Geology for Global Development

Heather Britton: Can Animals be Used to Predict Earthquakes?

One of the most common questions faced by the disaster risk reduction community relates to earthquake prediction (see this Geological Society briefing on prediction vs. forecasting). The disaster risk reduction community, however, would perhaps argue that improved buildings, reduction in poverty, and improved governance are a greater priority than predicting earthquakes. Even so, there are still many members of the international community focused on trying to identify ways to predict earthquakes, including through the study of animal behaviours.

Our understanding of where earthquakes are most likely to occur is improving, but our ability to predict when an earthquake will strike is lacking, often limited to the decadal scale at best. We also lack information on what the magnitude or size of an earthquake would be at that given point in time. If such a feat were possible, and an orderly evacuation could take place, lives could be saved. Many seismologists are of the opinion that the vast majority of earthquakes do not display early warning signals prior to the first p-waves reaching the surface, therefore earthquakes are likely to always remain stubbornly unpredictable. This does not mean that we will be unable to improve earthquake forecast, through probabilistic hazard assessment. It also does not mean that the disasters arising from earthquake are inevitable. We can still take significant steps to reduce exposure and vulnerability and reduce the impacts of earthquakes.

Other scientists disagree,  on the point of earthquake prediction, pointing to the anecdotal evidence which stretches back through historical archives around the world of animals predicting earthquakes far before modern technology would have us believe any indication of an earthquake existed. Is there any substance to these tales, and if so can it be used to support earthquake prediction?

Although devoid of substantial scientific evidence, the claim that early warning signs don’t exist fails to acknowledge the stories of animals abandoning their homes up to a month before an earthquake strikes. For centuries there have been reports of unusual animal activity prior to earthquakes: In 373 BC Greece it is documented that rats, weasels, snakes and centipedes abandoned their homes a month before a destructive earthquake struck, and in Italy toads disappeared from a pond where scientists were analysing their breeding patterns just days before a magnitude 5.9 earthquake killed over 300 people in 2009. Perhaps these animal behaviours can be used to predict the occurrence of earthquakes, but without knowing the nature of the signals which trigger their response it has limited applications in disaster risk reduction.

Figure 1- Frogs on logs. It has been suggested that aquatic organisms such as these may be able to predict earthquakes from changes in groundwater chemistry. (Source: https://www.pexels.com)

The problem with focusing so much on anecdotal evidence is that the stories are often augmented by the human imagination, an effect often seen in the game ‘Chinese Whispers’.  The result is that the unusual behaviour apparently displayed by the animals before earthquakes occur can become exaggerated and, in many cases, the reports only appear after the earthquake has struck. It is very well announcing a pet’s unusual behaviour after the disaster, but had the earthquake not occurred would the behaviour still have stood out as being so strikingly abnormal?

Animal behaviour is extremely complex and using this as a metric for earthquake prediction is not considered to be feasible because of the inconsistency of animal responses. This has not prevented at least one Chinese city from installing 24-hour surveillance on a snake farm with the intention of detecting unusual behaviour for the purposes of earthquake prediction. In 1975 officials successfully evacuated a city of one million people just before a 7.3 magnitude earthquake in Haicheng, China, purportedly based on abnormal animal behaviour. However, this has been rejected as substantial evidence for the power of animal foresight as this earthquake was one which was preceded by a number of low magnitude foreshocks which are thought to have given the governing body of Haicheng the confidence to evacuate the city, under the impression that a larger earthquake was on its way.

Figure 2 – Aftermath of an earthquake in 1971, San Fernando, California. Source:  USGS
Denver Library Photographic Collection.

As is almost always the case, the evidence from a number of different studies is contradictory and inconclusive, implying that the predictive signals, if present, may vary between earthquakes. Evidence for the ability of animals to predict earthquakes was found in a study in Peru – no animal movement was recorded by camera traps on the rainforest floor (an extremely unusual observation) five out of the seven days leading up to the magnitude seven Contamana earthquake that affected the area in 2011. Other studies, however, such as those performed in the 1970s by USGS, have found no correlation between earthquakes and the agitation of animals.

The evidence is patchy, but if there truly is a relationship between animal behaviour and earthquakes the identity of the signal that the animals are responding to remains a mystery. A paper released in 2011 describes a mechanism by which stressed rocks could release charged particles. These particles could then react with groundwater, producing chemical signatures which may be detected by aquatic and burrowing life. Other suggestions of potential signals include ground tilting, although this would have to be present only at miniscule levels not to be detected by current technology, or variations in the Earth’s magnetic field.

Currently research into the use of animals in earthquake detection is being led by Japan and China, two countries regularly affected by earthquakes and where a plethora of anecdotes relating to the powers of earthquake prediction by animals have originated. While earthquake prediction could help to reduce the impact of earthquakes on society, there are far more effective and immediate things that we can do. Ensuring properly constructed buildings and enforcing building codes, tackling the underlying social vulnerability (e.g., poverty, inequality) and improving governance structures and earthquake education are some examples.

Read more about disaster risk reduction in the UN Sendai Framework for Disaster Risk Reduction.

This guest post was contributed by a scientist, student or a professional in the Earth, planetary or space sciences. The EGU blogs welcome guest contributions, so if you've got a great idea for a post or fancy trying your hand at science communication, please contact the blog editor or the EGU Communications Officer to pitch your idea.