Why is this elephant running?

Earthquake prediction may run off in a new direction. We’ve tried electronics attached to seismic detectors (and made some progress), but there may be a new ally in the battle to give a warning before the next big one knocks you off your feet. The way things work now, geophysicists can’t tell you when an earthquake will strike – we can’t predict which hour, day, month, nor even year. We don’t know. But some people think that perhaps the elephants know.

The animal stories about quakes border on pseudo-science.  Elephants were much in the news after the 2004 Boxing Day quake and tsunami in southeast Asia. Eyewitnesses claim that very few animals were killed in the carnage that took hundreds of thousands of human lives. Flamingos and bats flew to high ground, dogs and cats fled together from seaside villages, and elephants screamed, then trumpeted as they ran, heralding the coming doom. Reports asserted that elephants were particularly keen to escape the lowlands during the moments before the tsunami. Ravi Corea, president of the Sri Lanka Wildlife Conservation Society said he was surprised by the lack of animal carcasses (other than two water buffaloes) following the tsunami, though about 50 humans died on a beach in his area. About an hour before the tidal wave hit, people noticed three elephants running away from the beach, said Corea. Another person claimed that his family dogs refused to go for their daily run along the water’s edge shortly before the tsunami struck. Three elephants running, two dogs hiding, and a flamingo flying uphill does not prove that animals have extra senses attuned to impending disaster. However, for centuries, folk legends have been telling us that animals know when woe is nigh.

The United States Geological Survey (USGS) decided to investigate the wisdom and predictive powers of animals. Is there real proof that creatures detect earthquakes and tsunamis before they occur? Or is it a case of hindsight on the part of observers? Humans have a brilliant tendency to recreate history and fill in gaps as part of our hard-wired system of explaining the world to ourselves. We are also excellent at extrapolating a few examples and turning them into a complete world-view. We are particularly susceptible to confirmation bias, our tendency to find explanations for odd phenomena that agree with our ideas of how the universe should work. (For example, a person may recall that just before a tremor, the cat was hissing. Perhaps it was. But that feeble old cat hisses thirty times a day, the tremor happened just once. Retrospective explaining that confirms a prior notion is confirmation bias.) Is this what is happening after a catastrophe when people report unusual activity of pets and wildlife that they noticed? The USGS wanted to know.

Tentatively, the USGS scientists concluded that a geophysicist-elephant’s skill at predicting earthquakes and tsunamis is not convincing. In their paper, Animals and Earthquakes, the investigators suggest that “We can easily explain the cause of unusual animal behavior seconds before humans feel an earthquake. Very few humans notice the smaller P wave that travels the fastest from the earthquake source and arrives before the larger S wave. But many animals with more keen senses are able to feel the P wave seconds before the S wave arrives.” Elephants are known to hear low frequency sounds (the rumbles of distant tribe members, the roar of an incoming tsunami) even when the source is 30 kilometres away. This might have something to do with those huge ears and big flat feet. But, as the USGS paper says, there might be senses other than auditory involved. Could animals feel the ground tilt, could they sense electric or magnetic atmospheric changes before an earthquake? The authors inform us that we don’t even know if electromagnetic field changes are involved and suggest that this is an area ripe for scientific inquiry.

The USGS paper cites three sources, all relatively old, dating from 1985, 1988, and 2000. The sources, especially the 1988 paper by UCLA’s Rand Schaal, looked for statistical correlation with commonly reported animal phenomena. In particular, Schaal reviewed the number of missing dogs reported in the local papers. (Schaal writes, “For a dozen years a theory has been advanced in the south San Francisco Bay area that when an extraordinarily large number of dogs and cats are reported in the “Lost and Found” section of the San Jose Mercury News, the probability of an earthquake striking the area increases significantly.”) Schaal’s statistics show that there were 62 large quakes during the study period but only 9 correlated with runaway dogs while 16 earthquakes were preceded by stay-at-home pooch activity. The paper’s bottom line: There is no relationship between missing dogs and impending quakes. Furthermore, the number of missing dogs “is not proportional to the quantity or magnitudes of quakes,” wrote Schaal.

OK, so dogs don’t make howling great seismologists. But there is still the nasty business of a story that has lasted at least 2400 years, a folk legend that tells us to mind our dogs, cats, and toads because they know when to run. For example, when the Greek city of Helike was flattened by a quake and tsunami in 373 BC, dogs ran three kilometres (2 miles) to Keryneia, a hill town about 300 metres (1000 ft) higher than coastal Helike. The destruction was preceded by flashes of bright light. Religious folks at the time said that the catastrophe was attributed to the vengeance of Poseidon (god of oceans, horses, and earthquakes) because the inhabitants of Helike had refused to give their statue of Poseidon to the Ionian colonists in Asia. Maybe they are right – I wasn’t there when it happened and the locals of the time swear it’s true. But I digress. And I glossed over the flashes of bright light. Maybe they caused the dogs to run.

