WEBVTT
Kind: captions
Language: en

00:00:00.840 --> 00:00:04.470
Good morning, everybody.
I think we’ll get going here.

00:00:04.470 --> 00:00:05.970
Couple of announcements.

00:00:05.970 --> 00:00:10.490
This Wednesday at our normal time,
the Earthquake Science Center

00:00:10.490 --> 00:00:14.610
will host Jay Patton from
Humboldt State University.

00:00:14.610 --> 00:00:19.270
He’ll present a talk titled
Sedimentary and Geophysical

00:00:19.270 --> 00:00:24.279
Investigations Offshore Sumatra –
Riddling the Seafloor for Evidence

00:00:24.279 --> 00:00:29.140
of Earthquakes and Submarine Landslides.
So please join us for that.

00:00:29.140 --> 00:00:34.280
Today, for this special joint seminar,
we are joined by Meredith Townsend.

00:00:34.280 --> 00:00:38.100
Meredith is a Ph.D. candidate
at Stanford University

00:00:38.100 --> 00:00:41.829
in the geological and
environmental sciences department.

00:00:41.829 --> 00:00:45.790
She did her undergraduate work
at Washington and Lee University,

00:00:45.790 --> 00:00:47.989
also in geology.

00:00:47.989 --> 00:00:52.079
And I recently learned,
while she was at Stanford,

00:00:52.079 --> 00:00:55.680
she completed something called
a Ph.D. minor in the department

00:00:55.680 --> 00:00:58.690
of feminist, gender,
and sexuality studies.

00:00:58.690 --> 00:01:01.000
So that’s pretty cool.
Pretty incredible.

00:01:01.000 --> 00:01:04.670
Today she will share
some of her graduate work

00:01:04.670 --> 00:01:07.539
on fracturing around
magmatic dikes.

00:01:07.539 --> 00:01:10.799
And I’ll now turn
things over to Meredith.

00:01:17.540 --> 00:01:19.620
- So thank you all
for being here today.

00:01:19.630 --> 00:01:25.670
It’s a real treat for me to get to
present my dissertation work here.

00:01:25.670 --> 00:01:31.270
So I just want to give a big thanks to
Ole and Brooks for arranging my visit.

00:01:31.270 --> 00:01:36.069
The title of my talk is
Fracturing Around Magmatic Dikes

00:01:36.069 --> 00:01:40.549
as a Precursor to the Development
of Volcanic Plugs.

00:01:40.549 --> 00:01:44.219
But as a quick summary,
this work is really about trying to

00:01:44.219 --> 00:01:49.880
understand deformation around
volcanic conduits using a combination

00:01:49.880 --> 00:01:54.590
of field observations and
structural geology along with

00:01:54.590 --> 00:02:00.369
models for the relevant
thermo-poroelastic processes.

00:02:00.369 --> 00:02:06.120
So the first half of the talk will be
mostly about the field observations.

00:02:06.120 --> 00:02:09.100
And then in the second half,
I’ll get into some of the modeling

00:02:09.100 --> 00:02:11.501
I’ve been working on.

00:02:14.500 --> 00:02:17.939
So I actually want to start
with acknowledgments.

00:02:17.939 --> 00:02:20.030
First and foremost,
I want to acknowledge

00:02:20.030 --> 00:02:22.740
my adviser at Stanford,
Dave Pollard.

00:02:22.740 --> 00:02:24.819
And I also want to
acknowledge my co-authors

00:02:24.819 --> 00:02:29.120
on the first part of this project,
Kendra Johnson, who is a Ph.D. student

00:02:29.120 --> 00:02:32.459
at Colorado School of Mines,
and also Cansu Culha,

00:02:32.459 --> 00:02:36.310
who is a graduate
student at Stanford.

00:02:36.310 --> 00:02:39.719
I also want to give
a shout-out to Josie Nevitt,

00:02:39.719 --> 00:02:42.829
who has helped me
tremendously in the field.

00:02:42.829 --> 00:02:47.629
And I also want to thank
these folks for constructive feedback

00:02:47.629 --> 00:02:51.320
on the project throughout
my time at Stanford.

00:02:51.320 --> 00:02:54.540
I also have to thank CoreLabs
for performing the

00:02:54.540 --> 00:02:57.840
permeability measurements
that I’ll be presenting.

00:02:57.840 --> 00:03:01.620
And I also have to acknowledge
John Shaw, who let me work

00:03:01.629 --> 00:03:07.189
in his lab last fall and use some
of the lab equipment at Harvard.

00:03:07.189 --> 00:03:11.780
Also, lately I’ve been receiving
lots of online technical support

00:03:11.780 --> 00:03:16.010
for Hydrotherm, and so I just
want to thank Steve Ingebritsen,

00:03:16.010 --> 00:03:18.640
Tom Driesner,
and Paul Hsieh

00:03:18.640 --> 00:03:22.439
for communicating with me
about those things.

00:03:22.439 --> 00:03:26.459
My work is funded through the
Department of Geological Sciences

00:03:26.459 --> 00:03:31.059
at Stanford, the Rock Fracture Project,
the DARE Fellowship,

00:03:31.060 --> 00:03:33.960
the McGee Grant,
and the MacDiarmid Grant.

00:03:33.960 --> 00:03:36.680
So with that,
let’s get started.

00:03:38.409 --> 00:03:44.620
So this work is couched in the
broader context of the evolution of

00:03:44.620 --> 00:03:51.139
basaltic eruptions, which have perhaps
been best documented in Hawaii.

00:03:51.139 --> 00:03:59.579
And eruptions tend to initiate as
nearly planar features called fissures.

00:03:59.579 --> 00:04:06.090
And most of these fissure eruptions
will freeze up within a few hours,

00:04:06.090 --> 00:04:10.829
but some of these eruptions
will localize into a discrete conduit

00:04:10.829 --> 00:04:13.099
through which the
eruption is sustained.

00:04:13.099 --> 00:04:18.109
So this is an example
of one such sequence of events

00:04:18.109 --> 00:04:21.079
at Kilauea Iki in 1959.

00:04:21.079 --> 00:04:25.470
The fissure eruption initiated and
grew to a length of about a kilometer,

00:04:25.470 --> 00:04:29.370
spewing magma about 30 meters
into the air at an eruption rate

00:04:29.370 --> 00:04:33.039
of 100 cubic meters
per second.

00:04:33.039 --> 00:04:38.419
And a few hours later, the eruption had
ceased along most of the fissure except

00:04:38.419 --> 00:04:45.370
for one location back here through – and
at this location, the eruption persisted.

00:04:45.370 --> 00:04:48.860
And four days later,
shown on the right,

00:04:48.860 --> 00:04:54.250
the conduit had substantially
widened and rounded out.

00:04:54.250 --> 00:04:57.289
And at this point, the eruption rate
had more than tripled,

00:04:57.289 --> 00:05:02.360
and the fountaining height
reached 350 meters.

00:05:02.360 --> 00:05:08.020
So the question I seek to answer,
at least in part, is what are the

00:05:08.020 --> 00:05:11.069
physical mechanisms that
lead to the growth of

00:05:11.069 --> 00:05:16.960
larger volcanic conduits
from dikes and fissures?

00:05:16.960 --> 00:05:23.110
And two of the main mechanisms
that have been proposed for this

00:05:23.110 --> 00:05:29.250
transition include thermal erosion –
that is, melting of the host rock –

00:05:29.250 --> 00:05:36.810
and mechanical erosion, particularly
via water – magma water interaction.

00:05:36.810 --> 00:05:42.090
And the work on thermal erosion
has primarily been theoretical.

00:05:42.090 --> 00:05:48.979
That is, no geologic study has yet
presented evidence that host rock

00:05:48.979 --> 00:05:54.379
melting plays a significant role in
this conduit geometry evolution.

00:05:54.379 --> 00:06:01.250
However, there are advection-diffusion
models for heat flow through fissures,

00:06:01.250 --> 00:06:08.300
and they generally show that,
for giant dikes, such as the ones

00:06:08.300 --> 00:06:12.690
that feed flood basalts that can be
hundreds of meters wide

00:06:12.690 --> 00:06:16.889
and thousands of kilometers long,
there likely is enough heat flow

00:06:16.889 --> 00:06:21.409
to cause some significant
melting of the host rock.

