Blog

Studying how people are injured inside buildings during earthquakes could improve safety and survival.

In 2023, the city of Antakya in southern Türkiye was one of the hardest hit by a M7.8 temblor, part of the Kahramanmaraş earthquake sequence that devastated the region and neighboring Syria. In Türkiye alone, tens of thousands died, millions were left homeless, and tens of thousands of concrete buildings were destroyed.

Now, scientists are studying this cataclysmic seismic event with the goal of devising new ways to reduce casualties in the future. According to Luis Ceferino, assistant professor of civil and environmental engineering, that work begins with understanding how people are injured inside buildings during major earthquakes.

This summer, Ceferino and a team of researchers traveled to Antakya to interview survivors of the Kahramanmaraş earthquake. Using this data, they aim to build models that can more accurately predict casualties and provide guidance for retrofitting seismically vulnerable buildings.

“By combining engineering and public health approaches, we hope to better understand how injuries happen during earthquakes, how people respond, and—most importantly—how to design policies that can save lives,” said Ceferino. “This includes soft interventions, like awareness and education. It also includes hard interventions focused on strengthening the infrastructure, such as retrofitting non-ductile concrete buildings, including the many in California built prior to the 1970s and considered at risk.”

According to Ceferino, while typical structural models allow researchers to understand the effects of an earthquake on a building, there is not enough data to produce models that can accurately predict earthquake fatalities. California, he noted, has not faced a truly catastrophic earthquake in the modern era, only moderate ones in the past century. He’s hoping the data from Türkiye will allow scientists to close the loop between structural models and fatality predictions.

“The data we have doesn’t tell us the whole story,” said Ceferino. “We need to understand the effects of large earthquakes and what will happen when the San Andreas Fault or the Hayward Fault ruptures. That’s why this work in Türkiye is important for the entire earthquake engineering community. It’s an opportunity to learn how catastrophic these events can be for people living and working inside concrete buildings.”

In addition to Ceferino, the U.S.-based team consists of co-principal investigator, Kimberley Shoaf, a professor in the Division of Public Health at the University of Utah; Yvonne Merino, a postdoctoral researcher in UC Berkeley’s Department of Civil and Environmental Engineering; and Megan Petersen, a Ph.D. student at the School of Medicine at the University of Utah.

In Türkiye, they partnered with Mustafa Kemal University, where professors Necati Ozpinar, Ali Utku Şahin and Bircan Kara led a team of student researchers conducting field work. To collect data, the research team developed a questionnaire that was then loaded onto tablets and used by the Mustafa Kemal University students as they interviewed survivors of the 2023 earthquake.

“As part of the survey, we asked people to draw where they were and why they moved to where they moved, so that we could then model that,” said Ceferino. “By mapping how people in an apartment behave and move during those one or two minutes of the earthquake, we want to understand how they get injured from collisions of different elements—as well as their likelihood of getting injured.”

The team now has enough data to reconstruct these two minutes for a hundred households, or about 400 people. Ceferino said collecting this data would not have been possible without the assistance of the students, who worked closely with the local community.

They will now need six months to a year to completely process their findings, but, so far, they have observed some patterns.

“We’ve learned that human behavior drives a lot of what happens during an earthquake, and it’s something that we haven’t documented before in a systematic way,” said Ceferino. “The responses of people tend to be super social, meaning they move around to go take care of others—their kids, their parents—instead of doing what we think they would do: drop, cover and hold on.”

He added that, given these findings, there is a lot to learn about how people follow protective measures in the moment. This improved understanding could lead to the development of more effective approaches and policies for earthquake safety.

“We need to understand that people are going to react in certain ways to protect their loved ones, no matter what we propose,” said Ceferino. “When you have an early warning signal, for example, people are not going to automatically seek a safe spot. Instead, they’re going to get their kids, and together they are going to seek a safe spot. So all these policies need to take that into account.”

Data collected from survivors also highlighted the seismic vulnerability of partition walls. These types of walls, typically built of masonry, are extremely heavy and, for many years, have been used to divide living spaces within concrete buildings. During the major earthquake in Türkiye, they often collapsed on people and blocked access to exits and stairs.

Next, the researchers will work to make this qualitative data on structural damage and human behavior more quantitative, so they can refine their models to better understand risk and provide specific recommendations for earthquake safety.

According to Ceferino, this last step also includes building computational models of concrete buildings in California to drive policy change—and save lives.

“At local levels, especially in California, we have a list of older concrete buildings that we need to make stronger. But this work is expensive, and we haven’t built enough momentum to put policies in place to retrofit them,” he said. “Using the data from Türkiye, we can model how people in those buildings are at greater risk of injury from a major earthquake than we originally thought—and potentially build that momentum.”