About those flashes of bright light. In second-year geophysics, I was taught something called the piezoelectric effect. If you squeeze particular minerals, ceramics, bones, money, proteins, or rocks hard enough, a bit of electricity is released. We are talking about enormous masses of rock and enormous earthquake pressures, so the amount of piezo- electricity could be significant. When visible, scientists call these discharges Earthquake Light. The flashes are reported to have shapes similar to those of polar auroras and are usually a white to bluish hue. The glow is reported to be visible for several seconds, sometimes even minutes. Earthquake Light has been reported in Hawaii, New Zealand, California, Japan, and ancient Greece – it seems unrelated to culture or geography, but is centered around earthquakes and has been seen before, during, and after quakes. But not every earthquake is accompanied by a glow show. However, it has been speculated that every earthquake may have some piezoelectric effect going on. If the right equipment is in place to monitor a fault zone, it may be possible to detect the buildup of rock stress and predict earthquakes.

One would think that the piezoelectric effect could become a potent tool for forecasting imminent earthquakes. There must be dozens of geophysicists investigating this. If there are, they are keeping pretty quiet about their work.

But one scientist willing to talk about this quake link is Friedemann Freund. Freund’s research into the way rocks under stress can release hundreds of thousand amperes may lead to a real breakthrough in earthquake forecasting. His work uses the subtle electromagnetic signals generated when stresses build up along fault zones. Freund, working with two post-doctorate researchers at NASA and later with a group of students, showed that air molecules become significantly ionized near rocks that are stressed. They even observed tiny sparks flying off the edges of the rocks – likely a scaled-down version of the bright light phenomenon observed since Grecian times in Helike.

This led Freund, working with Thomas Bleier at Stellar Solutions, to monitor the output of dozens of ultralow frequency sensors set along fault zones in California, Peru, Greece, Sumatra, and Taiwan. They say that they have found significant increases in ionization whenever there has been a moderate to large earthquake nearby.

“Changes in Animal Activity Prior to a Major (M=7) Earthquake in the Peruvian Andes”, appearing in last month’s Physics and Chemistry of the Earth, and written by Friedemann Freund, Rachel Grant, and Jean Pierre Raulin, correlates animal behaviour in Peru to a major 2011 quake. In their paper, the researchers

“…present records of changes in the abundance of mammals and birds obtained over a 30 day period by motion-triggered cameras at the Yanachaga National Park, Peru, prior to the 2011 magnitude 7.0 Contamana earthquake. In addition we report on ionospheric perturbations derived from night-time very low frequency (VLF) phase data along a propagation paths passing over the epicentral region. Animal activity declined significantly over a 3-week period prior to the earthquake compared to periods of low seismic activity.”

Freund believes that evidence indicates that ultralow frequency electromagnetic waves in the environment prior to an earthquake can have effects on animal behavior. Freund et.al. propose that the “multitude of reported pre-earthquake phenomena may arise from a single underlying physical process: the stress-activation of highly mobile electronic charge carriers in the Earth’s crust and their flow to the Earth’s surface… [These are] known to be aversive to animals.”

Freund is studying the potential link and is looking for ways of using pre-quake radiation to predict earthquakes. His key investigation is to determine if we can use the signals produced during the build-up of quakes and detect them at the Earth’s surface. Freund says his hope is that we will reach a point where we can forewarn of impending tremors in a way similar to severe weather alerts – we can not predict any particular lightning strike, but we can usually give warnings that extreme thunderstorm damage is likely in a region within a few hours. Such a warning may take the form of “Stresses at a particular fault seem to be building up deep in the earth’s crust and there is an increased chance of an earthquake within the next few days.” This would be a great improvement over our current predictive abilities which simply give a probability in the range of years. Parts of California’s San Andreas Fault, for example, have a 97% probability of being visited by a 6.7 or greater magnitude earthquake within the next 30 years. This, of course, is extremely helpful for construction plans and safety guidelines. But it is not the sort of information that will send you off on a two-week visit to Calgary or London.

What about the animals? It may be possible that dogs and cats and birds and lizards sense the buildup of electricity due to the overwhelming rock stress preceding an earthquake. But what would they do with such information? Is there an evolutionary advantage to fleeing earthquakes? Not likely. Devastating earthquakes may occur every century or two along an active fault zone. Even then, small surface animals would not likely be hurt. Most damage and death is structure-related. Humans fare poorly during earthquakes because our homes collapse. Burrowing animals may be similarly disadvantaged, but most surface-dwelling animals would simply be annoyed – not killed – during an earthquake. There would not be an inherited genetic predisposition that alerts animals to a brewing quake, nor would most animals have first-hand experience or handed-down legends to inform them of the best reaction to an earthquake.

If there is any animal-quake relationship, it may be due to the annoyance of the increasing electromagnetic field which some animals may find irritating enough to cause a quick jaunt towards anyplace away from the snap and crackle of atmospheric static. Relating charges to elevated earthquake threats makes an interesting hypothesis, but will take time to prove. However, such a study has life-saving potential and is worth the effort.