NARRATOR (BILL PATERSON): On February 23rd this year the houses and hotels that once stood here were destroyed by a massive avalanche. 31 people were killed and many more injured. The village was supposed to be safe, yet the Galtür disaster was the worst avalanche in the Alps for 30 years. Galtür has been not only a human tragedy but also a scientific mystery.

Scientists have studied this freak of nature to try to discover what extraordinary chain of events led to such a catastrophe. This is the home video of Christa and Helmut Kapellner from Düsseldorf, Germany. They visited Galtür on skiing holidays for the past 20 years. But their trip of February '99 was to end in tragedy. Helmut was killed by the avalanche. His wife Christa only just survived.

MAN: Look at the camera. Big smile, big smile.

MAN: Smile, smile, smile.

NARRATOR: On the day of the disaster they were returning to their hotel where they would have been safe. Nobody knew the avalanche was about to strike and they lingered to video the village. Helmut's last words were recorded.

HELMUT KAPELLNER: It's the 23rd of February and this is what Galtür's car park looks like. Even the back road to Wiel is closed.

CHRISTA KAPELLNER: On the way we took a video, it shows how much snow there was. If we hadn't stopped to take the video then we might have been safe.

NARRATOR: The final moments on the videotape, minutes before the avalanche hit.

CHRISTA KAPELLNER: There is Helmut, the snowman standing in front of Jesus. I haven't seen such snow in 20 years. The avalanche came from behind. Helmut shouted 'Christa!', but I couldn't answer. His legs were pulled from under him by the force and I was hit from behind.

LUGGI SALNER (Villager): I saw a huge wave coming towards us, it was like in those films about Hawaii. I thought this can't be happening, I'm in the wrong film!

GEORG WALTER (Villager): This cloud of snow came down like a steam roller, up and down like this. I saw it rise 40 to 50 metres over that hotel, maybe even higher.

LUGGI SALNER: I was horrified, my eyes widened with fear. Then I screamed, 'it's coming!'

NARRATOR: Christa was buried under a metre and a half of snow.

CHRISTA KAPELLNER: I tried to create an air pocket with my hands and get to the surface, wherever that was. I did not succeed and gave up.

NARRATOR: Jason Tait, an English tourist in the path of the village that wasn't hit actually videoed the avalanche from his hotel window.

JASON TAIT: I don't know whether you can see this but that's the snow. It looks like a snowstorm.

CHILD: It was a avalanche.

JASON TAIT: You couldn't see more than a few inches away from the pane of glass. It was like looking into a swirling washing machine, whatever and eventually when it subsided we made our way across to where casualties had been taken and tried to assist there. We spent 7 hours digging to see whether we could get anyone out alive or indeed someone that we might be able to revive, but the, the snow had set like concrete. Wherever you went it was solid and it was, it took some digging and of course you can imagine that, that there was very little chance of, of anyone surviving something like that.

NARRATOR: But Christa did live thanks to the efforts of the local doctor.

DR. FRIEDRICH TREIDL (Galtür Doctor): As quick as possible I took my rescue bag and went out to the avalanche. Some buildings I used to see I didn't see them, others I saw completely destroyed, some rooms are filled up with snow completely and there was many people digging for victims of the avalanche and we found a garage and there we tried to help those people.

NARRATOR: After nearly 2 hours buried under the snow, Christa was hypothermic. In the garage she regained consciousness.

CHRISTA KAPELLNER: I kept asking about my husband, Helmut. It was terrible to find out that he'd died. I had been hoping that he was alive.

FRIEDRICH TREIDL: My wife and my daughter were in a destroyed house, but they had good luck that they were in a room which was not destroyed and many of my patients died in that night and even whole families so it is terrible.

NARRATOR: It was a mystery as to how such a disastrous avalanche had struck this safe village.

LUGGI SALNER: Nobody could have predicted this, it was a catastrophe.

JASON TAIT: We felt that anything would come down a mountain would hit the valley bottom and stop, but of course when, when it came to light when we saw the big one we realised that that was no element of safety at all. It bounced and hit the other side.

GEORG WALTER: We had no idea that such a huge avalanche could happen. It was abnormal. Nothing like this has been recorded in the last 600 years. Why did it snow for so long?

