degrees C as sort of 00:26:14.679 --> 00:26:18.429 the temperature at which the mylonites were forming. 00:26:18.429 --> 00:26:21.370 And this will become really important as we move forward. 00:26:21.370 --> 00:26:25.549 Because we’re going to use this to base rheological modeling on 00:26:25.549 --> 00:26:30.179 in order to come back to the question of deep seawater circulation. 00:26:31.300 --> 00:26:35.700 So, now that we sort of set up the conditions for strain localization, 00:26:35.700 --> 00:26:39.740 it’s sort of important to understand what the mechanisms are. 00:26:39.740 --> 00:26:43.900 And so we did some work recently looking at actually 00:26:43.900 --> 00:26:46.930 a profile throughout the mylonite microstructure. 00:26:46.930 --> 00:26:52.260 So this is an electron backscatter diffraction map showing mineral 00:26:52.260 --> 00:26:57.039 phases as the color and also detecting crystallographic orientation 00:26:57.039 --> 00:26:59.900 to look at the mineral fabric. 00:26:59.900 --> 00:27:03.490 And so what you can see, as we’re going through these alternating zones 00:27:03.490 --> 00:27:08.600 of relatively coarse-grain olivine, which is in blue, and relatively 00:27:08.600 --> 00:27:13.610 fine-grain olivine that’s accompanied by tremolite, which is this cyan color, 00:27:13.610 --> 00:27:16.890 as well as orthopyroxene, which is in green. 00:27:16.890 --> 00:27:20.470 And then we transition back to one of these coarse-grain regions. 00:27:20.470 --> 00:27:24.600 So this is, again, equivalent to sort of moving across a profile 00:27:24.600 --> 00:27:28.730 like this in the mylonite microstructure. 00:27:28.730 --> 00:27:31.880 And so, as we move across that profile, we can see a couple interesting things. 00:27:31.880 --> 00:27:38.720 First, if we look at just the mineral fractions, and we look at the secondary 00:27:38.730 --> 00:27:43.110 mineral fractions – so tremolite and orthopyroxene in this case, we see 00:27:43.110 --> 00:27:49.370 some spikes, obviously, in this fine- grain region and this one down here. 00:27:49.370 --> 00:27:54.320 And those seem to be inversely correlated with the olivine grain size 00:27:54.320 --> 00:27:58.170 in each case. So when there’s a high amount of tremolite, 00:27:58.170 --> 00:28:01.220 the olivine grain size is actually significantly reduced compared to 00:28:01.220 --> 00:28:06.280 when you have none at all. And so that’s a fairly strong effect. 00:28:06.280 --> 00:28:10.090 The other thing that we see that’s interesting is actually the fabric 00:28:10.090 --> 00:28:14.159 strength of all the – the degree of crystallographic orientation, 00:28:14.159 --> 00:28:17.940 or crystallographic alignment, is actually variable across 00:28:17.940 --> 00:28:22.660 these two regions. So in the region of relatively 00:28:22.660 --> 00:28:27.580 coarse-grain pure olivine, we see fairly high fabric strength, 00:28:27.590 --> 00:28:31.730 which is quantified by this M-index here, which goes from zero to 1, 00:28:31.730 --> 00:28:36.700 1 being a perfect single crystal, and zero being completely random. 00:28:37.240 --> 00:28:42.080 And, in the regions of fairly fine-grained olivine, what we see 00:28:42.090 --> 00:28:46.890 is a very low fabric strength. And this is sort of represented by 00:28:46.890 --> 00:28:50.480 these pole figures on the right, which are contoured by multiples 00:28:50.480 --> 00:28:56.180 of a uniform distribution. So, again, 1 being completely random. 00:28:56.180 --> 00:29:00.740 So, if we just look at relatively coarse-grained olivine, we see 00:29:00.740 --> 00:29:04.289 a pretty well-defined fabric here. This is olivine greater than 00:29:04.289 --> 00:29:09.350 10 microns in size. But if we look at the – at the fine-grain stuff, 00:29:09.350 --> 00:29:13.679 the fabric is completely different and has been totally erased from 00:29:13.679 --> 00:29:18.990 that initial coarse-grain fabric, which we – which we saw. 00:29:18.990 --> 00:29:22.409 And this is quite fascinating because, you know, the scale over here 00:29:22.409 --> 00:29:26.270 is hundreds of microns. So these are, you know, grains which are right next to 00:29:26.270 --> 00:29:31.390 each other and are sort of deforming by completely different mechanisms. 00:29:31.390 --> 00:29:33.049 This is the pole figure for tremolite. 00:29:33.049 --> 00:29:36.560 I won’t really get into that at the moment. 00:29:36.560 --> 00:29:41.330 But what’s interesting here now is to really consider the different 00:29:41.330 --> 00:29:46.330 mechanisms of olivine deformation. So I’ll go through a quick primer of 00:29:46.330 --> 00:29:50.080 this stuff just to give some highlights. So there are a couple important 00:29:50.080 --> 00:29:52.979 mechanisms involving deformation to know. 00:29:52.979 --> 00:29:57.200 The first is dislocation creep. And this is really crystal-plastic 00:29:57.200 --> 00:30:01.880 deformation that’s accommodated by the movements of dislocations, or extra 00:30:01.880 --> 00:30:06.360 half-planes, inside the crystal structure. It’s temperature- and stress-dependent. 00:30:06.360 --> 00:30:10.080 And the important thing is that it tends to form a strong lattice 00:30:10.090 --> 00:30:13.110 preferred orientation because the dislocation movements 00:30:13.110 --> 00:30:17.320 are actually associated with the crystal structure itself. 00:30:18.300 --> 00:30:21.600 Generally, it’s also favored at a large grain size, and it is, 00:30:21.610 --> 00:30:24.590 in fact, insensitive to grain size. And that would be something 00:30:24.590 --> 00:30:28.270 very important in the rheological modeling. 00:30:28.270 --> 00:30:31.870 The next mechanism to think about is grain boundary sliding. 00:30:31.870 --> 00:30:38.299 So this is actually slip of grains that’s actually accommodated 00:30:38.299 --> 00:30:42.250 also by dislocations. And this is temperature-, stress-, 00:30:42.250 --> 00:30:47.620 and grain size-dependent and also forms a lattice preferred orientation. 