Within 1.3 seconds, the ground on either side of Myanmar’s Sagaing Fault moved 2.5 meters sideways activity that a security camera caught in real time, providing a unique look at earthquake rupture physics. The unusual video, studied by a Kyoto University team, presents a direct and quantitative observation of coseismic fault slip, dispelling entrenched presumptions and confirming decades of indirect inference in seismology.
The key to the discovery is the use of pixel cross-correlation, a digital image correlation method that enabled scientists to quantify fault displacement frame by frame. By monitoring movement of objects in the video, the group reconstructed the slip path with incredible accuracy. Their analysis indicated that the fault relocated at a maximum velocity of 3.2 meters per second, all 2.5 meters of displacement concentrated in a single burst a process described as a pulse-like rupture. As co-author Jesse Kearse explained, “The brief duration of motion confirms a pulse-like rupture, characterized by a concentrated burst of slip propagating along the fault, much like a ripple traveling down a rug when flicked from one end.”
The technical methodology was exacting. Scientists established 25 overlapping 26 × 26 pixel subsets of the image, targeting features 70–80 meters from the camera. By removing the displacement time series of stable features from those over the moving fault block, they removed camera shake and ground motion artifacts. Calibration with known-spaced fence posts, complemented with lens distortion correction and parallax correction, allowed them to convert pixel shifts into ground truth meters. The derived slip function, smoothed and differentiated with a 0.2-second moving average, yielded a high-resolution velocity profile of the rupture process Our results are best interpreted as coseismic slip on the Sagaing fault as a smooth ramp‐like slip function having an amplitude of 2.5 ± 0.5 m and a duration of 1.3 ± 0.2 s.
The novelty of this observation is the explicit visualization of a curved slip path. Though curved slickenlines fault plane striations are long suspected to indicate dynamic, non-linear slip during earthquakes, such evidence has consistently been post hoc and interpretive. Today, the video offers real-time confirmation: the slip path initially started obliquely, with a rake as high as 35°, before migrating toward very pure strike-slip motion as the rupture slowed. Kearse said, “Instead of things moving straight across the video screen, they moved along a curved path that has a convexity downwards, which instantly started bells ringing in my head, because some of my previous research has been specifically on curvature of fault slip, but from the geological record.”
This curvature, the researchers discovered, is not an oddity but a solution to rupture mechanics. Dynamic rupture simulations have also predicted that transient stresses within the cohesive zone ahead of the rupture front can cause oblique slip at the beginning, and the slip path becomes straight as velocity peaking and subsiding This kind of dynamic slip onset behavior is a characteristic found in a suite of dynamic rupture simulations. The orientation and sense of the curvature are also in accord with the sense of rupture propagation, as in the case of both the 2025 Myanmar and 2016 Kaikōura earthquakes.
The implications are not solely for academic interest. By comparing directly on-fault slip velocity to near-fault ground velocity measured at a strong-motion station 2.7 kilometers distant, the study provides a unique chance to test seismological models. Both records indicated excellent consistency in amplitude, length, and character, and implied that near-fault ground motion can be used as an effective proxy for on-fault slip during great earthquakes. A significant exception, though, was the estimation of the slip-weakening distance: 2.4 meters from strong-motion data versus 1.2 meters from on-fault observation. This discrepancy is probably a reflection of off-fault effect complexity and highlights the utility of on-fault measurements.
The larger context for this work is the development of digital image correlation and remote sensing for geophysical monitoring. What were laboratory analog experiments with DIC algorithms are now being used in the field in real time, high-precision tracking of displacement The application of digital image correlation algorithms to monitor earth flows’ displacement is a developing, new field. Their demonstrated reliability down to pixel-by-pixel accuracy places them on par with, and frequently ahead of, conventional GNSS or robotic total station procedures, particularly where rapid or complicated motion is involved.
Not only does this research fill the gap between geological records, numerical models, and instrumental observation, but it establishes a new standard for earthquake source physics. “We did not anticipate that this video record would provide such a rich variety of detailed observations. Such kinematic data is critical for advancing our understanding of earthquake source physics,” Kearse added. The second stage will apply physics-based rupture models and dynamic friction laws to investigate the underlying controls on pulse-like, curved slip a task now supported by direct observation as opposed to inference.