NARRATOR: Scientists were just as puzzled and would spend months trying to find answers to these questions. For years scientists have grappled with the complex physics that underlie all avalanches - what causes them, how they can be predicted and why they are so destructive. But only rarely are avalanches as massive and terrifying as the one that struck Galtür in February '99 and now science is attempting to understand what made it so much worse than other avalanches, the tragedy being that no avalanche should ever have reached the village.

It was this mountain that the avalanche came down. Most years a small avalanche follows the same route, but trickles out safely long before the village. It even has a name: the Wasser-Leiter, or water ladder. The avalanche had never hit the village before because Galtür is over 200 metres away from the base of the mountain. Jorg Heumader is responsible for the safety of villages in the region. He, like everyone else, had always assumed the Wasser-Leiter avalanche wasn't dangerous.

JORG HEUMADER (Avalanche and Torrent Protection Service): When I heard about the disaster, houses destroyed and people killed in this special area, the Wasser-Leiter area, I couldn't believe it. I heard it but I couldn't believe it. I could have told you 20, 20 places where I would have believed there was a disaster, but not this one.

NARRATOR: Snow fences had been built on many slopes elsewhere in the valley as a protection against dangerous avalanches. They hold the snow pack in place, but they weren't thought necessary in the Wasser-Leiter area. Historical records had never reported the avalanche to be a problem. Despite this, to be absolutely safe, the Wasser-Leiter had influenced the building zones of the village. No houses are built in the red area. In the yellow zone buildings must be reinforced to make them avalanche-proof. Only in the green zone are there no building restrictions at all, but it was here, in the safe zone, that the avalanche caused the most damage. There lay the mystery.

Nowadays computer simulations are used to create the hazard zones for villages such as Galtür. They are based on a worst-case scenario for the amount of snow which an avalanche might theoretically contain. They use a depth of snow so huge that it could only happen once every 150 years. In the Galtür simulation even this avalanche failed to reach the village. So now scientists must dissect every aspect of the Galtür disaster and put the findings into a new computer simulation to reveal how the avalanche caused so much destruction.

Over the last 6 months Horizon has followed the avalanche experts brought in to investigate Galtür. Two of the key scientists were Paul Foehn and Stefan Margreth. They believed it was a unique combination of natural elements, some quite ordinary, some extreme, that conspired to create the catastrophe. First they visited Galtür and the mountain ridge where the avalanche started from. They needed to work out how it could have travelled so far across the valley floor.

STEFAN MARGRETH (Swiss Snow and Avalanche Institute): Here the starting zone was up there in the fog on the cliffs on both sides of the gully. The avalanche jumped and flowed down and went into the direction of Galtür across these flat area and made the, the destruction of the houses.

NARRATOR: Studying how the buildings were destroyed would help them calculate the forces involved.

STEFAN MARGRETH: For the collapse of the house there are two possibilities. One is the main impact destroys the house, is not any more stable. The other possibility is that at the first impact not the whole building collapse but, but for example a window is broken, a door is broken, a room is filled up with snow, there's an over-pressure in the building and like an explosion, the walls can break or the roof can fly away.

(ACTUALITY CHAT BETWEEN STEFAN MARGRETH & PAUL FOEHN)

STEFAN MARGRETH: Another thing is you'll see impacts of stones or small projectiles carried at very high speed by the avalanche and so I think it's like the shots from a machine-gun.

NARRATOR: To help them understand Galtür, they would analyse every aspect of the avalanche and also compare events here with what they knew about past avalanches in the Alps.

ARCHIVE FILM NARRATOR: From the continent there comes the most tragic story of avalanche disaster for over 100 years. In Austria, Switzerland and Italy one incident after another has resulted in the cutting of communications, destruction of property and heavy loss of life.

NARRATOR: Until last year the worst series of avalanches in the Alps had been in 1951 when over 200 people were killed. There were many casualties because so little was then known about the physics of avalanches and weather prediction. The Austrians were determined it should never happen again and built a chain of automatic weather stations on mountain tops in an attempt to understand how avalanches happen. Rudi Mair is in charge of the automatic stations in the Tirol. Often the only way to reach them is by skis. The detailed information from these stations has been key in unravelling the weeks of weather leading up to the Galtür disaster, a series of events very different from normal.

RUDI MAIR (Avalanche Prediction Service, Tirol): Usually in February we have very fine weather here in the Tirol with a lot of sun but not much snowfall so this year this February '99 was a very special event.

