Did You Feel It? Expanding use of an earthquake crowdsourcing tool
The platform that rapidly maps earthquake damage can collect more global shaking intensity observations by supporting more languages and increasing social media traffic.
The U.S. Geological Survey (USGS) crowdsourcing platform Did You Feel It? (DYFI) rapidly transforms people’s earthquake shaking intensity experiences into detailed maps of damage extent. While the tool’s reach is global, language and technology barriers prevent participation in certain areas, according to a USGS and University of Michigan Engineering study published in Seismological Research Letters .
USGS has already begun implementing some of the study’s recommendations.
Within minutes of any earthquake measurable by sensors, DYFI responses help refine the USGS ShakeMap —a near-real-time map of ground motion and shaking intensity. About 20 minutes after the earthquake, these maps feed into the Prompt Assessment of Global Earthquakes for Response system, or PAGER for short, which combines shaking data with local population statistics to project fatalities and economic losses.
Related content: Turning data into resilience – measuring what truly matters
📊 To build a sustainable future, we need more than good intentions — we need strong data.
Domestic and international disaster response agencies rely on DYFI, ShakeMap and PAGER data. FEMA utilizes loss estimates and the Department of State’s International Disaster Response (IDR) Office turns to the tools when considering whether to send aid. In all, hundreds of NGO aid agencies, foreign governments and financial decision-makers rely on these products.
Tool engagement and barriers
In the first global assessment of DYFI engagement, the researchers used a combination of web analytics and a panel regression model. The team evaluated 52 variables, including earthquake parameters and socioeconomic data, linked to 1,758 earthquakes across 151 countries from 2009 to 2020.
The authors found that users predominantly accessed the platform through mobile devices—83% of international users and 78% of domestic users—often driven by social media referrals. In particular, the team was struck by the tremendous social media-generated engagement following a 5.9-magnitude earthquake in Chiba, Japan in 2021. A single tweet from a resident was largely responsible for the event’s 17,429 regional users, the greatest number for any event outside of the United States.
“It was impressive to see how much a single citizen in Japan influenced the response rate following the Chiba earthquake with an X post. Many people commented about their use of translators to fill out the form, and it was exciting to see their interest in documenting their experience,” said Eli Knodel , a research assistant in civil and environmental engineering at U-M, geospatial data scientist with the Eastern Band of Cherokee Indians, and lead author of the study.
When assessing the physical and socioeconomic drivers of DYFI participation, shaking intensity held a significant sway, as expected, but primary language spoken and access to broadband internet also impacted response rate.
“Crowdsourcing tools give the power to the people to help refine earthquake maps and damage estimates. Making the tool available in the local language can scale up accessibility to people in countries that need it most,” said Sabine Loos , an assistant professor of civil and environmental engineering at U-M and senior author of the study.
Expanding accessibility
To identify which countries would increase engagement the most with improved accessibility, the team developed a priority metric weighing six key factors: historical response rates, language, broadband access, shaking intensity, population exposure and seismic station density. The index prioritizes areas with low seismic station density, where crowdsourced observations are vital for accurate shake mapping.
The metric found that non-English speaking Asian and Central European countries are the highest priority areas for outreach, with China ranking highest followed by Germany and Indonesia. These countries have limited available physical sensors and large populations that are exposed to earthquakes, while lacking a supported local language on the platform.
Following the study, the USGS is already evaluating system performance on mobile devices and incorporating new language options including Chinese and German.
“The results of this analysis, in combination with ongoing bilateral relations, have allowed the USGS to prioritize which nations around the globe could most benefit from language options and collaborative outreach efforts,” said David Wald , a USGS research geophysicist and co-author on the study.
Moving forward, targeted outreach through social media could also help expand the use of DYFI or other crowdsourced citizen science tools designed for natural disasters.
“Improvements to the tool will benefit populations exposed to earthquakes as well as the scientific community’s understanding of ground motions,” said Knodel.