00:30:47.620 --> 00:30:52.140 The important part of grain boundary sliding is, because there’s slip between 00:30:52.140 --> 00:30:56.559 grains, it can actually lead to phase mixing and grain boundary switching. 00:30:56.559 --> 00:31:00.710 And so that would be something to keep in mind. 00:31:00.710 --> 00:31:04.380 The next two mechanisms to look at – one is diffusion creep, 00:31:04.380 --> 00:31:08.100 in which crystal-plastic deformation is not accommodated by the movement 00:31:08.100 --> 00:31:12.520 of dislocations, but actually the movement of atomic vacancies. 00:31:12.520 --> 00:31:15.980 And this is temperature-, stress-, and grain size-dependent but 00:31:15.980 --> 00:31:20.590 actually destroys any pre-existing lattice preferred orientation. 00:31:20.590 --> 00:31:24.480 And so this is – will be something really important that we’ll return to. 00:31:24.480 --> 00:31:28.260 And another thing I’ll mention is this thing – dynamic recrystallization, 00:31:28.260 --> 00:31:33.270 which again, is the process of crystal regrowth when you have things 00:31:33.270 --> 00:31:36.470 deforming under high pressure, temperature, and strain rate. 00:31:36.470 --> 00:31:41.539 And so this is really the process of strong deformed grains that have high 00:31:41.539 --> 00:31:46.860 dislocation densities being replaced by weaker, undeformed grains. 00:31:46.860 --> 00:31:49.779 And this is really important because it allows us to develop this 00:31:49.779 --> 00:31:54.559 consistent relationship between stress, temperature, and grain size that 00:31:54.559 --> 00:31:57.769 becomes this thing we call a paleopiezometer. 00:31:57.769 --> 00:32:02.269 And so this allows us to look at rocks from – that have deformed at various 00:32:02.269 --> 00:32:06.210 different times and actually look at their grain size and specifically 00:32:06.210 --> 00:32:10.039 relate that to conditions of temperature and stress. 00:32:10.040 --> 00:32:14.440 And that would be important, again, in establishing constraints in the future. 00:32:14.440 --> 00:32:18.470 So, again, when we look back at this, and we see that, you know, we have 00:32:18.470 --> 00:32:22.880 a fine-grain olivine here deforming without a fabric 00:32:22.880 --> 00:32:27.230 in the presence of tremolite. And we have relatively coarse-grain 00:32:27.230 --> 00:32:31.380 olivine here deforming with a strong fabric without any tremolite, 00:32:31.380 --> 00:32:36.669 it’s safe to say that we can interpret that either dislocation creep or 00:32:36.669 --> 00:32:40.370 grain boundary sliding is happening in these coarse grain regions. 00:32:40.370 --> 00:32:44.240 And diffusion creep is happening in these relatively fine-grain regions 00:32:44.240 --> 00:32:48.740 and most likely responsible for destroying this initial fabric 00:32:48.740 --> 00:32:52.020 that’s maintained in the coarse-grain minerals. 00:32:52.980 --> 00:32:58.420 And so all these deformation laws sort of have a form of the flow law 00:32:58.420 --> 00:33:03.929 that looks like this. And basically, everything on the right is equated 00:33:03.929 --> 00:33:08.610 to strain rate. And what we have is basically a pre-factor. 00:33:08.610 --> 00:33:12.960 This is something that is determined experimentally by doing experiments 00:33:12.960 --> 00:33:16.580 at different temperatures, strain rates, and stresses. 00:33:17.440 --> 00:33:23.160 Stress to a factor of n – in the case of diffusion creep, that factor is 1. 00:33:23.169 --> 00:33:28.580 And for both dislocation creep and grain boundary sliding, it’s close to 3. 00:33:28.580 --> 00:33:32.320 Grain size comes in. Of course, for dislocation creep, 00:33:32.320 --> 00:33:35.120 there’s no grain size dependence, and for diffusion creep, there’s 00:33:35.120 --> 00:33:39.100 a quite strong grain size dependence. And so that would be something that’s 00:33:39.100 --> 00:33:42.809 very important in actually the strain localization process. 00:33:42.809 --> 00:33:50.960 Because, as grain size reduce, the effect of this parameter raised to 00:33:50.960 --> 00:33:53.010 the power of p becomes greater and greater. 00:33:53.010 --> 00:33:57.409 And this tends to favor diffusion creep at smaller grain size. 00:33:57.409 --> 00:34:03.429 Also, water content is in this equation as the concentration of OH. 00:34:03.429 --> 00:34:06.710 And there’s also an exponential temperature dependence. 00:34:06.710 --> 00:34:11.700 And so we can use these different deformation laws that we’ve laid out 00:34:11.700 --> 00:34:15.780 to actually compare them in a – in a space where we think about 00:34:15.780 --> 00:34:20.649 evolution of the rock from the protolith all the way to the mylonite. 00:34:20.649 --> 00:34:23.210 And the way to do that is with these funky-looking 00:34:23.210 --> 00:34:25.710 deformation mechanism maps. 00:34:25.710 --> 00:34:30.020 And basically, what this is asking is that, at any temperature – 00:34:30.020 --> 00:34:37.010 any conditions of stress and grain size at a particular pressure-temperature 00:34:37.010 --> 00:34:41.860 condition, and also water content, what is the mechanism that is 00:34:41.860 --> 00:34:43.620 going to have the highest strain rate? 00:34:43.620 --> 00:34:48.800 Or, what is the mechanism that can deform the rock at the lowest stress? 00:34:48.800 --> 00:34:53.000 And so, if we look at the conditions for the protolith, we’re now assuming, 00:34:53.000 --> 00:34:56.220 you know, a fairly high temperature. You can call it a mantle potential 00:34:56.220 --> 00:35:02.060 temperature of about 1,300 C. Of course, dry conditions. 00:35:02.060 --> 00:35:08.599 We see that, for the grain size of the protolith, again, in the region of maybe 00:35:08.599 --> 00:35:12.920 1 to 5 millimeters in this case, we’re sitting at the boundary of grain 00:35:12.920 --> 00:35:17.380 boundary sliding and dislocation creep. And, again, the way to sort of read these 00:35:17.380 --> 00:35:24.390 plots is to use this piezometer – this relationship between stress 00:35:24.390 --> 00:35:28.350 and grain size – to actually track the evolution of the rock. 00:35:28.350 --> 00:35:33.500 And so just using this piezometer line and looking at the grain size conditions, 00:35:33.500 --> 00:35:36.240 we can sort of constrain the deformation conditions 00:35:36.240 --> 00:35:39.800 of the protolith to that little box right there. 