00:06:21.409 --> 00:06:26.750
However, for fissures of the
sizes observed in places like Hawaii,

00:06:26.750 --> 00:06:30.050
which tend to be only, you know,
1 or 2 meters wide,

00:06:30.050 --> 00:06:32.879
under typical flow conditions,
there doesn’t seem to be enough

00:06:32.879 --> 00:06:38.069
heat flow to cause significant
melting of the host rock.

00:06:38.069 --> 00:06:44.840
And so the other mechanism that’s been
suggested is magma-water interaction

00:06:44.840 --> 00:06:49.940
leading to brecciation of the host rock
and erosion of the conduit.

00:06:49.940 --> 00:06:55.360
And this mechanism has been treated
with less theoretical rigor but is

00:06:55.360 --> 00:06:59.580
perhaps better supported
by the geologic evidence.

00:06:59.580 --> 00:07:05.500
For example, breccias are commonly
associated with volcanic conduits.

00:07:05.500 --> 00:07:09.250
And so I got interested in
honing in on this question.

00:07:09.250 --> 00:07:13.280
What is the role of
magma-groundwater interaction

00:07:13.280 --> 00:07:17.860
in altering the geometry
of dikes and fissures?

00:07:17.860 --> 00:07:23.280
And I sought to tie together
the geology with robust models

00:07:23.280 --> 00:07:25.500
for host rock deformation.

00:07:27.000 --> 00:07:30.880
And it turns out there are
some other good reasons

00:07:30.889 --> 00:07:35.729
to study magma-water interaction.
We know it affects eruption dynamics.

00:07:35.729 --> 00:07:42.189
When water can directly invade the
conduits, we can get explosive phases.

00:07:42.189 --> 00:07:45.330
And as I’ve already mentioned,
and then as would be the focus

00:07:45.330 --> 00:07:53.699
of this talk, magma-water interaction
can affect host rock deformation.

00:07:53.699 --> 00:07:57.949
And hydrothermal processes are also
important at mid-ocean ridges

00:07:57.949 --> 00:08:03.159
where convection cells are set up
and relate to the heat budget

00:08:03.159 --> 00:08:06.860
at this very important
part of the crust.

00:08:06.860 --> 00:08:10.719
And magma-water interaction
is also important for understanding

00:08:10.719 --> 00:08:15.669
contact metamorphism.
And of course, hydrothermal processes

00:08:15.669 --> 00:08:20.889
are important for understanding
ore deposits and geothermal energy.

00:08:23.259 --> 00:08:27.020
So if you really want to learn
about deformation around dikes,

00:08:27.020 --> 00:08:32.409
one of the best things you can do
is go visit an ancient volcano

00:08:32.409 --> 00:08:36.830
where enough erosion has taken place
to expose the plumbing system.

00:08:36.830 --> 00:08:40.840
And so I visited the
Navajo Volcanic Field,

00:08:40.840 --> 00:08:45.590
which is located in the
central Colorado Plateau.

00:08:45.590 --> 00:08:55.250
And the most prominent feature in
this area is this plug called Ship Rock,

00:08:55.250 --> 00:09:02.130
which is about 500 meters high and
surrounded by dikes and smaller plugs.

00:09:02.130 --> 00:09:05.350
And just to give a little
more geologic background,

00:09:05.350 --> 00:09:09.320
the Navajo Volcanic Field is
30 million years old,

00:09:09.320 --> 00:09:13.440
and it’s composed of
a type of magmatic material

00:09:13.440 --> 00:09:18.290
that’s ultrapotassic,
and it’s called a lamprophyre.

00:09:18.290 --> 00:09:25.570
And this lamprophyric magma
intruded through flat-lying sediments

00:09:25.570 --> 00:09:31.210
of the San Juan Basin that were
deposited during the Cretaceous.

00:09:31.210 --> 00:09:36.060
And since the time of eruption,
erosion has cut down and

00:09:36.060 --> 00:09:39.160
exposed the plumbing system
at different levels.

00:09:39.960 --> 00:09:45.620
And at Ship Rock, it’s estimated
that about a kilometer of erosion

00:09:45.620 --> 00:09:49.680
has taken place since
the time of eruption.

00:09:49.680 --> 00:09:54.700
And the current land surface around
the dikes here is part of the

00:09:54.700 --> 00:09:58.430
Mancos Shale formation,
which despite its name,

00:09:58.430 --> 00:10:02.550
is actually composed of layers
with different grain sizes.

00:10:02.550 --> 00:10:07.290
And here it’s actually
more like a siltstone.

00:10:07.290 --> 00:10:11.660
The dikes are about
a few kilometers long.

00:10:11.660 --> 00:10:16.560
They dip vertically.
They’re a meter or two wide.

00:10:16.560 --> 00:10:20.570
And they’re typically broken
into echelon segments.

00:10:20.570 --> 00:10:27.900
And so the direction of propagation
inferred here is vertical at this level.

00:10:27.900 --> 00:10:32.350
And the northeastern dike, which I’ve
pointed out here, has been the focus

00:10:32.350 --> 00:10:39.090
of previous studies and is also
going to be the focus of my work.

00:10:39.090 --> 00:10:44.840
And so the analogous question
at these ancient volcanoes is,

00:10:44.840 --> 00:10:48.640
what are the mechanisms
for the growth of these larger

00:10:48.640 --> 00:10:53.420
volcanic plugs like Ship Rock
from these dikes?

00:10:58.220 --> 00:11:04.100
In the 1980s, Paul Delaney and
Dave Pollard used field data from the

00:11:04.100 --> 00:11:07.970
northeastern dike to try to learn about
the mechanics of dike propagation.

00:11:07.970 --> 00:11:12.850
And specifically they were interested
in whether propagation is controlled

00:11:12.850 --> 00:11:17.880
by the local structure
or whether propagation occurs

00:11:17.880 --> 00:11:21.700
through fractures
formed at the dike tips.

00:11:21.700 --> 00:11:28.260
And they measured the fracture patterns
around the dike and around the region.

00:11:28.260 --> 00:11:34.450
And they found that, at Ship Rock,
the orientation of regional fractures

00:11:34.450 --> 00:11:39.060
is not consistent with
the orientation of the dikes.

00:11:39.060 --> 00:11:43.580
However, what they did find is that
locally around the dikes,

00:11:43.580 --> 00:11:50.480
within about 15 meters, there was
a set of dike-parallel joints in the

00:11:50.480 --> 00:11:56.210
host rock that appeared to be related
to the propagation of the dikes.

00:11:56.210 --> 00:12:02.320
And so they sort of concluded that
the dikes in this area are propagating

00:12:02.320 --> 00:12:07.610
through their own fractures,
and they used this analogy

00:12:07.610 --> 00:12:13.230
of ground cracks that tend to
form ahead of propagating fissures.

00:12:17.150 --> 00:12:23.060
So when I visited Ship Rock,
I saw the dike-parallel fractures

00:12:23.060 --> 00:12:27.820
that were discussed in that paper,
which I’ve here outlined in blue.

00:12:29.440 --> 00:12:33.100
But I also saw another
set of joints in the host rock

00:12:33.110 --> 00:12:36.580
that Dave and Paul
had apparently missed.

00:12:36.580 --> 00:12:39.670
And much like
the parallel fractures,

00:12:39.670 --> 00:12:44.160
these joints are localized
within about a meter of the dikes.

00:12:44.160 --> 00:12:47.430
They’re not part
of a regional set.

00:12:47.430 --> 00:12:50.760
And they dip vertically,
and they’re consistently oriented

00:12:50.760 --> 00:12:54.800
perpendicular to
the dike contacts.

00:12:54.800 --> 00:12:59.450
So I got interested in what these
perpendicular fractures are,

00:12:59.450 --> 00:13:04.700
since they hadn’t been documented
before and since they seemed

00:13:04.700 --> 00:13:10.120
at odds with what I knew
about stress around dikes.

00:13:10.120 --> 00:13:15.110
So in an effort to be a good geologist,
I carefully mapped out the

00:13:15.110 --> 00:13:20.300
fracture patterns using a technique
called structure from motion.

00:13:20.300 --> 00:13:25.320
And this is where my colleague
Kendra Johnson was a really big help.

00:13:25.320 --> 00:13:28.530
This is the focus
of her Ph.D. thesis.

