Horizon (1964) s38e07 Episode Script

Volcano Hell

This is the story of a signal with the power to save lives.
It is a line on a graph that scientists had puzzled over for years not knowing what it meant.
Then a man announced that he had uncovered its secret.
The line, he said, was a volcano speaking to him.
He believed it would lead him to a scientific Holy Grail - understanding when a volcano would erupt, but not everyone was convinced.
What followed would shatter the worid of volcanology and destroy lives.
After 70 years of sleep the Mexican volcano - Popocatépetl has awoken.
A plume of gas rises from its 6,000m summit.
There are hundreds of volcanoes spread across Latin America.
Popocatépetl is one of the most deadly.
Two million people live in its shadow and their lives may now be under threat.
The question is: should they be evacuated, or is it safe to leave them in their homes? It is one of the toughest questions that scientists have to face.
To make the right call scientists have to predict when the volcano will erupt and that's the one thing that's been impossible to say, but there is a man who thinks he knows.
Bernard Chouet believes he may have found a way to predict when a volcano will erupt and he knows what's at stake.
When one looks at these kind of natural phenomena, volcanic eruptions, you always think of the impact on human lives.
The ultimate quest is to understand enough about the activity in that volcano to be in a position to make prediction, predict the occurrence of an eruption.
If Chouet is wrong thousands may die, but if he's right he will do more than save lives.
He may, at last, atone for one of volcanology's greatest failures - the tragedy of Armero.
The most tragic disaster in recent volcanic history began quietly enough.
It was Colombia in the winter of 1984.
High up in the Andes smoke was seen rising out of the summit of Nevado del Ruiz.
For the local people who lived on the flanks of this vast, 6,000m volcano, it was an alarming sight.
As far as they were concerned, Ruiz had never bothered anyone.
A small team of scientists travelled to the top of the volcano to investigate.
The smoke was vast quantities of volcanic gas.
Nevado del Ruiz had come alive for the first time in over a hundred years, but would Ruiz erupt and if it did, what would be the consequences? The scientists threw themselves into research.
The investigation was led by Marta Calvache.
She went to the National Library and tracked down old documents detailing previous eruptions on Ruiz.
Calvache read how the glacier had been melted by a volcanic blast causing huge rivers of mud and ice to pour down the valleys destroying everything in their path.
A lot of people died and they describe how they died.
Some of them they were in the mud and people were not able to help them for several days and they died there 'cos of the sun because of the lack of water and food.
The details of the description were very frighting.
The description so alarmed Calvache that she drew up a hazard map.
If the volcano erupted the blast would melt the glacier causing water to pour off the mountain.
What concerned her most was the eastern flank.
There was one valley but it was fed by two rivers.
She saw how the water would pick up mud and rocks and thunder down both rivers.
Then they would meet, doubling their power.
Calvache followed the path of the river still further.
There, 40 miles away, lay a town of 30,000 people Armero.
Armero was a peaceful place.
The people lived quiet lives, far away from the guerrilla war that had plagued the rest of Colombia.
Now they were being told by a group of scientists that they faced total destruction from a volcano that was 40 miles away.
They will say that is not going to happen.
It doesn't happen for more than a hundred years.
Why is it going to happen now? The authorities had a practical point.
Volcanoes can be active for years without erupting and to go to the crippling expense of uprooting people from their homes for an indefinite period just because a volcano had started to rumble seemed absurdly wasteful.
The authorities would only evacuate Armero when they had the answer to one question: when would Ruiz erupt? Next week, next month, next year? It was the one thing the scientists couldn't tell them.
The volcano is incredibly complicated and we were unable to give them a prediction.
That is we couldn't tell them how big the eruption would be and we couldn't tell them when it would happen.
On November 13th 1985 the scientists ran out of time.
It was very, very dark, it was raining, raining very strongly, it was really especially bad weather.
That night the people of Armero tried to sleep through the storm, but 40 miles away Ruiz had erupted.
Around nine, when we were about to go to sleep, we began to hear sounds coming from the volcano.
We went to the main door and looked towards the mountain.
It was overcast, but explosions like flashes of lightening could be seen.
The sound got louder, but all the time there was these muffled sounds.