00:35:39.800 --> 00:35:44.000 Now, if we look on the right at the mylonite, it’s a totally different story. 00:35:44.000 --> 00:35:47.320 Now we’re at much lower pressure-temperature conditions. 00:35:47.339 --> 00:35:50.609 Again, these were constrained by looking at the minerology 00:35:50.609 --> 00:35:53.820 and the composition of the amphibole in these samples. 00:35:53.820 --> 00:35:56.560 And we’re also looking at wet conditions. 00:35:56.560 --> 00:36:00.589 So this water content is just an equilibrium water content for what 00:36:00.589 --> 00:36:04.720 we expect for olivine at these pressure-temperature conditions. 00:36:04.720 --> 00:36:09.290 And, again, if we look at the piezometer in this case, now we’re looking at 00:36:09.290 --> 00:36:13.320 grain sizes related to the mylonite. So maybe hundreds of microns 00:36:13.320 --> 00:36:17.910 all the way down to 1 micron. We’re all the way over here now 00:36:17.910 --> 00:36:21.960 in the diffusion creep regime. So this sort of supports our initial 00:36:21.960 --> 00:36:26.080 findings from the mineralogical texture, which really showed us that the 00:36:26.080 --> 00:36:31.089 coarse-grained regions have a fabric that’s sort of representing diffusion 00:36:31.089 --> 00:36:34.190 creep or grain boundary – or, sorry – dislocation creep or grain boundary 00:36:34.190 --> 00:36:39.780 sliding, while the regions that are especially fine-grained and filled with 00:36:39.780 --> 00:36:45.540 tremolite have no fabric and likely are deforming in diffusion creep. 00:36:45.540 --> 00:36:48.980 One really interesting thing is, if we think about just the evolution 00:36:48.980 --> 00:36:59.380 from those relic grains – these guys right here – to these very fine grains, 00:36:59.380 --> 00:37:03.680 it maybe doesn’t make sense to continue to follow the piezometer. 00:37:03.680 --> 00:37:07.020 So the piezometer is something that’s developed based on the relationship 00:37:07.020 --> 00:37:11.910 between olivine grain size and stress. But, when we add in this secondary 00:37:11.910 --> 00:37:16.130 phase, that pins the olivine grain boundaries, forcing it to reduce 00:37:16.130 --> 00:37:19.710 grain size and deform faster, we actually maybe need to 00:37:19.710 --> 00:37:22.680 move straight away from the piezometer. 00:37:22.680 --> 00:37:26.940 And, being that these minerals are, you know, hundreds of microns 00:37:26.940 --> 00:37:30.880 away from each other, they’re likely deforming at the same stress conditions. 00:37:30.880 --> 00:37:36.700 But it’s clear that the fine-grain bands that are rich in tremolite are deforming 00:37:36.700 --> 00:37:39.990 at a much higher strain rate. And you can see these contours 00:37:39.990 --> 00:37:45.470 of strain rate in gray here on the right side of the diagram. 00:37:45.470 --> 00:37:50.241 And so, just from the conditions of the microstructure and the chemistry of 00:37:50.241 --> 00:37:54.150 these rocks, we have managed to sort of constrain the press-temperature 00:37:54.150 --> 00:37:58.980 conditions for which they underwent ductile deformation as well as some 00:37:58.980 --> 00:38:06.760 information about the grain size and strain rates that are relevant to 00:38:06.760 --> 00:38:10.860 the ductile deformation. And so what we want to do is actually 00:38:10.860 --> 00:38:15.320 use this information to now come all the way back and make some 00:38:15.320 --> 00:38:18.800 rheological constraints – some mechanical constraints – 00:38:18.800 --> 00:38:22.740 on how far seawater is actually delving into these rocks. 00:38:22.740 --> 00:38:26.220 And, to do that, one other piece of information we need is the 00:38:26.220 --> 00:38:30.720 thermal structure of the fault. So this was some nice work done 00:38:30.720 --> 00:38:36.150 by Monica Wolfson-Schwehr, who is at the Monterey Bay Research Institute. 00:38:36.150 --> 00:38:41.060 And she actually did a range of different thermal models that look like this 00:38:41.060 --> 00:38:47.460 on the left, basically establishing a potential temperature at the bottom 00:38:47.460 --> 00:38:51.599 and zero degrees at the surface. And then using the slip rate on the 00:38:51.599 --> 00:38:55.740 fault plus some information about how it was deforming – brittle 00:38:55.740 --> 00:39:00.060 deformation in the shallow crust, ductile deformation of the – 00:39:00.060 --> 00:39:03.780 similar to the laws we just looked at in the deep crust. 00:39:03.780 --> 00:39:08.960 And, again, hydrothermal circulation limited to the brittle-ductile transition. 00:39:08.960 --> 00:39:14.040 So, from this, we can actually get a nice temperature structure for the fault. 00:39:14.040 --> 00:39:18.220 So this is, again, the Shaka Transform Fault that we’ve been looking at. 00:39:18.220 --> 00:39:22.990 And you can see that, because it’s a fairly slow-slipping transform fault, 00:39:22.990 --> 00:39:27.960 you can go quite deep on this fault without actually reaching those 00:39:27.960 --> 00:39:31.740 high temperatures where we expect ductile deformation. 00:39:31.740 --> 00:39:35.820 And so, if you recall, when we were looking at the data on Gofar – 00:39:35.820 --> 00:39:40.150 the fast-slipping East Pacific Rise Transform Fault – 00:39:40.150 --> 00:39:45.530 this 1,000-degree isotherm was closer to 10 kilometers rather than 25. 00:39:45.530 --> 00:39:48.760 So we’re looking at sort of a different scale over here. 00:39:48.760 --> 00:39:53.150 And we’ll come back and revisit that. But now that we have sort of all 00:39:53.150 --> 00:39:56.780 the machinery we would need, let’s start to look at some 00:39:56.780 --> 00:40:01.370 actually rheological constraints on this fault. 