Mount St. Helens Rocked Our World! What we've learned since 1980.

Video Transcript
Download Video
Right-click and save to download

Detailed Description

What stories do rocks tell? What techniques do scientists use to study volcanoes? Dr. Heather Wright talks about the May 18, 1980 eruption of Mount St. Helens, provides an overview of volcanoes and how they erupt, and shows why scientists continue to monitor this active volcano, in this presentation to the Sno-Isle Libraries’ 2021 Whidbey Reads program.


Date Taken:

Length: 00:54:49

Location Taken: Vancouver, WA, US

Video Credits

Video edited by Liz Westby


The title of my talk today is Mount St. Helens rocked our world,

what we've learned since 1980.

This talk is part of a series of talks

that the Whidbey Reads program is helping to sponsor.

Each talk will give you a little bit of a different flavor of

activities that are happening at the USGS,

of ways that we can reflect back on the May 18th,

1980 eruption and things that we've learned.

My particular specialty at the USGS is to

focus on the rocks themselves and the stories that the rocks tell.

But there are a variety of different techniques that people use to study volcanoes.

You'll hear from people who specialize in things quite different from

my own specialty later on throughout this winter to spring.

By means of an introduction,

I want to acknowledge first that the volcano that

we call Mount St. Helens lies on native tribal lands.

It's now a designated traditional cultural property.

The volcano's known as Lawetlat’la by

the Cowlitz Indian Tribe and by the confederated tribes and bands of the Yakama nation.

This video that you're seeing here gives

a great introduction to the spectacular topography and

geology that's now exposed by the cataclysmic eruption in 1980.

You can see all of these gorgeous layers of rock that have

been excavated by a big landslide that happened in 1980.

Each of the layers that are shown here represent

different eruptions that through time, over thousands of years, have constructed

the big volcano that produced a conical peak that

existed in 1980 prior to that big eruption.

Before I get started,

I think the first thing I'll do is just give a little bit of

a primer on volcanoes themselves.

What happens when magma comes into the system?

Despite the fact that we cannot see the processes happening beneath the surface,

because they are beneath the surface,

there are remote proxies for magma ascent that we can use.

What I mean by this is that as magma

intrudes into some sort of reservoir beneath the surface,

it needs to make space to move its way towards the surface.

That breakage of a pathway can create

earthquakes in the brittle crust that surrounds the magma itself.

We can use seismometers to detect these earthquakes.

Magma also, as it moves its way towards the surface, starts to depressurize.

That is, the pressure at the surface is lower than that at depth.

Much like a bottle of soda when you take the lid

off and you see gas bubbles start to form,

in that case, they are carbon dioxide bubbles.

In magma, it's water, it's carbon dioxide,

and it's sulfur dioxide that's dissolved in magma at

depth starts to form bubbles as it moves its way towards the surface.

If those bubbles can separate from magma itself,

they can actually move their way through the overlying rocks,

and you might see gas escaping from the surface

before any magma makes its way all the way up to the surface.

Finally, the reservoir of magma may

increase or decrease in volume if new magma is being input into the system.

Again, there's a room problem.

Where does all that magma gets stored?

The volcano can deform or swell or in this case,

of Mount St. Helens in 1980,

actually make a big bulge accommodating all that magma coming into the system,

trying to find a place to reside.

Luckily for us, we have all these remote tools to detect

processes that we can't see at all since we cannot see beneath the surface.

We can look for earthquakes.

We can look at the surface to see if there's evidence of deformation.

We can measure escape of gases.

We use a variety of different tools that are shown in this slide here.

Tools to measure deformation.

Those include tilt meters installed on the volcano,

GPS sensors that measure

absolute position and then we look to see if that position changes.

Gas sensors that might be both airborne or ground-based.

Now we can use drones to fly gas sensors over volcanoes.

We can use remote sensing techniques,

that is satellites or cameras,

webcams that are online.

If you go to USGS pages,

you can see those cameras today.

We also use seismic instruments to detect earthquakes.

But what I'm going to focus on today is the part of

the story that is how we forecast eruptions,

how the eruptions were forecast before the May 18th,

1980 eruption, and then how that has changed or evolved through time.

What lessons we learned from the 1980 eruption

and ways that forecasting has changed to now.

Forecasts themselves involve a variety of different informational pieces,

including when the eruption's going to happen,

what type of eruption you might see,

and where that erupted material is going to travel.

This graphic here provides just a cartoon version of a rough way

we might look for changes in those monitoring data that could signal an eruption.

You see on the y-axis of this graph,

it says intensity of monitoring parameters.

Let's just take an example of a monitoring parameter that is seismicity.