What Fernando Gil was listening to was the sound of a blast of hot volcanic gas melting the glacier.
The river of mud was on its way down the valley.
BBC News at 6 o'clock.
Thousands of people, one estimate says 20,000, are thought to have been killed in South America after the eruption of a volcano in Colombia.
The town of Armero is reported to be buried under six feet of mud and water and many of its 23,000 inhabitants are missing.
20,000 people died that night and over the next few days as rescue workers struggled to dig the survivors out of the mud.
For the scientists monitoring the volcano the death toll was especially hard to take.
It was a very difficult thing to face.
It was absolutely devastating that we had warned the people that this could happen, but we were unable to tell them exactly when.
Science had failed the people of Armero that night.
There was now a real sense of urgency.
Someone had to find an effective way of predicting when a volcano would erupt.
The scientists involved with that, including me, were very impacted by that eruption and that failure and there was a group of us that were resolved not to let that happen again.
They went back to basics.
What actually made volcanoes like Ruiz erupt? It came down to something common to all active volcanoes - magma.
Magma is molten rock that pushes its way into a volcano through the Earth's crust.
The magma then rises up the volcano.
If, like Ruiz, the top of the volcano is sealed the magma has nowhere to go.
The pressure increases and eventually the volcano will blow.
That magma has to get to the surface and so volcanologists and those monitoring volcanoes have the, one of their priorities is to identify that magma and to spot it as it moves up towards the surface.
Scientists believe the key was to track the magma.
If they knew how close it was to the top of the volcano they hoped that would indicate when the volcano would blow, but the problem was they couldn't see inside volcanoes, so to track the magma they had to rely on external tools.
One method was favoured above all others.
It was seismology.
Seismology is the study of earthquakes, literally the recording of the sounds when rock cracks or breaks and in active volcanoes they may be small, but there are hundreds, sometimes thousands, of earthquakes.
Seismology has been studied on volcanoes since the middle part of the 19th century and it's critical because you don't get a volcanic eruption without seismic activity, without earthquakes.
For years seismologists had analysed the thousands of seismic signals from active volcanoes and one signal stood out above the others.
They called it the A-type.
It was instantly recognisable.
It had a clear beginning and tailed off quickly.
It was the sound of rock breaking.
Seismologists knew that only one thing could break rock in a volcano magma.
As magma churns its way up a volcano it breaks through the rock.
By tracking the A-types scientists hoped to detect a pattern which would help them see where the magma was inside the volcano and how fast it was rising to the surface.
And in Colombia scientists seized on the data from Ruiz convinced they would find a pattern of A-types.
Find that pattern and they'd cracked the problem of prediction.
The idea was that if we plotted those together that the trends of those numbers would tell us something about the eruptive potential of a volcano.
But the more they looked the more they found there was no pattern, not on Ruiz or on any other volcano.
For all their hopes the A-types had proved to be a dead end.
Every volcano created its own pattern and, and it was almost impossible to come up with a chart saying that this is the book that we're going to go by.
Seismology had fallen short.
It would take the arrival of a different kind of scientist to deliver volcanology out of its confusión.
He trained as an engineer, then as a physicist, then as a rocket scientist, but for five years Bernard Chouet had locked himself in his office attempting to solve what the rest of science said was impossible: understanding when a volcano will erupt.
I realised that volcanoes were, although they had been looked at for a long, long time and people had always been fascinated by them, they were relatively poorly understood and so this was a frontier that was worth exploring.
When Chouet arrived in Colombia he, too, went straight to the seismographs.
He saw how before the eruption the scientists had plotted the A-types desperately searching for a pattern.
They had paper records at the time.
I noticed that they had, of course, a lot of A-type earthquakes and they were all identified with a little sticker.
But something else had caught his eye.
Tucked in amongst the A-types was another signal.
Scientists knew about it.
They'd even given it a name: the B-type, but no one knew what it meant.
It was sort of a mystical sort of thing.
We didn't really understand fully what it was, nor did we understand that we could use it.
The B-types were a conundrum.
Unlike the A-types, they had no clear beginning and they tailed away slowly.
Often they would merge with other signals making them hard to see at all.
What was really difficult to separate them and say definitively well it was this type of event that we'd be hopefully forecasting.
They were too messy.
