Original subtitles

NARRATOR: A sudden landslide.

Nobody anticipated that it would cut loose

and bury that entire village.

NARRATOR: Some victims are buried alive.

Others narrowly escape.

ROBIN YOUNGBLOOD: I saw a gigantic wall of mud,

and then we were hit.

Houses were exploding.

911 CALLER: Oh my gosh!

911 OPERATOR: Are there any injuries?

911 CALLER: Yes, there are people yelling for help!

NARRATOR: A whole community is cut off.

What's going on with the highway?

I'm hearing rumors that there's a mudslide.

NARRATOR: A neighborhood, wiped from the face of the earth,

is invisible to rescuers.

BILL QUISTORF: All I could see was

just a vast wasteland.

There were no homes.

There were no remnants of homes.

NARRATOR: Volunteers risk their lives

to extract the missing from the mud.

DAYN BRUNNER: It was so eerie.

We could hear moaning.

We walked past four bodies.

YOUNGBLOOD: He was just drenched

and shivering and shaking.

NARRATOR: What could have triggered the tragedy?

ERNIE ZELLERS: Where's the baby?

I thought I was losing him,

and I'd say, "Stay with me, Bud!"

NARRATOR: Can science explain

how earth can turn to liquid in an instant?

Everybody feeling ready for go time here?

GEOLOGIST: Okay, we're ready for water.

Here it goes!

DAVID MONTGOMERY: Something like 22 inches of rainfall.

You know, that's almost two feet of rain in a month.

NARRATOR: Some of these disasters are triggered by rain,

but not all.

Around the world,

scientists fear that landslides could be on the rise.

Now, geologists are hoping to spot the next big one

before it strikes, to warn those in its path

and avert further loss of life.

There's no word for it other than "tragedy."

NARRATOR: How far out can we predict landslides?

The closer you get to an event, the easier it gets to predict.

It will be a very big one!

A huge one!

NARRATOR: The race is on to sound the alarm

and save lives from "Killer Landslides,"

right now on NOVA.

Major funding for NOVA is provided by the following:

We're committed to strong.ng.

And the Corporation for Public Broadcasting

and by PBS viewers like you.

Additional funding for "Killer Landslides" is provided by:

Furthering the values that contribute to a healthy planet.

Millicent Bell, through:

And:

NARRATOR: Late March.

Record-breaking rains have hit western Washington.

10:30 a.m.

Families are enjoying the first sunny weekend of spring.

It's been raining for months.

In an instant, this idyllic morning becomes a nightmare.

The saturated earth on a slope nearby starts to slip.

(rumbling)

A landslide comes plummeting across the valley

towards the unsuspecting families.

(rumbling and crashing)

AMANDA SKORJANC: The sound is like this horrible, rumbling, wet....

(rumbling)

I turned and I held Duke, and I did not let him go.

NARRATOR: Sitting at her table with a friend,

Reverend Robin Youngblood has no time to escape.

We heard this horrible sound.

It sounded like a 747 was going to crash in the valley.

NARRATOR: Half a mile away from Robin's home,

Amanda Skorjanc is in the kitchen

with her five-month-old baby, Duke.

SKORJANC: The sound is unexplainable,

and I will never get it out of my head.

YOUNGBLOOD: We ran to the window when we heard that sound,

and I looked to my right

and I saw a gigantic wall of mud,

and then we were hit.

SKORJANC: And it was like a movie.

Houses were exploding, and the next thing I remember is

our neighbor's chimney coming into our front door.

RICHARD IVERSON: Everything that was in its path, essentially, it picked up

and was throwing in front of it: trees, debris from houses.

It would be like being hit

with 1,000 battering rams at once.

It was just a completely unsurvivable sort of event.

YOUNGBLOOD: We were slammed, and we were underwater in mud.

The house was racing across the valley.

And then we stopped.

NARRATOR: Without warning, this hillside turned to liquid mud,

wiping out 49 homes.

The hillside collapsed, but it didn't stop there.

It dammed the river and continued on

to rampage through a neighborhood,

reached highway 530, and kept going.

All is now a wasteland of mud and debris,

one square mile buried under 15 to 75 feet of mud.

911 OPERATOR: 911, what is your emergency?

FEMALE CALLER: There's, like, a mudslide, and everything is gone!

The houses are gone!

There is a house on 530, and it is covering the road!

I was standing here,

and I heard this noise and I looked out,

it sounded like a big truck, and then my lights went out.

MALE CALLER: My neighbor's house has been completely taken out

and it's collapsed on several of them,

and they're trapped.

FEMALE CALLER: Oh my gosh!

So I walked out the front door,

and then I looked up the road and then I knew what it was.

I could see the stuff in the road

and then I knew it was a slide.

Took my phone out of my pocket and dialed 911.

JUPP (on phone): There was a big landslide at the bottom of Skaglund Hill.