If you think of this as number of earthquakes per day,

then as we move forward in time,

so from the left-hand side to the right-hand side of this figure,

you can see an orange line and it starts to climb.

Perhaps in this hypothetical situation,

we see the number of earthquakes getting more and more frequent through time.

At a certain point,

decision-makers have to issue some advisory or warning.

Scientists can provide a forecast to help decision-makers decide what to do.

Now the difficulty is that there is no consistent pattern

at volcanoes that always when seismicity ramps up,

it always leads to eruption.

Our hypothetical scenario, again,

if we look at the cartoon on the right,

what we imagine things might do is that as magma moves towards the surface,

earthquakes start to occur as the pressure builds in that magma reservoir,

earthquakes start to happen more and more frequently.

Everything accelerating, pressure building until finally

some critical threshold is reached and then magma erupts at the surface.

That would be like this path 1 here.

Unfortunately, it doesn't always happen this way.

In fact, many, many times,

an increase in activity can lead to nothing.

The volcano can go back to sleep with no magma reaching

the surface at all and then even more complex,

we might envision a scenario and this

has occurred in many different places around the world,

where activity increases, it plateaus,

it stays at an elevated level,

and then either it erupts or goes back to sleep.

All of these different cartoon paths,

just to say that a forecast that

a scientist issues is going to have a bunch of uncertainty

about when the eruption is going to happen because

there's no consistent behavior that we could always count on.

Now moving on to forecasting.

What might happen?

At any volcano, the type of eruption that might occur is very much dependent,

we learn the most about what might happen, based on the history of that volcano.

The past is the key to the future.

Mount St. Helens has its own personality,

just like every volcano has a different personality.

Mount St. Helens is a volcano that's over 275,000 years old and over this range of time,

it's erupted a wide range of eruption styles.

Some volcanoes are pretty consistent,

they really like one particular eruption style,

one composition, but Mount St. Helens has a wide range of different behavior.

I'm going to step through some cartoons that imagine some of

the stages of this volcanoes growth over largely the last

3,000 years and that's the interval that's produced most of what you see

when you look across the landscape at the big edifice that is Mount St. Helens.

Eruptions include formations of big domes and a broad dome field,

and those domes began to accumulate building more and more topography.

There's also been subsequent landslides or debris avalanches

that have caused or that are the result of collapses of part of that topography.

As we move forward in time,

the volcano then started to experience Strombolian eruptions,

these vigorous fire fountaining eruptions, also production of lava flows.

By the year 1479,

there were a sequence of very explosive eruptions that happened,

producing big ash columns that distributed ash and pumice across the landscape.

In fact, in the span of three years,

there were two consecutive Plinian eruptions,

in 1479 and 1482.

By the year 1520, again,

you can see the volcano was rebuilding itself after a big landslide,

producing a bunch of dacite domes and pyroclastic flows,

which are processes I'll tell you a little bit more about in several slides.

By 1720, the volcano was looking much more similar to its shape as of 1980.

Again, lava flows were produced,

a dacite summit dome,

that is a really sticky,

viscous pile of lava that makes a dome shape,

and again, pyroclastic flows.

Then this final image is from 1857,

products of the Goat Rocks,

dacite lava dome that are now cut and exposed by the crater at Mount St. Helens,

and the Floating Island lava flow that extends down the northwestern side.

We have a broad range of different things in

our forecasts that are possible for scenarios

of eruptive activity at the time of the 1980 forecasts.

Now forecasting where things might go,

depends not only on the eruption style,

so that's the eruptive history,

but also, the shape of the volcano.

This is a photograph of Mount St. Helens from before the eruption of May 18th,

and you can see it's a steep-sided cone with big valleys that extend off of the summit,

and any dense flows, lava flows,

or even pyroclastic flows,

might be funneled by this topography down the river valleys.

A lot of this information was known and

studied by two scientists named Crandell and Mullineaux,

who produced a publication in just two years before the 1980 eruption in 1978,

and the publication was called Potential

Hazards From Future Eruption of Mount St. Helens.

Importantly, I have just extracted a little bit of the text,

they said in that publication,

that an eruption is more likely to occur within the next 100 years,

and perhaps even before the end of the century.

They said this because they recognized that

the eruption history of Mount St. Helens showed fairly frequent activity,

much more frequent eruptions than any other volcano in the Cascades.

This was a pretty bold statement in a geologic publication,

but it was based on their solid studies of all of

the activity over the past hundreds to thousands of years.

Forecast in 1980, included eruptive scenarios that,

extended from dome formation at the summit,

to production of a vertical eruption column that might have dropped

pumice and ash across the landscape, also pyroclastic flows.

Pyroclastic flows are produced when that ash,

and pumice and hot gas that rises out of the volcano,

becomes denser than the atmosphere.

