快猫短视频

ASCENSION DAY, Thursday 8 May 1902. A day that should be written in blood. So
wrote Vicar-General Parel of the Caribbean island of Martinique, as he surveyed
the shattered ruins of the town of Saint-Pierre. No one had foreseen disaster on
this scale, despite the ominous rumbles of Mount Pel茅e, a volcano that
loomed over Saint-Pierre much as Vesuvius overshadows Naples. The town was
crammed, the numbers swelled for celebration of the feast day. At 7.50 am, a
black cloud appeared on the lip of the crater and began to speed towards the
town.

The last message to come out of Saint-Pierre, just two minutes later, was a
single word from a telegraph operator: 鈥淎llez鈥�. Of the 28 600 people in the
town, only two survived. The buildings were seared and blasted, thick stone
walls knocked over, wood charred and metal buckled. It remains the worst
volcanic disaster this century.

The geologist Alfred Lacroix, who was sent by the French government to
investigate the catastrophe, called the speeding mass of superheated material
that destroyed the town a 鈥渂urning cloud鈥�. Nowadays, scientists would call it a
pyroclastic density current.

Such currents have wrought havoc in many other eruptions: Mount St Helens in
the US, El Chich贸n in Mexico, Mount Pinatubo in the Philippines, Mount
Unzen in Japan and Soufriere Hills on Montserrat in the Caribbean. Their scale
can be awesome (see 鈥淲aiting for the Big One鈥�) and their behaviour is
astounding. They hug the ground for tens or hundreds of kilometres, travelling
faster than an express train, sometimes leaping over high ridges as they go.
Some even travel across the surface of the sea, devastating islands in their
path. The currents can be as hot as 800 掳C, and incinerate trees and houses.
Indeed, the heat can be so intense that the volcanic particles fuse to form a
solid sheet of black glass, welded onto the landscape.

Volcanologists are still trying to work out how pyroclastic currents perform
these tricks. They鈥檙e literally too hot to handle, and so our knowledge of them
is based on a patchwork of observation, inference and modelling. Much of it
depends the pumice and ash-rich deposits, or ignimbrites, that the currents
leave behind. However, different deposits point to different interpretations.
Now, though, new ideas are emerging that could yet reconcile the apparently
contradictory views and give us a clearer window into the burning clouds.

Not all volcanoes produce pyroclastic currents. It depends on their chemistry
and gas content. Most ocean island volcanoes, such as Hawaii, produce hot, runny
lava flows from which gases bubble harmlessly away. The lava may damage property
and look spectacular, but it moves so sluggishly that people can generally avoid
its path. Continental volcanoes such as Cotapaxi and Vesuvius, though, produce
viscous magmas in which gases become trapped, bottling up intense pressures that
may then be released in cataclysmic explosive eruptions. These pulverise magma
into ash and frothy pumice, and are the spawning grounds for pyroclastic
currents. Once unleashed, the currents move too fast to run from鈥攁nd are
too dangerous to study at close hand.

Therein lies the problem with trying to understand these violent events. The
downrushing cloud conceals the incandescent machine within. To get too close is
to court death. Only a few people have lived to tell the tale (see 鈥淐lose
Encounters鈥�). Others have paid the ultimate price. At Mount Unzen, small
pyroclastic currents coursed so regularly down one particular valley that, on 3
June 1991, 43 journalists and scientists waited on the valley side to watch the
next one go by. But that next current, without warning, engulfed the valley side
and the observers with it. All were killed instantly.

Even robots can鈥檛 help us here. No instrument yet devised can withstand the
heat and crushing impacts. Nonetheless, researchers have pieced together their
evidence and produced models of how the currents could be operating.

One approach considers pyroclastic currents as 鈥渁sh hurricanes鈥�, in which ash
and pumice particles are suspended and transported by turbulent whirlwinds of
gas. This was first suggested in the 1960s by Dick Fisher of the University of
California at Santa Barbara, and later elaborated by his former student Greg
Valentine and, more recently, by workers such as Andrew Woods at Bristol
University and Marcus Bursik at the University of Buffalo in New York.

