Estimating bounds on coherent acceleration and velocity pulses from video-recorded rocking of vehicles in Caraballeda during the Mw 7.5 Venezuela earthquake
The lack of accelerograms in the heavily damaged coastal towns of La Guaira state north of Caracas to record the motions of the Mw7.5 earthquake motivate the present inverse analyses to estimate plausible ranges of coherent ground-velocity and acceleration pulse amplitudes at a particular coastal site. To this end, we utilize an exceptionally revealing video recorded by two CCTV stations during the event in the devastated town of Caraballeda, located a couple of hundred meters from the seismogenic fault, and about 35 km from the epicenter. The video shows a parked, unoccupied Daihatsu Terios SUV undergoing strong lateral rocking oscillations with substantial wheel uplift — evidence of very strong excitation. Analysing the car as a block supported on tensionless springs and dashpots of the wheel-suspension system and using as base excitation idealized wavelets (derived from two motions recorded under comparable seismologic and geotechnical conditions), we constrain plausible pulse amplitudes by comparing computed and observed peak rotation angles. A second heavier car, visually recognized as a Jeep Cherokee/Liberty KK SUV, is also seen in the video to undergo rocking oscillations, but with much smaller amplitude and little uplift. Reproducing its response to the same excitation provides an additional consistency check on the inferred pulse characteristics. The investigation demonstrates the potential of video-recorded nonlinear rocking responses to provide quantitative constraints on (rather than unique predictions of) coherent near-fault seismic ground-motion pulses, at a location where no accelerogram is available.
Authors
- Evangelia Garini (ORCID: https://orcid.org/0000-0001-6972-5980)
- José Restrepo
- Rubi Garcia Gonzalez
- George Gazetas
- Baoning Wu
Publication Details
- Published
- 2026-11-15
- DOI
- https://doi.org/10.31223/x55f9v
- Primary Topic
- Seismic Performance and Analysis
- Type
- preprint