Earthquake lights are rare luminous phenomena that appear in the sky or near the ground before, during, or after seismic events. They have been reported for centuries, showing up as flashes, glowing orbs, sheets of light, or flame-like columns, often in colors ranging from white and blue to reddish-orange. Despite a long history of eyewitness accounts and, more recently, photographic and video evidence, the exact mechanism behind them remains unsettled. Several credible scientific hypotheses exist, most involving electrical processes triggered by extreme stress on rocks deep underground, but no single explanation has won universal agreement among geophysicists.
What People Actually See
Reports of earthquake lights vary widely, which is part of what makes them so hard to study. Some witnesses describe brief, lightning-like flashes near the horizon. Others report softly glowing spheres hovering near the ground or along mountain ridges. Still others see diffuse, aurora-like bands of color high in the sky. The lights can last anywhere from a fraction of a second to a couple of minutes, and their apparent size ranges from small, localized glows to expanses spanning hundreds of meters.
Some of the best-documented early observations come from Japanese earthquakes in the 1930s and 1960s. During the latter events, researchers captured both color and black-and-white photographs of bright, hemispherical, white luminescences based at ground level. These glows measured roughly 20 to 200 meters in diameter, lasted from about 10 seconds to 2 minutes, and appeared specifically on mountain summits composed of quartz-diorite faulted rock.1Bulletin of the Seismological Society of America. Earthquake lights: A review of observations and present theories That detail about rock type turned out to be an early clue pointing toward electrical mechanisms in the ground itself.
The lights do not always coincide with the moment of shaking. During the 2009 earthquake near L’Aquila in central Italy, witnesses reported luminous phenomena starting about nine months before the main shock and continuing up to five months afterward. These sightings were distributed over a large area around the city, extending as far as 50 kilometers to the north.2Copernicus Publications / Natural Hazards and Earth System Sciences (NHESS). The earthquake lights (EQL) of the 6 April 2009 Aquila earthquake, in Central Italy That timeline suggests the lights are not simply caused by the seismic rupture itself but may be linked to the slow buildup and release of stress in the crust over weeks or months.
Video Evidence From the Modern Era
For most of scientific history, earthquake lights were treated with skepticism because evidence was almost entirely anecdotal. That changed as security cameras, dashcams, and smartphones became widespread. One of the more compelling instrumental recordings came from the 2007 Pisco earthquake in Peru. Security cameras at a university campus in Lima, roughly 150 kilometers from the epicenter, captured luminous flashes that researchers were able to time-correlate with seismic ground accelerations recorded at a seismological station on the same campus.3Natural Hazards and Earth System Sciences. Co-seismic luminescence in Lima, 150 km from the epicenter of the Pisco, Peru earthquake of 15 August 2007 The lights and the shaking aligned in time, which helped rule out coincidental lightning or electrical infrastructure failures.
More recently, dramatic footage emerged from the 2017 Mexico City earthquake and the 2023 earthquakes in Turkey, with bright blue and green flashes visible over city skylines. These videos went viral on social media and renewed public interest, though they also raised a practical complication. In an urban setting, transformers blowing, power lines arcing, and gas ignitions all produce flashes that look remarkably similar to the geophysical phenomenon. Sorting genuine earthquake lights from electrical infrastructure damage is one of the persistent challenges in this field, and not every flash captured on video during a quake is actually an earthquake light.
The Leading Hypothesis: Stress-Activated Charge Carriers
The most developed scientific explanation centers on what happens inside rocks under enormous pressure. When tectonic stress builds in the Earth’s crust, it can activate a type of electrical charge carrier in igneous and metamorphic rocks. These charge carriers, called positive holes, are essentially spots where an electron is missing in the oxygen network of silicate minerals. Under normal conditions, they exist in a dormant, electrically inactive form. But when the rock is stressed, they wake up and start moving.