It follows topography, it goes down

away from the summit of the volcano as I've shown in the cartoon here.

They also recognized that a landslide was possible,

a debris avalanche, and also a lateral blast was possible,

although this was certainly not the subject of a lot of discussion

at volcanoes around the world prior to the May 18th, 1980 eruption.

Finally, mudflows, which were possible to be produced

by remobilization of water or melting of snow and ice,

that is liberated during the eruption,

and then again, flows down valleys.

Before I go on,

I just want to mention because,

there's so much in this book that's part of the URL reading for Whidbey Reads,

that it's not just about the science,

but it's about the much larger picture of how a crisis,

in this case, a volcanic crisis, is managed.

One part of that as science,

and the science that I'll talk about includes the monitoring network,

those different tools we can use to try to detect processes we can't see,

how we interpret the data that come from those tools,

and then how scientists produce hopefully an actionable forecast.

This importantly, then must lead off to

a collaborative partnership with emergency managers,

and cooperation of the population that might be affected.

All of these elements came into play at Mount St. Helens as they always do.

I'm going to show a timeline here,

this is starting from March of 1980, there's again,

a picture of Mount St. Helens,

the lake you see there's Spirit Lake which is off to the north of the volcano.

We'll start in March.

On March 15th, things might have looked something like this,

but by March 20th,

a large earthquake was detected on

a seismometer whose data was being sent back to the University of Washington.

This earthquake had a magnitude of 4.2,

sizable in terms of recognition on the seismogram,

but not felt, or at least not significantly felt locally,

by anyone who reported back to them.

It was concerning enough though,

that scientists at University of Washington took particular note of it,

and wanted to figure out where it was,

when they recognized it, it was beneath Mount St. Helens.

It drew their attention even more.

In subsequent days, this is a picture of a seismograph showing,

increase in the frequency of earthquakes,

that started with just a single event,

but then increased through time,

here's the number of earthquakes per day.

Again, I showed a single large earthquake on March 20th.

Seismicity actually started for much smaller events a couple days before that,

but ramped up to over 200 earthquakes by just the 22nd of March,

just two days after the

UW scientists recognized that 4.2 earthquake.

Indeed, things started to progress fairly rapidly,

here's a photograph from March 24th,

where you can see crack starting to form in the glacier near the summit of the volcano.

If we move forward in time,

here's a photograph from March 27th,

with the first explosive eruption.

This explosive eruption, excavated a bunch of material from near the summit,

that is old rock sitting on the volcano that was pulverized into ash,

and then fell across the snow.

It's turning the snow,

a darker gray color in this image here.

Now importantly, that ash was not new magma,

new magma, at least as identified at the time,

new magma was not recognized,

it did form a new crater at the top,

and many more cracks were forming near the summit.

In fact, interpretation of that ash and recognition of the hazards,

led to this interview with David Johnston,

who was killed in the May 18th, eruption.

>> This was not an eruption.

This is a steam explosion.

That is, just hot water very rapidly expanding,

shooting out, clay and mud and rock that's up here.

No magma has yet come to the surface,

but this means that it is heating up,

and so magma conceivably is rising,

because this is such a symmetrical volcano,

and because the crater is so high,

if there's an explosion,

it's probable that the debris,

the very hot incandescent ash that would come down,

will come down on all sides.

Right now, there's a very great hazard due to

the fact that the glacier is breaking up on this side of the volcano,

on the north side, and that could produce a very large avalanche hazard.

This is not a good spot to be standing.

>> I show this interview because David Johnston was

very familiar with Mount St. Helens,

but also with volcanoes from work up in Alaska.

If you've read the book, then you may have already understood that story.

There was very clear recognition though,

of an increasing hazard from magma rising,

even though it hadn't been recognized to have reached the surface yet,

this meant that scientists wanted to install many more instruments. More seismometers,

using different instruments to measure deformation of the surface, to detect gases,

to try to again use these various proxies for

processes that we can't see to try to figure out what was going to happen.

By April 27, the whole side of the volcano had created a large bulge,

and so I said that sometimes when magma ascends,

it causes deformation of the volcano.

In this case, the whole side of the volcano

bulged out in order to make space for magma rising behind it.

In fact, if we step back through images of the north side of

the volcano starting from March 21st with some cracks in the glacier,

you can see now in the image from April 15th,

many more cracks.

Perhaps you can also see that it's starting to deform or to bulge outward.

But by May 7th,

the bulge was significant and visible from across the landscape.

When you'd look at the volcano,

it was significant the changes that had occurred.

Simultaneously, there were activities happening on the response side.

The USGS issued this hazard map on the 30th of March.