NARRATOR: A month before the disaster snow was providing fine skiing at the resort, but unknown to the villagers the powerful forces of nature were already building up 4,000 kilometres away. Out in the Atlantic a storm was brewing. That storm was the first event that had to be analysed. It began on 20th January. A mass of turbulent warm air from the Tropics headed north. Being warm air it was heavily saturated with moisture which formed into towering clouds as a frontal system cooled and swung back around towards Europe. The storm it created was not abnormal. What made the situation exceptional was the relentless series of storms which followed over the next few weeks.

RUDI MAIR: The Atlantic is very, very important for the weather in the Alps. We call it the weather kitchen for the Alpine regions and we have a lot of humidity in this air coming in from the Atlantic Ocean and if this air comes to the Alps the air is raised at the Alps and therefore the temperature goes down and then precipitation or a snowfall will occur.

NARRATOR: The first storm might not have created so much snow, but for a terrible coincidence. At the same time it reached the Alps a separate stream of very cold air blasted in from the Arctic. The combination was key.

RUDI MAIR: These two air masses met at the north of the Alps and therefore we have a lot of snowfall and especially of very dry and light snow.

NARRATOR: By the beginning of February nearly a metre of snow had fallen on Galtür. More snow and things could become dangerous.

WEATHER FORECASTER: We expect snow and more snow...

WEATHER FORECASTER: The forecast is for three metres of snow...

WEATHER FORECASTER: The snowfalls have been immense...

NARRATOR: The first snow storm was followed by two even larger frontal systems. When they arrived over the Alps a further 3 metres of snow fell on top of the ever-burgeoning snow pack.

DR PAUL FOEHN (Swiss Snow and Avalanche Institute): In end of January the big snowfall started and 3 big snowfalls followed each other. This gave naturally a big amount of snow in these mountains which we didn't have for a long time.

NARRATOR: That February 4 metres of snow fell in the Galtür area breaking all records. But this didn't explain the first puzzle. When measurements were taken at the top of the Galtür mountain it was found that the snow pack was even deeper than could be accounted for just by snowfall. There could only be one answer.

RUDI MAIR: Wind is the most important of all meteorological perimeters concerning avalanches. The wind is the so-called build off the avalanches. With Galtür we had almost for 3 weeks very, very strong winds and winds that are, have 80-90 kilometres or even more than 100 kilometres. They can move tons of snows within some few hours, within 1 or 2 hours they can move 20/30 tons of snow from one side of a ridge to the other.

NARRATOR: Again fate intervened. If the wind had come from a different direction then the snow would have drifted onto the slopes facing away from the village, but the strong winds of February blew towards Galtür.

PAUL FOEHN: These north-westerly winds have blown the snow into the south-easterly slopes exactly the one which, which is situated above Galtür.

NARRATOR: This succession of extreme weather events had set the stage for the disaster, but the stability of the snow pack depended on the particular type of snow that fell.

STEFAN MARGRETH: In a total we have perhaps 20 different types of snow which are very different. The new snow has different shapes like star shapes, one of the most interesting. Then we have plates all hexagonal in shape. Then we have spikes and so many things I can't even tell all of them.

NARRATOR: A scanning electron microscope reveals the myriad of forms. Exactly which kind of snow formed during February would be crucial. The shape of the snowflakes, be they smooth or angular, determines the way they bind together or fall apart. The different snow types make up distinctive layers in the snow pack, essential in understanding how avalanches start. At the Swiss Avalanche Institute they have come up with an ingenious way of identifying the layers. Adding black ink to cross-sections of snow pack shows how well the crystals have bonded. Here the centre layer is loosely bound together, known as the weak layer compared to the more densely packed snow above and below. It is when this weak layer collapses that all the snow above slides off.

PAUL FOEHN: These two slabs represent two layers of a snow cover. Ordinarily they stick together by cohesive forces and now we tilt them up and it takes some angle until they slowly slide off after quite a lot of time. The ballbearings would represent a weak layer and afterwards if you have a weak layers in-between two layers the sliding off goes quite quickly and easily.