Laboratory experiments have shown that these positive holes propagate through rock as charge clouds rather than as classical electromagnetic waves. Before activation, they sit dormant as paired defects in the crystal structure of minerals.4Journal of Geophysical Research: Solid Earth. Time‐resolved study of charge generation and propagation in igneous rocks When stress breaks these pairs apart, the freed charges can travel considerable distances through the rock, flowing toward the Earth’s surface.
The hypothesis proposes that as an earthquake approaches, stress rates increase rapidly at certain points in the crust. High concentrations of these charge carriers accumulate, and when a critical density is reached, the charges form a kind of solid-state plasma that breaks out of the rock and causes electrical discharge at the surface. This discharge ionizes the air in the same way a spark plug ionizes fuel, producing visible light and high-frequency electromagnetic radiation.5Physics and Chemistry of the Earth, Parts A/B/C. Earthquake lights and the stress-activation of positive hole charge carriers in rocks Think of it as the ground itself building up a massive static charge that eventually discharges into the atmosphere, like a geological-scale spark.
The Piezoelectric Explanation
A competing and older hypothesis involves piezoelectricity, the ability of certain crystals to generate an electric charge when squeezed. Quartz is strongly piezoelectric, and quartz-rich rocks are abundant in the Earth’s crust. When tectonic forces compress or shear these rocks, the quartz grains produce a polarization of charge. If enough grains are suitably aligned, the rock itself develops a measurable piezoelectric response.6Tectonophysics. Piezoelectric effects in quartz-rich rocks
The idea is that rapid changes in seismic stress could generate large voltage differences across quartz-bearing formations, strong enough to ionize the air above and produce visible light. The early Japanese observations fit this model neatly: the lights appeared specifically on summits of quartz-diorite rock, exactly where you would expect piezoelectric effects to be strongest.1Bulletin of the Seismological Society of America. Earthquake lights: A review of observations and present theories
However, the piezoelectric hypothesis has limitations. For random rock to produce a large net charge, the quartz grains need to be oriented in a somewhat coherent direction, which does not always happen in natural formations. Critics have pointed out that in many geological settings the grains are oriented randomly enough that their individual charges cancel out. The positive-hole mechanism does not require any particular crystal orientation, which is one reason it has gained more traction in recent decades. That said, in specific geological contexts, piezoelectricity likely contributes to the overall electrical activity around faults.
Air Ionization and What It Does to the Atmosphere
Regardless of which rock-level mechanism dominates, the downstream effect appears to involve massive ionization of air near the ground surface. Research has shown that stressed rocks release charge carriers that ionize the air column above them, increasing its electrical conductivity. This ionization process may drive several observable phenomena at once: visible light, infrared emission detectable by satellites, and even perturbations in the ionosphere hundreds of kilometers overhead.7Journal of Atmospheric and Solar-Terrestrial Physics. Air ionization at rock surfaces and pre-earthquake signals
Measurements of the atmospheric electric field near earthquake zones have detected unusual disturbances before large earthquakes. Specifically, negative anomalous changes in the atmospheric electric field have been monitored multiple times before major seismic events.8Results in Geophysical Sciences. Study on the mechanism of atmospheric electric field anomalies before earthquakes These shifts in the local electric field are consistent with the idea that charge from deep in the crust is reaching the surface and altering the electrical properties of the lower atmosphere. The light people see may be a byproduct of this broader electrical disruption, essentially a visible symptom of an invisible electromagnetic storm brewing underground.