The idea is that turbulent gusts and eddies of gas drag the hot pumice and
ash particles along with them, while whirling vortices prevent the particles
from falling to the ground. Gravity powers this volcanic hurricane鈥攖he
particles of ash make it denser than the surrounding atmosphere, so it tumbles
down the slope like a swirling flood of water, gathering momentum as it goes.
Like water, it generally follows low ground and valleys, though it can leap over
ridges just as powerful floods can.

It doesn鈥檛 take much solid material to make a cloud of hot gas denser than
air: Brian Dade and Herbert Huppert at the University of Cambridge have
suggested that less than 1 per cent of particles is enough. The faster the
resulting ash hurricane travels, the more vigorously it is stirred and churned,
so the more particles it can carry. Nevertheless, the cloud remains just so much
hot air: the whirling particles are so widely spaced that they rarely
collide.

Evidence for this model comes from various sources. Fisher, for instance,
discovered that in ancient times huge currents had been light enough to travel
across the Bay of Naples in Italy, burying land on the opposite coast under
metres of ignimbrite. In the past couple of years, Sharon Allen, working with
Ray Cas at Monash University, found ignimbrites on Greek islands around Kos that
were left by pyroclastic currents that had travelled more than 35 kilometres
across the Aegean Sea.

More telltale evidence comes from the carefully sorted layers that some
pyroclastic currents leave behind. When the Taal volcano in the Philippines
erupted in the 1960s, rarefied, turbulent currents moved outwards like the
ground-hugging surges from nuclear explosions. They left behind fine volcanic
debris sculpted into dunes. Dig into these dunes and you see repeated thin
layers of fine and coarse ash, which built up one by one beneath the onrushing
current. The coarser layers were presumably deposited by more energetic eddies,
while finer material settled out between the passage of these eddies. In just
the same way, gusts in a gale can tear branches off trees, while leaves settle
to the ground in moments of intervening calm.

So far so good. The problem is that many deposits left by pyroclastic
currents don鈥檛 look like this: they choke valleys and bury plains under chaotic
debris up to several hundred metres deep. Rather than having fine layering, in
these deposits the finest debris is mixed haphazardly with pebbles and boulders.
It鈥檚 as though the whole mass had suddenly been dumped like so much concrete mix
out of a truck. These colossal deposits of ash, pumice and huge blocks of rock
don鈥檛 square easily with an ash hurricane, and prompted the development of
radically different ideas.

Could the rocks and boulders simply have bounced down the slope? If this were
all that were happening, friction would quickly bring the debris to a halt at
the bottom. But pyroclastic currents can carry boulders for long distances along
very shallow slopes. What could reduce friction enough to make this
possible?

Part of the answer could come from a phenomenon called granular flow.
Technologists working at, say, breakfast cereal factories, have long known that
when a mass of grains flows down a chute, the vigorously bouncing grains
transfer their momentum to each other. Billions of impacts, reverberating
through the moving mass, keep the whole mass suspended, allowing it to flow down
surprisingly gentle slopes.

Stuart Savage of McGill University in Quebec and Charles Campbell of the
University of Southern California use the term 鈥済ranular temperature鈥� to
describe the way these particle impacts and vibrations intensify as the
downslope speed increases, expanding the whole moving mass just as heating a gas
causes it to expand by increasing the random movement of gas molecules. However,
in breakfast cereals, the raisins, nuts and flakes obstinately separate
themselves during transport. The well-mixed nature of many ignimbrites means
that simple grain flow can鈥檛 be the whole story.

