In the early hours of 6 February 2023, a magnitude 7.8 earthquake struck southern Türkiye. Just 9 hours later, a magnitude 7.5 earthquake hit 90 kilometers (60 miles) to the north. For weeks afterward, thousands of smaller aftershocks rattled already frail buildings and disrupted relief work.
Before 2023, seismologists had warned that large earthquakes were likely in this area, but the scale of damage caused by the event was unforeseen. The quakes were among the most devastating the world has witnessed this century, inflicting terrible loss of life in Türkiye and neighboring Syria.
Scientists are now trying to understandhow the earthquakes in this cascade were seismically linked by exploring how the initial main shock triggered a chain of further shocks by offloading stress onto neighboring faults. The research points to where future aftershocks are likely to occur.
“Although we can’t predict earthquakes, we can identify faults that might be more likely to break,” said seismologist Ross Stein, CEO of the earthquake hazard modeling company Temblor, who is involved in the investigations.
Warnings from History
Eastern Türkiye sits at a seismically active crossroad, with two major strike-slip faults marking the junction between the Anatolian, Arabian, and Eurasian plates. The two major 6 February earthquakes (together referred to as the Kahramanmaraş earthquake sequence) occurred on separate faults branching off the 600-kilometer-long (380-mile-long) East Anatolian Fault, which runs through eastern Türkiye and south into Syria, separating the Anatolian and Arabian tectonic plates. Another major fault—the North Anatolian Fault—runs along the length of northern Türkiye.
“These two faults effectively hold Türkiye in a huge tectonic vise,” Stein said. The Anatolian plate, on which Türkiye sits, is slowly escaping from that grip, moving westward at about 20 millimeters (0.8 inch) per year. But this motion isn’t smooth, because fault surfaces aren’t smooth; as the plates slide past each other they get stuck, sometimes for centuries. When enough stress builds, the plates suddenly slip, releasing that stress and causing a large earthquake.
Türkiye’s national hazard map, published in 2018, highlights the North and East Anatolian Faults, signaling the high likelihood of seismic activity along them. The hazard was known, said Karin Şeşetyan, an earthquake engineer at the Kandilli Observatory and Earthquake Research Institute (KOERI) in Türkiye, but the recent double earthquake exceeded expectations.
The country’s last magnitude 7.8 (M7.8) earthquake happened in 1939, on the North Anatolian Fault. “That earthquake initiated a striking falling-domino sequence of earthquakes,” Stein said. Twelve very large quakes happened over the next 60 years, sequentially unzipping more than 1,000 kilometers (620 miles) of the North Anatolian Fault.
Southeastern Türkiye has been quieter over the past century. But “the East Anatolian Fault is still familiar with very large earthquakes,” Şeşetyan said. “You just have to look back through history to find them.”
Historic evidence from the past thousand years shows that this stretch of fault has hosted six very large temblors, with reconstructions placing their magnitudes between 7.0 and 7.5. The cities of Antakya in southeastern Türkiye and Aleppo in northern Syria were razed by several earthquakes over this period, including one of M7.0 in 1822 that caused more than 20,000 fatalities.
Damage of Staggering Proportions
The recent M7.8 earthquake struck just after 4:00 a.m. local time on the previously unrecognized Narli Fault, a minor branch of the East Anatolian Fault. Once an earthquake begins at an initial rupture point, it grows as a larger section of fault slips. The longer the stretch of fault that ruptures is, the greater the magnitude of the earthquake.
The initial rupture grew mostly northeastward, toward the East Anatolian Fault itself. “It could have been a much smaller earthquake,” said Jean-Paul Ampuero, a geophysicist from the Géoazur Laboratory, Université Côte d’Azur in Nice, France. If the rupture had stopped when it got to the East Anatolian Fault, he estimated that would have resulted in a roughly M6.7 earthquake, which is about 45 times smaller than M7.8. “Instead,” he said, “it became huge.” When the rupture reached the junction, it took off in opposite directions, growing to a staggering 350-kilometer-long (220-mile-long) span.
“The scale of the ruptures meant that damage was widespread.”
Hours later, the M7.5 shock hit directly to the north of the M7.8 epicenter, on a larger fault known as the Çardak-Sürgü fault zone, which branches off the East Anatolian Fault. That rupture spread for 150 kilometers (90 miles).
“The scale of the ruptures meant that damage was widespread,” said Luca Dal Zilio, a geophysicist from ETH Zürich.
Estimates from the U.S. Geological Survey (USGS) indicate that roughly 32 million people experienced a shaking intensity of “strong” or above as a result of the main and secondary shocks. Severe to violent shaking occurred along the length of both ruptures, but light shaking was felt even in Istanbul, roughly 800 kilometers (500 miles) away. In Türkiye, the provinces of Kahramanmaraş and Gaziantep suffered the most damage; in Syria the provinces of Idlib and Aleppo were the most affected.
Dal Zilio said that the M7.8 rupture was particularly unusual because it released three long segments of the East Anatolian Fault in one go. “Past earthquakes on the East Anatolian Fault have generally broken just one section,” Şeşetyan said.
Satellites observed roughly 6 meters of ground displacement along the East Anatolian and Çardak-Sürgü Faults, which ripped roads, buildings, and railways in two.
Reconstructions of how the faults slipped showed that the M7.5 earthquake ruptured extremely fast in one direction along the Çardak-Sürgü Fault. Previous observations of these so-called supershear earthquakes have suggested that they might be more destructive, said seismologist Diego Melgar of the University of Oregon. “The rupture speed could explain why ground shaking was so strong.”
“The quick turnaround between these two earthquakes of this scale would have been hard to foresee.”