NARRATOR: It's only minutes after the landslide buries

nearly a mile stretch of Highway 530.

Everything in its path, in a tiny community called Oso,

is altered forever.

Rescuers have no idea if there are any survivors.

JUPP: It didn't take very long for the guys to come from Oso.

They were here within ten minutes.

MIKE BLANKENSHIP: When you could see dirt

for a mile straight,

you knew you were in for the long haul.

DAVID MONTGOMERY: This slide apparently happened

in 138 seconds from start to finish

in two phases, it seems,

based on the seismic signal that was recorded

at the seismographic station at UW.

NARRATOR: If scientists converted these seismic waves

to sound waves and sped them up, this is how they would sound.

(rumbling)

The initial collapse registered on seismographs 170 miles away.

MONTGOMERY: Imagining all that material moving off that valley wall

all the way across the valley floor in about two minutes?

I mean, that's really hard to imagine.

PAULO FALCAO: Oh my gosh.

Look at this!

NARRATOR: The danger is not over for the victims or rescuers.

Another part of the hillside could break free

and come crashing down.

Some landslides strike twice.

Around the world,

landslides occur anywhere there's elevated terrain,

killing thousands of people every year.

In just seven months,

222 landslides worldwide caused loss of life,

resulting in nearly 1,500 deaths.

WOMAN: Oh my God!

NARRATOR: They occur in every U.S. state and cost billions annually.

A landslide is simply a downhill movement

of earth, rock or debris,

triggered when gravity overpowers

the binding strength of the materials that make up a slope.

Some landslides are slow and shallow.

The top layers of soil or rock slip over time,

just inches or feet a year.

Others, like mud flows and rock falls,

move rapidly and can flow long distances.

Rapid, deep landslides that travel far,

like the one in Oso, are the most dangerous

and difficult to predict.

MONTGOMERY: If you think about the distinctions

between landslide types,

the sort of deep-seated landslides

and shallow landslides,

shallow ones usually involve just the soil

or what we call the regolith,

the weathered stuff on top of rock.

Shallow landslides will often involve,

like, a meter or maybe a couple meters.

The one behind me is a deep-seated landslide.

The whole valley wall came down.

NARRATOR: The raw slope left behind after a landslide is called a "scarp."

David Montgomery is a geomorphologist

who studies landforms and the forces that shape them.

He begins his investigation at Oso

by surveying it from a distance.

MONTGOMERY: You can kind of get a sense

for how the old edge of that plateau

dropped down as part of the landslide,

but you can also see at the bottom

how the toe of the slide, the bottom of it,

spread out laterally across the valley bottom.

So this solid hillside started to fail

and then it fluidized, it turned into a debris flow.

This is a super big landslide that traveled super fast.

It's really hard to relate to that in human terms.

NARRATOR: Even now, days after the Oso landslide,

chunks continue to calve off.

GEOLOGIST: That's a big one.

NARRATOR: This slope has slid many times before.

GEOLOGIST: Whoa!

NARRATOR: To figure out if it could slide again in the future,

Dave needs to look into the past.

A technique called Lidar, laser-based altimetry

that can penetrate through the tree canopy,

reveals the scars left behind by past landslides.

MONTGOMERY: You basically take a laser and mount it in a plane,

fly it around and shine it down on topography

and measure the time it takes

for the returns to get back to the instrument.

So you're basically scanning the topography

below wherever you have the instrument.

NARRATOR: This Lidar image shows the contours of the Oso Valley

before and then after the March 22 landslide.

It also shows evidence of past landslides

up and down the valley.

The yellow landslides are the oldest,

followed by the tan and orange ones

which overlap them.

The red slides are the most recent,

with the crosshatched one

designating the March 22 event.

It wasn't the biggest landslide to strike the region.

The large orange landslide

that probably slid many centuries ago

is nearly twice the size of Oso's recent slide.

The challenge is to figure out how long ago

and how frequently landslides occurred.

That requires sort of good old-fashioned geology.

You need to go out and try and find something to date.

NARRATOR: David Montgomery and his team

found they could readily date the ancient landslides.

MONTGOMERY: This is one of the buried logs

that we carbon-dated at about 5,300 years.

It's buried in an old landslide deposit.

How do we date it?

Grab a piece of bark, take it back to the lab.

We got three carbon samples out of this margin of the slide,

all dated to about 5,300 years carbon age.

NARRATOR: Since 1933, aerial photos and satellite images

have provided even more clues about the history of the slope.

Pictures show the Oso slope slid repeatedly,

changing the course of the river at its base.

The regrowth of vegetation hid the scars of its unstable past.

MONTGOMERY: You have this pattern

with this slide on that slope of essentially failing

and then being quiescent for a while,

then failing again a decade or two later.

So this slope was a known actor in terms of landslides

that happened periodically

involving just the lower part of the slope

that eventually failed in 2014.

