WEBVTT
Kind: captions
Language: en-US

00:00:02.640 --> 00:00:04.296
Good afternoon.

00:00:04.320 --> 00:00:07.680
The topic of my talk today is
the magnitude 6.2 offshore

00:00:07.680 --> 00:00:12.216
Petrolia earthquake that took
place on December 20th, 2021.

00:00:12.240 --> 00:00:14.960
The subtitle, slip on the
Mendocino Fault Zone,

00:00:14.960 --> 00:00:16.400
has a question mark
attached to it.

00:00:16.400 --> 00:00:21.416
And the purpose of that question mark
will be made clear as the talk goes on.

00:00:21.440 --> 00:00:25.280
This talk contains contributions
from many other people at the

00:00:25.280 --> 00:00:28.800
Earthquake Science Center,
at the NEIC, and at UC-Berkeley.

00:00:28.800 --> 00:00:32.486
And I’m very grateful to
everyone who contributed.

00:00:33.360 --> 00:00:39.040
To begin, let’s start by looking at
the tectonics of the area where

00:00:39.040 --> 00:00:42.776
this earthquake took place –
the Mendocino Triple Junction.

00:00:42.800 --> 00:00:46.160
The Mendocino Triple Junction
is the intersection of the Pacific,

00:00:46.160 --> 00:00:49.096
North American,
and Juan de Fuca Plates.

00:00:49.120 --> 00:00:52.080
It’s one of the most seismically
active parts of California,

00:00:52.080 --> 00:00:55.576
and it has hosted several
significant earthquakes in the past.

00:00:55.600 --> 00:01:00.400
Some of these earthquakes include
the 1992 magnitude 7.2 Petrolia

00:01:00.400 --> 00:01:04.456
earthquake, which is shown here –
a reverse fault.

00:01:04.480 --> 00:01:10.536
And the 1980 magnitude 7.2
Trinidad earthquake, a strike-slip fault.

00:01:10.560 --> 00:01:13.200
Both of these earthquakes
interestingly produced aftershocks

00:01:13.200 --> 00:01:18.456
on structures other than
the main shock structure.

00:01:18.480 --> 00:01:22.960
An important plate boundary fault in
this area is the Mendocino Fault Zone,

00:01:22.960 --> 00:01:30.456
a right-lateral oceanic transform fault
separating the Pacific and Gorda Plates.

00:01:30.480 --> 00:01:35.840
So this area produces quite a few
earthquakes and people living in

00:01:35.840 --> 00:01:40.818
Humboldt County and Eureka are
no strangers to feeling earthquakes.

00:01:42.160 --> 00:01:46.160
So, on December 20th, 2021,
around midday,

00:01:46.160 --> 00:01:49.496
a magnitude 6.2 earthquake
took place.

00:01:49.520 --> 00:01:52.160
What we knew about the
earthquake initially is that

00:01:52.160 --> 00:01:57.120
its event location appeared to
be offshore, as shown here.

00:01:57.120 --> 00:02:02.000
And its focal mechanism appeared
to be a strike-slip earthquake.

00:02:02.000 --> 00:02:05.360
So, armed with these two pieces
of information, it seemed likely

00:02:05.360 --> 00:02:09.176
that the event took place on
the Mendocino Fault Zone.

00:02:09.200 --> 00:02:13.040
It was felt widely throughout the area,
and in some cases, the shaking

00:02:13.040 --> 00:02:17.360
was strong enough to be disruptive,
as shown here in a grocery store

00:02:17.360 --> 00:02:19.709
in Ferndale, California.

00:02:20.560 --> 00:02:26.000
For the next few slides of the talk,
I’m going to go into some of the

00:02:26.000 --> 00:02:31.976
geophysical data that was produced
during this earthquake and then

00:02:32.000 --> 00:02:36.720
move into the implications of
what that geophysical data implies

00:02:36.720 --> 00:02:39.736
for the source processes
of the earthquake.

00:02:39.760 --> 00:02:43.571
So first let’s look at
ground motion data.

00:02:44.400 --> 00:02:49.040
The ground motion coming in from
seismic stations throughout northern

00:02:49.040 --> 00:02:55.656
California shows shaking intensity
of up to MMI 7 near Petrolia.

