Earthquake emergency response: How can felt reports help?

A new method utilizes online testimonies collected after earthquakes to classify events as high or low impact in the immediate aftermath. The approach may provide emergency responders with valuable information, improving their ability to respond appropriately.
 

By Civan Yavas, Science Writer (@civan_yavas)
 

Citation: Yavas, C., 2023, Earthquake emergency response: How can felt reports help?, Temblor, http://doi.org/10.32858/temblor.313
 

In 2022, a magnitude-5.9 earthquake struck a remote region in Afghanistan, causing great destruction and killing more than 1,000 people. Although assessing the early impact of an earthquake can be straightforward when there are reliable data from a dense network of seismic stations, this was not the scenario facing authorities. Emergency responders underestimated the severity of the quake and the extent of its damage because the region’s sparse seismic network did not accurately reflect the level of ground shaking experienced at remote sites. Additionally, a high likelihood of landslides in these steep, mountainous regions compounded the risk, which wasn’t captured by risk calculations based only on data from seismic stations.
 

 

A recent study published in The Seismic Record shows how self-reported testimonies of earthquake experience can be used to rapidly classify events as high or low impact in the immediate aftermath, providing an important characterization of the quake that authorities can use to determine whether emergency action is necessary. The classification is based on residents’ testimonies or “felt reports,” and is independent from seismic instrumentation. The approach can provide emergency responders with qualitative quake characteristics before more sophisticated, quantitative estimates of shaking intensity and extent are available. This information could improve emergency services’ ability to respond appropriately.
 

Citizen-based seismology

While networks of seismic stations measure ground motion, felt reports use people’s own observations to gather information about the extent, duration and intensity of shaking. The data are inexpensive and rapidly available. The European-Mediterranean Seismological Centre’s (EMSC) LastQuake service has collected more than 1.5 million global testimonies from more than 10,000 earthquakes between 2014 and 2021. With LastQuake, the testimonies from earthquake witnesses are collected via the EMSC webpage or LastQuake mobile application. (This is similar to the U.S. Geological Survey’s “Did You Feel It?” program, also referred to as DYFI.)

Although the reporting tool has been in use for 12 years, researchers have underutilized the rich earthquake data it produces, suggest the authors of the new study.

“We kind of realized that felt reports were barely used in any quantitative study, and we wanted to see if we can utilize this data to make rapid preliminary assessments of an earthquake,” says Henning Lilienkamp, a University of Potsdam researcher who led the study. Lilienkamp and his co-authors focused on whether they could distinguish between high- and low-impact earthquakes because authorities responsible for emergency response depend on that distinction.

The researchers defined a quake as high impact if it caused at least one of the following: at least one destroyed building, at least 50 damaged buildings, at least two fatalities, or documented financial losses. If none of the high impact conditions were met, the quake was classified as low impact.

The team used LastQuake reports to develop a database of high impact and low impact earthquakes. In total, their database contains 254 high-impact and 1,994 low-impact events. In their analysis, the researchers saw a clear relationship between shaking area and impact, where strong shaking felt over large areas tended to have larger impact. The researchers then created a model that estimates the likelihood of future quakes being high impact based on LastQuake users’ reports. To ensure the accuracy of these forecasts, the researchers removed quakes with fewer than 50 reports from the database.

To test the effectiveness of the model, the researchers selected 11 validation earthquakes for which the impact levels were already known. They used the testimonies from these events to see whether the model classification matched the actual impact level. In particular, the strength of the model is its ability to correctly identify a large number of low-impact events with high confidence, allowing emergency responders to tailor their response accordingly — to avoid using too few or too many resources, and to arrive with the support a region really needs after a quake.
 

 

Inexpensive, rapid and crucial alternative

Using citizens’ testimonies to estimate initial damage from a quake could be especially important in remote regions, which tend to have no or sparse seismic monitoring networks or lack the funds to improve existing stations.

“We were pretty quickly aware that this data might fill the gaps, let’s say geographic gaps, where we don’t have expensive seismic instrumentation,” Lilienkamp says. While the authors’ method uses only testimonies to evaluate initial consequences, there are other rapid impact assessment tools (e.g., PAGER) that combine felt reports with recordings from seismic stations. However, these other methods rely on expensive, dense seismic instrumentation to maximize their accuracy, and may not be practical or available for all areas.

“This work is definitely a proof of concept of what can be done with community felt reports,” says Danielle Sumy, a project manager at the EarthScope Consortium who was not an author of the study. “Citizen seismology uses the opportunity that everyone can report their own information. Although people are not the greatest seismometers, [information reported by the community] does give us a lot of insight into earthquakes and their intensities, especially in areas where instrumentation is not all that dense,” Sumy says.

The information contained in the felt-it reports is “a new type of data that hasn’t been used before. We don’t think of our work as replacing the state-of-the-art technologies,” says Lilienkamp. “Our idea is to have a rapid, alternative source of information.”

The idea is to be able to bypass some of the things that other methods have relied upon, such as the use of ShakeMap, adds Sumy. That “may have a lot of impact in areas where we can’t get information for one reason or another.”
 

 

Collaboration is key

Secondary hazards like landslides or ground settlement were not incorporated into the study, but this may be a possible avenue of future research. Reports are also not always well-distributed around the affected region, which limits the method’s capabilities. Nevertheless, “if we were able to operationalize this in some way to see if something like this would work in real-time, I think that would be very valuable,” Sumy says.

Because earthquake impacts are highly variable from place to place due in part to differences in the built environment, creating a model that may predict the early impacts of earthquakes in specific regions would require working closely with local disaster managers to identify and plan appropriate emergency responses.

“We cannot really give a recommendation on how [the results of our study] will be implemented in practice without communicating with emergency responders,” Lilienkamp says.
 

Further Reading

Lilienkamp, H., Bossu, R., Cotton, F., Finazzi, F., Landès, M., Weatherill, G., & von Specht, S. (2023). Utilization of Crowdsourced Felt Reports to Distinguish High‐Impact from Low‐Impact Earthquakes Globally within Minutes of an Event. The Seismic Record, 3(1), 29-36.
 

Copyright

Text © 2023 Temblor. CC BY-NC-ND 4.0

We publish our work — articles and maps made by Temblor — under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license.

For more information, please see our Republishing Guidelines or reach out to news@temblor.net with any questions.