Like a fluid

But another analogy from industry might help: fluidisation. Blast gas through
powders and grains and they lose their friction and behave like a fluid.
Something similar may happen in a pyroclastic current, with the gas coming from
fizzing, or 鈥渄egassing鈥�, pumice fragments. But Ronnie Torres, a volcanologist in
the Philippines, has found that hot, fizzing pumice isn鈥檛 a vital part of the
recipe. In 1991, pyroclastic currents from the explosive eruption of Mount
Pinatubo filled the valleys around the volcano to the brim with loose, friable
volcanic ash. In a few terrifying hours, over 200 metres of ash and pumice were
laid down.

After the eruption, water from monsoon storms cut high, unstable cliff faces
into this hot ash; the cliffs collapsed, generating entirely unexpected
pyroclastic currents that travelled several kilometres, months and even years
after the volcano itself had stopped erupting. Torres showed that these
post-eruption pyroclastic currents travelled just as efficiently as the ones
that occurred during the eruption鈥攁nd left deposits that looked almost
identical. Because the pumice had finished degassing long before, this process
clearly could not have contributed significantly to the mobility of these
currents.

Hurricanes, grain flows, fluidisation: the list of suspected mechanisms is
growing longer. None seems to provide the whole answer, but might they all be
involved in some way? Clues to this riddle clearly lie within the layers of
pumice and ash. But these deposits are no longer moving. Their message might
become clearer if we knew just how they stopped.

Until recently, there was a general consensus that such a dense mass of
moving rubble decelerates, loses its gas and grinds to an abrupt halt,
transforming the flowing mass into a thick, stationary layer of debris. Thus, in
this view, the resultant layer of ignimbrite is a freeze-frame record of the
entire current. This is in sharp contrast to the way an ash hurricane builds up
an ash layer gradually as successive particles come to rest on top of each
other. But鈥攁nd this is crucial to making sense of the riddle鈥攃ould a
thick, chaotic rubble layer also have been laid down gradually, particle by
particle?

Neat layers

Intrigued by this thought, one of us (Mike Branney), working with Peter
Kokelaar of Liverpool University, is trying to reconcile the different models.
In Mexico, the Philippines and Tenerife, the ignimbrite left after the passage
of just one current can show all manner of mixes between neat layers and
disorganised, concrete-mix rubble. They鈥檙e hybrids, so perhaps most currents are
likewise hybrids of the various processes. Both the visible, turbulent shell and
the dense inner core could be intermeshing parts of the same machine.

What鈥檚 more, different processes probably operate at different levels within
the current. And the different levels might collaborate to keep yet more debris
powering downslope at high speed. Thus, turbulent billows propel dust and small
pumice pebbles high in the current while in the densely packed areas nearer the
ground, rock collisions and fluidisation take over. Indeed, the heavy rain of
rock and ash creates a kind of fluidisation all on its own as air and volcanic
gases are shouldered aside and forced upwards by the falling fragments (see
Diagram).

This updraught prevents the debris from settling and keeps it speeding
onwards. And the concentration of moving debris near the ground acts as a
smothering blanket, trapping some of the fine-grained dust and creating a
disorganised soup of all size fragments. As disorganised ash and pumice is
deposited out of the thick, soupy base of the flow, it is continuously
replenished from the increasingly turbulent levels roaring on overhead. From
time to time, turbulent vortices whirl downwards to the lower boundary of the
current, sweeping the pumice and ash into distinct layers which become
juxtaposed within the otherwise poorly organised ignimbrite.

If this view is right, volcanologists can no longer relate the thicknesses of
the ignimbrite to the thickness of the current that left it, as if the current
had suddenly frozen in its final resting place. A thin current could build an
enormous thickness of rubble, given enough time, just as a river channel can
slowly silt up: so a vertical strip though an ignimbrite simply represents a
record of the conditions at the base of the current as it passed over the
ground. The duration can usually only be guessed at. Some currents might have
flashed past in seconds, while others, fed by huge eruptions, may have lasted
for days.

