Yong-Gang Li and his colleagues at the University of Southern California,
Los Angeles, and Peter Malin at the University of California, Santa Barbara,
use seismic waves that are ‘trapped’ along a fault zone. The waves are confined
rather like light waves in an optical fibre.

This confinement occurs because of the nature of the fault zone. Along
a fault, the surrounding rocks are smashed into a jumbled mass of broken
fragments. In general, the bigger the fault, and the more it has moved,
the wider this zone. Seismic waves move more slowly through the smashed
rock than through the solid rock on either side. This low-velocity channel
can act as a guide for certain seismic waves, known as ‘trapped modes’.

Li and his colleagues have detected these ‘trapped modes’ for the first
time (Science, vol 249, p 763). They were able to achieve the feat chiefly
because they used seismic instruments which were close to the broken rock
of a fault zone, in boreholes between 300 and 400 metres below ground.

At the surface, seismic waves are contaminated by man-made noise, and
also because crumbled rocks, soil and topography generate all kinds of distortions
and echoes. But in a deep borehole, the slow ‘trapped mode’ waves are uncorrupted.
Geologists can observe them arriving shortly after the waves that travelled
through the intact rock on either side of the fault.

Li and his colleagues set up one experiment at Oroville, California.
They generated artificial seismic waves at the surface outcrop of an active
fault, and recorded the waves that travelled to a borehole, 305 metres deep.
When they compared the shapes of these ‘trapped mode’ waves with synthetic
seismograms made by fault zones of various widths, they were able to deduce
that the fault zone was 18 metres wide.

More importantly, the researchers have used recordings obtained from
Parkfield, where the San Andreas Fault fractures regularly. Parkfield has
been the site of moderate sized earthquakes about every 22 years since the
mid-19th century. The last shock was in 1966, so the next one is already
overdue. A programme of monitoring is under way to catch it.

Some of the seismic instruments at Parkfield are located in boreholes
that are close enough to the fault for the new method to be useful. Li and
his colleages have found that these instruments record ‘trapped modes’ of
tiny earthquakes on the fault at depths of between 4 and 10 kilometres.
By comparing these with synthetic seismic waves generated in a computer
model, they were able to map the properties of the fault zone that lies
between the earthquake and the recorder.

At the point where geologists expect the next Parkfield earthquake to
begin, the fault is locked tight – there is no relative movement of the
rocks on either side. Trapped modes show that this part of the fault zone
has a particularly low relative velocity and is about 100 metres thick.
In the next section of the San Andreas Fault to the north, where the fault
moves continuously without major earthquakes, the fault zone is wider and
has a higher velocity, indicating that the rocks are less smashed. The fault
zone also merges more gently into the surrounding rockmass.

In the build-up to the next Parkfield earthquake, it is likely that
the narrow part of the fault zone will change its internal properties before
finally breaking, say the researchers. Continuous monitoring of the ‘trapped
mode’ waves may finally provide a way to focus on the hidden source of a
major earthquake, before it happens.

AS A student it puzzled me how a geological time-scale, concocted in
western Europe, had been exported with so little resistance around the globe.
How did an ammonite in Adelaide tell a brachiopod in Barnstaple that the
Jurassic was finally over, roll on the Cretaceous? Great tracts of Earth
history were named after obscure Welsh tribes (the Silures and the Ordovices),
and 60 million years dedicated to a minor-league English county, better
known for its clotted cream. How did the Japanese or the Nigerians come
to accept this cultural imperialism of Earth history: why hadn’t they assembled
their own rival local chronology? Yet the fact that a layer of rock in Wales
could become a satisfactory world time-horizon is compelling evidence that
the history of the Earth has been punctuated by global events.

That we do now understand the significance of such synchronicity is
largely due to research during the last decade to crack the biggest unsolved
prehistoric murder mystery: the death of the dinosaurs. The time of death
looks less than coincidental, corresponding with some precision to the Cretaceous-Tertiary
(better known to streetwise stratigraphers as the K/T) boundary. Although
just one part of a global genocide, the fate of the dinosaurs has become
a powerful and adaptable allegory. Did they fall or were they pushed? Were
they the fallout of an interplanetary traffic accident, or victims of passive
smoking? Hypothermia or heatstroke: you name it, someone has written a paper
proposing it.

Something out of the ordinary happened at the K/T boundary: the biosphere
virtually shut down; there were extensive fires, dust storms, pea-soup smogs
and, while the sunsets must have been terrific, the smell (whose organic
signature has not yet been retrieved from the sediments) of the largest
pile of carcases in history must have been appalling. The question is, what
precisely happened, and why? Sixty-five million years after the last dinosaur
curled up its toes, the explanation for the K/T boundary has become the
biggest argument in the earth sciences since continental drift. Like the
First World War, what started off, 10 years ago, as a small incident (the
discovery of 9 parts per billion of iridium in a layer of claystone 10 millimetres
thick) in a far off country (Italy), has now expanded so far, there is hardly
an Earth scientist whose expertise is not in some way involved in disputing
one or other element of a rival theory. What, you might think, could vortices
in the Earth’s core have to do with the passing away of Tyrannosaurus; or
a highpressure analogue of quartz; or the flash point of a temperate forest;
or how rapidly magma poured out of a fissure in India? Claude Albritton
has written a book that attempts to provide the present debate with its
historical context. Beginning with the first theories of the Earth in the
17th century, he passes into the 19th-century battles between catastrophists
(who thought the past was short and cataclysmic) and uniformitarians (who
relied on vast quantities of time to explain it all). After 1900, catastrophists
had become an endangered species, derided by geologists, or marginalised
as disciples of Velikovsky. Yet now, in the guise of ‘neo-catastrophism’,
such views are once again acceptable.

Welding the 19th-century history onto a controversy about dinosaurs
that is still going on leaves an abrupt seam in the middle of the book.
The problem is the shift in perspective: the dispassionate historian, viewing
the action from the elevation of a hundred years, suddenly becomes the bewildered
(and partisan) war correspondent dodging the bullets. Until the furore has
died down, and it becomes obvious who is collecting the honorary degrees,
the result remains hard to predict. Supporters of an asteroid impact still
need to find the crater, while 13 per cent of British palaeontologists do
not believe that there was a mass extinction.

Whatever the outcome, the fight has inspired some very high-quality
research. Earth and atmospheric scientists have joined forces: research
on the K/T boundary has implications far beyond the dinosaur. It affects
theories from nuclear winters to runaway greenhouses. Until the dust, volcanic
or meteoritic, has finally settled, Albritton’s work will remain a useful
and comprehensive guide to the factions.

Robert Muir Wood edits Terra Nova, and is based in Oxford.