NARRATOR: Although rescuers in the region train for landslide disasters,

no one was prepared for this kind of devastation.

911 OPERATOR: We have requested a helo from the Navy

and from our Search and Rescue.

NARRATOR: The Snohomish County Helicopter Rescue Team

was the first air support on the scene.

Check!

NARRATOR: They use advanced technology

to search for survivors.

Scanning the mudscape

with an augmented reality mapping system,

pilots toggle between the mud-gray live video,

which is inset into the green pre-disaster aerial photography,

showing where the 49 homes once stood,

where streets were laid out just an hour earlier.

BILL QUISTORF: The partner that I was flying with,

our tactical flight officer, he told me right off the bat

that there were 30 homes in the area.

And I said, "30 homes where?" because all I could see

was just a vast wasteland.

There were no homes.

There were no remnants of homes.

NARRATOR: A thermal camera allows them to spot bodies

by the heat they give off.

QUISTORF: This tool was vital in helping us search

and determine that there were no live people in the mud.

NARRATOR: They expected to find people

near where their homes once stood,

but the instrument detected no heat signatures at all.

There was no longer anything there--

no houses and no people, dead or alive.

A geologist helped them to refocus the search.

RICHARD IVERSON: It took a bit of time for this to sink in with us

because we were thinking one way,

whereas the searchers were by and large thinking another way.

Being trained as first responders,

they had the mindset that the place you needed to get to

was the place where those people had been,

meaning where the houses had been

before the landslide occurred.

It w immediately clear

that there was going to be nothing there.

There would be no trace of civilization whatsoever

back where those houses originated.

NARRATOR: When the hill collapsed with such force,

it pushed all the debris, houses included,

to the far side of the valley.

It's what scientists call

"the distal end" of the landslide.

Iverson was able to map out the direction the landslide,

and anything caught up in it, would go.

IVERSON: This was an incredibly real manifestation

of something that I'd been studying

on a more abstract level for decades.

And to see it all play out right there

and witness firsthand the damage that it had wreaked,

it was really a very, very profound experience.

NARRATOR: The next time a landslide strikes,

geologists, first on the scene,

could be instrumental in saving lives.

Robin Youngblood's home

was rafted hundreds of feet from where it once stood.

YOUNGBLOOD: I was still conscious.

I couldn't breathe.

My nose, my mouth, my ears, my eyes,

everything was full of water and mud.

I knew I had to get out.

I don't know quite how I knew which way was up.

And I started pushing, and finally I got air.

RANDY FAY: Robin and her friend were the first ones we spotted.

They had gotten on top of the debris of Robin's home

and were waving and caught our attention.

The methodology we use for them

involves putting them in a hammock device.

So they don't have to move or really participate,

and as long as we're comfortable

there's no spinal injuries of any kind,

we can get them out of there pretty quickly.

ERNIE ZELLERS: I have her.

Give me some slack!

Are you okay?

YOUNGBLOOD: They dropped us off and said,

"We're going right back, we saw somebody."

FAY: Our crew chief spotted something moving in the mud.

NARRATOR: It was four-year-old Jacob Spillers,

struggling alone in the mud.

He was home with his father and three siblings,

Brooke, Kaylee and Jovon,

who were watching TV downstairs while their mother was out.

Jacob was on the second floor.

The landslide carried him over a quarter of a mile away.

All but Jacob were buried alive.

Jacob at that point was about waist deep in this

and covered head to toe with mud,

so it was just kind of the wriggling around.

So had we not come down lower and slower to watch that,

well, I don't think we would have spotted him, frankly.

NARRATOR: A neighbor miraculously made it through the mud to Jacob.

But now they're both stuck.

FAY: It was clear if he was sinking as a four-year-old,

a big, fat guy like me was going to have real trouble.

NARRATOR: Jacob is hypothermic.

His rescuer shields him from the helicopter's rotor wash.

STEVE KLETT: We attempted to come down and put a skid close to a log,

and there just was not enough rotor clearance

to get down that low.

FAY: We do a maneuver called low hover off-loading,

and we do this in the mountains a lot

if we're working against ledges

or areas where we can't land,

and it basically involves bringing the helicopter down

to a one-foot hover next to an obstacle,

and then we move on and off the aircraft

while the pilots hold it steady there.

I sank probably a little over ankle-deep

when I stepped off and was hesitant to move down the slope

for fear I'd get stuck.

So I had a hand line that I threw him.

And between the two of us,

we were able to pull them up out of there

and get them up with us.

We handed Jacob to the crew chief,

and then the other gentleman turned to leave

and immediately started sinking again.

So I motioned for him to come back up,

we put him on board, and he actually held Jacob

while we flew out of there,

which turned out to be a good thing

that he was on board with us.

YOUNGBLOOD: They brought Jacob to us in the ambulance,

and he was just drenched and shivering and shaking,

and I mean, I'm a grandma, right?