But Chouet could see something in them that no one else could see.
It stared you in the face.
Wow, this is obviously different.
Embedded in the record among all these A - type earthquakes were classic looking quasi-monochromatic harmonic signatures, beautiful textbook examples.
Or as he would put it more simply: a long period event.
Well the easiest way to visualise the difference between these two types of event is to draw them and so the A-type event is characterised by a sharp onset.
This sharp onset here is a signature, it's a sound of rock breaking.
In other words, we're seeing the brittle failure of rock material.
On the other hand, if you look at the long period event it's characterised by this slow onset, gradual build-up of energy in the signal and then slow decaying single tone which lasts for a while.
This emergent onset, emergent or slow onset and then gradual decay of that signal tone, what we're seeing here is resonance.
All around us we hear sounds that are resonating.
In scientific terms resonance is the sound of a gas or a liquid put under pressure.
We're here in this church because we have here a beautiful example of organ pipes.
It's an instrument, musical instrument that is well known to everyone and this is an excellent example of resonance and a resonator.
In that case you're talking about a pipe which is filled with air and you're pumping air in that pipe and the air vibrates and you're hearing the sound of this vibration, that tone.
As the air is pumped under pressure into an organ pipe it resonates, it vibrates and produces a single tone that gently fades away.
On a graph its signature is unmistakeable.
A long period event.
Each time I press the key I pump more air in that pipe and so you hear the result of pumping air in the pipe which is this resonance.
Each time more air is pumped into the pipe another note of resonance is heard, another long period event.
Chouet realised the same process was taking place in a volcano.
Just as air is under pressure in an organ pipe, so magma and its gases are under pressure as they travel up the cracks of a volcano.
They will resonate inside each crack producing a long period event.
With each injection of magma into the crack so another single tone of resonance will be heard, another long period event.
But there is one key difference between a volcano and an organ air can escape from an organ.
Chouet saw that a volcano is like an organ that has been sealed.
If the crater is closed at the top the magma and the gas has nowhere to go.
The volcano will pressurise and eventually blow.
It was the appearance of the long period events that Chouet believed signalled a volcano was pressurising.
This is a defining moment because suddenly you realise the volcano is speaking to you and you understand the language.
Bernard Chouet now believed that he was close to the ultimate goal of volcanology, but there also existed another very different approach to predicting when a volcano would erupt.
Stanley Williams's method could not have been a greater contrast to Bernard Chouet's.
He was a scientist who climbed into craters because he believed that to predict volcanic eruptions you had to get up close.
Stan's a very aggressive scientist.
He's put in some tremendous science.
He believes passionately in monitoring volcanoes.
It's his love.
He loves to do that.
What Williams was looking for was gas.
Gas scientists knew that as magma moves up the volcano so it releases more and more gas, which escapes through holes on the surface called fumaroles.
What we're looking at now is called a fumarole.
As you can see it's a fairly hostile environment.
As the volcano starts to heat up and magma comes closer to the surface we're going to get a component of gas coming out of here which is coming from the magma.
It's going to have increased sulphur dioxide, increased HCl, the temperature's going to go up here and all over here we're going to start to see a whole lot more gas come out.
It's not an easy place to work.
The theory was that if they could measure the amount of gas being produced then scientists would have an idea of how close the magma was to the top of the volcano and to an eruption.
Williams was convinced that gas could be used as a means of predicting volcanic eruptions.
Several times he had noted increased levels of gas before an explosión, but it would be back in Colombia that his methods would be tested against Chouet's.
The result would end in tragedy.
In the southern most part of Colombia stands another volcano: Galeras.
In its shadow is Pasto, home to 300,000 people.
For years they had lived with its quiet rumblings.
Then, in 1991, there were signs it was entering a more dangerous phase.
International scientists flocked to Colombia.
Among them Stanley Williams and his colleague John Stix.
They climbed into the crater of Galeras.
There were huge amounts of sulphur dioxide being, being emitted, on the order of thousands of tons per day, which is a pretty good indicator that there was magma underneath the volcano.
The gas increases told them Galeras was both active and dangerous, but they then discovered something that really alarmed them: a lava dome.
When you see lava domes well you think um, this isn't so good because pressure can build underneath lava domes and oftentimes lava domes get blown out, blown out of craters.