00:13:28.530 --> 00:13:32.990
So she came out with me,
and what we did was we filled

00:13:32.990 --> 00:13:37.580
this balloon with helium gas and
attached a camera to the bottom of

00:13:37.580 --> 00:13:43.230
the balloon and set the camera to
take pictures at a 5-second interval.

00:13:43.230 --> 00:13:47.320
And then the balloon was tethered
so that we could kind of control it,

00:13:47.320 --> 00:13:48.970
like flying a kite.

00:13:48.970 --> 00:13:53.620
And we walked up and down
one of the dike segments along

00:13:53.620 --> 00:13:59.740
the northeastern dike while the camera
took overlapping photographs.

00:13:59.740 --> 00:14:05.170
And we walked up and down once,
and it took about 30 minutes.

00:14:05.170 --> 00:14:09.300
And then the overlapping photographs
were uploaded into Agisoft Photoscan

00:14:09.300 --> 00:14:15.690
and automatically orthorectified with
the help of a few ground control points.

00:14:15.690 --> 00:14:23.200
And the result was a high-resolution
composite photograph of the dike,

00:14:23.200 --> 00:14:27.400
which here is this dark material
that’s curving around like so.

00:14:27.400 --> 00:14:32.370
And to give you a sense for the
resolution that we were able to achieve,

00:14:32.370 --> 00:14:38.980
this dike is about 150 meters long,
but now we’re going to zoom in

00:14:38.980 --> 00:14:45.250
on the red box, and we can start to see
a little more detail in the host rock,

00:14:45.250 --> 00:14:51.120
which is this buff-colored rock here that
you can see has been sort of baked

00:14:51.120 --> 00:14:57.970
up against the dike and is therefore more
resistant to erosion, so it stands out.

00:14:57.970 --> 00:15:01.370
But we’re going to
zoom in again.

00:15:01.370 --> 00:15:06.330
And now we can actually start to pick
out details of fractures in the host rock.

00:15:06.330 --> 00:15:09.950
So in the end, we achieved a
resolution of about 3-1/2 millimeters

00:15:09.950 --> 00:15:14.880
per pixel for the cost of $200,
which was the helium tank.

00:15:14.880 --> 00:15:18.970
And it took about 30 minutes
to gather the data and a few hours

00:15:18.970 --> 00:15:23.280
to orthorectify and make
the composite photograph.

00:15:23.280 --> 00:15:27.100
So this was a
very helpful technique.

00:15:27.100 --> 00:15:31.450
And then, using the photograph,
we mapped the dike contact,

00:15:31.450 --> 00:15:37.110
which is in white, and we mapped
the dike-parallel fractures in blue

00:15:37.110 --> 00:15:40.920
and the dike-perpendicular
fractures in red.

00:15:40.920 --> 00:15:46.260
And this kind of mapping was
useful because it allowed us to gather,

00:15:46.260 --> 00:15:50.110
quite rapidly, statistics
on the joint sets.

00:15:50.110 --> 00:15:56.740
For example, the joints’ lengths
and geometry, spacing,

00:15:56.740 --> 00:15:59.780
and relationship
to other fractures.

00:15:59.780 --> 00:16:05.980
And so, from this mapping, we
concluded that the dike-perpendicular

00:16:05.980 --> 00:16:10.860
joints are definitely associated with
the emplacement of the magma.

00:16:10.860 --> 00:16:14.910
Because wherever the contact of
the dike changes orientation,

00:16:14.910 --> 00:16:20.980
these joints in red also change
orientation to remain perpendicular.

00:16:20.980 --> 00:16:25.180
And we also found that the
dike-perpendicular joints

00:16:25.190 --> 00:16:28.950
tended to terminate
at the dike-parallel fractures.

00:16:28.950 --> 00:16:33.970
And so we inferred that
they’re younger than the parallel set.

00:16:36.400 --> 00:16:40.310
This is a regular
photograph looking up at a dike.

00:16:40.310 --> 00:16:45.410
And we can see the host rock baked
on and preserving the fractures.

00:16:45.410 --> 00:16:47.040
And so these are the
dike-parallel joints in red,

00:16:47.040 --> 00:16:49.040
and we see they’re
vertically dipping.

00:16:49.040 --> 00:16:54.630
And they cut through many of the beds
at once, which are highlighted in green.

00:16:54.630 --> 00:17:00.250
And there are also plumose structures
on the joint faces that indicate that

00:17:00.250 --> 00:17:07.159
the joints initiated at the dike contact
and then propagated up and away.

00:17:09.580 --> 00:17:14.860
The next key observation is that
there are xenoliths of the host rock

00:17:14.860 --> 00:17:22.480
that appear to have this rectangular
prismatic structure that appears to be

00:17:22.480 --> 00:17:28.350
formed by the combination of the two
orthogonal joint sets and the bedding.

00:17:28.350 --> 00:17:32.879
And sometimes you can see in places
where the rectangular blocks of

00:17:32.879 --> 00:17:38.819
host rock have just been rotated
out of place before the magma froze.

00:17:41.080 --> 00:17:44.940
And in some places, if enough
host rock has been removed,

00:17:44.940 --> 00:17:51.230
the dike has been widened into
a structure that we call a bud.

00:17:51.230 --> 00:17:56.360
And the host rock around these
buds tends to be highly brecciated.

00:17:56.360 --> 00:18:03.629
However, you can hopefully
pick out that the clasts in this breccia

00:18:03.629 --> 00:18:09.139
preserve that rectangular prismatic
shape, which suggests that the

00:18:09.139 --> 00:18:12.620
breccia is really just an area
where the dike-parallel and

00:18:12.620 --> 00:18:16.460
dike-perpendicular fracture
density is very high.

00:18:19.400 --> 00:18:24.720
So all together, the field observations
suggest that the two orthogonal

00:18:24.720 --> 00:18:29.570
joint sets, along with the bedding planes,
carve the host rock into isolated blocks

00:18:29.570 --> 00:18:35.090
that can be entrained by the flow of
magma to widen the dike and form buds,

00:18:35.090 --> 00:18:40.559
which represent the insipient stages
of volcanic plug formation.

00:18:40.559 --> 00:18:44.220
So since these dike-perpendicular joints
seem to be such a crucial step

00:18:44.220 --> 00:18:47.710
in this process, and since they
hadn’t been documented before,

00:18:47.710 --> 00:18:53.750
I was interested in trying to figure out
why they’re there. [chuckles]

00:18:56.560 --> 00:19:02.560
So now we’re going to turn our
attention to some fun problems

00:19:02.570 --> 00:19:06.399
related to deformation around
dikes in the shallow crust.

00:19:07.700 --> 00:19:11.140
And we’re going to try to see
if we can explain the formation

00:19:11.149 --> 00:19:13.779
of these dike-perpendicular joints.

00:19:13.779 --> 00:19:19.080
And I just first want to reinforce the
notion that the joints provide some pretty

00:19:19.080 --> 00:19:23.740
hard constraints on the state of stress
around the dike during emplacement.

00:19:23.740 --> 00:19:27.470
And in particular,
the joint orientations

00:19:27.470 --> 00:19:32.460
constrain the orientations of
the principal stress trajectories.

00:19:32.460 --> 00:19:36.570
And the fact that these joints
are tensile features suggests

00:19:36.570 --> 00:19:41.120
that the effective stress
was in the tensile regime.

00:19:41.120 --> 00:19:48.350
So we’re going to begin by
simplifying the problem to 2D

00:19:48.350 --> 00:19:51.200
such that the plane we’re
working in represents

00:19:51.200 --> 00:19:54.929
the horizontal slice
through one of these dikes.

00:19:54.929 --> 00:19:59.139
And before the dike is in place,
there is some initial state of stress

00:19:59.139 --> 00:20:05.999
characterized by two remote principal
stresses, denoted sigma-1r and sigma-2r.

00:20:05.999 --> 00:20:12.549
And we’ll take a tension-positive
convention so that sigma-1 is

00:20:12.549 --> 00:20:19.909
the maximum principal stress, or the
least compressive principal stress.

00:20:19.909 --> 00:20:25.499
Since the emplacement depth at
Ship Rock is estimated to be 1 kilometer,

00:20:25.499 --> 00:20:32.220
the lithostatic overburden contributes
to about negative 25 MPa.