So I just told him,

"I'm going to get those wet clothes off you,

"we're going to put you in a warm blanket,

"I'll take care of you,

we're going to sing songs, it's okay."

NARRATOR: The scope of the devastation caused by this landslide

is something that even a trained geologist like David Montgomery

has a hard time accepting.

MONTGOMERY: Wow.

You know, who would have been able to imagine that

on the very first nice day of spring,

the whole mountain could come down

and run all the way across the valley bottom?

NARRATOR: He wants to see for himself the distal end of the slide,

the place where houses, people, cars

and the remains of an entire neighborhood ended up,

bulldozed by the landslide.

Seeing ground zero is a shock.

MONTGOMERY: What can you say?

It's utter devastation.

We're at the distal end of the debris flow,

and you can see that

the jumble of wood and soil that was pushed

as the leading wave of the debris flow

that ran into the highway and everything

in between the slide with unimaginable force

completely torn up in a chaotic pile of debris.

But notice all the wood from the trees

that were entrained and pushed along within the flow

got rafted out to the front.

I've seen deposits from debris flows before,

but never anything this big and this extensive.

It's unimaginably large.

The scale of this is just overwhelming.

NARRATOR: What's surprising to the experts

is that a 600-foot-high slope slid for a mile,

many times further than this slope ever slid before.

It's a central mystery of the disaster,

something no one expected.

IVERSON: If the Oso landslide had behaved in an ordinary fashion,

it might've hit one or two houses.

That's of course not what happened at all.

Instead, the landslide mobilized

and very dramatically ran on the order

of five times further than you might've expected

based on typical behavior,

and that's what really lead to all the devastation.

NARRATOR: David Montgomery gears up

to investigate what's behind this deadly anomaly.

We are going over to the scarp of the slide,

which has been actively eroding since the failure.

And we want to be roped in and as safe as possible

getting close to the thing.

MAN: Dave, have you put on a harness before?

MONTGOMERY: Uh, I've had them put on me before.

NARRATOR: He's roped to four anchors for safety.

At any moment,

with his weight adding stress to the top of the landslide,

a chunk with Dave on it could peel away.

MONTGOMERY: You can hear the raveling off the face.

The sediments that are exposed

in the head scarp of the landslide

are sand and gravel.

It's fairly loose stuff,

and you can hear the individual particles being dislodged

and rolling down the face as we stand here.

You can actually see the stuff coming down.

You can see the plumes of dust rising up off of the scarp.

NARRATOR: The loose sand and gravel can't adhere to the scarp,

so it easily slides off the slope.

MONTGOMERY: This is safe enough

that we are roped in.

The head scarp has been retreating back

since the landslide happened.

It's fairly weak material.

We're obviously hoping that doesn't happen very soon.

NARRATOR: What was the top of the slope before the landslide

is now nearly 300 feet below the head scarp.

MONTGOMERY: The area downhill from here that has all the trees on it

that have been back-rotated is a big block of the landslide

that dropped out from under just in front of us

and was essentially the head of the slide

forming this head scarp.

NARRATOR: The Oso landslide came down in two pieces:

the lower half of the slope

that spread out onto the valley floor

and the heavily treed head scarp

that plummeted halfway down the slope.

Two events, just minutes apart.

The same thing happened in Afghanistan

just a few weeks after the Oso slide

while a wedding took place in Badakhshan.

(people yelling)

Two slides with long runouts came crashing down in succession

when rain turned a mountain to liquid mud.

The first slide crushed homes in an instant.

(yelling continues)

Eyewitnesses raced into the deadly mire,

desperately trying to dig out family members,

when the second slide hit minutes later,

much larger than the first.

Hundreds, perhaps thousands,

are entombed in more than 160 feet of mud.

The government declared the site a mass grave.

(translated): It's really hard for survivors to find their homes

using simple tools like shovels.

We haven't recovered any of the deceased in the mud.

The people are horrified.

They witnessed the tragedy with their own eyes.

They describe the landslide as if it were a monster

that devoured their loved ones.

(translated): My family, including my child and all my belongings, are gone.

We ask the government to help us here.

We don't have the necessary equipment.

(translated): Some people whose houses weren't destroyed

still won't return to their homes.

They'd rather live in temporary tents

than face the horror and fear of another landslide.

NARRATOR: Geologist Haji Mohammad Jamshid

found evidence of cracks at the top of the mountain,

a known first indicator of movement on a slope.

JAMSHID (translated): This shows an opening in the earth

that is ten meters deep,

and the width is between one

and three and a half meters wide.

NARRATOR: With better landslide awareness,

these obvious signs on a slope above a populated area

can raise red flags in time

for authorities to set up a monitoring system.

(translated): It's not possible for human beings

to prevent natural disasters.

But we can save people

by evacuating them from landslide-prone areas.

MONTGOMERY: The unfortunate reality is that sometimes,

if part of a slope fails,

it can destabilize the next piece.

