Scientists rely on seismometers and eyewitness accounts to identify an earthquake’s location, time and magnitude. A new study explores how the latter can be limited by socioeconomic factors, which can create biases in datasets that scientists use to characterize seismic hazards and coordinate emergency response.
Studying Earthquakes Through Eyewitnesses Can Be Tricky
As early American pioneers forged a long, arduous path across the country during the Westward Expansion, an earthquake hit what is now the State of Oklahoma on October 22, 1822.
“The trembling and vibrating were so severe as to cause door and window shutters to open and shut, hogs in pens to fall and squeal, poultry to run and hide, the tops of weeds to dip, [and] cattle to lowe [sic],” the Cherokee Advocate reported.
Almost two centuries later, scientists determined that this earthquake had an approximate magnitude of 4.8 and occurred within the Ouachita Fold and Thrust Belt in the southeastern part of Oklahoma. The Choctaw Nation likely felt it most intensely and according to the Arkansas Gazette on October 31, 1882, the tremor was strong enough to topple chimneys. These accounts helped alert scientists to the quake and guide their research.
Gathering information about an historic earthquake from people rather than scientific instruments may seem like a task for Sherlock Holmes, but even today scientists rely on eyewitness reports to help determine the location and shaking effects of tremors in regions with and without seismometers.
Researchers use this information to understand seismic hazards, guide emergency responders and inform policies that could potentially save lives and property. However, self-reported data can be spotty and hindered by socioeconomic and geopolitical factors, said Susan Hough, a geophysicist with the U.S. Geological Survey, and Stacey Martin, a graduate student now at Australia National University and native of Pune, India, in their recently published paper.
To characterize biases in datasets that include information from people who experienced shaking, Hough and Martin studied the 1822 Oklahoma earthquake, three earthquakes between 2011 and 2015 in Bihar, India, and three earthquakes between 1989 and 2019 in California.
The importance of sharing what happened
The 1822 Oklahoma quake’s geographic location was initially misplaced by scientists, with estimated locations in three different states, but after examining eyewitness accounts scattered in newspapers and studying the region’s underlying geology, researchers were able to more precisely pinpoint the location to near Fort Gibson, Oklahoma.
Apart from this earthquake, which was large enough to be felt in surrounding areas, there are no other known records of small earthquakes in Oklahoma’s Native American territories through the 19th century, in part because most Native American communities shared stories orally.
The 1822 Oklahoma quake is a unique and illustrative case study that called attention to an otherwise overlooked potential earthquake hazard area in the central part of the country, Hough and Martin say.
Although the world is a much different place than it was two centuries ago, lacking eyewitness accounts can still lead to poor characterizations of even relatively strong shaking, which can hinder immediate emergency response and affect tools that are used to estimate losses in life and property.
However, Hough did note a bright spot. “Although there are disparities in reporting between places like California and India, over time, eyewitness reporting systems in California appear to have become more inclusive,” Hough said, in part because more people are aware of the systems.
Did you feel it?
Immediately after feeling the ground tremble and seeing light fixtures sway, most people want to ask their neighbors, “Did you feel it?”
In 1999, the USGS decided to ask the world the same thing, introducing the official Did You Feel It (DYFI) system, which collects user experiences through a webform. DYFI data support critical USGS products like ShakeMap, which provides near real-time maps of ground motion and shaking intensity following significant earthquakes, and the Prompt Assessment of Global Earthquakes for Response (PAGER) system, which provides timely fatality and economic loss impact estimates for significant earthquakes worldwide.
Both ShakeMap and PAGER are used for emergency response efforts. In California, for example, the USGS partners with the California Department of Transportation to share detailed ShakeMap information at overpasses and bridges across the state with an application called ShakeCast. The app gives the state’s engineers an immediate look at the severity of shaking at key structures.
In a previous study from 2016, scientists Sum Mak and Danijel Schorlemmer, then with the Helmholtz Center Potsdam, confirmed an expected positive correlation between the number of DYFI responses and three variables - an earthquake's magnitude, person's closeness to the epicenter and the affected region's population size. Furthermore, residents of California and the Central and Eastern U.S. states were equally likely to report feeling an earthquake, despite how many more earthquakes hit California.
For earthquakes in California, there is some tendency for people from relatively affluent areas to contribute more reports, Hough said, but DYFI is usually able to gather enough felt reports from a wide range of socioeconomic areas to map out an earthquake’s intensity in detail. In India, however, contributed reports are overwhelmingly submitted from affluent urban areas, with very few submissions from less affluent rural villages.
Hough and Martin attribute that disparity to higher income levels and more ubiquitous access to internet and smartphones in California compared to Bihar. Education is also critical. “Do people in a specific region of India know that ‘Did You Feel It’ exists?” asked Sara McBride, a research social scientist at the USGS. “And even if people know about it, they would need to want to submit that information to a U.S. government agency.”
“We need to be mindful about why people are giving us information,” McBride said, and provide compelling reasons for them to share their experiences. For instance, she added, sending out teams of local researchers with physical DYFI forms could be a way to build trust with the community and emphasize why data collection is important.