Plate Nº 33 · recorded November 16, 2026
Earth & ClimateReported finding
CCTV Cameras Reconstructed Venezuela's Earthquake Shaking at 2 m/s
No seismometer recorded Venezuela's June 24 magnitude 7.5 quake where damage peaked, so researchers rebuilt the shaking from CCTV — and found one of the strongest motions ever measured.
By Nathan Brooks5 min read918 words
In brief
- The magnitude 7.5 earthquake struck Venezuela's coast on June 24, 2026, collapsing hundreds of buildings with no seismic instrument recording the worst-hit area.
- Researchers reconstructed a peak ground velocity of roughly 2 m/s from CCTV footage — stronger than any record from the 2023 Türkiye earthquakes.
- The estimate runs on average nearly twice Venezuela's seismic design code for an extreme earthquake across mid- and high-rise building periods.
- Only two recordings of quakes above magnitude 7.4 existed within 5 km of a fault before 2023, per the NGA-West2 database.
- The study, published in The Seismic Record, is the first video-based ground motion reconstruction verified by an independent object check.
A magnitude 7.5 earthquake struck Venezuela's northern coast on June 24, and no seismic instrument recorded how hard the ground shook where the damage was worst. So researchers rebuilt the shaking from security camera footage — and found a peak ground velocity of roughly two meters per second, one of the strongest values ever measured.
That figure, published in The Seismic Record, comes from an international team including Dr. Miguel Bravo-Haro, senior lecturer in structural engineering at City St George's in London. The paper, led by José Bustos and colleagues, is titled "Ground-Motion Intensity in Caraballeda During the 24 June 2026 Mw 7.5 Venezuela Earthquake Estimated from CCTV Video."
Bravo-Haro developed the numerical model — a bidirectional sliding block simulation — and the inversion algorithm, the math that works backward from observed motion to the shaking that caused it. This is the first reconstruction of the ground motion that hit Caraballeda, in La Guaira state, during the quake.
How do you measure an earthquake with no instruments?
The June 24 rupture tore along the San Sebastián and Boconó faults, passing almost directly beneath La Guaira's communities. Hundreds of buildings collapsed and many more were damaged in one of the most violent earthquakes to hit a densely populated area in decades.
Yet the region still lacks a network capable of measuring a quake of this size near the fault. History repeated itself eerily: when the 1967 Caracas earthquake damaged many of the same neighborhoods, the era's few instruments failed or were never properly installed — one accelerograph in Caracas had no battery. For nearly 60 years, scientists had to infer shaking intensity indirectly, from damage.
That gap matters because large, shallow earthquakes that rupture within a few kilometers of a city are the events engineers fear most. They can produce a strong, long pulse of motion that pushes buildings far beyond their design limits. They're also the world's least-recorded quakes. According to the NGA-West2 database — the international dataset engineers use to build design codes — only two recordings of earthquakes above magnitude 7.4 existed within 5 kilometers of a fault before 2023. The 2023 Türkiye earthquakes added 20 more.
What did the cameras actually capture?
The team turned to two ordinary security cameras on a commercial building in Caraballeda that happened to be filming the street. In the footage, a plastic traffic delineator post slides across the pavement, and an 11-meter (36-foot) lamppost sways side to side.
Neither object is a measuring device, but the physics of how each responds to shaking is well understood. The researchers tracked both objects frame by frame, then calculated the ground motion that must have produced those exact movements. To make the numbers reliable, team members visited the site, surveyed the objects, and measured the friction between the post's base and the pavement on the spot.
Two features strengthen the result. First, the objects respond to shaking in very different ways — one sliding, one vibrating. Second, the researchers calculated the ground motion from the sliding post alone, then checked it against the lamppost, which played no part in the original calculation. The fact that one result explains both objects gives high confidence in the estimate.
Four analysts also tracked the post independently, and the final figure falls within the range of their traces. To the authors' knowledge, this is the first video-based ground motion reconstruction verified independently this way.
How strong was the pulse?
Extraordinarily strong. The ground received a pulse lasting less than one second, directed mainly north-south, roughly perpendicular to the fault.
For context, only two recordings in the NGA-West2 database — both from the 1999 Chi-Chi earthquake in Taiwan — exceed the roughly two meters per second observed at Caraballeda. The value tops the strongest record from the 1995 Kobe earthquake in Japan and anything recorded during the deadly 2023 Türkiye earthquakes.
The intensity also exceeds what engineers' worldwide models predict, even after correcting for proximity to the fault. Across the range of vibration periods typical of mid- and high-rise buildings in the area, it runs on average nearly twice what Venezuela's seismic design code specifies for an extreme earthquake.
The same cameras caught something else: eight buildings beginning to collapse at the same moment, just as the pulse ended. The authors believe it's the first time video has captured that exact instant.
Why does this matter beyond Venezuela?
In practice, this is the only public, quantitative estimate of how the ground moved in one of the worst-hit areas. Engineers need it to understand why buildings failed, update seismic hazard maps and Venezuela's design code, and decide how to repair and rebuild — all urgent tasks.
The method travels, too. Security cameras now sit on almost every street in earthquake-prone cities, and ordinary surveillance video can become a reliable engineering tool where instruments don't exist.
Bravo-Haro put it this way: "Working with outstanding colleagues and students across four countries made this project as stimulating as it was productive. Together we've shown that CCTV cameras, found on almost every street, can help us measure earthquake shaking, and that opens up exciting new directions for research."
The findings are preliminary in one sense — a single site, reconstructed from video rather than instruments — but the independent verification gives the estimate real weight. Cities that can't afford dense seismic networks may already have the next best thing mounted on their buildings.
via phys.org/rss-feed/ (Source)
More from Nathan Brooks
Show full bio
Market editor covering companies, deals and strategy in general science news at SciBeat.
402 articles
Nearby plates
- Lab Tests Overestimate Volcanic Tsunamis, Stromboli Study Finds
- Southern California Faults Hit 1,000-Year Stress High, Study Finds
- Cascadia megaquake could shake northern Oregon 17% harder than expected
- Satellite Radar Can Track Lava Flows — With Caveats, Study Finds
- Satellite radar tracks Kīlauea lava and crater changes