The hazard, it presents the science part of the story,

the scenarios for where things might travel,

all this stippled area and the brown colors,

you can see are largely directed down

river valleys or on the flanks of the volcano itself.

Here's Mount St. Helens,

here's Highway 504 as that existed at the time.

I'll just point out Elk Rock and Spud Mountain as

points of reference because you'll see these

appear in later maps that show distribution of the deposits.

But as we move forward in time,

now I'm also showing images of graphics from proposed restricted land use areas,

and that's distinct from a hazard map because it

includes consideration about land ownership and land management.

Here including two different zones,

a red zone and a blue zone.

Now those zones, shapes and distributions changed through time

for a variety of different reasons that I don't really have time to get into today,

but include competing interests and the understanding of the hazards.

I'll again move through these fairly quickly and just point out

the red gate and the blue gate on this map.

Here's an image from one of those gates.

The road became closed as it was identified that

the hazard was increasing and that bulge was growing,

especially on the northern side of the volcano.

There was a proposal,

the sheriffs had proposed to extend the zones to a much greater area,

but this proposal hadn't passed by the time the 18th arrived.

I'm moving on. Now we're to May 17th,

I said before that in the ideal case,

any sort of monitoring parameter, that is seismicity or deformation,

what we might look for is an acceleration that

indicates a runaway process moving towards an eruption threshold.

Now, the two graphs on the left-hand side are both measuring deformation at the volcano.

You can see that bulge was growing at the enormous rate of

1.4 meters per day over an interval in 

early- to mid-May, and if anything,

that wasn't seen to be increasing through time,

it started to slow a little bit as we started to move towards May 18th.

Similarly, if we look at this plot,

it's a little bit busy, sorry about that, but again,

March till May in 1980,

and this time this is a seismic events per six hours.

Ignore the red curve here,

look at the other curves.

You can see that there was a big increase in

seismic activity towards the first steam eruption that I showed was in late March.

Again, the number of events or the frequency of those events was not

increasing steadily or even in an accelerating way towards May 18th.

If anything, some of the seismicity had plateaued and leveled

off and didn't show any clear precursors before the May 18th eruption.

Again, this creates uncertainty to the forecast of the timing of any particular eruption,

because there's no clear signal of acceleration towards an event.

In fact, because of some of these slow,

gradual changes without abrupt increases,

the response from the community was varied.

Some people, like the classic curmudgeon,

Harry Truman refused to leave his dwelling along

Spirit Lake because he felt

such an affinity to the volcano that he thought it wouldn't hurt him.

In fact, people wanted to get back to their summer homes.

35 people signed liability waivers just a day before the big eruption, went in

and collected belongings, and left again.

But on May 18th, a Sunday,

which luckily was a day that all the loggers

working on the volcano were not there working.

At 8:32 AM, a remarkable event happened,

and that event included an earthquake with a magnitude of 5.1.

This large earthquake coincided with triggering of a landslide that

I'm showing a video that reconstructed from images taken by Gary Rosenquist.

You can see the whole northern side of the volcano collapsing away from that bulge area,

and almost immediately an explosion

that's happening right at the site of that big collapse.

This explosion produced an eruption cloud that went vertically

traveling up because it contains so much hot gas and ash in a low density mixture.

But imagine again, like our example of a bottle of soda that's been shaking up,

if we poke a hole in the side of that bottle of soda.

Now, the stream of bubbly soda mixture is going to head out sideways, not straight up.

That's essentially what happened with Mount St. Helens.

The landslide excavated a big hole in the side of the volcano,

and so you can see this cloud being directed north

is just to the right-hand side here of this image.

That is what is termed the lateral blast from this volcano.

On May 18th, David Johnston,

a scientist with the USGS,

was sitting on a ridge north of Mount St. Helens,

at a relatively high elevation compared to the valley bottom.

But the velocity of this both collapse from the landslide and the blast that

followed, was high enough that material was

able to cover significant distances and even climb ridges,

surmounting up and over those ridges and reaching

even valleys that extended at much greater distances away.

The debris avalanche of the landslide reached speeds of up to a 155 miles

per hour and reached distances of almost 14 miles to the northwest.

Again, like I said, it was moving so fast,

it didn't just go down the Toutle River valley.

It climbed up and over what is now known as the Johnston Ridge.

The debris avalanche again,

I showed you the pictures of that column was followed almost immediately

by an explosion that produced what we call a lateral blast.

This is a mixture of pyroclastic or broken fragments of magma,

ash and pumice, and gas that traveled up to 220 miles per hour.

It reached great distances,

several ridges away from the volcano itself.

In fact, it was moving so fast that it actually outran the landslide,

before the landslide had even made it down to the valley floor.