00:40:01.370 --> 00:40:07.040 And so, to do that, I’ll try and build up a strength-depth profile from sort of 00:40:07.040 --> 00:40:12.130 the first principles. Maybe it’ll work, and maybe it won’t. We’ll see. 00:40:12.130 --> 00:40:16.000 But the first thing to do is obviously look at frictional strength in the 00:40:16.000 --> 00:40:21.060 upper crust. And this is something we have sort of known about for a while. 00:40:21.060 --> 00:40:26.640 Just using Byerlee’s law at fairly low pressures, friction coefficient 00:40:26.640 --> 00:40:32.100 is about 0.85, and at high pressures, is closer to 0.6. 00:40:32.100 --> 00:40:36.850 So that’s a good basic assumption to start out with. 00:40:36.850 --> 00:40:41.120 Through Margaret Boettcher’s work, we know that this factor, tao-zero, which is 00:40:41.120 --> 00:40:47.640 the cohesion – this little bit right here – is about 21 MPa. So we can add that. 00:40:47.640 --> 00:40:52.020 Maybe we also want to throw on this diagram the friction coefficient for 00:40:52.030 --> 00:40:56.210 relatively weak hydrous minerals, which, for serpentine, 00:40:56.210 --> 00:41:02.130 can be in the range of 0.3 to 0.5, and for talc, maybe it’s closer to 0.2. 00:41:02.130 --> 00:41:07.020 And so that gives you sort of a bounds on what the 00:41:07.020 --> 00:41:10.400 frictional behavior in the shallow crust can be. 00:41:10.400 --> 00:41:14.180 And then we need to add in, of course, the laboratory data 00:41:14.180 --> 00:41:16.730 from Margaret Boettcher. And you can see right now, 00:41:16.730 --> 00:41:20.560 those circle points don’t really fit very well with just the simple 00:41:20.560 --> 00:41:25.770 friction curves that we have set out. But we’ll revisit that after finishing 00:41:25.770 --> 00:41:30.360 to build up this diagram and see where everything is at. 00:41:31.250 --> 00:41:35.880 So the next thing we need to do is add the ductile deformation information. 00:41:35.880 --> 00:41:39.770 So the ductile deformation in the protolith we saw from this deformation 00:41:39.770 --> 00:41:45.829 mechanism map at grain sizes about 1 to 5 millimeters, the strain rate was 00:41:45.829 --> 00:41:50.850 maybe in the neighborhood of 10 to the minus 10 to 10 to the minus 12. 00:41:50.850 --> 00:41:56.740 So here we have a curve for 10 to the minus 12 at 5-millimeter grain size. 00:41:56.740 --> 00:42:01.000 Again, here we’re using the grain boundary sliding law, which is, again, 00:42:01.000 --> 00:42:05.510 stress and temperature-dependent and also weakly dependent on grain size. 00:42:05.510 --> 00:42:11.090 And this is, of course, at dry conditions. This is sort of the base of this 00:42:11.090 --> 00:42:15.599 brittle-ductile transition which we’re trying to outline. 00:42:15.599 --> 00:42:20.240 One thing you can see is just small effects of changing things here. 00:42:20.240 --> 00:42:25.990 If we increase the strain rate by a factor of 2, this deepens the intersection of 00:42:25.990 --> 00:42:30.150 these points, which we’re going to call the brittle-ductile transition. 00:42:30.150 --> 00:42:34.530 And if you decrease grain size, it actually increases that depth. 00:42:34.530 --> 00:42:37.320 So that’s something to keep in mind. 00:42:38.589 --> 00:42:42.180 If we just draw simple constraints here, we can say, okay, you know, 00:42:42.180 --> 00:42:46.890 the deepest point at which the flow laws and the friction intersect, 00:42:46.890 --> 00:42:49.930 we’ll call that – you know, everything below is ductile 00:42:49.930 --> 00:42:53.900 and everything above is brittle. But things are probably 00:42:53.900 --> 00:42:57.220 a little bit more complicated than that. And we know that because there’s 00:42:57.220 --> 00:43:01.760 actually evolution of the fault from this coarse-grain protolith 00:43:01.760 --> 00:43:07.500 to this fine-grain mylonite. And so, in blue, if we outline the 00:43:07.500 --> 00:43:11.040 deformation conditions of the mylonite, again, from this deformation 00:43:11.040 --> 00:43:15.940 mechanism map, if we look at grain sizes maybe 1 to 10 microns, 00:43:15.940 --> 00:43:22.230 that implies strain rates of 10 to the minus 7 to 10 to the minus 9. 00:43:22.230 --> 00:43:26.450 And so that gives us a much shallower range for the potential 00:43:26.450 --> 00:43:29.890 brittle-ductile transition. So, over here, we have this 00:43:29.890 --> 00:43:35.650 coarse-grain protolith that’s slowly evolving to this fine-grain mylonite. 00:43:35.650 --> 00:43:39.170 And they represent completely different conditions rheologically 00:43:39.170 --> 00:43:42.710 in terms of the transition from brittle to ductile behavior. 00:43:42.710 --> 00:43:46.500 And, again, these are at wet conditions with 00:43:46.500 --> 00:43:49.280 a quite grain size- sensitive rheology. 00:43:49.280 --> 00:43:53.820 Which, again, was implied by looking at the textures of the rock. 00:43:55.720 --> 00:43:59.000 The next thing we can do is maybe just sort of outline a brittle-ductile 00:43:59.000 --> 00:44:04.589 transition zone that spans, you know, everything from the protolith 00:44:04.589 --> 00:44:08.089 all the way to the mylonite. And this looks pretty good. 00:44:08.089 --> 00:44:12.230 And maybe gives us some new information about what’s going on. 00:44:12.230 --> 00:44:14.750 But maybe we can do a little bit better than that. 00:44:14.750 --> 00:44:18.910 And I got an idea from that by actually looking at some work done 00:44:18.910 --> 00:44:24.720 by Greg Hirth and Nick Beeler. So, if we return to this idea of slip 00:44:24.720 --> 00:44:29.940 on a fault being movements of two rough surfaces against each other, 00:44:29.940 --> 00:44:35.240 and these asperities having a particular strength – this sigma-y – 00:44:35.240 --> 00:44:40.099 the yield strength, we can think of how pore fluid affect – pore fluid 00:44:40.099 --> 00:44:45.670 pressure affects things as we get down to the brittle-ductile transition. 