00:20:32.220 --> 00:20:38.120
And in hydrostatic conditions, the fluid
pressure would be about 10 MPa.

00:20:38.120 --> 00:20:41.019
And so the effective stress
at this depth was

00:20:41.019 --> 00:20:45.079
about 15 MPa of compression.

00:20:47.620 --> 00:20:50.820
When the dike is emplaced,
there is a stress perturbation

00:20:50.830 --> 00:20:56.440
associated with the elastic
distortion of the host rock.

00:20:56.440 --> 00:21:02.409
And there’s also heat transfer to the host
rock, which causes thermal expansion of

00:21:02.409 --> 00:21:09.320
the grains and the pore fluids, which
we’re going to be assuming to be water.

00:21:09.320 --> 00:21:14.409
And as a side note, the thermal
expansion of water is about

00:21:14.409 --> 00:21:19.420
two orders of magnitude greater than
the thermal expansion of quartz.

00:21:19.420 --> 00:21:22.120
So we’re really going to be
focusing primarily on the effects

00:21:22.129 --> 00:21:25.709
of heat transfer
to the fluids.

00:21:27.230 --> 00:21:31.710
So our failure criteria is that
all of these stresses have to

00:21:31.710 --> 00:21:38.259
combine such that the effective stress
reaches the tensile strength,

00:21:38.259 --> 00:21:43.299
which for most rocks is
only a few MPa, or effectively zero.

00:21:43.299 --> 00:21:51.820
So we need about an
additional 15 MPa of fluid pressure

00:21:51.820 --> 00:21:54.899
to overcome
the lithostatic stress.

00:21:56.680 --> 00:22:01.700
So we’re going to start by
looking at the stress perturbation

00:22:01.700 --> 00:22:05.820
due to the elastic displacements
around a dike.

00:22:05.820 --> 00:22:10.980
And we’re here just using the
solution from Pollard and Segall.

00:22:10.980 --> 00:22:17.090
And what I’ve plotted here,
the blue tick marks are the

00:22:17.090 --> 00:22:24.059
orientations of the most compressive
principal stress trajectory, which in 2D

00:22:24.059 --> 00:22:31.870
represents, more or less, the orientation
that hypothetical joints would form.

00:22:31.870 --> 00:22:37.019
And the difference between the
left and the right plots is this ratio

00:22:37.019 --> 00:22:40.809
of the difference in the
two remote stresses

00:22:40.809 --> 00:22:44.870
relative to the driving pressure
for dike opening.

00:22:44.870 --> 00:22:50.830
And so this left plot essentially
represents an isotropic state of stress.

00:22:50.830 --> 00:22:56.870
And we see that joints
that might form at the dike contact

00:22:56.870 --> 00:23:00.779
are perpendicular
to the dike.

00:23:01.740 --> 00:23:05.440
However, when this ratio
increases to a critical value,

00:23:05.440 --> 00:23:11.980
which happens to be about 0.6,
the stress trajectories swap

00:23:11.980 --> 00:23:16.679
and become parallel
with the dike contact.

00:23:16.679 --> 00:23:22.490
And so from this simple analysis,
we can conclude that most likely

00:23:22.490 --> 00:23:26.460
the remote stress at the time
of dike emplacement was isotropic,

00:23:26.460 --> 00:23:32.840
or nearly isotropic.
However, the elastic stresses are

00:23:32.840 --> 00:23:38.789
compressive, so we still don’t have a
mechanism to induce tension in the rock.

00:23:38.789 --> 00:23:43.450
So now we’re going to focus on the
thermal expansion of pore fluids

00:23:43.450 --> 00:23:46.559
and see how that might
elevate the pore pressure

00:23:46.559 --> 00:23:51.179
and perhaps bring
the host rock into tension.

00:23:51.179 --> 00:23:55.470
So Paul Delaney was also interested
in thermal pore fluid pressurization.

00:23:56.580 --> 00:24:01.159
And he envisioned a model for an
instantaneously emplaced dike

00:24:01.159 --> 00:24:05.220
within a fluid-saturated host rock
where the temperature change

00:24:05.220 --> 00:24:08.539
at the contact
was about 500 C.

00:24:08.539 --> 00:24:14.889
And in his model, he assumes no
fluid exchange across the contact.

00:24:14.889 --> 00:24:18.429
And for now, we’re also going to
adopt that assumption since

00:24:18.429 --> 00:24:23.490
there’s not any obvious evidence
that fluids were exchanged.

00:24:23.490 --> 00:24:28.679
So at the margins of the dikes at
Ship Rock, there tends to be

00:24:28.679 --> 00:24:33.339
a glassy rind in most places
that makes the contact pretty sharp.

00:24:33.340 --> 00:24:39.440
So we’re going to stick with this
no-flow boundary condition for now.

00:24:39.450 --> 00:24:46.009
And so in this paper, Delaney
presented an analytical solution

00:24:46.009 --> 00:24:50.460
to the coupled heat and fluid flow
equations in porous media.

00:24:50.460 --> 00:24:55.700
And at short times, heat flow
is dominated by conduction.

00:24:55.700 --> 00:25:04.070
And thermal expansion can increase the
pressure and set up a Darcy-type flow.

00:25:04.070 --> 00:25:12.320
And the estimate for the increase
in pore fluid pressure is this number,

00:25:12.320 --> 00:25:16.830
where the first term is essentially
a driving term that describes

00:25:16.830 --> 00:25:20.919
the competition between
thermal expansion of the fluids

00:25:20.919 --> 00:25:25.289
and their ability to compress
within the pore space.

00:25:25.289 --> 00:25:31.409
And then that driving term
is sort of regulated by this ratio

00:25:31.409 --> 00:25:35.519
of the thermal-to-hydraulic
diffusivity.

00:25:35.519 --> 00:25:40.629
So for very permeable rocks
with high hydraulic diffusivity,

00:25:40.629 --> 00:25:46.869
pressure can diffuse faster than it builds.
So there’s not a huge pressure build-up.

00:25:46.869 --> 00:25:50.279
But for rocks with a
low enough permeability,

00:25:50.279 --> 00:25:53.759
the pressure builds because
it can’t diffuse quickly.

00:25:53.759 --> 00:26:01.440
So using this solution, we find that
at Ship Rock, the pressure change

00:26:01.440 --> 00:26:04.220
due to thermal expansion
of the fluids is anywhere

00:26:04.220 --> 00:26:07.990
from hundredths of MPa
to hundreds of MPa.

00:26:07.990 --> 00:26:11.700
And as a side note, obviously
the pressure would not actually

00:26:11.700 --> 00:26:15.139
increase to 500 MPa.
Fracturing would occur first.

00:26:15.139 --> 00:26:21.269
But this is just sort of a
hypothetical range of pressures.

00:26:21.269 --> 00:26:26.320
And so this wide range is arising
because the permeability of rocks

00:26:26.320 --> 00:26:30.740
can span 10 orders of magnitude,
even within the same lithology.

00:26:30.740 --> 00:26:35.110
So this just highlights the need
to constrain the permeability

00:26:35.110 --> 00:26:42.269
if you want to evaluate the effects
of thermal pressurization.

00:26:42.269 --> 00:26:47.549
So I collected samples of
the fresh, unaltered Mancos shale

00:26:47.549 --> 00:26:52.070
from an outcrop a few tens of
meters away from the dike.

00:26:52.070 --> 00:26:54.649
And I had core flooding
measurements performed

00:26:54.649 --> 00:27:00.200
that showed that the permeability
of the unaltered host rock is

00:27:00.200 --> 00:27:03.259
on the order of 10 to the
minus 13th meters squared,

00:27:03.259 --> 00:27:06.549
which corresponds to hydraulic
diffusivity of about 10.

00:27:06.549 --> 00:27:10.869
And when we input this into
that analytical equation,

00:27:10.869 --> 00:27:15.549
we find that the pressure increase
is only a few tenths of MPa.

00:27:15.549 --> 00:27:18.889
And if you recall, we need
something more like 15

00:27:18.889 --> 00:27:24.129
to induce failure,
so this is not enough.

00:27:24.129 --> 00:27:28.100
However, if you recall,
the host rock is baked up

00:27:28.100 --> 00:27:32.580
against the dike and therefore
is less permeable.