Lava domes are a sign volcanoes are about to pressurise.
Fresh magma breaks through the crater and creates a dome of lava.
While the dome stays open the gas can escape, but if it seals the gas is trapped and the volcano pressurises.
An explosión is guaranteed.
Williams and Stix confirmed the volume of gas was a sign that fresh magma was still being forced up the volcano and would seal the dome.
They said an explosión at Galeras was a virtual certainty, but they were not the only scientists in Pasto.
Bernard Chouet was there too.
He agreed with Williams and Stix that Galeras would erupt when the dome sealed, but he went further and refined their prediction.
He told the Colombians they would know the dome had sealed by looking at the seismographs.
When the volcano began to pressurise a new seismic signal would appear.
For a while nothing appeared on the seismographs.
No long period events, nothing out of the ordinary at all.
Life in Pasto continued as normal.
Then a new signal did appear.
Once, sometimes twice a day, the seismographs recorded a new and ominous trace.
It was a long period event and if Chouet was right then the countdown to an eruption had begun.
This was an indication that you were pressurising the dome and that you're moving toward an explosión that would blow the dome apart.
The area around the volcano was evacuated.
Then, four days after the appearance of the long period events, Galeras erupted.
No one was hurt.
The only casualty: a police station that had been built on the crater rim.
It seemed both methods for prediction had been proved right Chouet's seismology and William and Stix's gas but what happened next would be tragically decisive to the future of volcanic prediction.
Six months after the eruption Stanley Williams was back in Pasto.
He was hosting an international conference on volcanoes.
With Galeras still active it was an ideal case study for the 50 scientists who had gathered there.
It was a good conference.
I have to say I was enjoying the conference very much, people were presenting interesting papers, it was lively.
There's a sense of camaraderie, especially in a small field like that.
There aren't that many volcanologists in the worid.
For Williams the highlight of the conference was his field trip.
He told local reporters that he planned to take a group of scientists into the crater.
Although Galeras was still active, Williams was confident it was not dangerous.
Just before the conference Williams had checked on the volcano's activity.
He had climbed into the crater and measured the gas emissions.
This time they were low, to Williams a clear sign that no eruption was imminent, but Galeras was anything but quiet.
For the previous few days a familiar signal had occasionally inked its way across the graph pape a long period event.
The two methods seemed to contradict each other.
Gas measurements suggested Galeras was quiet.
The long period events spelt danger.
The night before the field trip Williams met with other scientists at their hotel to talk through the options: to go into the crater or not.
We were faced with a dilemma in a sense.
Here was an active volcano, but a very quiet active volcano.
Where were we going? They talked about the long period events.
Some expressed reservations about going into the crater when these signals were appearing.
We were concerned by these long period events and what had happened when we'd seen them before.
But Williams and Stix were not seismologists and Chouet's long period events were still relatively untested.
There was a concern, but we didn't really understand what those events were, were telling, were telling us.
The one scientist who could have helped was missing.
Bernard Chouet had been unable to come to Colombia.
Both Williams and Stix decided to rely on the science they knew.
Gas levels were low, a clear sign, they believed, that Galeras presented no immediate danger.
There was a possibility that there could be explosive eruptions in the next weeks or months and yet the activity compared to previously was extremely low.
A decisión had to be made about the field trip the next day.
Every volcanologist who works on monitoring active volcanoes puts himself, or herself, in that situation at some point.
They have to.
It's the nature of the beast.
The field trip would go ahead as planned.
Early the next morning the scientists set off for the volcano.
It was an international team made up of Russians, Britons, Americans and Colombians, all eager to work in the crater.
Williams led the party.
By 9.
30am they were on the summit of Galeras overlooking the active crater.
We made different groups going up to the summit and most of the people were very happy.
12 scientists, led by Williams, left the rest of the party and headed down into the volcano.
They were joined by three Colombian tourists who had come to watch the volcanologists.
Then at 9.
47 the seismograph twitched into life.
For four hours the scientists worked quietly in the volcano, but at 1.
30 that afternoon they heard something.
I remembered at that time hearing three rock falls in the space of about a minute.
After a couple of 'em I asked Stan, you know, you hear those rock slides and as I remembered it was right after that that the thing blew up - boom.