Signals in the Ionosphere
One of the more striking recent findings is that the electrical disturbances associated with earthquakes reach far beyond the visible lights at the surface. Researchers using GPS satellite data have detected changes in the ionosphere, the electrically charged layer of the upper atmosphere, in the minutes and hours before major earthquakes. Before the massive 2011 Tohoku-Oki earthquake in Japan, the total electron content in the ionosphere deviated by as much as 10% from normal patterns, starting about 40 minutes to over an hour before the rupture. These deviations appeared simultaneously over a wide area and did not propagate like a wave, suggesting they originated from below rather than from solar or atmospheric activity.9Journal of Geophysical Research: Space Physics. Apparent ionospheric total electron content variations prior to major earthquakes due to electric fields created by tectonic stresses
Similar patterns have turned up before other earthquakes. Ionospheric changes were detected about 23 minutes before the first of the devastating 2023 Kahramanmaraş earthquakes in Turkey, with the anomaly reaching roughly 2% of the background electron content.10Advances in Space Research. Ionospheric changes immediately before the 2023 February Kahramanmaras earthquakes, Turkey And before the 2025 Myanmar earthquake, total electron content above the fault changed its trend about 36 minutes before the rupture, with the positive anomaly reaching about 1% of the background.11Geophysical Journal International. Ionospheric changes immediately before the 2025 March 28 Mw7.7 Myanmar earthquake
These ionospheric perturbations do not prove any particular mechanism for earthquake lights, but they strongly suggest that tectonic stress generates electric fields powerful enough to influence the atmosphere all the way up to the ionosphere. If an electric field can shift electron densities at altitudes of hundreds of kilometers, it is entirely plausible that the same field could ionize air at ground level and produce visible light. The ionospheric data and the visible lights may be different symptoms of a single underlying electrical process rooted in crustal stress.
Why Some Scientists Remain Skeptical
Despite the accumulating evidence, earthquake lights remain contentious in mainstream seismology. The core problem is that they are sporadic and unpredictable. You cannot set up instruments in advance to measure something you do not know is coming, in a location you cannot predict. Most evidence still comes from eyewitness testimony, security footage, or chance recordings, not from controlled scientific observations.
There are also persistent confounding factors. Electrical storms sometimes coincide with seismic events. Utility infrastructure routinely fails during earthquakes, producing flashes that are easy to mistake for geophysical phenomena. Atmospheric conditions before earthquakes can include fog, dust, and unusual cloud formations that may trick observers. The 1973 review of earthquake lights noted that “great difficulties and uncertainties accompany any attempt to explain the phenomenon,” and that assessment has not entirely changed in the half-century since.12GeoScienceWorld / Bulletin of the Seismological Society of America. Earthquake lights: A review of observations and present theories
The positive-hole hypothesis has been tested in laboratory settings with rock samples, but scaling those results up to the size of a real fault system involves assumptions that have not been fully validated. Critics point out that the charge densities observed in small rock samples may not translate directly to the enormous, heterogeneous rock formations that exist in real earthquake zones. The phenomenon sits in an uncomfortable scientific space: too well-documented to dismiss entirely, but too variable and poorly controlled to confirm with the rigor the field demands.
Landslides and Fractured Rock
Some researchers have explored whether the light produced during earthquakes might come not from deep crustal processes but from the violent surface effects of shaking, particularly landslides. When large rock masses slide against each other during an earthquake, frictional heating and fracture produce their own light through processes similar to what you see when you crack certain crystals in the dark. Laboratory experiments simulating impact and shear fractures in various rock types have highlighted multiple physical and chemical processes tied to the frictional heating generated during these impacts.13Earth, Planets and Space. Laboratory investigation of earthquake lightning due to landslide This mechanism could explain some reports of lights observed along mountain ridges and slopes during earthquakes, though it would not account for pre-earthquake lights seen days or months before the shaking starts.
The reality is that earthquake lights likely do not have a single cause. Different mechanisms may dominate in different geological settings and at different times relative to the earthquake. Charge carriers from deep stress may explain pre-seismic lights, piezoelectric effects in quartz-rich formations may explain co-seismic lights on certain ridges, and frictional fracture luminescence may explain lights associated with landslides during strong shaking. The phenomenon might be a family of related effects rather than a single unified process.