00:27:32.580 --> 00:27:38.029
And so I was curious if there was any
way that this thermal alteration process

00:27:38.029 --> 00:27:45.899
was contributing to changing the
pressurization and flow around the dikes.

00:27:45.899 --> 00:27:51.169
And so, in the SEM,
we can get a sense for

00:27:51.169 --> 00:27:57.580
what the effects are of alteration
of the Mancos shale.

00:27:57.580 --> 00:28:02.909
So compared to the unaltered Mancos
shale on the left, the altered rock

00:28:02.909 --> 00:28:09.090
on the right has clearly experienced
some quartz recrystallization.

00:28:09.090 --> 00:28:15.450
And furthermore, the space between
the quartz is filled with this

00:28:15.450 --> 00:28:21.989
dark amorphous cement, which is
rich in aluminum, magnesium, and iron,

00:28:21.989 --> 00:28:27.450
but which doesn’t take on any
consistent mineral formula.

00:28:27.450 --> 00:28:33.710
And then, within this cement,
there are tiny microclasts

00:28:33.710 --> 00:28:37.629
of quartz floating around.

00:28:37.629 --> 00:28:43.659
And the effect of this alteration
on the hydraulic properties

00:28:43.659 --> 00:28:47.220
was to reduce the porosity
from something that was

00:28:47.220 --> 00:28:51.369
initially about 22%
to something like 8%.

00:28:52.900 --> 00:28:56.720
And there was a corresponding
reduction in permeability.

00:28:56.730 --> 00:29:00.720
So we used a combination of
pulse decay and core flooding

00:29:00.720 --> 00:29:08.080
experiments to find the
permeability of the host rock

00:29:08.080 --> 00:29:11.739
as a function of
distance from the contact.

00:29:11.739 --> 00:29:15.379
And what we found is that,
right at the contact,

00:29:15.379 --> 00:29:18.159
the permeability of the
host rock is on the order of

00:29:18.159 --> 00:29:20.309
10 to the minus
18th meters squared,

00:29:20.309 --> 00:29:27.139
which is a reduction of five orders of
magnitude from the fresh host rock.

00:29:27.139 --> 00:29:31.249
And the permeability
consistently increases

00:29:31.249 --> 00:29:35.570
with distance away
from the dike.

00:29:35.570 --> 00:29:40.669
And this relationship of
permeability with distance

00:29:40.669 --> 00:29:45.330
suggests the possibility that
permeability and porosity change

00:29:45.330 --> 00:29:49.539
as a function of the
amount of heat experienced

00:29:49.539 --> 00:29:53.529
as a function of
position near the dike.

00:29:53.529 --> 00:30:01.489
And so I decided to entertain that idea,
and I began with looking at

00:30:01.489 --> 00:30:07.789
what the temperature profiles
look like around a dike.

00:30:07.789 --> 00:30:14.110
And so these are just results for
a simple heat conduction model

00:30:14.110 --> 00:30:21.059
for a meter-wide dike at 1,000 degrees
in a host rock that’s initially zero.

00:30:21.059 --> 00:30:29.600
And the blue line
is the temperature in time

00:30:29.600 --> 00:30:32.360
for a position basically
right at the contact.

00:30:32.360 --> 00:30:35.200
And what we see is that
the temperature initially spikes

00:30:35.200 --> 00:30:39.019
to about 500 and then
decays back down.

00:30:39.019 --> 00:30:42.909
And then, a half a meter away
from the dike, in red,

00:30:42.909 --> 00:30:46.700
the temperature spikes
to about 350 C.

00:30:46.700 --> 00:30:49.679
And then, at a meter away,
the maximum temperature reached

00:30:49.679 --> 00:30:57.999
is only about 275.
And for reference, at these pressures,

00:30:57.999 --> 00:31:03.139
quartz recrystallization
would start at around 300.

00:31:03.139 --> 00:31:10.119
So I don’t think it’s any coincidence that
the width of the baked zone corresponds

00:31:10.129 --> 00:31:16.709
to the width of the host rock that achieved
a temperature or 300 or greater.

00:31:18.480 --> 00:31:23.300
So fueled by this potential
relationship between temperature

00:31:23.309 --> 00:31:30.739
and hydraulic properties,
I plotted the porosity and permeability

00:31:30.739 --> 00:31:36.200
as a function of this maximum
temperature reached in the host rock.

00:31:36.200 --> 00:31:40.690
And what I found was that the
porosity and max temperature

00:31:40.690 --> 00:31:43.590
have a more or less
linear relationship,

00:31:43.590 --> 00:31:49.389
and the permeability with temperature
have a log linear relationship.

00:31:49.389 --> 00:31:53.210
And this checks out when we realize
that porosity and permeability

00:31:53.210 --> 00:32:00.669
vary together exponentially in the
form of a Kozeny-Carman relationship.

00:32:00.669 --> 00:32:07.730
So I used these empirical relationships
to motivate a suite of models

00:32:07.730 --> 00:32:11.989
for thermal pore pressurization
near a dike using

00:32:11.989 --> 00:32:15.520
temperature-dependent
hydraulic properties.

00:32:15.520 --> 00:32:20.360
And I did this using Hydrotherm,
which is a finite difference code

00:32:20.369 --> 00:32:25.840
written here at the USGS
and simulates a multi-phase

00:32:25.840 --> 00:32:29.399
groundwater flow
and heat transport.

00:32:29.399 --> 00:32:35.559
So I used this to model an
instantaneously emplaced dike

00:32:35.559 --> 00:32:40.809
a meter wide, initially 1,000 degrees C
within a fluid-saturated host rock

00:32:40.809 --> 00:32:51.259
at 25 C and 10 MPa of fluid pressure to
represent the conditions at 1 kilometer.

00:32:51.259 --> 00:32:57.820
So for now, I’ve just been
solving 1D heat and fluid flow.

00:32:57.820 --> 00:33:03.179
And I’m solving for the pore pressure
change at the dike contact.

00:33:03.179 --> 00:33:06.570
And for each simulation,
I assigned hydraulic properties

00:33:06.570 --> 00:33:09.999
to the host rock, and I was able to
make them a function of temperature.

00:33:09.999 --> 00:33:15.320
And in Hydrotherm, how you do this
is you assign a value for the

00:33:15.320 --> 00:33:20.109
permeability at one temperature
and a value for another temperature.

00:33:20.109 --> 00:33:26.909
And the relationship is interpolated
between those two values.

00:33:29.330 --> 00:33:35.389
So it was not obvious to me
what the individual effects would be

00:33:35.389 --> 00:33:39.450
of changing permeability
and changing porosity.

00:33:39.450 --> 00:33:45.659
So to be systematic, I first ran a suite
of models where I fixed the porosity

00:33:45.659 --> 00:33:49.259
and made the permeability
a function of temperature.

00:33:49.259 --> 00:33:51.509
And then I did a suite of models
where I fixed the permeability

00:33:51.509 --> 00:33:54.399
and made the porosity
a function of temperature.

00:33:54.399 --> 00:33:57.509
And then finally, I made
both a function of temperature

00:33:57.509 --> 00:34:01.730
to see how they
add or don’t add.

00:34:01.730 --> 00:34:08.460
So to start, here are the
results for fixing porosity at 25%

00:34:08.460 --> 00:34:11.000
and then changing permeability.

00:34:11.000 --> 00:34:15.280
So the colors represent the
change in pore fluid pressure, where the

00:34:15.280 --> 00:34:21.540
blue is only a few MPa, and the dark red
is getting into hundreds of MPa.

00:34:21.540 --> 00:34:26.400
And the permeability at 25 C,
or the initial permeability

00:34:26.410 --> 00:34:28.790
of the host rock,
is on the X axis.

00:34:28.790 --> 00:34:34.070
And the permeability defined
at 500 C is on the Y axis.

00:34:34.070 --> 00:34:39.120
And so this black line represents no
change, or no temperature dependence.

00:34:39.120 --> 00:34:42.980
Below the line represents
a decrease in permeability,

00:34:42.980 --> 00:34:47.080
you know, due to something
like mineralization.

00:34:47.090 --> 00:34:49.320
And above the black line
would represent permeability

00:34:49.320 --> 00:34:53.190
enhancement due to
something like fracturing.