00:44:45.670 --> 00:44:49.450 So if we just think simply about an effective pressure, P, which is the 00:44:49.450 --> 00:44:55.920 difference of the normal stress on the fault, times this pore fluid factor, alpha, 00:44:55.920 --> 00:45:02.859 and the pore fluid, P-f, in most cases, in the shallow crust, this alpha factor 00:45:02.859 --> 00:45:06.590 is somewhere around 1. Because this ratio of real area 00:45:06.590 --> 00:45:09.690 of contact to the area of contact is very small. 00:45:09.690 --> 00:45:14.440 And that’s been shown in quite a lot of shallow frictional experiments. 00:45:15.520 --> 00:45:20.030 What Hirth and Beeler did was actually try and connect the relationship 00:45:20.030 --> 00:45:25.380 between the real area of contact to this asperity yield strength, sigma-y. 00:45:25.380 --> 00:45:30.990 And using this ratio, and combining that with the effect of pore fluid, 00:45:30.990 --> 00:45:34.250 what you get is, actually this alpha factor becomes the ratio of that 00:45:34.250 --> 00:45:39.780 effective pressure to the yield strength. And so what you get is a different 00:45:39.780 --> 00:45:43.510 expression for effective pressure that, again, depends on the difference 00:45:43.510 --> 00:45:47.110 between that normal stress and the pore fluid pressure, but also is 00:45:47.110 --> 00:45:52.150 related to this yield stress – the strength of the asperities. 00:45:52.150 --> 00:45:55.829 One other thing – important thing that I’ll use is this pore fluid ratio would 00:45:55.829 --> 00:46:00.520 just be the ratio of the pore fluid pressure to the effective pressure. 00:46:00.520 --> 00:46:05.310 And so, if we just think about the yield strength for olivine, and olivine 00:46:05.310 --> 00:46:10.990 asperities in this case, we can combine the laws for a low-temperature plasticity 00:46:10.990 --> 00:46:14.450 and dislocation creep at higher pressures and temperatures and 00:46:14.450 --> 00:46:18.730 add them in parallel to come up with an expression for the yield stress. 00:46:18.730 --> 00:46:24.690 And this was something that Hirth and Beeler did in the continental crust, 00:46:24.690 --> 00:46:29.980 but this is sort of the first time it’s been applied to oceanic settings. 00:46:29.980 --> 00:46:35.120 And what we find when we do that is actually that the shape of the friction 00:46:35.120 --> 00:46:39.540 line changes a bit and actually tends to roll over as we get closer to this 00:46:39.540 --> 00:46:43.210 brittle-ductile transition. And that happens because the 00:46:43.210 --> 00:46:47.970 yield strength actually increases as we get to lower pressures and 00:46:47.970 --> 00:46:52.170 temperatures because, of course, the creep is exponentially dependent 00:46:52.170 --> 00:46:56.750 on temperature. And so what we find is, actually, as this alpha factor 00:46:56.750 --> 00:47:02.310 goes from 1, in which pore fluid has a large effect on effective pressure, 00:47:02.310 --> 00:47:07.020 to zero, where it has a very small to negligible effect on pressure, this 00:47:07.020 --> 00:47:11.390 affects the shape of our friction line. And so this upper bound here, 00:47:11.390 --> 00:47:16.190 you can consider that all the nominal area of the fault is in contact, 00:47:16.190 --> 00:47:18.640 and you basically have ductile deformation. 00:47:18.640 --> 00:47:22.500 And here – oops. And here, a very small amount 00:47:22.500 --> 00:47:28.180 of that nominal area is in contact, and you have frictional slip. 00:47:29.100 --> 00:47:33.720 So if we incorporate this into our existing stress – strength-depth 00:47:33.730 --> 00:47:38.599 diagram, of course we have the mylonite at lower pressure-temperature 00:47:38.599 --> 00:47:42.349 conditions, and the protolith at higher pressure-temperature conditions. 00:47:42.349 --> 00:47:46.220 And now we have this diminishing effect of pore fluid pressure 00:47:46.220 --> 00:47:50.630 as we go further in depth. And what we see is that slightly 00:47:50.630 --> 00:47:54.599 shallows the depth of the brittle-ductile transition just compared 00:47:54.599 --> 00:48:00.190 to using this normal friction line. And we see that also increasing 00:48:00.190 --> 00:48:06.140 pore fluid pressure, or this factor, lambda, increases this effect. 00:48:06.140 --> 00:48:12.540 So if we go to high pore fluid pressures, you know, greater than hydrostatic, 00:48:12.540 --> 00:48:15.980 it actually increases that effect. And then the effect decreases 00:48:15.980 --> 00:48:19.079 if we go to relatively low pore fluid pressures. 00:48:19.079 --> 00:48:24.480 So what we see from this is quite interesting, in that there’s not only 00:48:24.480 --> 00:48:29.619 just this static brittle-ductile transition, but actually a brittle-ductile transition 00:48:29.619 --> 00:48:34.500 zone that’s dynamic, not only in the sense of fault zone evolution, 00:48:34.500 --> 00:48:38.030 but also in the sense of strain rate. You know, increasing strain rate 00:48:38.030 --> 00:48:41.760 going from low to high strain rate actually has a significant effect 00:48:41.760 --> 00:48:44.309 on deepening the brittle-ductile transition. 00:48:44.309 --> 00:48:47.880 So this points to the fact that there might be connections between 00:48:47.880 --> 00:48:51.609 the seismic cycle in these transform faults and actually the depth at which 00:48:51.609 --> 00:48:55.960 seawater can percolate into the crust. 00:48:55.960 --> 00:48:59.520 Now we’ll return to this lab data. And we can see that, 00:48:59.520 --> 00:49:03.920 in the shallow part, we have maybe a little better fit to the points. 00:49:03.920 --> 00:49:07.010 But then we have these points, you know, all the way down here 00:49:07.010 --> 00:49:11.690 in no-man’s-land, and, you know, how can we really represent them? 00:49:11.690 --> 00:49:15.620 So if we think about the laboratory data a little bit, rather than a 00:49:15.620 --> 00:49:18.640 coarse-grain protolith, which we’re trying to model here, 00:49:18.650 --> 00:49:22.829 the lab data was actually made on measurements of fine-grain olivine 00:49:22.829 --> 00:49:30.050 aggregates and at very high laboratory strain rates with pore fluid pressure. 