Sometimes two happen in very rapid succession,

like the most recent Oso slide and like the Afghanistan slide,

and that can be really heart-breaking

if you have people doing the very human thing

of rushing to help

and then those people get caught in the second bit.

There's no word for it other than "tragedy."

NARRATOR: Understanding why a landslide behaves the way it does

is almost like investigating a crime scene.

Geologists have to search for clues,

like the composition of the soil that slid.

Dave Montgomery peels back the layers of the Oso slope.

It all dates back to when glaciers were here

15,000 years ago.

MONTGOMERY: This material is the glacial lake deposits

that are at the bottom of the geological pile here.

And it's a mix of silt and clay

that I can just crumble with my fingers.

NARRATOR: When wet, the very fine material

that makes up this base layer of clay

slips like a banana peel.

The layer above it is a porous, silty sand.

Water drains through it like beach sand.

MONTGOMERY: These gray pillars around me are all made out of glacial till

which came off of the head scarp.

And the stuff is pretty strong, it's pretty cohesive,

it holds together well.

It survived the trip down to here in fairly large chunks.

NARRATOR: Glacial till is an upper layer of hard-packed material

left behind by the compressing action of a glacier

from the last ice age.

It's the layer just beneath

the very loose sand and gravel at the top.

And we know it's till because there's boulders,

there's gravel, there's pebbles, there's sand, and there's clay.

It's everything.

It's like a dog's breakfast of a deposit.

NARRATOR: Add to that a regular flow of groundwater.

MONTGOMERY: How much of that water was actually flowing into

and helping to saturate the material

that failed in the 2014 landslide is an open question.

But there was certainly groundwater coming in

from the margins of the slide.

NARRATOR: How much water the layers on each slope can hold

is a factor landslide geologists try to measure.

IVERSON: The contributing factors in the great majority of landslides

have to do with water.

In the case of this event at Oso,

there had been prolonged heavy rain.

The wetter it is, the more likely it is

that the groundwater pressures will tend to build.

And it's that increase

that eventually triggers the landslide.

NARRATOR: Layers of clay, sand, glacial till,

and a loose, gravelly top layer

are typical materials found worldwide on mountain slopes.

This combination of ingredients, mixed with a wet climate,

paves the way for a long runout landslide.

MONTGOMERY: You have the geological setup of a naturally unstable slope,

and then we get a very, very wet winter.

The proximal cause of this landslide is

we had the wettest March in history,

something like 22 inches of rainfall in this area.

It's almost two feet of rain in a month.

BLANKENSHIP: I live in Washington.

(laughs)

So, it rains a lot in Washington.

NARRATOR: The Pacific Northwest is landslide country.

More rain falls here than any other part of North America,

as much as 118 inches annually

on the western slope of the Cascades.

Research suggests climate change could result in an increase

in the number and intensity of showers.

This could cause more frequent landslides on unstable slopes.

But Oso experienced a slide beyond expectation,

and scientists want to know what caused the mountain

to turn to liquid so quickly and flow so far.

Richard Iverson has studied debris flow for 30 years.

IVERSON: When we first arrived on the site

and looked out across the area of devastation,

it really didn't make sense.

You don't expect a landslide originating from that height

to travel that kind of distance.

So immediately, that poses

a big question in your mind scientifically

as to why this thing traveled as far as it did.

NARRATOR: The unique physics of the landslide

was the result of a phenomenon Iverson knows well,

when the ground turns to liquid in what's called liquefaction.

Usually when people hear the word "liquefaction,"

they think of something that occurs during earthquakes

because when the ground shakes and you have loose, wet soil,

that soil can liquefy

and buildings can become destabilized.

But it turns out that landslides can also cause liquefaction.

NARRATOR: Iverson's models help illustrate just how fast

a rapid, deep landslide, now liquefied, can move.

He believes Oso's 7,600,000 cubic meters of mud

was moving at 60 miles per hour.

IVERSON: It took less than one minute

for the debris to move from the source area

all the way across the flood plain.

That's a distance of roughly one kilometer.

Here's the scale.

And then for some extended period afterwards,

it gradually oozes out along the margins.

NARRATOR: Virtual Oso provides some answers,

but what if we could recreate a real landslide

to learn how far and fast it'll go?

In Oregon, Richard Iverson oversees experiments

at the world's largest landslide flume,

a 310-foot-long concrete chute.

Here, the U.S. Geological Survey

slops together truckloads of soil materials

to track how they behave when water is added.

This is how geologists here study the forces of nature.

They want to figure out how much rain is too much

on slopes like Oso.

IVERSON: The experiments are actually very pertinent

to what happened at Oso.

We put a pile of loosely packed sediment

in the top of our flume,

and we simply water it until failure occurs.

The landslide will liquefy

because of the loose packing configuration

or the high porosity of that soil.

NARRATOR: Porosity is a measure of how much open space there is

between the individual particles that make up soil.