00:34:53.190 --> 00:34:59.420
And what we find from the trend of
the contours is that the initial –

00:34:59.420 --> 00:35:05.440
the initial permeability is what really
dominates the pore fluid pressure.

00:35:05.440 --> 00:35:10.090
And that the change in
permeability essentially has no effect

00:35:10.090 --> 00:35:14.730
unless that change gets to be
really extreme and ends up

00:35:14.730 --> 00:35:18.090
at something like
10 to the minus 21st.

00:35:19.380 --> 00:35:22.160
So then we fix the permeability
and make porosity

00:35:22.160 --> 00:35:25.510
as a function of temperature.

00:35:25.510 --> 00:35:29.530
And what we find is that the
trend of the contours are now

00:35:29.530 --> 00:35:32.800
a little different and increase
to the bottom right,

00:35:32.800 --> 00:35:39.340
which means that the porosity change
actually does affect the fluid pressure.

00:35:41.460 --> 00:35:48.980
And the significance of this porosity
reduction depends on the permeability.

00:35:48.990 --> 00:35:56.250
So here on the left are values for
pore pressure, changing porosity,

00:35:56.250 --> 00:36:01.200
but for a permeability of 10 to the
minus 18th, which is fairly low.

00:36:01.200 --> 00:36:06.120
And on the right is for a permeability
of 10 to the minus 13th.

00:36:06.120 --> 00:36:10.450
And what I want to
point out to you is that

00:36:10.450 --> 00:36:13.590
the values of the pore pressure
are obviously very different.

00:36:13.590 --> 00:36:22.340
But more so, for the low permeability,
the maximum pore fluid pressure change

00:36:22.340 --> 00:36:28.050
is about 6 times what it would be
if porosity had not changed.

00:36:28.050 --> 00:36:33.430
Whereas, in the high permeability case,
the max pore fluid pressure change

00:36:33.430 --> 00:36:39.290
is only about twice what it would
be without porosity reduction.

00:36:39.290 --> 00:36:46.300
So this was sort of an interesting result
for me that inelastic porosity reduction

00:36:46.300 --> 00:36:51.120
can increase the fluid pressure by
decreasing the pore volume.

00:36:51.120 --> 00:36:54.540
And that the significance
of this process depends on

00:36:54.540 --> 00:37:00.540
both the rate of change of the
porosity and also the permeability.

00:37:02.920 --> 00:37:07.990
So when both permeability and porosity
are a function of temperature, we find

00:37:07.990 --> 00:37:15.730
that, once again, the initial permeability
really dominates the effects.

00:37:15.730 --> 00:37:23.050
And however, the values of
pore pressure change are higher

00:37:23.050 --> 00:37:30.530
when we have the porosity changing,
which here was from 25% to 5%.

00:37:32.180 --> 00:37:38.060
And so I plotted the permeability and
porosity data from Ship Rock onto this

00:37:38.060 --> 00:37:45.710
solution space and found that, at most,
the pressure increase is only a few MPa.

00:37:45.710 --> 00:37:52.860
So that is still not enough to
explain the presence of these joints.

00:37:52.860 --> 00:37:57.080
So in the end, I sort of
concluded that, you know,

00:37:57.080 --> 00:38:02.460
although thermal pore pressurization
is perhaps not really a major contribution

00:38:02.460 --> 00:38:06.520
to the stress around dikes at Ship Rock,
because the initial permeability

00:38:06.520 --> 00:38:15.670
is just very high, many of these
values for permeability and porosity

00:38:15.670 --> 00:38:20.060
are realistic values for other
environments in the Earth.

00:38:20.060 --> 00:38:23.580
So perhaps these processes
might be important in other

00:38:23.580 --> 00:38:26.920
hydrothermal environments.

00:38:29.600 --> 00:38:32.420
So conclusions from the work.

00:38:32.420 --> 00:38:36.890
We found that fracturing and
entrainment of the host rock

00:38:36.890 --> 00:38:40.850
is a mechanism for
changing conduit geometry.

00:38:40.850 --> 00:38:45.610
That dike-perpendicular joints at the
dike contact are promoted by

00:38:45.610 --> 00:38:51.650
a small difference in the remote
stresses relative to the driving pressure.

00:38:51.650 --> 00:38:56.180
At Ship Rock, fluid pressure
rose at least about 15 MPa

00:38:56.180 --> 00:38:59.590
as evidenced by the
dike-perpendicular joints.

00:38:59.590 --> 00:39:04.500
That thermal pressurization alone
does not explain these joints.

00:39:04.500 --> 00:39:10.100
But as an aside, I found that
inelastic changes in porosity –

00:39:10.110 --> 00:39:15.080
that is, inelastic porosity reduction
can significantly increase

00:39:15.080 --> 00:39:17.860
fluid pressures if the
permeability is low enough.

00:39:17.860 --> 00:39:25.000
And this is an independent effect
independent of thermal pressurization.

00:39:26.490 --> 00:39:33.250
So [chuckles] when all of your models
fail to explain the geologic phenomena,

00:39:33.250 --> 00:39:38.260
you have to revisit some of
the initial model assumptions.

00:39:38.260 --> 00:39:41.880
And so for example, you know,
we assumed emplacement depth

00:39:41.880 --> 00:39:44.840
of 1 kilometer and
hydrostatic conditions.

00:39:44.840 --> 00:39:51.100
But maybe if the dike was actually
emplaced at a shallower depth,

00:39:51.110 --> 00:39:55.330
or if the fluids were over-pressured,
then the effective stress

00:39:55.330 --> 00:39:58.430
would already be
closer to failure.

00:39:59.880 --> 00:40:04.540
This is pretty difficult
to constrain in the field.

00:40:04.540 --> 00:40:09.280
But another assumption that we made
that I think is worth questioning

00:40:09.280 --> 00:40:14.010
was that there’s no fluid exchange
between the dike and the host rock.

00:40:14.880 --> 00:40:17.880
However, at these shallow depths,
we know that water will

00:40:17.890 --> 00:40:21.400
start to exsolve
out of the magma.

00:40:21.400 --> 00:40:25.260
And so – and actually,
the magmatic material in the dikes

00:40:25.260 --> 00:40:29.910
at Ship Rock have
about 15% porosity.

00:40:29.910 --> 00:40:35.800
So one direction for future research
that I’m very interested in pursuing

00:40:35.800 --> 00:40:40.570
is understanding this process
of vesiculation and trying to

00:40:40.570 --> 00:40:46.570
quantify the amount of
outgassing to the host rock.

00:40:46.570 --> 00:40:50.750
So I’m also interested in learning
about how these hydrothermal processes

00:40:50.750 --> 00:40:53.610
contribute to deformation
in other environments,

00:40:53.610 --> 00:40:58.650
such as at mid-ocean ridges
or in metamorphic aureoles.

00:40:59.860 --> 00:41:05.220
And I also want to take this time to
put in a plug for another project that I’m

00:41:05.220 --> 00:41:11.820
working on with Dave Pollard related
to the stability of magmatic dikes.

00:41:11.820 --> 00:41:16.780
So we’re using new analytical
solutions for some elastic boundary

00:41:16.780 --> 00:41:23.810
value problems to analyze things
like the effects of gravity and

00:41:23.810 --> 00:41:32.460
crustal density layering on the size
and propagation of magmatic dikes.

00:41:32.460 --> 00:41:36.820
And personally, I’m also interested
in extending some of those models

00:41:36.820 --> 00:41:41.870
to think about how vesiculation –
vesiculation rates might affect

00:41:41.870 --> 00:41:46.580
vertical propagation
or stability.

00:41:46.580 --> 00:41:51.140
So I’ll be around today,
and if anyone wants to discuss

00:41:51.140 --> 00:41:54.590
any of these projects in addition
to the work I presented,

00:41:54.590 --> 00:42:02.350
I would encourage you to catch me
or email me at my Stanford address.

00:42:03.900 --> 00:42:06.220
So thank you all
for listening.

00:42:06.220 --> 00:42:12.460
[ Applause ]

00:42:16.980 --> 00:42:18.140
- Check.

00:42:20.920 --> 00:42:23.520
- Meredith, that was a really
nice talk and a – and a –

00:42:23.530 --> 00:42:26.130
you know,
nice body of work.

00:42:26.130 --> 00:42:30.880
But I’m really a little skeptical
about how representative the

00:42:30.880 --> 00:42:34.780
permeability values you’re getting
from the near surface are.

