At 6:04 p.m. local time on June 24, as millions of Venezuelans watched Scotland play Brazil in a World Cup match, the ground began to shake. In Yumare, a town 90 kilometers (56 miles) west of Caracas, a magnitude 7.2 earthquake ruptured at a depth of 20 kilometers. Thirty-nine seconds later, a second, more powerful magnitude 7.5 earthquake struck 5 kilometers (3 miles) to the east, at a shallower depth of 10 kilometers.
The two quakes, a doublet, in seismological terminology, were generated by different faults, rupturing in rapid succession along the Bocono and San Sebastian fault system. More than 900 people have been confirmed dead, with 4,500 injured and over 50,000 displaced according to the latest reports. At least 250 residential buildings collapsed in the coastal city of La Guaira, and a 22-story building in the Altamira district of Caracas was completely destroyed. The Simon Bolivar International Airport was heavily damaged and remains closed.
The disaster was devastating. But for the scientists who had studied these faults, it was not a surprise.
Faults that were waiting
The Bocono fault is a major strike-slip boundary running approximately 500 kilometers (310 miles) through the Venezuelan Andes. It accommodates about 10 millimeters per year of relative motion between the Caribbean Plate and the South American Plate. The last time it ruptured in a major earthquake was 1812, when a magnitude 7.1 event destroyed Caracas, killing 15,000 to 20,000 people.
In 2017, Franck Audemard, a geologist at the Central University of Venezuela, published a study in the journal Tectonics measuring slip rates on the Bocono fault. Audemard’s team found that 5.0 to 11.2 millimeters per year of strain had been accumulating since 1812, and that the Yaracuy Valley segment of the fault had accumulated a slip deficit of 1 to 4 meters, enough to produce an earthquake of magnitude 7.0 to 7.6.
“The 1812 earthquake’s brother,” Audemard told Science after the June 24 quakes.
Just one week earlier, Machel Higgins, a geophysicist at Florida International University, had presented findings about the San Sebastian fault system at an ESA-organized conference. Using satellite radar data, Higgins showed that the San Sebastian fault had been partially locked since approximately 1900 and had accumulated strain capable of producing a magnitude 7.1 earthquake. ESA radar images taken after the quake confirmed that the rupture had occurred along the San Sebastian fault, at a magnitude exceeding Higgins’s calculation.
The area was, in Higgins’s word, “overdue”, though he cautioned that earthquakes cannot be predicted.
A doublet in seconds
Double earthquakes, two comparable-magnitude events occurring close together in time and space, are unusual. The 2023 Kahramanmaras earthquakes in Turkey were a doublet, but they struck seven hours apart. The 1997 Kagoshima doublet in Japan was separated by 48 hours. The 39-second gap between the two Venezuelan quakes is exceptional.
Germán Prieto, a seismologist at the National University of Colombia, told Science that such a short interval makes separating the seismic signals “very complicated.” The likely mechanism, according to Higgins, was stress transfer: the first rupture on the Bocono fault system changed the stress on the San Sebastian fault segment, triggering the second, larger event.
The Italian INGV institute, using Sentinel-1 satellite data, modeled the event as a combined magnitude 7.6 complex rupture with slip propagating eastward over a length of 210 kilometers (130 miles).
A warning from history
The 1812 Caracas earthquake may have been a doublet itself, two sub-events of magnitude 7.4 and 7.1 on the Bocono and San Sebastian faults respectively. The 1900 San Narciso earthquake, magnitude 7.6 to 7.7, was the largest in Venezuela until this month’s events.
The infrastructure to catch such events has degraded. Venezuela once had the most advanced seismic network in the region. Now it is, in Higgins’s words, “spotty to nonexistent.” Audemard relied on instruments from the United States and Europe to interpret the wave intervals. He believes the second quake may have ruptured much closer to Caracas than the USGS initially reported.
The USGS PAGER system issued a RED alert for the mainshock, estimating a 43 percent probability of between 10,000 and 100,000 deaths, a warning level that acknowledges the potential for much worse outcomes than the current official toll.
What comes next
The earthquake doublet validates the hazard assessments that Audemard and Higgins had published before the event. But it also challenges existing hazard models, which treat fault segments independently. The rapid stress transfer between the Bocono and San Sebastian faults suggests that multi-fault rupture scenarios must be incorporated into future assessments for northern Venezuela.
For now, the region continues to experience aftershocks. Hundreds of thousands of people are displaced. And the scientists who warned about these faults are left with the grim satisfaction of having been proven right.
Source: Science AAAS by Laura Martin Agudelo. Additional data from USGS, INGV, PAHO/WHO, and published studies by Audemard et al. (2017, Tectonics) and Higgins et al. (FRINGE 2026).