Earthquakes are caused by extensive fracture phenomena that occur in the brittle part of the Earth, several kilometers below its surface. Seismic fractures are not instantaneous, instead they develop along fault surfaces during a time that can vary from a few seconds to several minutes depending on the extent of the fractured area, which in turn determines the magnitude of the earthquake.

The speed with which the fracture propagates along the fault from the nucleation point is one of the most relevant earthquake parameters, as it governs the evolution and stopping of seismic ruptures. Rapid changes in velocity during the fracture process can result in the radiation of seismic waves at high frequencies and with amplitudes that can cause strong ground shaking and potential damage to buildings.

However, the measurement of fracture velocity from seismic records is difficult to obtain and is generally affected by large uncertainties. In fact, its estimation requires detailed knowledge of the mechanical properties of the medium in which seismic waves propagate and constraints on source geometry, which are also affected by uncertainty.

Until the 1960s, based on theoretical models and laboratory experiments, the scientific community believed that the fracture velocity in rocks could not exceed that of the shear waves (S-waves) which travel through the Earth with velocities about half of that of compressional waves (P-waves). Subsequently, new experimental evidence at the scale of earthquake fractures has instead revealed the existence of super-shear events, i.e., earthquakes in which the whole or a part of the fracturing process occurs with velocities greater than that of S waves. This class of events is particularly relevant because of the potential destructive impact of the seismic radiation emitted during the fracturing process and whose amplitude is greatly amplified by the propagation of the fracture at high velocities.

In this sense, being able to accurately determine the propagation velocity of the rupture front during an earthquake is important for identifying where and when the emitted waves can produce strong ground shaking and potential damage to buildings.

In the article just published in Nature Communication Earth & Environment, Mauro Palo and Aldo Zollo propose a novel approach to track the rupture front based on the polarization of the first seismic waves recorded at Turkey's dense national accelerometer network, within a few hundred kilometers of the fault that originated the 7.8-magnitude earthquake in Turkey and Syria on Feb. 6, 2023.

The polarization of seismic waves represents the direction of the ground motion oscillation during the seismic wave propagation from the earthquake source to the Earth's surface. In the case of P-waves, this direction is parallel to the wave trajectory, so using advanced triangulation techniques it is possible to reconstruct the position of the fracture front and thus its instantaneous velocity.

Using this approach, the authors reconstructed the location of the rupture front of the first 25 seconds of the Kahramanmaras earthquake, which originated on a branch  of the East Anatolian Fault on the Syrian-Turkish border causing enormous damage and about 60000 casualties.

The study reconstructed the kinematics of the main rupture that caused the Kahramanmaras earthquake. This event originated on a secondary fault, then the rupture propagated northeast wards and reached after about 15 seconds the East Anatolian Fault, that separates the tectonic plates of Anatolia and Arabia and is about 700 km long. During the following 20 seconds, the seismic rupture propagated along the northern segment of the East Anatolian Fault and then asynchronously activated the bi-lateral rupture of the southern segment, for a total fault length of about 340 km. From the position of the rupture front every second, the authors determined the fracture velocity, highlighting that it reached the highest super-shear values in the portions of the fault where the co-seismic fault slip was maximum raching values of about 8 meters.

This and other experimental evidence suggest the presence, along the same fault structure, of segments that exhibit different frictional regimes during the fracture process; in particular, some segments promote ultra-rapid fracture propagation and significant co-seismic fault slip. Both phenomena probably contributed to the radiation of highly destructive seismic waves in  the areas surrounding the slipped fault.