Loose soil is more porous than tamped-down earth.

If the soil were more densely packed,

then it would have much less inclination to liquefy.

NARRATOR: Iverson and his colleagues will measure how long it will take

to turn this solid mass to liquid.

Everybody feeling ready for go time here?

GEOLOGIST: Okay, we're ready for water?

NARRATOR: Playing Mother Nature,

the team adds rain to Mount Flume.

IVERSON: It's very much like an Oso-type landslide

being triggered by long-term accumulating rainfall

and building groundwater pressures.

NARRATOR: Cracks, showing slippage at the top of the slope,

soon form in the loose rock and soil mixture.

The porous slope takes on water like a sponge.

IVERSON: Rocks and soil are an amazing thing

in terms of how they respond to water.

One slope can take on water faster than another one can.

NARRATOR: Instruments relay 7,500 measurements per second.

IVERSON: We're going to have something like 70 electronic sensors

deployed in each one of these experiments,

and some of those will be measuring

deformation of the slope internally

as it begins to creep and fail,

and others will be measuring the water pressure,

some will be measuring the surface displacement,

so it's really quite a chaos of cables when we run one of these.

NARRATOR: This is one instance where "failure" means "success."

92 minutes after applying five inches of simulated rain,

the solid slope liquefies.

IVERSON: We've learned a great deal

from being able to witness these things hundreds of times over

in our flume.

There's really no substitute for seeing things firsthand

to spur your thinking

and make you understand on an intuitive level

things that are otherwise quite abstract

in the way of physical concepts.

It's a real big part of doing physics, frankly,

to have a deep intuition about things

in addition to simply doing the math.

NARRATOR: After the first mini slide of liquid mountain,

Iverson doubles the rainfall

to trigger an even bigger landslide.

IVERSON: First, we had a relatively shallow slope failure

that's typical in many mountainous areas.

But then after that,

we continued to add water to the slope,

eventually triggering a much deeper-seated failure

where the material liquefied when it failed,

and then we got a much larger debris flow mobilizing.

GEOLOGIST: Here it goes!

NARRATOR: Heavy rain and porous soil

makes the particles behave like ball bearings.

They lose their cohesive strength

and become liquid earth.

IVERSON: And that, in effect, is our attempt to simulate

what happens when you have prolonged rain

that sets the stage for landsliding.

NARRATOR: Flume experiments like this one could help refine models

to predict the length and speed of future landslides

and even guide search-and-recovery teams.

The immense area affected by the Oso landslide

requires an extraordinary effort.

More than 900 people were involved,

many working alongside 123 rescue dog teams

to recover those presumed dead,

sifting through every inch of the one-square-mile mudscape.

BOB VANDER YACHT: There are still seven humans that have not been located.

There's not a worker out there

that is going to want to stop doing a search

until they've found everybody.

BROOK ALONGI: At this point, it's not everybody

on their hands and knees,

digging through the mud and the sticks.

NARRATOR: Search zones are delineated, like an archaeological dig.

VANDER YACHT: The efforts are very meticulous.

They're using a grid pattern.

It has an excavator with an operator,

dog teams, spotters.

ALONGI: And when they pick it up, they don't just drop it;

they gently drop a little bit

and we watch every little thing that falls out.

NARRATOR: It's a fine-tuned protocol using proven search tactics.

ALONGI: In the areas that I was searching in,

most of the two-by-fours were reduced to pieces this big.

And then you find a couch that's in a million pieces,

and then you find a tractor that's broken in half.

Trucks were ripped into five, six pieces.

Diesel trucks!

So you can imagine what that did to people, right?

NARRATOR: Many residents in Oso believed they were safe,

unaware that a landslide there could go so far.

The last landslide here, in 2006, stopped at the river,

but in steeper terrain like Nepal,

landslides are an even more common occurrence.

The world's most landslide-prone roads

wind their way through Earth's highest mountains.

Annual monsoons and earthquakes

bring mountainsides crashing down

with devastating results.

Nearly one-third of the world's landslide-related deaths

happen in the Himalayas, where, in 2007,

falling earth killed a thousand people.

In Nepal, 29 million people are at imminent risk.

RANJAN KUMAR DAHAL:

NARRATOR: Geologist Ranjan Kumar Dahal

is investigating the impact of human activity on landslides.

Many Himalayan roads are cut into the base of hillsides.

When the slopes above them fail, the road is blocked.

Villages are cut off from the world.

Homes are often built on the site of past landslides.

NARRATOR: Terraces cut into previous landslides,

and heavy crops, like water-rich rice,

destabilize the slopes.

NARRATOR: It's a system that is a recipe for disaster.

Homes are destroyed

when the saturated terrain fails around them.

Ranjan travels up

one of the most landslide-prone roads in Nepal

to search for signs of villages at risk.

He visits families to raise awareness

so they will know when to evacuate.

