Les Animaux Peuvent-ils Prédire les Tremblements de Terre? Ce Que la Science Dit

The Myth: Animals Can Sense Earthquakes Before They Happen

Stories of dogs barking, snakes fleeing burrows, and fish behaving erratically just before major earthquakes have circulated for centuries. In ancient Greece, historians recorded unusual animal behavior before the 373 BCE destruction of Helice. In China, a 1975 earthquake evacuation was partly triggered by reports of restless animals. These accounts feel compelling — surely there must be something to them. But when scientists apply rigorous scrutiny, the picture becomes far more complicated than the folk wisdom suggests.

What People Claim to Observe

The reports are remarkably consistent across cultures and centuries. Cats and dogs become agitated or run away from home. Cows refuse to enter barns. Snakes emerge from hibernation in winter. Birds take flight en masse and refuse to return to roosts. Fish leap out of water or school in unusual formations. Bees abandon hives. In some accounts, the unusual behavior precedes the earthquake by hours, days, or even weeks. The apparent pattern is tantalizing: if animals can sense something humans cannot, could they serve as a biological early warning system?

The Science of Animal Perception

Animals do possess sensory capabilities that exceed human limits. Dogs can hear frequencies up to 65,000 Hz, compared to the human ceiling of around 20,000 Hz. Some fish detect changes in electromagnetic fields. Snakes sense ground vibrations through their jawbones with extraordinary sensitivity. Elephants communicate through infrasound and may detect ground vibrations through their feet. These abilities have led researchers to hypothesize that animals might detect P-Wave (Primary Wave)The fastest seismic wave, traveling through both solid rock and liquid at 5-8 km/s. P-waves compress and expand material in the direction of travel, like a slinky. They arrive first at seismograph stations. arrivals — the fast-moving but less destructive seismic waves that precede the stronger shaking. P-waves travel ahead of the damaging S-waves and surface waves, giving a few seconds to tens of seconds of advance notice at most. If animals detect P-waves and react visibly, this would explain the most credible short-term observations (seconds to a minute before shaking) but cannot account for reports of behavioral changes hours or days in advance.

What the Research Actually Shows

Multiple systematic studies have attempted to document pre-earthquake animal behavior under controlled conditions. A landmark German study monitored farm animals — cows, sheep, and dogs — with GPS accelerometers before two Italian earthquakes. The animals did show elevated activity, but the most significant changes occurred in the days before the quake, not in the final hours, and the researchers could not rule out other environmental factors. A USGS review of the historical record found that most dramatic animal behavior accounts come from after earthquakes, when people selectively remember any unusual behavior they noticed and ignore all the times animals acted strangely without a quake following.

This is the critical problem: ForeshockAn earthquake that occurs before the mainshock in the same region. Foreshocks can only be identified in retrospect — there is no reliable way to distinguish them from ordinary earthquakes beforehand. sequences — small earthquakes preceding a major event — do cause ground motion that sensitive animals might detect, but most foreshock sequences are only recognized as foreshocks after the mainshock. Animals may also respond to changes in groundwater chemistry, radon gas emissions, or electromagnetic anomalies that some researchers hypothesize precede earthquakes, but none of these precursors has proven reliable enough for Earthquake Prediction vs ForecastingPrediction claims to specify exact time, place, and magnitude of a future earthquake — currently impossible. Forecasting provides probabilistic estimates of earthquake likelihood over time periods..

Confirmation Bias and the Memory Problem

Human memory is selective. After a major earthquake, people naturally search their recent memories for any anomalous event that might have been a warning — including unusual animal behavior. Studies have found that the same proportion of people report unusual animal behavior before earthquakes as before randomly selected non-earthquake days. The difference is that post-earthquake reports are memorable and retold, while the countless false alarms (animal weirdness not followed by earthquakes) are forgotten. This confirmation bias makes anecdotal evidence essentially useless as a scientific signal.

Why Official Agencies Don't Use Animal Monitors

National earthquake agencies including the USGS have evaluated animal-based monitoring programs and consistently concluded they are not reliable enough to serve as predictors. The fundamental challenge is specificity: even if animals could detect some precursor signal, you need to know what baseline normal behavior looks like for a specific animal in a specific location, you need to account for weather changes, reproductive cycles, predator presence, illness, and dozens of other behavioral triggers, and you need a signal that distinguishes "earthquake in 24 hours" from "thunderstorm incoming" or "breeding season stress." No study has successfully developed such a protocol.

What Animals Can Legitimately Detect

It is entirely plausible — and supported by some evidence — that animals can detect P-waves a few seconds before humans feel the stronger shaking. This is consistent with the physics of seismic wave propagation. Dogs that suddenly whine and run to their owners right before shaking begins may genuinely be responding to P-wave arrivals. This is interesting scientifically but does not constitute earthquake prediction in any useful sense — a few seconds is not enough time to take protective action beyond what Earthquake Early Warning (EEW)A system that detects an earthquake and sends alerts to people and systems before strong shaking arrives. Can provide seconds to tens of seconds of warning, enough to take protective action. electronic systems already provide through ShakeAlertThe US earthquake early warning system operated by USGS and university partners. Covers the West Coast (California, Oregon, Washington) and sends alerts through Wireless Emergency Alerts. and similar networks.

The Bottom Line

Animals are fascinating seismic sensors in a narrow sense, but they are not reliable earthquake predictors. The romanticized notion of nature's creatures warning humanity of impending disasters does not withstand scientific scrutiny. The best early warning systems are electronic networks of SeismographAn instrument that detects and records ground motion caused by seismic waves. Modern digital seismographs can detect movements smaller than a nanometer. instruments that detect P-waves and instantly calculate the likely shaking intensity at distant locations, delivering alerts seconds before S-waves arrive. For genuine preparedness, focus on Earthquake PreparednessThe ongoing process of planning and preparation to minimize earthquake impact, including securing furniture, creating communication plans, maintaining emergency supplies, and practicing drills. plans and structural safety rather than watching your pets for earthquake warnings.