00:42:34.780 --> 00:42:37.920
That permeability of 10 to the
minus 13th meters squared is sort of

00:42:37.920 --> 00:42:42.250
like a clean sand permeability.
- Mm-hmm.

00:42:42.250 --> 00:42:46.300
- I think a typical Mancos shale
permeability at depth

00:42:46.300 --> 00:42:51.280
would probably be less than 10 to the
minus 17th meters squared.

00:42:51.280 --> 00:42:54.720
I mean, if you look at Chris Neuzil’s
review papers of the permeabilities of –

00:42:54.720 --> 00:42:58.880
on the permeabilities of shales,
I think you’ll find 10 to the

00:42:58.880 --> 00:43:01.970
minus 13th just practically
doesn’t exist.

00:43:01.970 --> 00:43:06.010
It’s somewhere off of the …
- Yeah.

00:43:06.010 --> 00:43:09.080
- So that may be part of the issue,
that you’re just – you’re assuming

00:43:09.080 --> 00:43:13.920
those modern surficial samples are 
representative of 1-kilometer depth.

00:43:13.920 --> 00:43:20.400
And …
- Yeah. So I can respond.

00:43:20.400 --> 00:43:24.120
So – right, so we did
collect surface samples.

00:43:24.680 --> 00:43:27.000
I’ll say a couple things.

00:43:27.010 --> 00:43:31.980
One is that we did test the –
how the compressibility of the rock

00:43:31.980 --> 00:43:35.020
affects the permeability.
So we simulated the permeability

00:43:35.020 --> 00:43:42.590
measurements at the confining stress
of 25 MPa and found that this

00:43:42.590 --> 00:43:47.690
didn’t change the permeability by
more than an order of magnitude.

00:43:47.690 --> 00:43:52.730
The second thing is that the Mancos –
the Mancos Shale is the name

00:43:52.730 --> 00:43:56.250
of the formation, but it’s
over a kilometer thick in this area

00:43:56.250 --> 00:43:59.450
and it quite heterogeneous
going up section.

00:43:59.450 --> 00:44:03.080
And the section that’s
exposed at Ship Rock

00:44:03.080 --> 00:44:10.020
where these joints are occurring
is a coarser-grained section.

00:44:10.020 --> 00:44:14.510
And it’s pretty homogenous throughout
the entire Ship Rock area.

00:44:14.510 --> 00:44:20.580
And that, you know,
I am saying it’s homogenous,

00:44:20.580 --> 00:44:25.660
both because the permeability results
were homogenous and also that

00:44:25.660 --> 00:44:31.430
the joints are pretty consistently
spaced along the dikes.

00:44:31.430 --> 00:44:40.140
And so – yeah, so I guess, you know, the
question here is, you know, how much –

00:44:40.140 --> 00:44:47.370
how different could the permeability
have been 30 million years ago?

00:44:47.370 --> 00:44:53.650
And, you know, I don’t suspect the
grain size really changed that much.

00:44:53.650 --> 00:44:58.510
And we have already tested
for conditions and compaction.

00:44:58.510 --> 00:45:03.740
So unless there was some massive
regional cement that could seal

00:45:03.740 --> 00:45:09.090
these things up, I’m not sure
that I can really make this –

00:45:09.090 --> 00:45:12.100
you know, suggest that the
permeability was different

00:45:12.100 --> 00:45:15.799
by four or five orders
of magnitude at the time.

00:45:15.799 --> 00:45:19.070
- There’s been a lot of
groundwater models that include

00:45:19.070 --> 00:45:23.230
the Mancos shale as a –
as one of the units that’s simulated.

00:45:23.230 --> 00:45:25.800
What sort of permeabilities
do they invoke?

00:45:25.800 --> 00:45:29.780
- Yeah. So there are a lot of models
for the Mancos shale in the parts of

00:45:29.780 --> 00:45:33.780
the formation where – you know,
that are gas-bearing shales.

00:45:33.780 --> 00:45:38.840
And those tighter shales are quite –
have very low permeability, like,

00:45:38.840 --> 00:45:43.690
10 to the minus 18th and 19th.
Yeah.

00:45:43.690 --> 00:45:46.370
- I would just suspect that those
kind of values are more representative

00:45:46.370 --> 00:45:48.480
of kilometer depth.
Although it might be a really

00:45:48.480 --> 00:45:51.600
unusual pocket of the Mancos that
you’re – that you’re studying there.

00:45:51.600 --> 00:45:54.720
- Yeah.
I think it is, actually.

00:45:59.430 --> 00:46:03.750
- Well, the most amazing thing
to me about walking around

00:46:03.750 --> 00:46:10.440
Ship Rock is the fact that
the alteration – there’s no alteration.

00:46:10.440 --> 00:46:14.690
You get right on up to the dike,
and there’s just no alteration.

00:46:14.690 --> 00:46:20.890
So that I’m surprised the thermal
models are even relevant.

00:46:20.890 --> 00:46:25.700
And by the way, there’s an
extreme helium field just

00:46:25.700 --> 00:46:29.490
very close by, which –
whether that’s relevant or not,

00:46:29.490 --> 00:46:36.310
I don’t know about,
but heat or the fluid content …

00:46:37.660 --> 00:46:42.160
I know there’s a little –
Gene Shoemaker drew a profile

00:46:42.160 --> 00:46:46.920
of Ship Rock for me, and it does
show a little sag on the top.

00:46:46.920 --> 00:46:50.920
And I don’t know if he used the word
“minette” or whatever it is to describe it.

00:46:50.930 --> 00:46:55.530
Well, that teeny little structure
has any relevance.

00:46:55.530 --> 00:46:59.820
And I did map
a gravity low that’s quite –

00:46:59.820 --> 00:47:03.810
rather symmetric
with Ship Rock.

00:47:03.810 --> 00:47:08.280
You would think that would
enter into interpretation,

00:47:08.280 --> 00:47:13.440
but I haven’t studied it myself
since quite a while ago, so I don’t know

00:47:13.440 --> 00:47:22.940
if others have taken that into
account since that – since that time.

00:47:24.120 --> 00:47:29.100
There are a lot of – of course,
[inaudible], Agathla, and various

00:47:29.100 --> 00:47:33.630
other plugs, but Ship Rock
is in this great isolated place

00:47:33.630 --> 00:47:37.790
that you can really
try to pin it down.

00:47:37.790 --> 00:47:44.510
But I don’t see how thermal
enters into the discussion.

00:47:44.510 --> 00:47:54.920
Because this – the source – the magma
source down below is of consideration.

00:47:54.920 --> 00:47:58.460
And how this dike whipped its
way through [chuckles] –

00:47:58.460 --> 00:48:03.560
dramatically through the Mancos
is merely the top-most layer.

00:48:03.560 --> 00:48:09.210
And then all that layered to the tip of the
Ship Rock, that’s not exposed anymore.

00:48:09.210 --> 00:48:11.360
It’s kind of interesting.

00:48:11.360 --> 00:48:15.400
I don’t know if I have a specific
point that you’d like to address.

00:48:16.120 --> 00:48:21.260
- Yeah. Well, I’m fascinated by
the detail in which you can

00:48:21.260 --> 00:48:24.450
describe these outcrops.
That’s really awesome.

00:48:24.450 --> 00:48:29.310
They are wonderful outcrops
for looking at these processes –

00:48:29.310 --> 00:48:35.680
that is, deformation around dikes
and the growth of buds and plugs.

00:48:35.680 --> 00:48:41.390
So you’re right that there has not
been regional thermal alteration

00:48:41.390 --> 00:48:47.100
the way that there is in, you know,
larger plutonic environments.

00:48:47.100 --> 00:48:55.920
But I just – we’ll come back to this
picture to show that the amount of

00:48:55.930 --> 00:49:02.390
alteration is definitely low, but it does
happen, and it’s really localized, though.

00:49:02.390 --> 00:49:07.230
It’s localized to this zone, you know,
less than a meter wide around the dike.

00:49:07.230 --> 00:49:08.740
- Oh, I see.
- Yeah.

00:49:08.740 --> 00:49:10.910
So that’s the alteration
that I’m referring to.