He's developed a simple model to educate his people

so they can read the signs of an impending landslide:

fragile soil cover, cracks in the earth,

slumping land, and small slides that grow over time.

His main concern is that a massive landslide will follow

when the next big earthquake occurs.

NARRATOR: Rivers and roads do great damage

because they cut into already unstable slopes.

Ranjan monitors Nepal's highways and remote roads,

looking for signs of slope failure.

NARRATOR: With each monsoon,

slopes slough their materials onto the roads below.

Constant vigilance is critical,

but Nepal, one of the poorest nations in the world,

can only put a Band-Aid on the problem.

NARRATOR: Bioengineering with rocks can be effective for shallow slides,

but for deep ones like this,

gabions won't hold the hill back.

NARRATOR: In a country where trails and roads

are carved from mountains by hand,

there's little funding to protect people

from natural hazards.

Yet the construction of new roads continues on,

without landslide planning.

NARRATOR: Two months after Ranjan's prediction,

in the heart of the monsoon, a well-known unstable slope

turned into a rapid, deep landslide

and buried Nepal's major highway to Tibet.

It dammed the Sun Kosi River and killed about 150 people.

There were very few survivors.

People around the world

share in the tragedy brought about by landslides.

In the first hours after the Oso landslide,

11 people are rescued from the debris field,

all far from their homes.

911 OPERATOR: Are there any injuries?

FEMALE CALLER: Yes, there are people yelling for help!

JUPP: And we could hear screaming,

and I said, "There is somebody screaming over there,

and it sounds like a baby."

FIREFIGHTER: A woman with a baby.

JUPP: Oh my god!

BLANKENSHIP: When we heard the baby cry, it changed the thought process.

That just sped everything up a little bit.

I held onto that baby like that was the only purpose that I had.

NARRATOR: Minutes after the landslide struck,

firefighters and EMTs arrive, but the quicksand-like mud

renders ground rescues nearly impossible.

BLANKENSHIP: When we initially got out there,

it was almost like a milkshake or pancake batter.

If you dropped something in it, it was gone,

and we knew it was four to five feet deep.

NARRATOR: Thrown 750 feet from her house and now stuck on a rubble pile

in the confines of her crumpled couch,

Amanda Skorjanc clung to five-month-old Duke

until she heard a sign of hope.

SKORJANC: As soon as I heard that voice, I screamed,

and then that caused Duke to cry.

She was in bad shape,

and I just kept her talking to me the whole time.

BLANKENSHIP: I remember asking Steve,

I said, "Steve, check the baby out."

And next thing I know,

Steve's checking the baby out and he said, "The baby's fine."

The baby got handed off and the baby was packed out.

The baby's fine.

ERNIE ZELLERS: Okay.

BLANKENSHIP: At that point, we knew the baby was fine,

and so we're moving on to our next patient, which was Amanda.

NARRATOR: Extracting Amanda from her couch took a chainsaw and many hands.

BLANKENSHIP: We just started ripping

at the fabric and the insulation,

got all that off,

and I remember getting to the wire

from the back of the couch,

and we didn't have a tool for that,

so four guys just started pulling on it really hard.

And then we heard, in the distance, the helicopter.

We all worked together,

and then up, up and away she went.

NARRATOR: The search for possible survivors

who were driving on State Route 530 was grueling,

especially for those who were at the center of the landslide,

the most powerful point of impact.

Dayn Brunner's sister, Summer,

was headed west on the highway when the landslide struck.

BRUNNER: My sister, she left the house at 10:30.

That would have put her right at the worst part

of the slide at 10:45, 10:47.

My mom called me that morning.

She goes, "Well, what's going on with the highway?"

And I said, "All we know is

there's a mudslide down by Skaglund Hill,"

and she goes, "Summer's in it."

It was just that mother's intuition,

and she was just adamant about it.

She said, "No, Summer is in there.

You need to go get your sister."

And I figured if she was in her car

and she had an air pocket or something,

she's waiting for me to come get her.

Me and my two sons walked in

and we were in there for almost six hours.

It took us two and a half hours to walk a half a mile.

You're in mud that's up to over your waist,

and I'm sinking down.

We're looking through cars.

We walked past four bodies.

Two of them were covered up, two of them weren't.

It was so eerie.

We could hear moaning, but we couldn't decipher

if it was a person that was trapped

or if it was an animal.

I wasn't trying to be selfish

if there were other people that needed help,

but I was going to find my sister.

NARRATOR: What if predicting landslides were possible

so that roads like this one could be closed?

That will require a system of meticulous monitoring

of places likely to slide.

In Switzerland, scientists are trying to develop technologies

that will save lives.

This device uses radar beams to measure minuscule movements

of the Earth's surface.

In the Swiss village of Preonzo,

this scar is what remains of a landslide

physicist Lorenz Meier was able to predict

right down to the hour.

He uses a system called InSAR

to monitor a mountain known to be dangerous.

