The following text is part of the “Data, Displacement, & Disaster Response: Two Philadelphia Libraries Take Action in the Aftermath of Hurricane María” presentation given at the Penn in Latin American and the Caribbean (PLAC) 4th Annual Conference: Climate Change, Resilience, and Environmental Justice in Latin America and the Caribbean. The presentation was conducted by Penn Libraries Digital Strategies Librarian, Coral Salomón Bartolomei, and Free Library of Philadelphia Community Organizer, Tania Ríos Marrero. The post below pertains to the work performed by Penn Libraries in collaboration with Humanitarian OpenStreetMap.
Data, Displacement, and Disaster Response: Two Philadelphia Libraries Take Action in the Aftermath of Hurricane Maria (view the full presentation)
On September 2017, hurricane Irma and María raged across Puerto Rico, dismantling the archipelago’s infrastructure and energy distribution systems. What began as a natural disaster quickly escalated into a humanitarian crisis as days, weeks, and then months passed without restoration of essential services.
Delivery of food and other necessities in PR was hampered by a variety of ills, including lack of telecommunication services and what has been deemed as a slow and insufficient response from the local and federal government.
Another issue slowing recovery was the lack of geographic data.
Proprietary mapping platforms are more comprehensive in the United States than in PR. Juan Saldarriaga, research scholar at the Center for Spatial Research at Columbia University, explains that this is because companies like Google create their maps for commercial reasons; therefore, areas that have less economic activity are given less attention.
Another issue that responders faced was a lack of centralized geographic data. According to José Oliver-Didier, software engineer and Puerto Rican OpenStreetMap collaborator, geographic information in PR is maintained at the municipal or township level. This regional data is descriptive and in Spanish, instead of geographic coordinate data, complicating coordinated response efforts.
The way Penn Libraries created map data for responders in PR was through the HOT (Humanitarian OpenStreetMap Team) Tasking Manager tool.
Responders in PR issued requests for map data through the tool and then volunteers traced satellite images to create the data–essentially turning static images into usable map data.
The Penn Libraries OpenStreetsMaps response was part of a larger network of university and academic libraries responding to the Red Cross requests in the Humanitarian OpenStreetMap tool.
Libraries hosted “mapathons” events by providing resources, such as wifi, laptops, space, and the ability to train students, staff, and the public. These are resources that we take for granted, but proved to be vital during the crisis. Momentum for these “mapathons” spread through word of mouth, social media, and news coverage.
In the first two weeks and a half since the Red Cross issued its initial mapping request, volunteers mapped more than 700,000 buildings. According to the OpenStreetMap wiki by December 4, 2017 (~2.5 months) all the requested data for PR was mapped and validated by volunteers.
The Red Cross printed the digital data into physical maps that were disseminated to aid workers on the ground in PR every day. FEMA and even the US Army used these open-sourced map. The Puerto Rican government also created applications using the crowdsourced OpenStreetMap data.
According to Oliver-Didier, an example of how this crowdsourced open map data was applied, included determining which residents needed to be evacuated during the breach of the Guajataca dam.
Furthermore, the map data was used in the seminal study, Mortality in Puerto Rico after Maria, published in the New England Journal of Medicine.
According to the study’s primary author, Nishant Kihore, OpenStreetMap offered the most actualized geographic data of PR. Since it is free and accessible, it allows for transparency and seamless integration with the study’s toolkit, which is openly available on GitHub.
Some of the lessons learned, as shared by those interviewed for this presentation, is that it is essential to insert geographic data, such as identifying baseball fields (where helicopters can land) and community centers (where mobile dialysis centers can be set up), before disasters.
These are locations that might not be labelled in a commercial map, but that are indispensable during moments of crisis, especially in places that have scarce map data.
For those looking for takeaways of how this effort can be applied to other research in Latin American and the Caribbean:
- Make sure that if you’re gathering data that it is open and accessible. Publish it on GitHub (it doesn’t have to be fancy).
- Publish your methodology. Make it free, scalable, and replicable. Publish it in Spanish too, or the language spoken in the applicable country, so others in the region can use it.
- Support open data efforts and independent data gathering institutions in Latin America and the Caribbean. This past year in Puerto Rico, we saw how the support of institutions inside and outside of the archipelago prevented our Instituto de Estadísticas (Institute of Statistics) from being shut down by the local government. It is important to be a vocal ally of fact- and data-gathering entities in the region.
Open data is essential. Having authoritative data is important, but it also needs to be centralized and open. Open, high-quality, centralized data saves lives.
Aman Kaur, Girmaye Misgna, & the Penn Libraries team
If you wish to donate to the Free Library of Philadelphia’s campaign for full and just funding for all neighborhood libraries, click here. Your help will support the equitable distribution of knowledge and education across the city.
If you wish to support efforts in PR, please consider donating to ISER Caribe (non-profit focused on environmental conservation, restoration, and education in the Caribbean) and the Brigada Solidaria del Oeste along with its network of mutual aid centers (BSO is a community organization that is currently rebuilding houses destroyed by hurricane María).