It reached distances of up to 17 miles to the north-northwest,

but it covered a broad swath that extended both from

the northeast all the way around to the north and then to the northwest as well.

The lateral blast material that was moving fast was very hot.

It reached these distances in a very short time.

Took less than 10 minutes to do all of this destruction,

knocking over heavy equipment,

knocking over all these trees,

pointing them in the direction of flow,

and covering the landscape.

One of the very first lessons that we learned here

was that the scale of something that had been recognized before,

that is a landslide and avalanche and the potential for

material to be directed in one direction or make

a lateral blast was just much larger than had been

anticipated because it hadn't been well recognized at other volcanoes.

This eruption also produced a vertical eruption column.

Here you can see pictures of that vertical eruption column from around the volcano.

This material reached up to 30 kilometers high.

The eruption lasted for hours.

The ash circled the earth in 15 days and it produced thick deposits of ash and pumice,

10 inches thick of material at 10 miles away.

Again, more images of the way that this ash in

the atmosphere turned day to night in several different locations,

including near Ritzville, off in Eastern Washington into Idaho.

The scene was very eerily covered with ash and difficult for cars,

for people to see,

to drive, to navigate.

Motorists were stranded.

Many encountered these things

scientists call accretionary lapilli,

which is ash that concentrates around a little moisture in the atmosphere,

accumulates, making up a ball and then drops out.

Some people noticed that these things look something like mud balls.

In personal accounts, they called them mud balls landing on their vehicles.

Pyroclastic flows also accompanied these eruptions.

Those traveled again, at speeds of up to 80 miles per hour,

they traveled down the flanks of the volcano.

These are very hot,

over 700 degrees Celsius.

If you were to go to Mount St. Helens now,

you could see the products of these pyroclastic flows.

Here's a person.

Actually, there's 1, 2, 3, 4 people in this photo.

For scale, you can see that there are tens of meters of pyroclastic flow deposit,

and this is all from the May 18th, 1980 eruption.

This has been cut by erosion from

the stream that we're walking along at the time I took this photo several years ago.

Lahars also traveled down the volcano after the eruption on the north and south flanks,

but the lahars were much larger to the north.

Here you can see the landscape created by the landslide or the debris avalanche.

It produced all these bumps or what we call hummocks - bits

of the mountain that are distributed along the path of that landslide.

The channel is also filled in by a lahar or

a volcanic mudflow where the landslide itself.

All these bits of what was originally the volcano traveled down in

there with all the groundwater that was part of that rock mixture,

dewatered or got squeezed out of these hummocks and concentrated in the channel.

It traveled down with a lot of fine ash material and was moving at significant speeds,

able to carry large trees that had been blown down in the blast,

then was able to destroy large sections of bridges.

These lahars, reached all the way to the Columbia River, 100 kilometers downstream,

stranding 31 ships in upstream harbors, upstream of the point that the Cowlitz reaches the Columbia River.

It destroyed logging camps, homes,

and caused a lot of both human life damage,

so it killed people,

but also a lot of property damage,

stranding people and producing a lot of the havoc that required

a huge amount of work to try to help rescue people in the days to weeks following.

The human life toll from this eruption was significant.

Here's a map showing victims from the eruption itself,

and also the distribution of the phenomenon that I've just talked about.

The brown colors here,

this is the volcano in the center of the image,

but all the brown colors show where you can now still see

the hummocks from the debris avalanche or landslide.

But the big thing I want to point out is the wide swath of

the blast deposit that reached huge distances away from the volcano.

Not only in a direct path to the north,

but across a 180 degree arc in the northern direction.

This is why it was impossible for people to escape.

It traveled so fast and it covered a wide swath.

Some of the people that are represented by these numbers probably couldn't

even see the mountain at the time that the eruption happened.

Here's another map showing distribution of deposits.

This one also shows, in brown,

the path of lahars that made it all the way

to the Columbia River near Kelso and Longview.

The devastation and recovery was a significant effort,

one of the biggest recovery efforts that the US has seen.

It included a visit to the area by Jimmy Carter,

and significant helicopter-based operations to try to

find missing people and survey the extent of damage.

The problems that were created by all this material

moving its way down the river are not over.

In fact, there's long-term sediment management problems, that is,

problems with all of the ash and pumice that's still clogging up the Toutle River.

Dredging on the Columbia River and the Cowlitz and Toutle happened

right after the eruption to re-establish the shipping traffic at the time.

There were also structures that were created in

order to trap sediment in the upper reaches of the volcano,

including a sediment retention structure on the upper Toutle River.

Here, that sediment's trapped behind the retention structure to try to prevent

a problem from continued accumulation of material that would require more dredging.

You can see as you drive on I-5 past Castle Longview area and along the Columbia River,

big terraces that are created from the dredge spoils that

had to be dredged after this eruption.