00:49:30.050 --> 00:49:36.359 So, to really represent those, we need to actually throw on diffusion creep lines 00:49:36.359 --> 00:49:41.900 at a little bit higher grain size and higher strain rates to actually 00:49:41.900 --> 00:49:44.220 encapsulate the strength of those points. 00:49:44.220 --> 00:49:48.060 So that’s really just putting those things in context. 00:49:49.360 --> 00:49:52.810 So now that we have sort of some broad constraints on sort of what 00:49:52.810 --> 00:49:57.230 the range of brittle-ductile behavior is, we can think 00:49:57.230 --> 00:50:01.280 a little bit about uplift of these mylonites to the surface. 00:50:01.280 --> 00:50:06.780 And specifically about how uplift may be related to a thermal cracking process. 00:50:06.780 --> 00:50:10.260 So this is some work by Brian deMartin and others. 00:50:10.260 --> 00:50:14.620 And, on the left, he’s basically showing a grain of olivine that’s unloading. 00:50:14.620 --> 00:50:17.420 It’s getting lower in pressure and temperature. 00:50:17.420 --> 00:50:22.140 And these lines in the center of the grain basically show differential 00:50:22.150 --> 00:50:26.690 or anisotropic thermal expansion, where thermal expansion is 00:50:26.690 --> 00:50:30.170 much greater in this direction and lower in this direction. 00:50:30.170 --> 00:50:36.120 And what that causes, actually, is stresses on the grain in which it’s, 00:50:36.120 --> 00:50:39.880 you know, contracting in this direction and actually compressing. 00:50:39.880 --> 00:50:42.530 And it’s in tension in the other direction. 00:50:42.530 --> 00:50:46.589 And so, just the fact of bringing these rocks up to the surface and cooling 00:50:46.589 --> 00:50:52.250 them can actually impart stresses that may be able to actually crack the rock. 00:50:52.250 --> 00:50:56.829 And deMartin and others did some theoretical work to actually establish 00:50:56.829 --> 00:51:00.750 this temperature constraint, which they called the viscous to elastic transition 00:51:00.750 --> 00:51:04.740 temperature, which they really thought represented the transition point from 00:51:04.740 --> 00:51:08.270 which those stresses build up to be high enough that you can 00:51:08.270 --> 00:51:12.500 actually crack the rock by thermal unloading. 00:51:12.500 --> 00:51:16.960 And so you might think, you know, how do these mylonites, you know, 00:51:16.960 --> 00:51:19.820 get to the surface anyway? We’re talking about transform 00:51:19.820 --> 00:51:24.010 rocks and – you know, transform faults and rocks coming from, 00:51:24.010 --> 00:51:27.210 you know, 20, 30 kilometers in depth. 00:51:27.210 --> 00:51:31.190 And one way might be just a simple rotation of the spreading direction. 00:51:31.190 --> 00:51:34.990 So this is some work on the Vema Transform Fault from the Mid-Atlantic 00:51:34.990 --> 00:51:38.790 Ridge in which the authors showed that actually a small change in the spreading 00:51:38.790 --> 00:51:43.309 direction caused a trans-tensional environment that caused deformed 00:51:43.309 --> 00:51:46.240 rocks to come up from the surface. 00:51:46.240 --> 00:51:50.860 Here’s another example from St. Peter St. Paul’s Rocks, also on 00:51:50.860 --> 00:51:54.369 the Mid-Atlantic Ridge, in which a different orientation of rotation 00:51:54.369 --> 00:51:58.070 actually caused a transtensional environment, pushing rocks up to the 00:51:58.070 --> 00:52:02.819 surface in this sort of flower structure, which is often seen in continental 00:52:02.819 --> 00:52:06.660 settings. And so that’s just some broad constraints on how these 00:52:06.660 --> 00:52:10.790 mylonites might actually get to the surface in the first place. 00:52:10.790 --> 00:52:17.070 And so we used that theory developed by deMartin et al. to actually calculate 00:52:17.070 --> 00:52:21.640 this viscous to elastic transition temperature as a function of grain size. 00:52:21.640 --> 00:52:25.500 And one thing that’s important to note is you’ll see that grain size here, 00:52:25.500 --> 00:52:29.690 D, comes in to the factor of 3. So there’s a very strong grain 00:52:29.690 --> 00:52:33.069 size dependence in this. And this is, of course, because, 00:52:33.069 --> 00:52:37.150 as the size of grains grow, the size of flaws in those grains also grow. 00:52:37.150 --> 00:52:41.220 So, as you decrease grain size, you end up with these very small, 00:52:41.220 --> 00:52:45.860 very stiff flaws in the grain, making it harder to crack. 00:52:45.860 --> 00:52:48.680 And we see that over here with the shape of these curves. 00:52:48.680 --> 00:52:53.260 So if we just look at the cooling rate for the Shaka Transform Fault, just 00:52:53.260 --> 00:52:59.680 estimated by broad constraints on uplift rates, from studies like the one in Vema. 00:52:59.680 --> 00:53:04.110 I’m looking at the thermal gradient from the thermal model that we established. 00:53:04.110 --> 00:53:12.109 We can see that, over the range of grain sizes that we consider, there is cracking 00:53:12.109 --> 00:53:15.660 that we expect in the protolith, for the protolith grain size, somewhere in the 00:53:15.660 --> 00:53:20.589 neighborhood of 800 to 900 degrees C. And then, in the mylonite, this goes 00:53:20.589 --> 00:53:24.560 all the way down to 600 degrees C. So, again, this is sort of just kind of 00:53:24.560 --> 00:53:29.550 an independent way to cross-check our rheological constraints on the mylonite 00:53:29.550 --> 00:53:34.730 and actually say, you know, what stresses are imparted on it 00:53:34.730 --> 00:53:37.470 as it is actually coming to the surface? 00:53:37.470 --> 00:53:41.829 And this will be something that we’ll integrate now in the full constraints. 00:53:41.829 --> 00:53:46.109 So I’m sort of going to summarize everything in a diagram that looks 00:53:46.109 --> 00:53:51.520 like this. And so the way to start to read this is just temperature 00:53:51.520 --> 00:53:54.470 as a function of depth, and so this is just our 00:53:54.470 --> 00:53:57.599 geotherm right here from the Shaka Transform Fault. 