Community science builds a seismic network in Haiti
Small, inexpensive seismometers are capable of sharing high-quality data in real time—and were put to the test during an August 2021 earthquake.
In August 2021, a devastating magnitude 7.2 earthquake ripped through Haiti’s Tiburon Peninsula. The closest conventional seismic stations were 120 kilometers from the earthquake’s epicenter, too far away to capture near-field seismic information, which is important when trying to fully understand the main shock and its aftershock sequence. Luckily, these stations were supplemented by a network of small devices called Raspberry Shake (RS) seismometers, tucked inconspicuously inside people’s homes.
Raspberry Shakes cost a fraction of the price of conventional seismometers. They readily connect to the Internet (unlike conventional seismometers, which can require manual data downloads), making them perfect for a community-run seismic network, said Eric Calais, a geophysicist at the École Normale Supérieure in Paris. He helped initiate this project in 2019 in collaboration with researchers at Université d’État d’Haïti. The team demonstrated the devices’ utility in a paper recently published in Science.
Steeve Symithe, an assistant professor at Université d’Etat d’Haïti and a researcher with URGéo Research Laboratory, was the person responsible for installing 13 of the 15 operational RS devices. “When I realized that every single piece of equipment that I had installed worked perfectly,” he said, “I was very excited.”
The network captured the initial shock and uploaded the data to an online data bank in real time. In the days immediately after the quake, Symithe braved aftershocks, gang-related insecurity, and a tropical storm to install three more Raspberry Shakes and 12 conventional seismic sensors near the earthquake’s epicenter. The additional devices provided the network with data to better forecast the earthquake’s continuing aftershocks.
Community Is Key
A reliable community-run seismic network is an attractive alternative in seismically active places that lack the government resources needed to maintain conventional seismic stations, said Emily Wolin, seismic network manager at the U.S. Geological Survey’s Albuquerque Seismological Laboratory, who was not involved in this work in Haiti. Wolin pointed to a similar project in Nepal. “I think that these kinds of projects are a really good way to do real-world seismology under all of the constraints that exist in the real world,” she said, “and I think it’s awesome to get this kind of data out.”
Wolin has seen issues in international development projects following earthquakes in which outside agencies contribute new instrumentation and monitoring systems. “Everyone loves to come in and build new stuff,” she said, but “it is much harder to convince funding agencies to keep paying for telemetry and maintenance.” The network in Haiti, in contrast, was developed and is maintained locally and hinges on teamwork between the public and scientists—a collaboration both Symithe and Calais hope will grow.
A Seismic Network More Than Its Instruments
Now that researchers have demonstrated that RS devices can create a reliable seismic network, their next ongoing objective is to help community members become ambassadors for seismic risk in Haiti, Calais said. Bringing sociologists on board helped the team of seismologists provide meaningful information to people hosting a Raspberry Shake, and these experts will continue to help participants moving forward.
It’s vital to understand the people involved in your project, said Laure Fallou, a sociologist at the European-Mediterranean Seismological Centre who is helping expand the project beyond seismology. She and other sociologists on the team interviewed hosts to assess their expectations of the researchers as well as of the project, which helps ensure that the cultural context of the work is not overshadowed by seismic data.
“A seismic network is [composed not only] of the sensors but also of the people,” said Fallou. “We really created a network of people and sensors that are working together to make seismic risk just a little bit less.”
These data also have been used to inform future policy. The data collected from the August earthquake showed that the local ground movement was more than current construction codes could handle. The existing codes were quickly written after the devastating 2010 earthquake and were based on data from regional seismic stations available at that time. Twelve years later, the community network will provide more precise data to improve these codes.
“After the 2010 earthquakes, we had to rely on international scientists to come and tell us what happened,” Symithe said. “But this time, 2 days after the main event, it was local scientists in the field doing the same work. That’s a big improvement.”