NARRATOR: Paul Foehn uses a strain gauge to measure the strength of the different layers. All avalanches are due to a weak layer, but they're never normally as catastrophic as at Galtür. What must have made this avalanche so large was that the weak layer didn't collapse soon enough. Instead it held on as more and more snow built up on top before it finally gave way. So why did the weak layer hold out so long? The answer would like right at the top of the mountain inside the fracture line itself. When Paul studied the snow crystals of the weak layer more closely he discovered frozen melt water between the grains. The weak layer was a special type known as a melt crust. Melt crusts are created when the snow thaws the top layer of snow during the day which then refreezes at night. Any snow that then falls on top only binds partially to the snow below. Knowing this allowed Paul to look through the recent weather records to pinpoint when it was warm enough for a melt crust to form.

PAUL FOEHN: The melt crust formed around end of January before the snow falls and then we had quite often quite warm weather naturally until the snowfalls started end of January till 23rd February when the catastrophic happened.

NARRATOR: The crystals of the melt crust were bonded together with ice which must have made it firmer than other weak layers. It was created a whole month before the disaster. At any time the weak layer could have given way forming just a small avalanche, but instead more and more snow piled on top. Galtür was sitting below a time bomb.

(RUDI MAIR GIVING WEATHER WARNING)

NARRATOR: Rudi Mair doesn't only monitor the weather stations. He's also head of avalanche prediction in the Tirol. To assess the risks he combines information from the automatic stations with the weather forecast. By mid-February things were grim.

ANNOUNCER: And now here is Rudi Mair with an avalanche warning

RUDI MAIR: Good morning, we have a widespread level 5 warning. This is a critical situation.

NARRATOR: In all the years prior to last winter level 5 was only used 3 times. In the month of February 1999 there were 16 level 5 warnings.

RUDI MAIR: I was in my office about 18-24 hours a day because a lot of people are called, wanted information about how the situation will be in the future and therefore it was a, a little bit frustrating because I knew that the situation will get worse and I couldn't do anything.

NARRATOR: Rudi Mair's warnings were for large regions of the Tirol. He can't give predictions for individual valleys.

RUDI MAIR: The whole Tirol is a mountain region. That means in the Tirol we have thousands of mountains and every mountain has hundreds of slopes and therefore you have hundreds of thousands of slopes in the whole Tirol and therefore it's absolutely impossible to give an exact degree of hazard or danger for every slope.

NARRATOR: For a more detailed assessment Galtür ski guides go out into the mountains to search for dangerous, weak layers.

(SKI GUIDE'S ACTUALITY CHAT)

NARRATOR: At the time tests confirmed that there was a high risk of avalanches in the area. One test involves breaking off the snow above a weak layer. However, because of the history of the valley they assumed no avalanche would ever reach the village. If Paul and Stefan were to find out why the Galtür avalanche did hit the village then they'd have to calculate the exact weight of snow that was in the starting zone. Paul compared photos of the fracture line with detailed maps to measure its length. This was combined with the depth and density of the snow to give the total weight that was released.

PAUL FOEHN: The starting zone is large. It's about 500 metres and the depth of the snow which got loose is about 2½-3½ metres and finally the, the mass of the whole snow is about 170,000 tons which broke loose. It's a huge amount of snow.

NARRATOR: With this massive amount of snow, along with the steepness of the slope, they hope to recreate the disaster on the computer. But even with so much snow the avalanche stopped short of the village. Something else was involved. Stefan knew the moisture content of the snow was important and this would be the last piece of the puzzle. Wet snow forms slow-moving avalanches called slab avalanches. They don't travel far and their movement is easy to predict. Drier powder snow, like that at Galtür, behaves very differently.

For years scientists have tried to analyse the turbulent motion of powder avalanches through complex mathematics and chaos theory. Much of the work is being done on physical models. Tiny particles of glass released into tanks of water. Powder avalanches in miniature. Scientists have always known that these models aren't accurate enough to truly represent what goes on. They needed the real thing. But how to study full-scale avalanches without getting killed?

Just 2 weeks before the Galtür disaster a team of scientists carried out a bold experiment which would have a profound impact on Paul and Stefan's analysis of Galtür. The scientists had built a reinforced concrete bunker at the base of a mountain. An array of instruments was installed to monitor every aspect of the planned event. Inside the bunker scientists would collect unique data as a massive avalanches thundered over them. The radar dishes would even allow them to peer into the seething mass of powder. All they needed now was an avalanche set off with dynamite. If their sums were wrong the bunker would be flattened and everyone killed. The avalanche was bigger than they'd intended.