00:02:55.680 --> 00:03:00.560
This is quite strong for an offshore
event, suggesting almost immediately

00:03:00.560 --> 00:03:04.240
that the event was either closer
to shore than was initially reported

00:03:04.240 --> 00:03:07.976
or that it had an eastward
rupture directivity, perhaps.

00:03:08.000 --> 00:03:12.960
As an example, in the town of Ferndale,
California, the PGV was up to

00:03:12.960 --> 00:03:19.176
25 centimeters per second,
and the PGA was about 29% g.

00:03:19.200 --> 00:03:22.936
So next we can look at the
seismic data from the event.

00:03:22.960 --> 00:03:27.016
And shown here are two
types of record sections.

00:03:27.040 --> 00:03:33.976
On the left is a record section with
local stations from David Shelly.

00:03:34.000 --> 00:03:39.595
And on the right is a more teleseismic
record section from Paul Earle.

00:03:40.560 --> 00:03:44.320
Relatively quickly,
we were also able to pull

00:03:44.320 --> 00:03:49.496
a solution from the
GCMT catalog for the event.

00:03:49.520 --> 00:03:54.776
Now, these seismic data have
some unusual characteristics.

00:03:54.800 --> 00:03:59.416
I highlighted them here in green.
Some of the unusual features are

00:03:59.440 --> 00:04:05.256
an arrival between the P wave and the
S wave in the local record section,

00:04:05.280 --> 00:04:10.080
and in the case of the GCMT catalog,
a parameter called centroid time

00:04:10.080 --> 00:04:15.096
minus hypocenter time ended up
being 12 seconds in this solution.

00:04:15.120 --> 00:04:20.320
12 seconds is unusually long because
other events in this area of this

00:04:20.320 --> 00:04:25.520
magnitude have centroid time
minus hypocenter time more typically

00:04:25.520 --> 00:04:29.736
between 1 and 4 seconds.
So 12 seconds is quite long.

00:04:29.760 --> 00:04:35.280
Another interesting feature is that
the location of the GCMT solution

00:04:35.280 --> 00:04:39.782
is farther to the east than the
initial location of the event.

00:04:40.800 --> 00:04:45.600
Another interesting seismological
observation is that the aftershocks

00:04:45.600 --> 00:04:49.600
fall into two distinct clusters.
One of these clusters is aligned

00:04:49.600 --> 00:04:54.160
with the Mendocino Fault Zone here,
and the other is onshore and is a little

00:04:54.160 --> 00:04:59.360
bit deeper – around 25 kilometers deep.
These two distinct clusters of

00:04:59.360 --> 00:05:04.216
aftershocks became apparent almost
immediately, in the first day even,

00:05:04.240 --> 00:05:10.240
but the aftershock activity has continued
over the few weeks after the earthquake

00:05:10.240 --> 00:05:17.576
and remains relatively vigorous for an
event in the Mendocino Triple Junction.

00:05:17.600 --> 00:05:23.736
So next we will look at some of
the geodetic data from this event.

00:05:23.760 --> 00:05:29.600
Initially, on the first day, and in the
very few days after, it appeared that

00:05:29.600 --> 00:05:39.496
there were relatively small, perhaps
unable to be determined, offsets.

00:05:39.520 --> 00:05:44.640
So, from a Sentinel-1 interferogram,
shown here on the left, this is a 12-day

00:05:44.640 --> 00:05:52.000
interferogram spanning the event.
It is difficult to see any offsets at all.

00:05:52.000 --> 00:05:57.016
And, in the case of the rapid –
the high-rate GNSS solutions,

00:05:57.040 --> 00:06:02.144
it also appears that the offsets
are within the level of the noise.

00:06:02.880 --> 00:06:08.608
This is not terribly surprising for
an event with an offshore location.

00:06:09.120 --> 00:06:12.720
However, in the days after the
earthquake, as more data came in,

00:06:12.720 --> 00:06:18.160
it became clear that there may be some
small offsets in the GNSS time series.

00:06:18.160 --> 00:06:24.320
So, on the left, I’m showing plots of east
and north position at GNSS stations.

00:06:24.320 --> 00:06:28.696
This is daily position at
GNSS stations in this area.

00:06:28.720 --> 00:06:36.695
And, on the right, we’re seeing a plot
of offsets produced by Jerry Svarc,

00:06:36.720 --> 00:06:40.696
solved from the GNSS time series –
the daily time series.

00:06:40.720 --> 00:06:46.856
The offsets are quite small –
around 2 to 4 millimeters, typically.

