午夜福利1000集合

TRAVEL through the arid south-west of the US, or across the dry fringes of
the Gobi desert of western China, or over the mountain deserts of the Argentine
Andes, and you could see a strange sight. The apparently bare rocks of these
parched places are often coated with a mysterious, brightly coloured film. It is
as if someone, or something, has painted the desert.

Chip away at this coating, and you will see that is exactly what has
happened. The rocks are covered with a wafer-thin 鈥渧arnish鈥�. It is created by
colonies of bacteria that live on the surface of the rocks and extract energy by
oxidising minute quantities of metals such as manganese, blown in on airborne
dust and in the debris falling in occasional rain showers.

This secret landscape, often as pervasive as grass on pasture, is beginning
to tell a remarkable story. It is helping to unlock the mystery of past climates
in places where, till now, scientists believed there was nothing to hold such a
record. It鈥檚 early days, but this varnish could ultimately tie down one of the
central mysteries of climatology and warn us if abrupt climate change lies
ahead.

Rock varnish used to be an arcane study for a dedicated few. But Tanzhuo Liu
and Wally Broecker from the Lamont-Doherty Earth Observatory in New York have
transformed it into a potentially vital tool. Liu is the world鈥檚
greatest鈥攁nd close to only鈥攅xponent of analysing desert varnishes.
Broecker is an oceanographer and pioneer of uncovering past climate. He is
interested in Liu鈥檚 varnishes because he believes it could help unravel the role
of the deep circulation system of the world鈥檚 oceans in triggering sudden
climate flips.

Broecker thinks that the Earth鈥檚 ocean circulation system has two stable
states. The one operating now is driven by water descending from the surface to
the ocean floor around Greenland. In the other, water descends predominantly
around Antarctica. For reasons that are not yet clear, the ocean suddenly flips
between the two.

These flips seem to trigger changes in global climate. In the early 1980s
Hartmut Heinrich of the University of G枚ttingen discovered that every 8000
years or so during the last ice age, sudden extra-cold pulses sent armadas of
icebergs floating from Canada out into the Atlantic as far south as Bermuda.
Later it emerged that these 鈥淗einrich events鈥� coincided with cold phases in
another climatic pulse also found in the North Atlantic鈥攙iolent swings in
temperatures recorded in Greenland ice cores, and known as the
Dansgaard-Oeschger cycles.

There seems to be a distinct pulse to the Earth鈥檚 climatic metabolism, says
Broecker (New Scientist, 2 September 2000, p 30). Things tend to happen
in a series of discrete flips that seem to be driven by the ocean conveyor. It鈥檚
an idea that could be vitally important. Could the world be on the verge of
another flip through natural forces, or maybe due to the stresses we are placing
on the climate?

To answer these questions, Broecker needs to know if the climate pulses were
truly global. True, the Heinrich events seem to coincide with a series of other
climatic flips, blips and pulses that show up as imprints in spots such as lake
sediments, ice cores and ocean coral. Now Broecker dearly wants to know if
deserts are also part of the process.

Despite their seeming persistence, many deserts are on a knife-edge. Small
changes in their climate can cause major shifts鈥攕uch as the emergence of
dense bush and even forests in place of sand dunes. The Sahara, for instance,
was covered in bush just 6000 years ago, but within a few decades switched to
desert.

鈥淭he Sahara has two potentially stable states: deserts and heavily vegetated.
It could flip between the two, and only a little increase in rainfall could do
it,鈥� says Martin Claussen of the Potsdam Institute for Climate Impact Research
in Germany. Once vegetation is established, it recycles moisture in the air,
creating clouds and rain and perpetuating and exaggerating any rise in rainfall.
Conversely, when vegetation disappears, the skies clear and the land dries out.
Global warming is expected to produce more rainfall in coastal areas but less in
continental interiors. Some say the result could be that tropical deserts like
the Sahara will become even drier; others think the coastal rain could be enough
to trigger the return of vegetation to many deserts.

This sensitivity makes deserts particularly important pieces of the climate
jigsaw. The trouble is that they are also largely black holes in the climate
record, offering no ice cores to drill, and few lake sediments or tree rings to
inspect. Almost all you have to work with is rock and sand. Though deserts cover
a substantial fraction of the globe, we know next to nothing about their past,
and how they are likely to behave in the future. Only the few that contain lake
basins have revealed any of their climate history. Here at least, researchers
can look for the remains of old lake shorelines, estimate the size of the lakes,
and from that calculate past rainfall.

But most deserts have none of these lakes. So could rock varnishes begin to
tie desert climate to the global events being unmasked by Broecker? Liu thinks
so. The chemical composition of the varnishes is highly variable. While iron
levels remain more or less constant, the concentrations of other ingredients
such as manganese, calcium and barium vary. And this chemistry determines the
colour of the varnish. Rocks richest in manganese are black and smooth, while
the colour changes to brown as manganese levels fall. Scratch the surface of the
varnish and you see other colours, particularly orange, beneath. Peer through a
microscope and you鈥檒l see a rich array of reds and yellows and even blues. Liu
believes the sheer variety of the varnish composition could be the key to
interpreting past climates.

Liu and Broecker have been collaborating on interpreting the varnishes for 7
years. Analysing them is a time-consuming business. 鈥淲e look at many, many rocks
before selecting the most stable ones,鈥� says Liu. In the lab he carefully cuts
thin sections around a micrometre across. But no varnish contains a complete
record, so Liu painstakingly assembles different samples and correlates their
patterns to provide an overall record of a desert鈥檚 varnish.