00:53:57.599 --> 00:54:02.200 And the first constraint we sort of have on hydration is the occurrence of 00:54:02.200 --> 00:54:06.750 those fluid inclusion planes. And so we assume that those start 00:54:06.750 --> 00:54:12.619 coming into place when you cross this ductile deformation line for 00:54:12.619 --> 00:54:15.020 the protolith, which, again, is defined by grain boundary 00:54:15.020 --> 00:54:20.580 sliding at relatively low strain rates and a coarse grain size. 00:54:20.580 --> 00:54:24.980 As you come down this curve, you start to hit the stability field for tremolite, 00:54:24.990 --> 00:54:30.119 which allows hydrous phases to form that are really important in the strain 00:54:30.120 --> 00:54:34.560 localization process – the process of grain size reduction. 00:54:36.380 --> 00:54:39.890 As we go through, there is – we have these rheological constraints – 00:54:39.890 --> 00:54:47.630 the temperature of thermal cracking for grain sizes – 5 to 0.5 millimeters, 00:54:47.630 --> 00:54:52.040 and 1 to 100 microns for the mylonite. And so these rheological constraints, 00:54:52.040 --> 00:54:57.119 both from the flow laws from these thermal cracking calculations and from 00:54:57.119 --> 00:55:01.200 the mineralogical constraints – the thermometry and the presence 00:55:01.200 --> 00:55:05.490 of tremolite – they all give us sort of this self-consistent picture that there’s 00:55:05.490 --> 00:55:08.910 this brittle-ductile transition zone that’s occurring somewhere in the 00:55:08.910 --> 00:55:14.930 neighborhood of 600 to 900 degrees C. And so this suggests that, in this region, 00:55:14.930 --> 00:55:18.880 you can both have earthquakes – there can be seismicity, but you 00:55:18.880 --> 00:55:23.420 can also have fluid flow. And there is an interrelation between 00:55:23.420 --> 00:55:28.300 those things that actually causes you to develop this localized fine-grain shear 00:55:28.300 --> 00:55:34.230 zone that’s represented by the mylonites and this sort of broad zone in 00:55:34.230 --> 00:55:40.630 which you can have both brittle-ductile deformation and also fluid flow. 00:55:40.630 --> 00:55:44.060 The reason this is particularly interesting is, if we go back to the 00:55:44.060 --> 00:55:48.220 Gofar Transform Fault – again, this is the fast-slipping transform that 00:55:48.220 --> 00:55:53.410 was instrumented by ocean-bottom instruments – we can see that 00:55:53.410 --> 00:55:59.600 earthquakes not only are fairly deep in terms of the thermal structure – again, 00:55:59.600 --> 00:56:04.660 10 kilometers here corresponds to about 1,000 degrees C, but they 00:56:04.670 --> 00:56:13.220 also vary significantly along strike. So there’s not only these differences 00:56:13.220 --> 00:56:15.960 in the depth of earthquakes, but also their character. 00:56:15.960 --> 00:56:20.250 And this is something that I’m going to look into a little bit more in the future. 00:56:20.250 --> 00:56:28.150 But if we just throw up the Gofar [volume too low to hear] actually, 00:56:28.150 --> 00:56:30.620 you know, the region that the earthquakes are being sensed that 00:56:30.620 --> 00:56:34.930 this is – this is just the maximum depth of earthquakes that were located Gofar – 00:56:34.930 --> 00:56:38.799 it’s extremely deep in terms of the thermal structure. 00:56:38.799 --> 00:56:41.490 You know, even deeper than our constraints here of a 00:56:41.490 --> 00:56:46.079 brittle-ductile transition zone from 600 to 900 degrees C. 00:56:46.079 --> 00:56:50.750 So there’s really still a little bit of mismatch, but what we sort of 00:56:50.750 --> 00:56:54.910 establish is that, yes, there are fluids very deep on these faults. 00:56:54.910 --> 00:57:00.240 And likely, it is earthquakes that are actually emplacing those fluids. 00:57:00.240 --> 00:57:03.980 And so there’s likely some feedbacks between what’s happening in the 00:57:03.980 --> 00:57:05.849 seismic cycle and what’s happening rheologically 00:57:05.849 --> 00:57:10.460 on these faults in terms of hydration and strain localization. 00:57:10.460 --> 00:57:15.839 And so – I’ll skip this for now – but, in the future, we sort of have 00:57:15.839 --> 00:57:19.549 an exciting opportunity coming up in November to actually go to this 00:57:19.549 --> 00:57:23.839 Gofar Transform Fault, which has high-resolution bathymetry, 00:57:23.839 --> 00:57:27.140 high-resolution seismic recording, and now we’re finally going to be 00:57:27.140 --> 00:57:31.160 able to actually dredge the fault to actually pick up rocks from 00:57:31.160 --> 00:57:34.280 these different sections and actually try and relate them to 00:57:34.280 --> 00:57:40.260 the seismic properties of the fault. So, sorry, I’m a little bit over time, but 00:57:40.260 --> 00:57:46.380 I’ll just end here on this diagram that sort of summarizes our observations. 00:57:47.540 --> 00:57:48.760 Thanks for your attention. 00:57:48.760 --> 00:57:53.140 [Applause] 00:57:53.980 --> 00:57:56.820 - Are there any questions for our speaker? 00:57:59.100 --> 00:58:06.620 [Silence] 00:58:06.620 --> 00:58:10.100 - I enjoyed your comprehensive talk. That was really good. 00:58:10.100 --> 00:58:13.880 I was wondering, does just the slip rate of the Gofar – because it’s so much 00:58:13.880 --> 00:58:18.460 faster than the other one, does that make the difference, or are there many other 00:58:18.460 --> 00:58:21.271 things that are possibilities? - There are definitely some 00:58:21.271 --> 00:58:24.790 other possibilities. That’s probably the main factor. 00:58:24.790 --> 00:58:30.200 So, you know, the difference in slip rate gives you sort of this – you know, 00:58:30.200 --> 00:58:34.700 difference in the thermal structure. What that really does is, you know, 00:58:34.710 --> 00:58:38.680 each of these constraints, a lot of which are, you know, fairly vertical 00:58:38.680 --> 00:58:42.340 in this diagram – for example, the tremolite stability field and 00:58:42.340 --> 00:58:46.910 things like that, they don’t change that much in terms of the 00:58:46.910 --> 00:58:50.980 temperature conditions. But the depth range and sort of 00:58:50.980 --> 00:58:54.420 the width of that brittle-ductile transition zone might change a bit. 00:58:54.420 --> 00:58:58.609 And, of course, that can impact things that, you know, I didn’t really touch on, 00:58:58.609 --> 00:59:01.900 like, you know, mineral kinetics and sort of the details of the 00:59:01.900 --> 00:59:05.430 hydration reaction, which actually impact the fault rheology. 