00:06:46.880 --> 00:06:51.200
They’re small enough that it’s quite
difficult to see them visually in the

00:06:51.200 --> 00:06:58.776
time series at all, but they are spatially
coherent, and they do make sense for

00:06:58.800 --> 00:07:02.720
some of the – for some of the source –
estimates of source processes

00:07:02.720 --> 00:07:04.320
for this event.
So we think that they are

00:07:04.320 --> 00:07:09.128
real offsets even though they
are only 2 to 4 millimeters.

00:07:10.080 --> 00:07:14.216
So, when we look at all of these
geophysical observations in total,

00:07:14.240 --> 00:07:17.920
some of them, as I mentioned,
are unusual for an event of the

00:07:17.920 --> 00:07:22.349
initial location – the strong on-land
shaking, the second phase arrival,

00:07:22.349 --> 00:07:26.545
the parameters of the GCMT solution,
the aftershock pattern, and some of the

00:07:26.545 --> 00:07:33.656
 offsets. These are a little bit unusual.
And, as a result of these observations,

00:07:33.680 --> 00:07:37.496
several people at the USGS and
UC-Berkeley got together

00:07:37.520 --> 00:07:41.840
and determined that there was
another solution for the origin

00:07:41.840 --> 00:07:45.976
of the earthquake that made more
sense in light of these observations.

00:07:46.000 --> 00:07:51.016
And that solution
was two earthquakes.

00:07:51.040 --> 00:07:55.520
They are separated by about
11 seconds, and they are separated

00:07:55.520 --> 00:08:01.200
by 30 kilometers in spatial distance.
One of them – the first was a smaller

00:08:01.200 --> 00:08:06.056
event of magnitude 5.7, and the second
was larger – about magnitude 6.2.

00:08:06.080 --> 00:08:11.680
So the solution that is now presented
in the USGS catalog has two events –

00:08:11.680 --> 00:08:17.148
5.7 being here offshore
and the 6.2 onshore.

00:08:18.160 --> 00:08:25.040
The moment tensor for this event
is likely dominated by the

00:08:25.040 --> 00:08:28.696
moment release of the
larger of the two events.

00:08:28.720 --> 00:08:35.120
And this new solution has more success
at explaining some of those unusual

00:08:35.120 --> 00:08:40.080
observations, such as the second phase
arrival and the strong on-land shaking.

00:08:40.080 --> 00:08:43.600
It also has success in
explaining the GNSS offsets.

00:08:43.600 --> 00:08:47.336
This is ongoing research.
It’s very preliminary.

00:08:47.360 --> 00:08:52.880
But we’re able to forward model
two point sources. It’s not an inverse.

00:08:52.880 --> 00:08:57.360
It’s simply a forward model of two
point sources – one located onshore

00:08:57.360 --> 00:09:02.376
and one located offshore – and forward
models the GNSS displacements.

00:09:02.400 --> 00:09:08.160
What we find is quite consistent
with the observations in the quadrants

00:09:08.160 --> 00:09:12.992
and the overall pattern of
the displacement field.

00:09:13.520 --> 00:09:18.160
This forward model also allows us
to understand why the InSAR

00:09:18.160 --> 00:09:23.280
displacements from the observations
were so low because the predicted

00:09:23.280 --> 00:09:28.080
vertical here in this forward model
is only plus or minus 2 millimeters,

00:09:28.105 --> 00:09:33.417
which is very difficult
to observe using InSAR.

00:09:34.400 --> 00:09:37.920
So let’s look at the implications
of these events and place the

00:09:37.920 --> 00:09:42.616
December 20th events in the context
of the Mendocino Triple Junction.

00:09:42.640 --> 00:09:47.360
Typical seismicity in this area
includes strike-slip earthquakes

00:09:47.360 --> 00:09:51.279
on the Mendocino Fault Zone,
as you can see here.

00:09:52.240 --> 00:09:54.960
And it also includes strike-slip
earthquakes within the

00:09:54.960 --> 00:09:59.680
Gorda Plate itself. There have been
large earthquakes, up to magnitude 7 –

00:09:59.680 --> 00:10:02.080
strike-slip earthquakes within
the Gorda Plate, and their

00:10:02.080 --> 00:10:06.500
orientations are in a northeast
or southeast trend.