We are left with a patchwork of incomplete models, and with the notion that
pyroclastic currents are extremely complex animals. And this view of a
pyroclastic current, which does away with the neat compartmentalisation of its
different parts, makes the modellers鈥� task even more challenging.

Could we ever check what really does go on? Might we one day build a
sear-proof, shockproof capsule armed with sensors to measure pressure,
temperature, shear force and impact? To sit at a crater lip, waiting for the
next explosion, then, a white-knuckle ride later, to be dug out and let the
world know how a pyroclastic current really works?

Given the forces that a seriously hellbent volcano has at its disposal, such
an adventure may never be possible. Volcano-watchers will simply have to go on
developing their insights as best they can鈥攁nd, by comparing their models
with real eruptions, try to prevent the populations of huge cities such as
Naples and Manila suffering the same fate as the inhabitants of
Saint-Pierre.

This century鈥檚 pyroclastic currents include events impressive in scale and
devastating in human and social terms. The eruption of Mount Pinatubo in the
Philippines in 1991 was one of the biggest this century, spewing out nine cubic
kilometres of ash and pumice in just five hours. It buried an entire landscape
beneath up to 200 metres of ash and incinerated hundreds of square kilometres of
forest.

But this is dwarfed by ancient examples. About 35 000 years ago, the
Campanian pyroclastic current covered 32 000 square kilometres in and around the
Bay of Naples to a depth of up to 100 metres, with a total volume of 500 cubic
kilometres. At Yellowstone in the US, 1000 cubic kilometres of ash and pumice
spewed out 600 000 years ago, and even this was outdone when 2500 cubic
kilometres were vented two million years ago. Researchers from the US Geological
Survey have found that volcanic calderas in the San Juan Mountains in the
Southern Rockies contain ignimbrites up to 3 kilometres thick.

If such eruptions happened today, the devastation would be unthinkable. The
Yellowstone pyroclastic currents engulfed large parts of three states, and the
airborne dust covered most of the country in a grey ash blanket. Industry,
transport and agriculture would collapse overnight. Eruptions on this scale
occur once every 100 000 years on average. Let鈥檚 hope the next one is a long
time coming.

On the morning of 18 May 1980, Charles McNerney and John Smart went to Mount
St Helens to watch the promised eruption. Along with other spectators, they got
past three roadblocks and stopped 6 kilometres from the crater to get a good
view. At 8.32 am, the blast came. It was much more than they had bargained
for.

Horrorstruck, they watched a black cloud leap a ridge and come their way. It
was the leading edge of a pyroclastic current. A warm wind began to blow towards
them. They leapt into their car and raced away for nearly 20 kilometres at up to
140 kilometres an hour, the heat from the cloud on their backs. They only just
outran it.

Dale and Leslie Davies and Albert Brooks were on another track, over 15
kilometres from the crater. Their truck stalled and they watched helplessly as
the approaching cloud tossed trees into the air. Then the cloud reached their
truck and engulfed it. The cab became pitch black and very hot. The van was
pounded by chunks of rock and wood. One window was smashed, letting ash seep
through into the cab. They received scattered burns on their arms and legs,
which took months to heal. 鈥淟ike microwave burns,鈥� a doctor said later.

The heat lasted only for a few minutes and then it became light, briefly,
before thick, cooler ash began falling out of the sky. Having abandoned their
truck, they were rescued later in the day, when the sky had cleared enough for
helicopters to operate. People out in the open, including campers and loggers,
suffered more, with fatalities caused by falling trees and inhalation of the
brief burst of intensely hot gas and fine ash.

The Mount St Helens accounts resemble those of other survivors of pyroclastic
currents, most famously that of the murderer Augustus Ciparis, who was locked in
a windowless cell at Saint-Pierre in Martinique. He was badly burnt in 1902 when
Mount Pel茅e erupted, but lived to benefit from his experience. Not only
did he receive a free pardon, but he joined a travelling circus, earning his
living thereafter by recounting his singular tale of survival.