Geoscience Reference
In-Depth Information
in swarms of many parallel and intersecting cracks or faults. Most of them suffer frequent
minor earthquakes and any of them could be the centre of a big one. The most famous fault
of all, effectively the plate boundary itself, is the San Andreas fault. It can be traced from
the south of California, curving round inland of Los Angeles and running northwest straight
for San Francisco and the sea. It achieved its notoriety in 1906 when San Francisco was
devastated by a major earthquake and the terrible fire that followed it through the wooden
houses.
Between Los Angeles and San Francisco, the landscape is arid and the fault can be fol-
lowed easily through the bare hills. Sometimes it is marked just by a slight change in the
slope. Sometimes it can be seen cutting through the landscape as though some great hand
had run a knife across the map. It seems to run straight for 100 miles. I followed it along
a rough farm track midway between Los Angeles and San Francisco. To the east lie low
eroded hills of the Temblor range; to the west, the dry Carrizo plain slopes gently towards
San Luis Obispo and the Pacific. Coming down from the hills are a number of dried-
up stream beds. As they reached the base of the slope, something strange seems to have
happened to them. Instead of flowing straight on to the west, they all take a sharp right-
hand bend through 90 degrees, follow the base of the hills north for a few tens of metres,
and then turn sharp left again to continue to the sea. One of the biggest of them, Wallace
Creek, named after Robert Wallace of the US Geological Survey, is cut deep into the soft
hillside. As it crosses the fault, it is displaced by 130 metres. Originally it must have flowed
straight down the slope, cutting its course. In a series of earthquakes, the plain to the west
has lurched north, taking the stream bed with it. The winter floods could not cut a new
course through the high banks, so they followed the fault until they found their old bed
again. It did not move all at once. Using a combination of excavation and radio-carbon dat-
ing, Robert Wallace and his colleagues have worked out the stages. The only one recorded
in the history topics took place in 1857 and accounts for the last 9.5 metres of slip. Its two
immediate predecessors, both of them prehistoric, moved the stream courses a further 12.5
and 11 metres. Averaged out, the San Andreas fault has been slipping at a rate of 34 milli-
metres per year for the past 13,000 years. If it keeps up this rate, in 20 million years' time
Los Angeles will be as far north of San Francisco as it is south today. But, as Californians
know to their cost, the passage is not smooth.
Measuring the movement
It used to be almost impossible to measure motions of metres or centimetres across the
scale of continents, but now it is comparatively easy. Fault zones such as California and
Japan bristle with instruments. In particular, receivers linked to the GPS can keep a con-
tinuous tab on their position on the surface of the globe. If they are linked into an automatic