00:10:07.440 --> 00:10:10.400
Some of these earthquakes, based on
their aftershock patterns, such as the

00:10:10.400 --> 00:10:16.856
2014 earthquake, appear to be on
this northeast-trending lineation.

00:10:16.880 --> 00:10:19.840
However, there’s also historical
precedent for some of these earthquakes

00:10:19.840 --> 00:10:26.474
taking place on the southeast trend
of this focal mechanism.

00:10:27.120 --> 00:10:30.560
And then, within the Triple Junction
itself, the seismicity becomes

00:10:30.560 --> 00:10:34.160
a little bit more of a diffuse cloud,
but there are some identifiable

00:10:34.160 --> 00:10:39.256
bands and structures
within the Triple Junction.

00:10:39.280 --> 00:10:45.040
So the events of December 20th suggest
that the first one – first event may have

00:10:45.040 --> 00:10:48.856
been on the Mendocino Fault Zone,
as shown here – a strike-slip event.

00:10:48.880 --> 00:10:51.440
And that the second,
with its deeper depth –

00:10:51.440 --> 00:10:58.207
about 27 kilometers,
is likely within the Gorda slab.

00:10:58.894 --> 00:11:00.880
There haven’t been
too many of these

00:11:00.880 --> 00:11:04.056
strike-slip earthquakes right
around this area, though.

00:11:04.080 --> 00:11:08.000
And the location and the timing
of the two events do suggest

00:11:08.000 --> 00:11:11.859
a dynamic triggering relationship
between the two of them.

00:11:12.880 --> 00:11:17.576
Another interesting feature about this
area is that the Mendocino Fault Zone

00:11:17.600 --> 00:11:23.520
has some indication of partial creep
from repeating earthquakes and from

00:11:23.520 --> 00:11:30.480
accounting of moment budget.
But it is important to remember

00:11:30.480 --> 00:11:35.120
that even partially creeping faults
are still capable of magnitude 6

00:11:35.120 --> 00:11:38.936
and larger earthquakes.
And this event shows the

00:11:38.960 --> 00:11:45.576
seismic potential of a fault like
the Mendocino Fault Zone.

00:11:45.600 --> 00:11:49.280
So what have we learned?
We have seen a moderate-sized

00:11:49.280 --> 00:11:53.200
earthquake on the Mendocino Fault
Zone and a larger inter-slab earthquake.

00:11:53.200 --> 00:11:57.360
These events, in general, are consistent
with historical seismicity in the area,

00:11:57.360 --> 00:12:02.080
but the magnitude 6.2 is of
a mechanism and a location

00:12:02.080 --> 00:12:07.096
that we have not seen
too many of in recent years.

00:12:07.120 --> 00:12:10.960
The two events were separated by
about 11 seconds and 30 kilometers,

00:12:10.960 --> 00:12:15.816
and they produced a vigorous
onshore aftershock sequence.

00:12:15.840 --> 00:12:20.080
The geodetic displacements are quite
small, but they are consistent with the

00:12:20.080 --> 00:12:26.136
story of the source processes that
we’re deriving from seismic data.

00:12:26.160 --> 00:12:28.080
And what is there
still to learn?

00:12:28.080 --> 00:12:33.200
There are a lot of ongoing research
directions generated by this earthquake.

00:12:33.200 --> 00:12:37.840
For example, where exactly was the
moment released, and how much?

00:12:37.840 --> 00:12:42.776
What is the connection between
the offshore and onshore faults?

00:12:42.800 --> 00:12:47.096
And interestingly, what structure
is hosting the onshore events?

00:12:47.120 --> 00:12:49.760
These events underscore that the
Mendocino Triple Junction is a

00:12:49.760 --> 00:12:54.296
complex area with a history of
earthquakes on multiple faults.

00:12:54.320 --> 00:12:57.680
Modeling using multiple data sets
will be needed to understand the

00:12:57.680 --> 00:13:01.280
connections between these events,
specifically the structures that hosted

00:13:01.280 --> 00:13:05.176
them, the aftershocks sequences,
and the major plate boundaries.

00:13:05.200 --> 00:13:10.080
And lastly, the events raise some
interesting questions about seismic

00:13:10.080 --> 00:13:14.160
and aseismic slip on faults and
about dynamic triggering relationships

00:13:14.160 --> 00:13:19.250
between earthquakes in the
Triple Junction. Thank you.

00:13:20.609 --> 00:13:27.187
[silence]