But this wasn't the end of activity.

In fact, volcanic activity continued for quite a long time after May 18th.

Following very quickly on the heels of these recovery efforts was

a continued monitoring effort to figure out if the eruption was going to continue.

I mentioned that in the 1400s,

there were two big eruptions,

one in 1479, one in 1482.

Both of those happened just within three years and both had eruption sizes,

that is, big columns that were bigger than the May 18th, 1980 eruption.

Scientists were aware they needed to keep their eye

on the situation and the evolution of activity.

Again, moving to our timeline, we're here.

This is a cartoon I created to step us through

the sequence of events starting with May 18th,

the cataclysmic eruption, the landslide, lateral blast,

a vertical eruption column,

pyroclastic flows, lahars, and mudflows.

As we move forward in time,

we can see another eruption happened on May 25th.

That eruption was followed less than a month later by another on June 12th.

This activity was lesser in vigor,

but still produced both vertical eruption columns and pyroclastic flows.

A dome then started to grow in the newly formed crater on June 12th,

or after June 12th,

that was destroyed in a July eruption,

another explosive eruption in August,

continued dome growth, another explosive eruption in October,

and then dome growth again.

In fact, growth of a dome in the crater of Mount St. Helens

continued from 1980 all the way through 1986.

Here is a photograph of how that dome looked.

We're looking from the north to the south at that big dome with

the new crater wall on the south side in the background.

One of the big things from a forecasting perspective

from the May 18th eruption was that there was no clear escalation of activity.

The thing that triggered the eruption on May 18th,

was the big earthquake and landslide.

It was like taking the lid off a bottle of soda, that itself,

maybe scientists would have made much easier to

forecast an explosion if they'd put a bunch of instruments on it,

and they saw pressure increasing,

they saw the bottle stretching,

they saw a little hole starting to form.

But in this case, something different happened May 18th,

the lid was taken off,

or the side of the bottle was taken off.

In contrast, we had real success,

by we I really mean not me,

I was three in 1980.

Other scientists at the USGS with

all the increased instrumentation that was installed on the volcano.

In 1981 and 1982,

they had great success with a combination of looking at seismic patterns,

deformation, and gas emission.

If you look at all these curves,

although the details don't matter,

you can see that all of them begin to steepen towards the dashed line.

The dashed line is the point of an eruption.

This consistent pattern over several different cycles allowed scientists to

issue some of the most successful forecasts.

Here they would even call them predictions,

although I'd say that in general in

volcano science, predictions are pretty difficult to impossible to do.

In this case, the predictive windows or

forecast windows were more and more narrow through

time and we're very accurate with respect to the actual eruptions.

There's real improvement and real success in those later periods.

What did we learn about eruption forecasting?

We learned more about eruption triggers,

how triggers are not just internal,

not just building of pressure on the inside of the volcano,

but we can actually get something outside to trigger an event.

But we learned that with a lot of instrumentation and recognition of patterns,

there are other cases where we can do quite a good job of understanding and

forecasting the timing of eruptions that are triggered by internal processes.

We learned that the scale of hazards that themselves, although recognized,

they were not well understood and they were

underestimated in scale with respect to what was actually possible,

and especially in terms of the blast.

Now I'm not really going to get much into this,

we've learned a lot more about other parts of the system

since the eruption by using a variety of other tools.

Here you can see a cartoon.

This is a cartoon I drew of Mount St. Helens with the dome in the center.

You can see that I've drawn a shape of a magma reservoir at depth that's largely

informed by information from a variety of different tools, including seismology.

All of the stippled pattern area here shows the distribution of

earthquakes that have been located beneath the summit of the volcano.

Since earthquakes form in the brittle crust that, for example,

could include breakage of rock surrounding the magma chamber,

we can actually use the shadow zone where there are

no earthquakes to infer where magma might lie.

Here I've just drawn a cartoon of where earthquakes are distributed beneath our

hypothetical Mount St. Helens looking from the north side.

When we compare that distribution and the shape of

the hypothetical magma chamber with information we gained from the products themselves,

that is rocks that have crystals inside them,

we see that they correspond quite well.

The way we get information from crystals about depth of

storage is we use these big experimental apparatuses.

This is not the one that was used at the time,

but it is quite similar to the type of equipment that can take a sample of rock,

it can squeeze it to very high pressures,

take temperatures up to magmatic temperatures,

so something over 800 degrees,

and then replicate a variety of different pressure and temperatures to try

to match the composition of crystals we see in the actual products.

As soon as they get a good correspondence between experiments and what we see in nature,

then we can better understand the depth at which these crystals grew,

and indeed, it matches quite well with the picture that I'm showing here on the left.

