THENEWYORKER,MARCH23, 2020 23
sink beneath bigger ones.) RAMMS seeks
to predict the outcome of this churn.
The software was validated on his-
torical avalanches—especially on data
about whether trees had been knocked
down, and, if so, how old they were. “Trees
are wonderful mechanical sensors,” Bar-
telt said. If an avalanche takes down a
seventy-year-old stand of trees, you know
that the avalanche has a return period of
at least seventy years. Fine-tuning the
model would require more precise data,
which are hard to come by. Gathering
this information would require taking
readings inside, or under, an avalanche.
F
or this purpose, the S.L.F. maintains
an avalanche test site in the Vallée
de la Sionne—a steep, mountainous area
about two hundred miles from Davos.
Hearing the phrase “test site,” one might
imagine a bunny slope. Actually, it is an
enormous mountain, improbably re-
served for science.
The site’s chief scientist is Betty So-
villa, a hydraulics engineer. When we
met at S.L.F., she was wearing red-
framed glasses, a black cardigan, jeans,
and red boots. “RAMMS is a very sim-
plified model,” she said. The goal of the
test site was to develop a more realistic
version, by correlating detailed measure-
ments of the snow cover with the ava-
lanches it created. She was particularly
interested in glide avalanches: there were
more of them every year, but they were
elusive. “You cannot predict when they
are released,” she said. “This is really the
avalanche of the future.”
One morning, Pierre Huguenin, a
forty-nine-year-old mountaineer and
snow scientist, drove me to the site in
a white Mitsubishi Pajero. “You see the
flakes. You see the crystals,” he said, ges-
turing out the window. There had been
a storm the previous night. He stopped
the car where the road ended, and we
changed into snowshoes.
Outside, there was about a foot of
pristine powder. I stooped and ran my
hand through it. Bone-dry, it was the
pure bright white of confectioner’s sugar,
with the texture of sea salt. Huguenin
pulled out his phone. The avalanche fore-
cast for the area had us covered in or-
ange. “We are in the third degree,” he
said—the risk category in which the most
avalanche deaths occur in the Alps, equiv-
alent to the American “considerable.”
He pulled out two avalanche beacons—
transmitters that would relay our loca-
tion to rescuers—and set them to Send.
We strapped them under our jackets.
“My job before working at the S.L.F.
was at a salmon plant,” Huguenin said,
as we set out. (He was an engineer there.)
“It was so loud.” Now we could hear the
river as we walked. Beneath the blue sky,
ours were the only tracks. After twenty
minutes, the site came into view: a broad,
bare mountainside, eight thousand feet
high. Between two couloirs—the main
avalanche paths—a half-dozen chalets
huddled near a small wood.
“They are not allowed to live here in
the winter,” Huguenin said. Two days
earlier, there had been a naturally occur-
ring glide avalanche at the site. I asked
whether it had been dangerous. “You
would be dead,” he said. “No chance.”
The site was built in 1997; in the win-
ter of 1999, the snow was the heaviest
it had been since 1951—perfect condi-
tions for an experiment. Using explosives
dropped from a helicopter, the S.L.F. trig-
gered three avalanches in the course of a
month. They were so massive that they
destroyed most of the institute’s equip-
ment. If you had been skiing on the moun-
tain during the last avalanche, you might
have heard a soft exhalation: air releas-
ing from a crack in the slab. Upslope, it
would have looked as though someone
had slit the mountain’s forehead. Now its
face was falling off; the break, nine foot-
ball fields across, was as deep as eleven
feet in places. Blocks of snow would begin
leaping up prettily, breaking like roiling
water. In the quiet, you might feel some-
thing lapping at the back of your legs be-
fore being swept off your feet.
The slide generated a powder cloud
nearly two hundred feet high. It seemed
to move in slow motion, like dry ice bil-
lowing, but it levelled the trees. Under-
neath, the core was formed by four hun-
dred thousand tons of snow. Huguenin
asked me to visualize the test peak, two
kilometres distant, and the peak of the
mountain on which we stood as the two
sides of a half-pipe. With a deep roar,
he said, the avalanche had run through
the valley like a skateboarder, with
enough speed to climb the other side.
“It came all the way up there?” I asked,
pointing to the top of our peak, three
hours’ hike away.
“Yup, and there is a trail there. One
of the wards was on it. The guy at that
time saw a huge amount of snow jump-
ing the top here”—he motioned toward
the ridgeline above us—“and falling on
the other side.” As the snow poured over
“If they turned off when you clapped, they probably
weren’t the northern lights.”
• •