(ACTUALITY CHAT)

DR. DIETER ISSLER (Swiss Snow and Avalanche Institute): We were really surprised by the force by which the avalanche hit the shelter. We did not expect such a big thing to come down. This is a very massive and solid bunker. I think the walls are about 40 centimetres thick and that was the last you could feel the vibrations of the whole building. We heard a strong noise. That was because the door broke open and the snow came in. Then the pressure in the shelter goes enormously. It was like diving into 2 metres of water. Well after the avalanche had hit us we had first tried to try to get out because the shelter was completely covered by snow and so we started digging a tunnel out, then we had to break out. It was really very compact snow.

NARRATOR: Once out of the bunker they could begin to analyse the data. Video revealed the avalanche to be travelling at over 100 miles an hour. Stefan needed all the details.

URS GRUBER (Scientist): ...this area.

STEFAN MARGRETH: What was the height here of the powder cloud?

URS GRUBER: You see here the big pile, that's 20 metres, so it is about 50 metres, but this is not the maximum height, but I think the, the maximum height is about 200 metres.

DIETER ISSLER; Here also we see that the height is growing rapidly as the avalanche gets jammed into the valley. The, it is, gets decelerated at the front because it has, has to climb up so steeply and more material is coming from the back.

NARRATOR: The key thing about a powder avalanche is that it's not only fluffy powder snow. At the bottom is a solid layer of much denser snow. Over this rides a huge cloud of the lighter powder. In-between these two layers scientists now believe there's a third layer, a tumbling mixture of light and heavy snow. It's called the saltation layer and it would play a deadly role at Galtur. How it forms is only now being understood.

DIETER ISSLER: This is contains particles of different sizes, many small ones like the fine snow in the avalanche and there are a few bigger ones that are, have not broken up that much and now as I shake it, I get the big particles on top and here we have the small particles.

NARRATOR: The presence of a saltation layer hidden inside avalanches had been suspected. Now it was confirmed by the radar data from the bunker experiment. This screen shows a cross-section actually inside the avalanche with huge blocks of snow thrown upwards. The saltation layer could travel further and faster than expected.

DIETER ISSLER: You can see here the darkness of the colour corresponds to the particle size, the darker the dot the bigger the particles, or to the density of the snow, so we see a big particles riding on top of a less dense layer. We interpret this as the lower layer being strongly sheared so the, the particles are rolling over each other and very strongly agitated and the big lumps are travelling piggy-back on the, the smaller ones.

NARRATOR: The radar work also confirmed something else remarkable. The avalanche increased in volume as it descended. As the avalanche hit the layers of untouched snow here on the left it ate into the snow and gathered it up as it passed, getting bigger all the time. These discoveries from the bunker experiment turned out to be decisive. Stefan applied this latest research to his computer simulation of the Galtür avalanche. He knew the starting volume of snow wasn't enough to cause all the building damage and so he back calculated how much extra snow was gathered up on the way down the mountain.

STEFAN MARGRETH: During the fall in the track the avalanche picked up two times the initial volume of snow in the avalanche and I think it was only possible through this picking up that such a disastrous avalanche could grow up.

NARRATOR: The avalanche now weighed over a third of a million tons. In most powder avalanches the dense layer, because it contains the heavier snow, causes the damage. Thanks to the bunker experiment, Stefan knew that this might not necessarily be the case for Galtür.

STEFAN MARGRETH: We found out that the dense parts didn't travel into the village and were surely not responsible for the destruction of the houses.

NARRATOR: Only the powder cloud and saltation layer reached the village. While the saltation layer destroyed the buildings it was the powder cloud that suffocated the victims.

GEORG WALTER: The snow is like fine dust. I had excperienced avalanches before, so I protected my face with my hands.

FRIEDRICH TREIDL: If you are in a room and the avalanche comes in this avalanche had such a power that the whole room was completely filled up with snow and all those people who have no space to breath they die of asphyxia.

NARRATOR: After 6 months the research was nearly complete. With all the weather data, the field measurements from Galtür, the findings of the bunker experiment Stefan had the information he needed. Once this raw data was fed into the computer, the simulation would finally reveal how an avalanche could have hit the village. The computer showed a wave of snow over 100 metres high. It reached 186 miles an hour.

