Shaken, rattled, and rolled

Featured image: an aerial photograph of the Capitolias/Beit-Ras theater, courtesy of the Aerial Photographic Archive of Archaeology in the Middle East (APAAME), CC-BY-NC-ND 2.0

Paper: Two inferred antique earthquake phases recorded in the Roman theater of Beit-Ras/Capitolias (Jordan)
Authors: M. Al-Tawalbeh, R. Jaradat, K. Al-Bashaireh, A. Al-Rawabdeh, A. Gharaibeh, B. Khrisat, and M. Kázmér

One of the biggest questions in earthquake seismology is whether we can see into the future, to forecast seismic activity based on what we know about faults and how they behave. We’re about as likely to accurately predict earthquakes as we are to see the future in a crystal ball, but one way we can improve our forecasts of seismic hazard actually involves looking in the other direction: back into the past.

Seismologists use historical records of large earthquakes to assess the future earthquake potential of specific faults. The two main quantities of interest are the recurrence interval, the average amount of time between large events, and the magnitude of the largest earthquake a fault is capable of producing. For example, the Dead Sea Transform fault system (DST) on the eastern side of the Mediterranean has hosted sizable earthquakes in the past, including a magnitude 6 event in 1927 and a M~7 event in 1837. Based on that information, we might estimate that the DST generates a M6+ event approximately once a century. But with only two data points, how can we tell if that hundred-year interval is actually representative of the fault’s long-term behavior? One way to answer that question is to look farther into the past, reconstructing the seismic history of the DST over millennia using information from archives, geological features, and even archaeological evidence.

A new study by Al-Tawalbeh et al. uses archaeoseismological techniques to estimate the timing and the magnitudes of two periods of earthquake activity on the DST circa 200-400 CE. The authors analyze features like cracked stones, tilted walls, and dropped arches in a Roman theater to date and measure earthquakes that hit the ancient city of Capitolias in northern Jordan.

A view inside the Capitolias theater showing a collapsed vault and, on the left, an arch with a dropped keystone. Photo by Wikipedia user Zachchan2015, CC-BY-SA 4.0

The theater in Capitolias is a particularly rich site for archaeoseismology, as a dated inscription on a construction repair can be used to separate two distinct phases of earthquake damage. Comparisons with archival records enable the authors to connect the first round of structural damage to earthquakes in either 233 or 245 CE. Following some repairs completed in 261 CE, the theater suffered a second round of damage that likely corresponds to an earthquake in 363 CE. The theater was later filled in with debris, holding the stones in place and preventing subsequent earthquakes from causing further damage prior to the building’s excavation.

Archaeological evidence can also be used to estimate earthquake magnitudes. Specific types of damage – stones with chipped corners, arches with their keystones shifted a few centimeters out of place – are known to happen when shaking reaches certain thresholds. Shaking intensity doesn’t correspond directly to what we think of as an earthquake magnitude, but we can convert between intensity and magnitude using relationships between the two for modern earthquakes. For the events that hit Capitolias, the archaeological evidence suggests the earthquake magnitudes were between 6.2 and 7.5.

As a seismologist, I’m accustomed to looking at direct recordings of ground motion to gauge earthquake magnitudes; or, if those aren’t available, turning to the rock record for geological evidence of how faults have slipped and shifted over time. Archaeoseismological research in places like Capitolias reminds me that records of natural processes are everywhere, in the built environment as well as in nature, if we only remember to look.

Shaken, rattled, and rolled by Hannah Mark is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

Leave a Reply

Your email address will not be published. Required fields are marked *