The eruption hurled blocks of rock the size of refrigerators more than a mile up in the air.
I was looking up and you know I could see the plume going up and it was grey and the fog around it was grey and the blocks were grey and I'm looking up trying to see something and I hear like three or four heavy impacts around me, big blocks coming in that I'm not seeing and I just think well this isn't going to work, I'm just going to have to run for it.
Some of the group managed to escape out of the crater.
Others were still being bombarded by the blast.
You know I was running and falling, getting up, running and falling.
I ran, I passed José Arles.
He was obviously dead.
He was lying face down, not moving and I fell down close to Stan.
He had blood down the side of his face and sort of lifted his leg at me and said, "My leg is broken, it's broken, it's severed.
" I tried, I thought well there's only one thing I can do for him that's going to be meaningful and that's to pick him up.
I tried to reach him and my legs weren't responding very well and I, you know I was too weak and I just thought, you know, if I try and do this we're both going to get killed and so I had to leave him there and that was really awful because I thought, at the time I thought, you know, I'm probably leaving him to die 'cos I don't think, I don't think I'm going to make it and he can't move.
They found Andrew Macfarlane at the bottom of the crater deep in shock and with a fractured skull.
Stanley Williams was also pulled out alive.
It would be two years before he would walk again, but nine others had died in the explosión.
In going into the crater a team of experts had made the wrong decisión.
Some do believe the tragedy of Galeras could have been avoided if Williams had relied less on his gas methods and had better understood Chouet's long period events.
I think we can be a little myopic at times where we focus in on the information that we feel that we're experts on.
We don't feel it's the most important area, but that we're experts on and in a science like volcanology you have to be able to look at things collectively.
It's always easy in hindsight.
If I'd been there of course I'd looked at the records and I would have seen immediately the parallel between July and December and I would have started sounding bells, alarm bells in there.
I know I would have done that.
Whether I would have been successful in preventing people from taking a trip to the crater on that day would have depended on how how they would react to what I was saying.
Chouet's long period events had proved correct, but at great human cost.
Only when his work was taken up by others would volcanology make the advances it desperately needed to if it was actually going to save lives.
It happened at Popocatépetl.
The giant Mexican volcano had been active since 1993.
For the two million people living on its flanks its daily rumblings had become part of their lives.
Then the scientists saw a change.
The seismographs recorded a new signal long period events began to appear.
The scientists had heard about Bernard Chouet's ideas, so they asked him to come and take a look.
I told them what significance of long period events which they weren't aware of at the time.
From then on they began to track Chouet's long period events.
It's like a red light flashing.
When you see these signals something important is happening.
And in December 2000 the long period events began to increase dramatically.
Popocatépetl was pressurising.
The volcano was singing its song.
I mean actually this is a little like chirping if you want with these sustained waves from the long period event.
The question was: when did the people need to be evacuated? Valdés knew the consequences of a false alarm.
If people move from their homes and nothing happened then there would be no way they would move the next time the volcano threatened to erupt.
You have to keep your mind in the scientific work and say look, other volcanoes have done this, the potential of this volcano is that these particular villages could be in danger.
If Valdés was to convince people to move out he had to predict exactly when the volcano was going to erupt.
On 16th December the rate of Chouet's long period events escalated still further.
The volcano couldn't continue at this rate for much longer.
This is a siren song so to speak because it's telling you well, OK, I, I'm under pressure here, I'm going to blow at the top.
Valdés had to make the decisión.
We could clearly see that it would be between four and six in the afternoon of 18th.
The order was given.
Two thousand soldiers raced to the most vulnerable areas in an effort to get the people out in time.
30,000 people had to be evacuated in 24 hours.
Popocatépetl blew, as predicted, on 18th December.
It was the biggest eruption for a thousand years.
Chouet's work had provided the key to the accurate prediction at Popocatépetl.
Scientists had warned people of an impending eruption with confidence.
No one was hurt in the eruption.
It takes somebody to say look, we can really do this, we can go after this and understand what's going on in a volcano.
Science goes in steps, and this is a big one.
In volcanology this is a biggie.
It would be dangerous to suggest we're now safe from future eruptions, but the work of one man has at least offered us a chance to avoid future tragedies.