NARRATOR: Geologists also use lasers, mirrors and crack meters

to measure the mountain's movement.

The instruments detect signs

that the slope will soon give way,

but how can scientists tell exactly when?

NARRATOR: The crack meters

measured the growing gaps between crack walls,

an ominous sign that the top of the mountain

will soon crash into the valley below.

Mattia Soldati, a forestry engineer,

continues to monitor them.

SOLDATI (translated): We decided to use these automatic extension meters

to measure the growing gap.

One end of the instrument is mounted

on the stable side of the mountain,

the other on the sliding side.

The extending stick in the middle

provides the measurement of the growing gap.

MEIER:

NARRATOR: The laser stopped working when clouds set in,

another advantage of radar,

which penetrates through weather.

InSAR allowed Lorenz to make a precise prediction

because of the refined,

millimeter-accurate measurements of movement.

MEIER:

NARRATOR: Lorenz plotted his measurements onto a digital terrain model

to show the village authorities

just how fast the top of the mountain was moving.

The growing areas of red and blue

indicate the top of the slope is moving not millimeters per day,

but now centimeters per hour.

The whole slope is about to fail.

MEIER:

NARRATOR: In the early morning hours on May 15, 2012,

precisely when Lorenz and his team predicted,

the slope at Preonzo collapsed.

(rumbling)

MEIER:

NARRATOR: 300,000 cubic meters of mountain

crashed 1,000 meters down the slope.

No one was hurt, the highway was closed,

and no buildings were affected.

The Swiss predicted the landslide in Preonzo,

but they can't monitor every slope.

Roads and railways are particularly at risk

in high mountain valleys.

In late summer 2014, a train car plunged into a ravine,

and others were left hanging from the rails

as a result of a landslide.

At Val Parghera, another high Swiss valley,

radar has determined a shallower landslide is moving slowly,

but constantly.

Lorenz has set up a timelapse camera

to document a year's worth of earth

creeping down the mountain.

An alarm system is in place

in the event rain causes dangerously high debris flows

into the valley.

NARRATOR: Over the last 30 years in Switzerland,

landslides, not uncommon,

have caused a death toll of just one person per year.

Close monitoring of every slope

in the world's mountainous areas is impossible,

but bringing potential hazards to light,

like the dangerous slope in Oso, is clearly needed.

MONTGOMERY: The evidence for large landslides

that could run across the valley

was there in the Lidar data.

But just because you have data that shows you something

doesn't mean anybody's looked at it, processed it,

or the information has gotten out to people

who might need to or want to know it.

We need to understand these failures as well as possible

with all the tools and tricks and models in our trade

and get that kind of information

out to the people who could actually use it:

the homeowners who may be making decisions about where to live,

the agencies who may be making decisions

about where to permit different kinds of development,

where to place highways.

NARRATOR: Five days after the Oso landslide,

Dayn Brunner and his family finally reached closure

in their search for his sister, Summer.

BRUNNER: I got a call from my sister's best friend,

and she goes, "We've found Summer's car."

I grabbed my son.

(crying)

I said, "Are you ready for this?"

And he goes, "Dad, I've been ready for five days.

Let's go get her."

By the time we got there, there was 15 people digging.

Her face, neck up, was exposed,

and we spent the next hour and a half digging by hand.

When we were ready to pull her out,

I just reached down and wrapped my arms around her,

around her upper body, and two guys were on her legs,

and I pulled up and they pulled up,

and we pulled her out and put her on her tarp,

and they just let you as a family say your goodbyes.

She was in pretty good shape,

because I had seen about 15 bodies come out

and they were horrible-looking.

They had suffered suffocation

and other various kinds of death,

but she was whole, and in fact, her hands

were still on the steering wheel when they found her.

Her foot was on the gas pedal,

and she was sitting strapped in her seat.

The ME's office said that she had died from percussion trauma,

so it was the sound waves hitting her car,

basically blowing all the glass out,

blew the roof off of her car, the trunk lid, the hood,

but her car, it didn't keep her alive,

but it kept her in one piece.

If there's any comfort to it at all, she didn't suffer.

She never saw it comin'.

NARRATOR: Summer was number 17 of the 43 people

that have all now been recovered

from beneath the one-square-mile stretch of liquid earth.

The devastation at Oso is a wake-up call

to the urgency of identifying hazardous slopes,

because the more we understand their unique physics

and exactly when they might slide,

the more we'll be able to protect ourselves

against one of nature's most powerful forces.

You could wish that it would have happened

on a Friday morning

where everybody would have been at work

and the fatality count would have been way down.

But, you know, it's nature.

Nature runs its course.

Captioned by Media Access Group at WGBH That's o access.wgbh.orgr man...

This NOVA program is available on DVD.

To order, visit shopPBS.org, or call 1-800-PLAY-PBS.

NOVA is also available for download on iTunes.