A reliable method of predicting earthquakes would be a tremendous lifesaver, and would certainly garner the developer kudos from around the world. At the moment we can predict temblors about as accurately as predicting a Gabor sister marriage used to be: we knew it would happen, just not when.
Although the damage to housing and infrastructure would still occur, the thousands who died in the recent Pakistani temblor would have been spared. The new method highlighted in this New Scientist
article is far from producing a working quake predictor (anything worth while would need to warn at least two days in advance).
WITH refugees still huddling in tents across Kashmir after tens of thousands died in October's earthquake there, the need for earthquake prediction systems is once again thrown into stark relief. Knowing that the geologically restless Himalayas will produce more, stronger quakes is no use: what people need to know is when and where a quake will strike next.
So far, however, earthquake prediction has proved an elusive art: no one has worked out how to read Earth's vital signs to provide accurate warnings. But there is hope. Among the welter of dead ends - from monitoring animal behaviour to measuring radioactive gas emissions or the flow of groundwater - a new bellwether is coming to the fore: electromagnetic radiation.
Prior to some recent quakes, scientists have detected electromagnetic pulses emanating from the ground and electromagnetic disturbances in the ionosphere, the planet's tenuous envelope of charged particles extending from about 80 to 1000 kilometres up. "There are definitely hints of something [electromagnetic] happening in the region of earthquakes before the earth moves," says Colin Price, a geophysicist at Tel Aviv University in Israel.
Price and others have been working in quake-prone regions in California, Japan and Russia. At a meeting of the International Union of Radio Science (URSI) in New Delhi, India, in October, he and his colleagues speculated that as underground stresses build up, rocks containing magnetic particles begin to fracture, generating ultra-low-frequency (ULF) radio waves - below 1 hertz - as they are torn apart. Detect these radio waves, suggest the researchers, and you might have the makings of a prediction system.
Some research groups are already tunnelling underground to pick up radio pulses in the ULF range, while others are using sensor-stuffed satellites to measure radio disturbances in the ionosphere above quake-prone regions. Because there have been many false dawns in earthquake prediction, Price is cautious. "But if the chances are one in a hundred that we succeed, the huge benefits of success make this research worth continuing," he says. [...]
Since then, others have tried to make similar measurements in seismically active regions. Groups in Japan and Russia have observed similar signals to Fraser-Smith's, but for up to one or two months before a quake. Could this be the long-sought early warning of seismic catastrophe?
Minoru Tsutsui at Kyoto Sangyo University in Japan is trying to find out. His team has bored a hole 100 metres deep and 10 centimetres wide in a back lot on the campus. They placed one directional ULF antenna at the bottom of the hole and another above the ground. The relative strengths of any ULF signals detected at the two antennas allow the team to work out which direction the pulses come from.
On 4 January 2004 the system began detecting ULF radio pulses coming from the south-east. Two days later, a magnitude-5.5 quake struck the area, with an epicentre 130 kilometres away - to the south-east. Six hours after the quake, the ULF signals spread out, arriving from both the south-east and south-west, and died off the next day.
Since then, the Kyoto team has discovered that this effect is only detectable above a certain threshold quake strength (Geophysical Research Letters, DOI: 10.1029/2005GL023691). Tsutsui now wants to investigate the mechanism that produces these ULF radio pulses. Until we know this, he says, "we cannot easily predict where the epicentre will be". [...]
They may not be measurable in any case, says Masashi Hayakawa, an electronics engineer at the University of Electro-Communications in Tokyo, who has strong reservations about Tsutsui's conclusions. He points out that all kinds of natural phenomena produce ULF signals - thunderstorms, solar activity and meteors among them - and thinks Tsutsui won't be able to pick out ULF signals caused by imminent earthquakes from the noise.
Hayakawa thinks the atmosphere holds the answers: he measured ionospheric disturbances between 3 and 30 kilohertz a few days before Japan's 1995 Kobe quake. And Michel Parrot of France's National Council for Scientific Research (CNRS) in Orleans agrees. He points to preliminary data from a European Space Agency satellite called the Detector for Electromagnetic Emissions Transmitted from Earthquake Regions (DEMETER). Using a battery of sensors that measure the temperature, density and composition of the ionosphere, DEMETER measured an increase in ion density and temperature of the ionosphere seven days before a quake of magnitude 7 hit Japan's Kii peninsula on 5 September 2004.
This year it observed similar disturbances two days before the 23 January quake in Indonesia and five days before a quake on 30 August near Japan, and last November two days before a quake close to New Zealand.
As luck would have it, DEMETER was turned off during the 26 December 2004 Asian tsunami quake off Sumatra and the quake in Kashmir on 8 October this year, so it captured no data on these two events, Parrot says.
Ian Main, a seismologist at the University of Edinburgh, UK, believes the ULF and ionospheric findings are intriguing, but not yet convincing enough to establish a link to earthquakes. To do that, a far larger number of quakes must be examined, he says.
This is indeed intriging. Although even if this turns out to work, I suspect it will be decades before it begins functioning reliably in high risk areas.