Original title: Co-seismic surface fault displacement captured in videos: image tracking across three earthquakes
Shiro Hirano · August 2026
Standard earthquake tools sit far from faults, leaving gaps in what we know about how ground breaks right at the surface. To solve this, researchers analyzed security camera videos from three earthquakes. Cameras shake violently during quakes, so the team stabilized the footage by aligning each video frame to keep fixed reference areas steady. This visual tracking showed how the ground moved second by second. Across all three events, the tracked motion lasted approximately 2 s, even though the quakes differed widely in size. In each case, peak speeds were on the order of 1 m/s. The motion in one quake also included compressed ground. Finally, differences in how faults slip seem to change how complex the movement becomes over time.
You film a crack in a sidewalk as a truck drives past. The truck shakes the camera so much that you cannot tell if the crack widens. By aligning each frame so a nearby building stays still, you cancel the shake and see the crack move. In this paper, the crack is the fault, the truck's vibration is the earthquake's shaking, and the building is a still part of the scene used to keep the background fixed.
Seismometers far from a fault miss the exact motion of the ground breaking. These CCTV videos capture it frame by frame, revealing that surface displacement lasts just a few seconds regardless of earthquake size. That helps engineers design buildings that can survive the quick slip, and helps scientists understand how faults actually move.
The study only examines three earthquakes, so it is not clear if the 2-second duration holds for all quakes. The method relies on reference regions in the video that stay still, which may not exist in many CCTV recordings. The velocity estimate comes from field surveys, not direct video measurement. The analysis does not explain why the Kumamoto earthquake had a more complex velocity history.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Co-seismic surface fault displacement captured in videos: image tracking across three earthquakes — Shiro Hirano
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.