00:49:10.910 --> 00:49:15.500
It’s just that very local alteration
around the dikes that causes the rock

00:49:15.500 --> 00:49:20.740
to be sort of baked up and –
against the dike and preserved.

00:49:22.820 --> 00:49:31.880
[ Silence ]

00:49:33.040 --> 00:49:37.600
- Early in your talk, you alluded
to an interpretation that these

00:49:37.610 --> 00:49:44.470
radial dikes are fed vertically rather
than laterally from the Ship Rock center.

00:49:44.470 --> 00:49:48.680
What’s the evidence for that?
- Yeah. So ...

00:49:48.680 --> 00:49:51.700
- And how might it bear
in your interpretations?

00:49:51.700 --> 00:49:54.580
- Mm-hmm. Yeah.
Good question.

00:49:54.580 --> 00:50:05.580
So the northeastern dike, at least,
is broken up into about 35 echelon

00:50:05.580 --> 00:50:11.140
segments across this entire length,
which is about 2 kilometers.

00:50:11.140 --> 00:50:16.020
So, you know, one constraint
is just that the segmented nature

00:50:16.020 --> 00:50:20.560
of the dike would prevent
significant lateral flow.

00:50:21.580 --> 00:50:30.700
And there are also cusps where the
dike is sort of rotating where you can

00:50:30.710 --> 00:50:35.910
measure the dip of the cusp, which
indicates the propagation direction.

00:50:35.910 --> 00:50:40.420
And that is vertical
for the northeastern dike.

00:50:41.620 --> 00:50:45.840
However, if, for some reason,
we were thinking about

00:50:45.850 --> 00:50:54.540
lateral propagation, then much of
the same things would actually apply.

00:50:54.540 --> 00:51:01.290
So the parallel joints would still
be parallel with the dike contacts.

00:51:01.290 --> 00:51:07.450
And, you know, I don’t know about
the perpendicular joints, but …

00:51:07.450 --> 00:51:14.230
- Okay. Two questions and a comment.
If the – if you had a third dimension

00:51:14.230 --> 00:51:17.390
of exposure in any length,
might you find that some of the

00:51:17.390 --> 00:51:24.090
on-echelon segments really connect
at depth and are fingering out

00:51:24.090 --> 00:51:27.290
vertically rather than …
- Mm-hmm.

00:51:27.290 --> 00:51:31.520
- … but connected so that they
could have flowed laterally?

00:51:31.520 --> 00:51:37.180
And I just wondered whether you had
seen any fluting or grooving in the walls

00:51:37.180 --> 00:51:41.240
that would indicate vertical directions
or whether anybody had ever tried

00:51:41.240 --> 00:51:44.980
magnetic and isotropy
measurements.

00:51:44.980 --> 00:51:50.440
Because in other radial dike systems
of different composition that have

00:51:50.440 --> 00:51:56.550
minerals like amphibole
that are linear, it’s very common

00:51:56.550 --> 00:52:00.960
to see sub-horizontal directions.
- Mm-hmm.

00:52:00.960 --> 00:52:07.700
Yeah. So there are not
well-preserved grooves.

00:52:07.700 --> 00:52:13.640
And to my knowledge, no one
has done magnetic anisotropy.

00:52:13.640 --> 00:52:17.110
In some fin sections that I looked at,
I cut some fin sections in

00:52:17.110 --> 00:52:19.040
different orientations
of the dike.

00:52:19.040 --> 00:52:25.840
And I noticed that some of the bubbles
were sort of deformed vertically.

00:52:25.840 --> 00:52:30.600
So that’s sort of an indication also that,
at least at the last moments, you know,

00:52:30.600 --> 00:52:34.650
things were moving
either up or down.

00:52:34.650 --> 00:52:38.330
But yeah, the segments
probably do connect at depth.

00:52:38.330 --> 00:52:40.800
- The bubbles would rise
vertically regardless of

00:52:40.800 --> 00:52:44.580
[how the magma was in place].
- Sure. But there are phenocrysts

00:52:44.580 --> 00:52:52.980
of the phlogopite that are also oriented –
sort of imbricated vertically.

00:52:55.340 --> 00:52:57.500
- I know this is kind of – kind of –
certainly beyond the scope of

00:52:57.510 --> 00:53:01.120
what you’ve been doing, but would you
like to speculate on what we might look

00:53:01.120 --> 00:53:06.260
for in real-time data to see if – see –
maybe see this process going on?

00:53:06.260 --> 00:53:10.740
- Yeah. [laughs]
That’s a great question.

00:53:12.960 --> 00:53:18.200
Of course, active volcanoes are, you
know, really difficult to get close to.

00:53:18.200 --> 00:53:21.690
And, you know, any deformation
that’s this localized

00:53:21.690 --> 00:53:25.030
would get covered up
by eruptive products.

00:53:25.030 --> 00:53:29.290
I suppose one thing that could be –
you know, I don’t know if this

00:53:29.290 --> 00:53:32.010
would be fruitful or not,
but I know that sometimes

00:53:32.010 --> 00:53:37.470
a lot of fissure eruptions will
sort of drain back into the conduit.

00:53:37.470 --> 00:53:42.920
And if there were somehow, I guess,
a way to get into the conduit and, like,

00:53:42.920 --> 00:53:46.340
look along the walls for
any evidence of, like,

00:53:46.340 --> 00:53:50.930
deformation and – I don’t know,
maybe something like that.

00:53:50.930 --> 00:53:54.670
- [inaudible]
[laughter]

00:53:54.670 --> 00:53:59.620
- Maybe a few years after everything
has cooled off. [laughs]

00:53:59.620 --> 00:54:02.740
Send a drone down there.

00:54:02.740 --> 00:54:05.540
- I had one final question,
Meredith.

00:54:05.540 --> 00:54:10.760
I was curious about the
timing of the dike-parallel joints

00:54:10.760 --> 00:54:12.800
and how that fits in
with your model.

00:54:12.800 --> 00:54:17.000
It just seems like you need those to be
later than the perpendicular joints

00:54:17.000 --> 00:54:19.970
because it might affect the permeability?
- Ah.

00:54:19.970 --> 00:54:21.980
- Next to the dike?
- Yeah.

00:54:21.980 --> 00:54:28.890
So the dike-parallel joints,
we pretty much know are a result

00:54:28.890 --> 00:54:32.730
of the concentration at the
tip of the dike as it was propagating

00:54:32.730 --> 00:54:38.060
and that they occurred prior to
the dike-perpendicular joints.

00:54:38.070 --> 00:54:42.080
So the timing on that
is pretty well-constrained.

00:54:42.080 --> 00:54:46.880
The parallel joints would
affect the permeability.

00:54:46.880 --> 00:54:51.010
But not in every direction.

00:54:51.010 --> 00:54:57.610
And the direction that I’m interested in
is perpendicular to the dike contact.

00:54:57.610 --> 00:55:03.840
So if these joints are parallel with
the contact, then, you know,

00:55:03.840 --> 00:55:09.500
assuming there’s enough compression
on those parallel joints due to the

00:55:09.500 --> 00:55:13.940
magma pressure, then I don’t –
I suspect that the permeability

00:55:13.940 --> 00:55:19.019
wouldn’t be terribly altered
in the direction perpendicular.

00:55:19.019 --> 00:55:24.840
- Maybe there’s mineralization
in those joints as well

00:55:24.840 --> 00:55:32.360
that might inhibit permeability?
- Yeah. I don’t see any, but I –

00:55:32.360 --> 00:55:35.240
you know, I don’t know
what it looked like at the time.

00:55:35.240 --> 00:55:39.580
I don’t know how fast
mineralization takes place.

00:55:39.580 --> 00:55:42.140
I suspect these – you know,
these joints formed, you know,

00:55:42.140 --> 00:55:44.940
one after the other,
like, pretty quick.

00:55:44.940 --> 00:55:47.060
- Okay. Thank you very much.

00:55:47.080 --> 00:55:51.040
And if you’d like to join us for lunch,
I think we’re going to meet out front

00:55:51.040 --> 00:55:54.300
by the flagpole area in five or 10 minutes
and walk over to the patio.

00:55:54.300 --> 00:55:56.640
Let’s thank Meredith one more time.
- Thanks.

00:55:56.640 --> 00:56:00.340
[ Applause ]

00:56:00.340 --> 00:56:32.100
[ Silence ]