STEFAN MARGRETH: It took about 50 seconds until the avalanche crossed the road and about nearly 2 minutes until the avalanche was completely stopped. You have perhaps only 10/20 seconds to react to run away and that's a very short time. You didn't look up then I think you have no chance to realise that an avalanche is flowing down.

NARRATOR: On the day of the disaster the roads out of Galtür had been closed for almost a week because of the risk of avalanches in the rest of the valley. But the village itself was still seen as a safe haven. They'd even organised a fun race down the main street.

JASON TAIT: The atmosphere in the village was strange. There was a lot of milling around, people couldn't actually go anywhere, there was no skiing to be done, people were wondering what to do with themselves. We felt very trapped, not knowing where it would all end really.

GEORG WALTER: The roads were closed because of avalanches. But of course no-one realised that such a big avalanche could happen here.

NARRATOR: The avalanche scientists now know the intricate chain of events that led to that fateful Tuesday. Throughout February 3 huge weather fronts had dropped more snow on the mountain than ever recorded before. The wind piled it up into even deeper drifts. Deep below the melt crust supported an ever increasing weight of snow. A graphic reconstruction reveals what happened next. Just before 4pm the melt crust collapsed. In the 50 seconds it took to come down the mountain the avalanche doubled in size. Its speed was nearly 200 miles an hour. The saltation layer and powder clouds sped on ahead causing all the destruction. Seven buildings were demolished, 31 people killed. It was all over in a couple of minutes. The people who faced the brunt of the disaster were now only too aware how violent nature could be and how little time there was to rescue the injured.

FRIEDRICH TREIDL: It was not possible that help from outside could come in because helicopters couldn't fly because of the snow storm, so it was good luck that there was a rescue dog here in the village.

NARRATOR: Heiko turned out to be a hero. Here he is on a training exercise searching for someone deliberately buried under the snow, but during the first night of the disaster he was the only rescue dog in Galtür.

(ACTUALITY CHAT)

NARRATOR: One of the people he saved that night was Christa Kapellner buried under the avalanche. Christa knew a little about how she was rescued or how her husband died. This summer she returned to Galtür to find out more

SILVIA KATHREIN (VILLAGER): You were found here under two cars, one was on top of the other. The snow was very high.

CHRISTA KAPELLNER: So I must have been thrown from the bend in the road to here.

SILVIA KATHREIN: Yes, but your husband was found down there by the fence.

CHRISTA KAPELLNER: But we were walking only a metre apart.

SILVIA KATHREIN: Goodness, I don't believe it!

CHRISTA KAPELLNER: That's over fifty metres away!

SILVIA KATHREIN: Your husband was hit by a falling roof but the pressure threw you to here. Heiko sniffed all around and then he came here and started digging. This is Heiko who saved your life.

CHRISTA KAPELLNER: You saved my life and gave me a second chance, but you don't even understand what I'm saying.

NARRATOR: At Galtür changes have already begun. The avalanche hazard zones have been extended and they are building new defences, defences that up to now have never been thought necessary.

JORG HEUMADER: To prevent the avalanche from starting up there on the top of the mountains avalanche snow fences made of steel will be constructed for the settlement and all buildings built here now under construction now have to be avalanche-proof buildings with reinforced concrete walls, no windows on the avalanche side and such. The other part down here on the valley floor is an avalanche dam and an avalanche wall for maybe 300 metres to the settlement here and the combination will give safety for the settlement.

INTERVIEWER: You hope.

JORG HEUMADER: No, I am quite sure. Of course you have some remaining risk in high mountain area, nevertheless, but of course a high degree of safety. Not 100%.

NARRATOR: Throughout the world scientists are facing the reality of how unpredictable the forces of nature can be. This makes any assessment of hazard zones for floods, hurricanes, earthquakes and avalanches uncertain. Experts now recognise that whatever equations are used to estimate the worst case scenario on very rare occasions it will inevitably be exceeded by the extreme event. Despite all the advances of modern science, nature will always have the last word.

STEFAN MARGRETH: It's not possible to take the very extreme event into account otherwise it's not possible to settle in the Alps because you have, everywhere you have rock falls, you have slides, you have big avalanches and it's necessary to take a certain risk into account.

CHRISTA KAPELLNER: I kept asking myself, why me? Why did we have to be on the road at that time? But I've stopped asking why because there is no answer.

Return to programme summary