Then these patterns have to be dated using any available independent
evidence. Conventional methods such as carbon dating simply do not work because
there is no carbon in the varnishes themselves. Ultimately, Broecker is pinning
his hopes on isotopes formed by the impact of cosmic rays. Meanwhile, he has to
try and connect varnishes with other physical features in the landscape whose
age researchers already know. These features, such as former lake shorelines,
are rare, but enough of them exist to begin to piece together a chronology.

Some varnishes date right back through the last glaciation, at least into the
previous interglacial era more than 100,000 years ago. They are still only about
200 micrometres thick, but that may be about the limit. Rocks can take only so
many layers of varnish before the whole lot peels away.

What do these records show? Broecker and Liu are convinced that the chemical
make-up of desert varnishes, and in particular the amount of manganese, is
determined by rainfall. The evidence is circumstantial but compelling. It hinges
on the data from a handful of lakes in the American West whose past shorelines
reveal sharp changes in rainfall between the end of the last glaciation, some
10,000 years ago, and the Holocene era that followed. Typical is Lake Lahontan
in Nevada, which shrank by 90 per cent during this period.

These changes coincide perfectly with the colour patterns in varnishes found
in the region. Liu鈥檚 analysis of Death Valley varnishes show manganese levels
that flip from 40 per cent during most of the last glaciation to as low as 4 per
cent in layers created during the dry and warm Holocene era. Similar patterns
occur right across the deserts of the American West. Changes in the varnish turn
up during other known switches in rainfall patterns and, says Liu, there is a
鈥済enerally positive correlation鈥� between manganese content in the top layers of
varnishes and today鈥檚 rainfall.

But this is all going a bit too far for some, including Steve Reneau of Los
Alamos National Laboratory in New Mexico. When Reneau analysed surface layers of
a series of varnishes from the Mojave desert he found their manganese levels
varied widely. Since they were on the surface, he assumed these layers all came
from the dry Holocene era. Yet by Liu and Broecker鈥檚 reckoning those with a high
manganese content must have originated in a very wet climate. Liu believes this
anomaly is easily explained: the high-manganese samples did indeed come from a
wet period at the end of the last ice age. Reneau found them on the surface
because more recent layers from the Holocene had been eroded away, Liu suggests.
Whether or not this is true, Reneau says it is vital that Liu and Broecker鈥檚
work be confirmed by independent analysts鈥攁 view the pair themselves
heartily support.

Several theories have been put forward to explain how the change in varnish
composition comes about. The bacteria that are best at concentrating manganese
from the air seem to prefer wetter conditions. And under dry conditions,
manganese may remain locked up inside dust and unavailable to the bacteria.
Whatever the mechanism by which manganese accumulates, Broecker is convinced
that its concentration reveals the deserts鈥� climatic past. 鈥淰arnish samples have
the potential to yield wetness records back to the last interglacial period for
each of the world鈥檚 desert regions,鈥� he says.

Many of the fragments of varnish assembled by Liu appear to show not just a
colour change at the end of the last glaciation but also a series of alternating
black and orange bands during the ice age. Where dating is possible, the black
manganese-rich bands in varnishes from the US coincide with the dates of cold
Heinrich events. Liu鈥檚 forays into Asian and Latin American deserts suggest that
the same may hold true worldwide. In Argentina a cold, wet period some 5000
years ago shows up. In western China they fit two wet periods, 3000 and 8000
years ago. And in the Middle East they reflect the ups and downs of the Dead
Sea.

Could varnishes show not only crude swings but also how much rain fell in
past millennia? Broecker thinks it鈥檚 possible. Correlation with records of lake
sediment in the American West suggests that if there is no rainfall, varnishes
typically contain 4 to 9 per cent manganese. Every extra 100 millimetres of
rainfall triggers an increase of around 4 per cent. The residual manganese comes
from some source other than rainfall鈥攑erhaps dust or aerosols, or even
dew.

To try and establish how much of the manganese comes from changes in
rainfall, Broecker and Liu have taken over a corner of the Biosphere 2
complex鈥攖he artificial ecosystems built in giant greenhouses in the
Arizona desert and now run by Columbia University. They created a 鈥渧arnish
garden鈥�, bringing in varnished rocks from Panamint valley in California and
McDowell Mountain in Arizona. Some were exposed to the elements, while the rest
were equipped with automatic umbrellas that opened to protect them whenever it
started to rain.

After eight months, the varnished rocks were analysed for an isotope of
beryllium produced in the atmosphere by cosmic ray bombardments. The varnishes
protected from the rain contained around 40 per cent as much beryllium as those
exposed to the rain鈥攕uggesting that this could also be the proportion of
other elements delivered by sources other than rain. Broecker and Liu hope that
separating the components of the varnish out in this way will help them work out
how changes in the amount of manganese relate to changes in wetness.

Eventually, these results might make it possible to find out exactly how much
rain fell in the deserts during the last ice age, during the climate flips, and
perhaps during previous warm periods. Much of the world鈥檚 land surface has now
been mapped for its past climate, thanks to tree rings, coral and much else.
Desert varnishes now hold out the hope of filling the remaining gaps. They could
help us predict whether global warming could cause the deserts to expand and
retract. And they could be the vital piece in Broecker鈥檚 jigsaw of how our
planet鈥檚 climate functions. Just shows how a touch of paint can complete the
picture.