February 18, 2026 — A new study by researchers at the Massachusetts Institute of Technology (MIT) challenges traditional assumptions about how earthquakes propagate along faults, showing that so-called “boomerang” or back-propagating ruptures may occur more often than previously recognized — even on simple, straight faults.
Earthquakes typically start at a point underground and rupture the fault surface as stress is released between fractured rocks. Seismologists have occasionally recorded cases where the rupture front seems to reverse course shortly after initiation — a phenomenon dubbed a “boomerang quake.” Past interpretations held that such behavior was limited to complex fault systems with multiple branching surfaces. However, the MIT team’s simulations suggest that this ricocheting effect can also arise along uncomplicated fault zones under the right conditions.
Published in the journal AGU Advances, the study used physics-based computer models to represent an idealized elastic crust with a single straight fault. Researchers varied factors such as the length of the fault, the depth of earthquake initiation, and whether the rupture propagated unilaterally (in one main direction) or bilaterally (in two directions). They observed that when a quake travels in only one direction and the friction along the fault drops, rises, then drops again, a secondary rupture front can develop that moves back toward the origin — hence the “boomerang” effect.
“Our simulations suggest that these boomerang quakes may have been undetected in a number of cases,” said MIT graduate student Yudong Sun, pointing out that traditional seismic analysis methods might miss subtle back-propagating fronts. The simulations indicate that large earthquakes, which travel long distances along faults, are especially prone to this effect due to dynamic variations in friction and stress.
Co-author Camilla Cattania noted that even though most people wouldn’t be able to perceive the backward rupture in real time, the directional focus of shaking could alter seismic hazard assessments. Because conventional methods of analyzing seismic data assume rupture propagation in a single direction, boomerang behavior may have gone unnoticed historically.
The findings have implications for understanding earthquake physics and for assessing seismic risk more accurately in regions dominated by simple fault structures, such as segments of California’s San Andreas Fault. The researchers believe their work motivates further observational studies to search for evidence of back-propagating rupture fronts in real seismic events.