Shaking was also magnified by the shallow depth of the earthquakes. Vertical ground motions were unusually strong close to the faults, according to Eser Çaktı, an earthquake engineer at KOERI. “This kind of shaking could be particularly damaging for high-rise constructions,” she said.
The many staggering dimensions of this earthquake sequence were all the more devastating because they played out over the course of just a few hours. For Şeşetyan, that’s what made the Kahramanmaraş earthquake sequence scientifically unique. “The quick turnaround between these two earthquakes of this scale would have been hard to foresee,” she said.
Identifying how these two events were connected could add to scientists’ understanding of cascading aftershock sequences, which can worsen the damage done by the initial main shock and hamper rescue and relief efforts.
Earthquake Communication
Scientists have known for some time that earthquakes, particularly large ones, can trigger or inhibit additional tremors on neighboring faults. “Stress relieved during an earthquake doesn’t just fade away into Earth’s crust,” Stein said. It can get transferred to other faults in the area; even a subtle increase in stress can trigger another tremor. In this way, “earthquakes are effectively in communication with one another,” Stein said.
By mapping out patterns of stress change during the Kahramanmaraş earthquake sequence, scientists tested whether the first shock caused the M7.5 event.
“We could see that the one-two punch of earthquakes in close succession was pretty unusual,” said Matthew Herman, an earthquake modeler at California State University, Bakersfield and one of the authors of a stress model developed by the USGS. “Stress transfer helps us explain that sequence.”
Two groups developed independent stress transfer models in the days after the earthquake sequence: One was published by Temblor, and the other was published by the USGS 2 weeks later. Both models treat Earth’s crust as if it were a stiff block of rubber with faults scored through it. Running the models, the researchers simulated how the rocks grated past each other to identify patterns of stress change. To check their models, they compared the locations of forecast stress changes against those of actual recorded aftershocks.
Both models showed that the M7.8 quake likely triggered the M7.5 quake by shifting stress from one fault to another. “That stress transfer unclamped a section of the Çardak-Sürgü—the same section where the second quake nucleated,” Stein said.
“That little fault gave a kick to the East Anatolian Fault, which was already primed to go. Then it just kept unraveling.”
The stress transfer calculations can explain other elements of the sequence, such as why the M7.8 quake started on an unmapped minor fault and propagated along the larger fault, Herman said. “That little fault gave a kick to the East Anatolian Fault, which was already primed to go. Then it just kept unraveling.”
Some aspects of the Kahramanmaraş sequence can’t be explained by stress transfer alone. One peculiarity is that the M7.5 earthquake nucleated on a fault that was oriented almost perpendicular to the East Anatolian Fault. Stein explained that the odd arrangement of the Çardak-Sürgü Fault should make it harder for tectonic forces to get it moving.
“Even if I’d known that section was unclamped, I wouldn’t have foreseen that fault going next,” Stein said. Herman agreed: “Why didn’t the East Anatolian Fault take all of the deformation?” He added, “Other factors must also influence which faults end up rupturing.”
Though stress transfer from the M7.8 earthquake seems to have triggered the M7.5 shock, the short time between the two events still puzzles scientists. Investigations by Ampuero and his colleagues have so far yielded no smoking gun indicating that seismic activity was building along the Çardak-Sürgü Fault during the intervening hours.
Aftershocks in the Shadows
By overlaying patterns of stress released by the M7.8 and M7.5 quakes, the researchers developed an idea of where future aftershocks might strike. According to both models, zones of high stress now radiate from both ends of the East Anatolian Fault and the Çardak-Sürgü Fault.
Seismologists generally agree that once a fault has ruptured, the chance of another large quake along the same stretch is lessened. That’s because earthquakes tend to release nearly all their pent-up stress, and as such, there is promise of calm to follow. “Some people might say that the immediate area is out of the woods for a couple of centuries,” Stein said, “but it’s important to remember that other faults in the vicinity will have been stressed, and aftershocks will occur.”
More than 10,000 aftershocks struck the region in the 3 weeks following 6 February. Most of them were too small to be felt, but roughly 2,000 registered at M3 or greater. The aftershocks dot the faults that ruptured in the main event and reach farther into the zones of calculated stress increase.
Several larger aftershocks occurred at either end of the East Anatolian Fault, within these high-stress zones. An M6.4 quake on 20 February in Antakya, Türkiye, struck at the fault’s southwestern limit. At least three M5.2–M5.7 quakes have so far occurred at the ruptured section’s northeastern limit, near Malatya, Türkiye. Aftershocks will continue for several years, Stein said, becoming less frequent but not necessarily smaller.
Stress transfer calculations could help delineate areas where seismic risk might be higher, Herman said, but “as with any earthquake forecast, we can’t say what magnitude that earthquake will be or when it will happen.”
The Kahramanmaraş earthquake sequence is a potent reminder of the challenges to earthquake forecasting. Stress transfer can cause earthquakes to happen more frequently in some locations for a time or even inhibit them in others. These interactions are difficult to build into earthquake forecasts, which often assume that quakes will follow the same tempo they kept in the past.
The Kahramanmaraş earthquake sequence broke the mold in terms of scientists’ expectations, but the hazard was known nonetheless. Social and economic factors, including poor management of building quality, conspired to make this a human tragedy.
Ampuero and Dal Zilio have several ideas to help build resilience to future earthquakes in this region. They suggested that beyond improved regulation of construction practices and enforcement of building codes, awareness campaigns could help empower citizens to verify the quality of buildings they rent or buy.
Çaktı agreed that the priority should now be on building design and the control of construction. “If you can get all of that right, then your structure is still standing, and you can save lives.”
—Erin Martin-Jones (@Erin_M_J), Science Writer