Connections to Unusual Animal Behavior
One of the more intriguing threads connects earthquake lights to another longstanding puzzle: reports of animals behaving strangely before earthquakes. Dogs howling, birds fleeing, fish surfacing in unusual numbers. These reports, like earthquake lights, were long dismissed as folklore. But recent research has proposed a physical link between the two phenomena.
The same electromagnetic signals that produce visible lights and ionospheric perturbations may also be detectable by animals. One proposed mechanism suggests that the seismic electric signals emitted before earthquakes can affect electro-sensitive ion channels on cell membranes, disrupting the electrochemical balance of cells. Sensitive animals could experience this as discomfort or distress, explaining the panicked or unusual behavior reported before major earthquakes at distances of up to 500 kilometers from the epicenter.14Science of the Total Environment. On the biophysical mechanism of sensing upcoming earthquakes by animals
Laboratory experiments have tested this more directly. When animals including rats, gerbils, hamsters, guinea pigs, and small birds were placed on a wet conductive floor and exposed to increasing ground electric fields, they initially displayed grooming and nervous behavior, then tried to avoid the field, and finally ran and jumped in panic as the field intensity rose from 1 to 1,000 volts per meter. Researchers calculated that an electromagnetic model of a fault under stress could produce field intensities sufficient to generate the kind of body current that triggered these responses.15Japanese Journal of Applied Physics. Ground Electric Field Effects on Rats and Sparrows: Seismic Anomalous Animal Behaviors (SAABs) If the same electrical processes that light up the sky also ripple through the ground as detectable fields, it would make sense that the two anomalies, lights and animal behavior, share a common cause.
Earthquake Lights in Mythology and Religion
Long before seismologists debated charge carriers and piezoelectricity, people witnessed these lights and folded them into their understanding of the world. In the Greek Orthodox tradition, legends about the founding of certain shrines are intertwined with luminous phenomena that modern researchers now interpret as probable earthquake lights. A study of these legends examined two cases: the founding legend of St. George Monastery near Cape Fiolent in Crimea, and the legend of the Panagia Tripiti Church in Aigion on the Peloponnese. In both cases, the perception and interpretation of the lights depended heavily on religious and cultural traditions, with the luminous phenomena being understood as divine signs rather than geological events.16Mediterranean Archaeology and Archaeometry. EARTHQUAKE LIGHTS IN LEGENDS OF THE GREEK ORTHODOXY
This pattern repeats across cultures. Japanese historical records going back centuries describe mysterious fires and glowing skies associated with earthquakes. In Andean traditions, lights in the mountains before tremors were attributed to spirits. In each case, the phenomenon was real enough that people built narratives around it, but the narratives obscured the physical cause. The history is useful to scientists today as a form of long-duration observational record: if people in seismically active regions reported lights for centuries, the phenomenon is unlikely to be purely a modern artifact of electrical infrastructure or camera artifacts.
Could Earthquake Lights Be Used for Prediction
The question that inevitably follows any discussion of pre-earthquake phenomena is whether they could serve as a warning system. Earthquake lights, ionospheric anomalies, and unusual animal behavior all share a tantalizing quality: they sometimes appear before the shaking starts. But “sometimes” is the key word. Not every earthquake produces lights. Not every luminous event near a fault zone is followed by a quake. The false-positive rate would be enormous if you tried to use lights as a prediction tool, and the consequences of false alarms in earthquake preparedness are serious.
The ionospheric data is somewhat more promising because it can be monitored continuously with existing GPS satellite networks, and the anomalies appear to follow a pattern, showing up in a roughly 20-to-40-minute window before major ruptures. But even those signals are small, on the order of 1 to 10 percent of background electron content, and distinguishing them from normal ionospheric variability caused by solar activity and weather is an unsolved problem. Earthquake prediction remains one of the hardest challenges in earth science. The electrical phenomena surrounding earthquakes are a genuine lead, but they are far from becoming a reliable forecasting tool. For now, the lights remain a fascinating window into the extreme physics happening beneath our feet, rather than a practical early warning signal.