Original title: Time-Resolved Surface-Fault Displacement During the 2026 Kumamoto Earthquake From Near-Fault Video
Bogdan Enescu, Shinji Toda, Yuji Yagi, Kazuo Oike · August 2026
During an earthquake, the ground breaks and slips along a crack called a fault. Scientists want to know how fast this slip happens, but camera shakes and video glitches make videos tricky to measure. Researchers studied a copy of security camera footage from the 28 July 2026 Kumamoto earthquake. To track the motion, they used two methods. One tracked specific visual features, while the other tracked matching patches across video frames. Both techniques tracked target ground motion against stable reference spots and filtered out overall camera shake. The team matched their video measurements to field data showing 1.383 m of total ground offset. They found that the main slip, going from 20% to 80% of total motion, took approximately 0.9 s. The ground moved at an average rate near 0.9 m/s. An apparent later shift might be extra ground movement or video distortion.
Imagine filming a friend doing a long jump while your hand shakes. The whole video jiggles, so you cannot tell how far the sand moved from your unsteady camera. In this paper, a security camera recorded the ground during an earthquake, and the researchers had to separate the camera's own shake from the actual ground movement. They tracked two spots on opposite sides of the crack and timed how quickly one moved relative to the other, finding the main jump took about 0.9 seconds.
Earthquakes don't just shake; they can permanently shift the ground along a fault. This video shows that shift can happen in under a second, which is exactly the kind of timing that matters for understanding how earthquakes break the ground.
The native recording was unavailable, so the analysis used a secondary copy, which may introduce artifacts. The later apparent motion could be either continued ground displacement or the geometry of the secondary recording, so the permanent endpoint and physical overshoot cannot be determined. The timing estimates are more sensitive in spatially restricted tests. The robust conclusion is limited to the central part of the surface displacement.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Time-Resolved Surface-Fault Displacement During the 2026 Kumamoto Earthquake From Near-Fault Video — Bogdan Enescu, Shinji Toda, Yuji Yagi et al.
Abstract licence: CC0
Explanation generated by Synthesis from the abstract — it is not the original paper. This is a preprint, so it hasn't necessarily been peer reviewed.