We learned more since the eruption about where magma is actually stored.

We can actually learn a lot more

about the system from the stories that crystals can tell.

Here's a picture on the right-hand side of a crystal

in a piece of pumice from May 18th, 1980.

You can see that this crystal,

which is the bright gray,

roughly rectangular shape is not

just one color and where the grayscale value

here corresponds with the composition of the crystal,

in fact, it has zones.

Those zones give us really great information,

especially how these colors change from one zone to another,

gives us information about the timing of when changes happen in a magma reservoir.

When we compare that timing information with timing of earthquakes before May 18th,

we see that they correspond really well.

Just the days prior to eruption,

the crystals are telling us that there were

changes in the magmatic system that correspond with those earthquakes.

So we learn more about the system by studying crystals as well.

Then our final example without getting into the details,

if crystals were the products of cooling or of

decompression of magma in the May 18th eruption,

the products, the pumice is produced

from rapid decompression or a rapid ascent of magma towards the surface.

In contrast, later in the eruption sequence,

that is the material that made domes in late 1980 or continuing to 1986,

magma ascended very slowly towards the surface and

sometimes produced relatively small eruption columns.

We can tell that there was this very different rate of ascent of magma in

the two cases by looking again at the pumice clasts or at the dome material.

These two images are slices of volcanic rocks where on the left,

the black, those are bubbles in a piece of pumice,

the gray is the glassy material that surrounds the bubbles,

and you see no crystals,

no rectangular-shaped objects in here because

there simply was not enough time to grow crystals,

and that's as expected for rapid ascent toward the surface.

In contrast, in a magma that ascends very slowly,

like in a dome, look at all of these rectangular shaped crystals.

Those had enough time to grow and they tell us a bit more about

the story of how material made its way towards the surface to make those eruptions.

We learned that crystals tell the story of how fast magma ascends.

There are so many more things we learned from scientific studies,

but I'm going to continue on through the sequence of events that followed 1980-86.

I'll mention that there were a series of eruptions in

1989-1991 that didn't include any new magma at the surface,

but did produce some pretty spectacular plumes.

This photo was sent to me by Larry King,

and it's from a hike he took on December 20th of 1990.

If you were out there,

you would be quite surprised and nervous about this kind of eruption,

even though it didn't contain any new magma.

And these were from superheated steam from explosions

that simply blew out parts of the summit crater area.

They were relatively short-lived and relatively short eruption columns in

comparison with many of those in the 1980-86 sequence.

Another magmatic eruption happened though starting in 2004, lasting to 2008.

In fact, that ramped up again very quickly between September 23rd,

recognition of the first earthquake,

and September 27th, which is shown at the top of this graph.

Here you can see a rapid increase over just days in the number and

size of earthquakes detected on seismometers at Mount St. Helens,

where here the size of the event correlates

with the size of these little squiggles on the seismograph.

Our picture of understanding both what and

where eruptive material might travel in the 2004-2008 eruption,

again informed by the same background studies of volcanic history.

But now the volcano looks very different,

so there isn't a big edifice,

a big mountain capable of producing the same kind of landslide.

At the time that the eruption of 2004 started,

there was not a new dome that

would have been likely to collapse and form a lateral blast.

Although certainly construction of new domes was possible after that time.

Indeed by October 1st,

just again days later,

now we're looking down at the crater of Mount St. Helens,

the 80-86 dome is the gray colored material on

the center of this image and the south crater wall is on the left-hand side of the image.

You can see the Crater Glacier is cracked and

deformed and an explosion occurred at the summit.

In fact, this led to construction of a much larger dome.

Here you can see again a picture from inside

the crater of Mount St. Helens where now that slightly cracked

and discolored glacier has grown to form a huge dome that whole very light-colored,

spiny surface there that was termed the whale back,

is a new dome that was constructed starting in 2004 and extending to 2008.

The largest explosive eruption of this sequence happened on March 8th of 2015,

distributing a very minor amount of ash across the landscape.

But if we look now at this image from 2007,

you can see more of the context of the 1980-86 lava dome,

the glacier that was surrounding it that had been pushed

out of the way by growth of the new lava dome behind it,

in-between the 1980-86 dome and the south crater wall.

The Crater Glacier itself,

which in this image in 2007,

had not met on the north side of 1980-86 dome now has and continues to advance.

It's one of the only advancing glaciers around.

We can look at this cartoon to look at the size of that new dome that grew,

the 1980-86 dome shown on the right-hand side.

But the new dome that was growing from 2004-2008.

Showing both the Space Needle and the Empire State Building for scale,

you can see that it's much bigger than at first,

you might recognize, a huge dome that rose above

the height of the '80-'86 dome and was butted up against the south crater wall.