00:59:05.430 --> 00:59:08.390 So things could be significantly different over there. 00:59:08.390 --> 00:59:11.980 And that’s why it’s sort of important to have, you know, both the rocks 00:59:11.980 --> 00:59:15.900 and the seismic observations [volume too low to hear]. 00:59:15.900 --> 00:59:17.660 - Thank you. 00:59:19.240 --> 00:59:21.140 - Any other questions? 00:59:26.080 --> 00:59:28.380 And can you adjust your microphone so it’s not being blocked by your collar? 00:59:28.380 --> 00:59:31.040 - Oh, sorry. - I think that’s what’s happening. 00:59:33.740 --> 00:59:35.980 - Thanks, Arjun. That was really neat. 00:59:35.980 --> 00:59:39.800 So when you started out, and you said that the fluid inclusions, which is 00:59:39.820 --> 00:59:44.260 sort of one of your first constraints, those were olivines, I guess? 00:59:44.260 --> 00:59:47.940 - Yeah, that’s right. - Right. But you didn’t – you didn’t look 00:59:47.940 --> 00:59:53.460 at the composition of the inclusions? - No. Unfortunately, 00:59:53.460 --> 00:59:59.220 the inclusions are quite small. So even using [volume too low to hear] 00:59:59.220 --> 01:00:03.100 very hard to actually see what’s inside those inclusions. 01:00:03.109 --> 01:00:06.240 - Right. - The sort of cross-check to that 01:00:06.240 --> 01:00:12.600 is looking at the composition [volume too low to hear]. 01:00:12.600 --> 01:00:16.780 - Right. So – but those are the very deepest. 01:00:16.780 --> 01:00:21.160 So the thing – is the thing that limits the depth at which the water 01:00:21.160 --> 01:00:25.810 goes basically the limits of brittle deformation? 01:00:25.810 --> 01:00:29.660 That’s how – the mechanism in which it gets down there. 01:00:29.660 --> 01:00:31.760 - That’s right. And, you know, it’s really 01:00:31.760 --> 01:00:37.440 something that’s kind of modulated by strain rate. 01:00:37.440 --> 01:00:42.300 So, you know, the faster you push these things, the deeper you can sort of 01:00:42.300 --> 01:00:46.450 extend that brittle behavior to. And, you know, if you think about – 01:00:46.450 --> 01:00:50.549 you know, if you’re a rock sitting at depth on the transform, and you have 01:00:50.549 --> 01:00:55.559 these quasi-repeating events as you tend to see in these settings, you know, 01:00:55.559 --> 01:00:59.950 over and over again, you’re constantly putting stress perturbations into 01:00:59.950 --> 01:01:02.559 the ductile field in the base of the seismogenic zone. 01:01:02.559 --> 01:01:07.600 And so, in my mind, that depth to the brittle-ductile transition is sort of being 01:01:07.600 --> 01:01:12.559 modulated by the earthquake cycle. And maybe also that’s modulating 01:01:12.560 --> 01:01:18.380 how fluids are entering the rock and eventually affecting the rheology. 01:01:19.480 --> 01:01:26.400 - Right. Right. So you’re imagining that, let’s say, over the seismic cycle, 01:01:26.400 --> 01:01:33.720 that dynamically there’s some kind of communication with the surface 01:01:33.730 --> 01:01:37.930 over the long-term. But, at any particular moment, 01:01:37.930 --> 01:01:42.030 the pore pressure at great depth, for example, where these fractures 01:01:42.030 --> 01:01:47.740 are forming, it could be essentially lithostatic pore fluid pressure. 01:01:47.740 --> 01:01:51.810 So it is – it’s definitely lithostatic below the brittle-ductile transition, 01:01:51.810 --> 01:01:55.240 wherever that brittle-ductile transition happens to be at this particular moment. 01:01:55.240 --> 01:01:59.800 But it – right? That’s what you’re thinking? 01:01:59.800 --> 01:02:03.440 - There’s an – maybe an interesting point in that. 01:02:05.130 --> 01:02:10.520 You know, we sort of tried to visit this issue on whether the fluid 01:02:10.530 --> 01:02:14.230 inclusions themselves could be a result of lithostatic pore fluids. 01:02:14.230 --> 01:02:18.130 You know, instead of, you know, fracturing due to stress perturbations 01:02:18.130 --> 01:02:21.619 from earthquakes, you could just have fluids in place in a 01:02:21.619 --> 01:02:25.180 ductilely deforming rock. And, as it strains, those fluids have no 01:02:25.180 --> 01:02:29.870 place to go, and pore fluid builds up, and eventually you get these fractures. 01:02:29.870 --> 01:02:36.130 But I’m sort of trying to do this stress analysis, and maybe it’s a little curse, 01:02:36.130 --> 01:02:40.980 but, you know this stress direction here is, you know, sort of normal 01:02:40.980 --> 01:02:45.240 to the strike of the fault. And that’s also sort of the 01:02:45.240 --> 01:02:49.869 orientation we see those, you know, healed tensional cracks. 01:02:49.869 --> 01:02:54.960 And so that definitely is possibly another interpretation of those. 01:02:58.320 --> 01:03:00.220 - Any last questions? 01:03:02.500 --> 01:03:05.240 Okay. Let’s thank our speaker again. - Thank you, everyone. 01:03:05.240 --> 01:03:09.220 [Applause] 01:03:09.230 --> 01:03:11.059 - There are plans to take the speaker out to lunch, 01:03:11.060 --> 01:03:14.600 so if you’d like to join, please come up and do so. 01:03:14.600 --> 01:03:21.109 Or if you’d like to meet with Arjun after lunch, I think he’ll be staying around. 01:03:21.109 --> 01:03:25.400 So please come let us know. Thank you. 01:03:31.340 --> 01:03:33.580 - Thank you very much. Sorry I couldn’t figure out the mic. 01:03:33.580 --> 01:03:35.360 - No worries. It’s finicky. 01:03:36.760 --> 01:03:40.180 - Do I need to [inaudible]? - Nope. I’ll take care of it. 01:03:40.180 --> 01:03:41.820 - Thank you so much. - Yeah. 01:03:41.820 --> 01:03:43.000 [Silence]