Again, showing our timeline of eruptive events starting in 1980,

we learned from crystals and from seismic studies, also from deformation,

the distribution of where ground was moving,

that a variety of different processes have happened at Mount St. Helens since 1980.

There was a big eruption that pulled material from a reservoir that extended down to

about 12 kilometers depth and that erupted quite a significant amount of material,

but the eruption continued and it erupted mostly shallower material.

You can see the little arrow I've drawn here,

this tapping the top of a magma reservoir.

By '89-'91, although magma still layed beneath the surface,

it did not reach the surface,

but was transferring enough heat that allowed steam explosions to happen.

No new magma reaching the surface.

By 2004-2008 yet again,

we have some perturbation to the system.

Something disturbed the system may be input of new magma,

maybe just changes in the storage region itself,

that caused an eruption, tapping again the top of the magma reservoir.

I've drawn an extra little reservoir here on the right though,

to note that just because these are their time intervals that we've seen things happen at

the surface at Mount St. Helens doesn't mean these are

the only times things were happening in the magma reservoir.

In fact, a great example,

I just note one from 2014 here that was

highlighted in an information statement that the USGS issued, where there

we said that the magma reservoir beneath Mount

St. Helens had been slowly re-pressurizing since

2008 and this is caused by the arrival of

a small amount of additional magma at 4-8 kilometers depth.

It doesn't indicate that the volcano is likely to erupt anytime soon, and in fact,

the seismic activity that caused this statement to be

written has happened at other times throughout the history of Mount St. Helens.

One of the other talks in the series focuses much more on seismicity.

But just to point out that there are continually processes happening in the system.

What lessons did we learn at Mount St. Helens since 1980?

These are just a tiny sampling of some

of the lessons we learned where here we're focusing

on the eruption forecasting part of the story and the stories crystals can tell.

We learned much more about the nature and size of debris avalanche deposits.

Here you can see an image of that hummocky topography that's so

classic from volcanic debris avalanches,

where the hummocks are distributed across

the topography and the direction of transport of

the landslide and we recognize these at volcanoes

around the world since the Mount St. Helens eruption,

including there's a spectacular example at Mount Shasta down in California,

where as you're driving down I-5,

you can see all these hummocks from an old eruption of that volcano.

We also learned that the size and distribution of blast hazards are much more

significant than we might have originally

recognized, covering this 180 degree distribution.

Recognition of the deposits allowed us to recognize

blast deposits at a variety of different other volcanoes since that time too.

We learned more about triggers for eruptions,

we learned that eruptions don't always happen because of

a rapidly pressurizing system but there can also be external triggers,

like the big landslide that triggered the May 18th eruption.

There will always be uncertainty because of

the variety of paths that any system can take.

We learned a lot more about the stories that crystals can tell,

they tell us about where magma sat before eruption,

they tell us more about how fast it arrives towards the surface,

and we're starting to combine information from the crystals with monitoring data to

tell us about the timing of events that might be triggering eruption.

These are just some of the many lessons

learned that help us to better forecast eruptions.

This is a simplified version of the hazard map for Mount St. Helens.

Here you can see a variety of different colors that encompass a multitude of hazards,

including lava, pyroclastic flows,

some of the phenomena that I mentioned earlier in my talk.

What does the future hold for Mount St. Helens?

Well, the future for sure holds a construction of

more lava domes like the two that have already been produced since the May 18th eruption.

It might also include

more mobile or longer traveling lava flows

of a slightly different composition, and together,

these lava flow formers and the pyroclastic eruptions that might

accompany them will eventually rebuild Mount St. Helens as has been done in the past.

In order to detect these processes,

we continue to maintain and install equipment,

monitoring tools to detect the processes that

we cannot see that are buried beneath the surface.

We also make campaign measurements that is, go out into the field to make

measurements of gases or of deformation to look for changes in the volcano.

This forms just one piece of the much larger puzzle,

which is volcanic crisis management which we work so hard to try to maintain

partnerships with emergency managers and provide

information to anyone with interest in these volcanoes.

This includes coordination plans and includes working with partners on

signage of directions towards safety or evacuation if there were to be an eruption.

It includes social media and web presence,

it includes hazard maps and signs at national parks

or forest service lands and things like alert levels for

alerts issued to aircraft to advise them to take flight paths out of

the direction of a potential explosive eruption

that would be devastating for the aircraft.

I'll just finish with this final video.

This was taken last year at

the still continuing to steam 2004-2008 dome

where some colleagues of mine are making measurements.

This is a drone that is being flown, taking these videos,

but also making measurements of

gases that are being emitted from the volcano still today.

With that, I think I will finish and take some questions. Thanks.