Satellites map environmental vulnerabilities in U.S. prisons

Geoscientists are using remote sensing to gather data on risks including increased exposure to air and soil pollution, excessive heat, wildfire, and flooding. 

Incarcerated people in the United States face disproportionately more environmental hazards than the overall population, a pattern of environmental injustice referred to as prison ecology.

Although efforts to recognize environmental injustice have grown in recent years, prisons have been largely overlooked, said Ufuoma Ovienmhada, a graduate student in remote sensing science at the Massachusetts Institute of Technology in Cambridge. “Most people don’t think about prison, period, and even those who think about prison have not always heard about the issue of environmental injustice,” she said. “It’s still underreported and underdocumented.”

Now, researchers are using remote sensing data, supplemented by on-the-ground reports from currently and formerly incarcerated people, to uniformly define the scope of environmental vulnerability across the carceral landscape.

Multiple Axes of Marginalization

In the United States, more than 6,300 carceral facilities imprison almost 2 million people. For years, reports have trickled in from facilities across the country about adverse environmental conditions experienced by incarcerated people, prison staff, and corrections officers. For example, excessive heat in Texas prisons is a recurring issue, air pollution in prisons near coal ash sites has been litigated, and carceral facilities built on landfills such as Rikers Island in New York City report abnormally poor health conditions for prisoners and staff alike.

Prisons are environmental justice communities, said Caitlin Mothes, a geospatial data scientist at Colorado State University in Fort Collins. “They have an overrepresentation of both people of color and people of low income,” she said, as well as few options to move away from unhealthy environmental conditions.

The U.S. EPA began including prisons in its Environmental Justice Screening and Mapping Tool (EJScreen) in 2017. Researchers and policymakers use this tool to identify areas and groups of people that may have a greater need for environmental conservation efforts or have been exposed to undue environmental burdens. However, efforts to uniformly gather data on environmental conditions across the entire carceral system have been slow to gain speed.

“For a long time, prisons were marginalized even within environmental justice research,” Ovienmhada said, “and environmental justice research has been marginalized within the traditional Earth sciences.” This marginalization has led to a lack of data needed to understand the scope of environmental issues facing incarcerated populations.

Mothes said she became aware of this gap through colleagues who were studying forced prison labor in agriculture. Together they realized that they lacked geospatial data on the environmental conditions that forced laborers experience. The group sought to “examine it at all prisons across the U.S. to build a foundational dataset that we…can use and build off.”

Ovienmhada said that the 2020 murder of George Floyd and the national reckoning with racial justice prompted her to shift her research focus. “I learned about this intersection of prisons and environmental injustice, and I felt like the remote sensing skills that I already had could support documenting these hazards,” she said.

Redrawing Pollution Maps

Mothes, Ovienmhada, and their colleagues separately turned to satellite data to characterize the problem. They partnered with currently and formerly incarcerated people and their families, who shared their experiences and testimonies to help the researchers understand which environmental conditions were most prevalent or put people at the most risk.

Ovienmhada’s team homed in on particulate matter (PM2.5) exposure to describe the environmental vulnerability around prisons.

The researchers used a satellite-derived model to track PM2.5 exposure around carceral sites at the 1-kilometer scale from 2000 to 2018 and determine which facilities exceeded EPA’s National Ambient Air Quality Standards. For comparison, EJScreen uses data with a 12-kilometer resolution to evaluate air pollution exposure.

“We found dozens of facilities that would not have been recognized as experiencing that type of air pollution burden when using the more coarse federal data set,” Ovienmhada said. In their analysis of data from 2017, the Baltimore City Correctional Facility, New Jersey State Prison, and FCI Morgantown in West Virginia scored in the 90th percentile for PM2.5 exposure, but were below this threshold in EJScreen. That might affect which areas receive attention under recent environmental policies like the Justice40 Initiative, she said.

Ovienmhada and her colleagues are also currently working with land surface temperature information from several satellite data sets to show how many days per year different prisons experience extreme heat, which has been tied to severe health risks. Ovienmhada will present this research at AGU’s Annual Meeting 2023 on 11 December.

Big Data for Environmental Justice

Mothes’s group combined several indicators into a single environmental vulnerability index that describes conditions at prisons across the country. “This kind of big data hasn’t been around forever,” she said. “We’re in an era where we can now do large-scale analyses for not just scientific research, but also social and environmental justice work.”

They gathered publicly available data on 11 environmental metrics going back around 10 years from several satellites and government databases, focusing on the areas surrounding almost 1,900 state- and federally run prisons.

A climate metric considered heat exposure, canopy cover, wildfire risk, and flood risk. Facilities in California, Texas, and along the Pacific and Atlantic coasts scored poorly because of worsening wildfires, heat, and hurricanes.

Environmental exposure included ozone and PM2.5 levels, traffic volume, and pesticide risk. Prisons in New York City and along the Mississippi River scored poorly in this category—the former because of poor air quality and the latter because of agricultural pesticide use.

The third category included proximity to superfund sites, hazardous waste sites, and nuclear and chemical plants. Facilities across the country scored poorly on this metric because carceral facilities are often built on or near these undesirable sites, for example, Rikers Island and the now closed Northwest Detention Center in Tacoma, Wash.

Mothes’s team combined these three categories into a single environmental vulnerability index to better compare prisons across the country. Thirty of the 50 facilities with the worst overall score were in California, she said.

“We hope this will be useful to other researchers or activists to highlight which prison we should focus on, write stories on, investigate further,” Mothes said. “But it’s also important to look at individual environmental burdens, too. Some prisons could be high-risk because they’re located close to so many different toxic facilities, whereas others might be high-risk for heat and climate risks.”

For example, the facilities that scored poorly in California did so because of climate risk and poor air quality. When the team examined the 50 facilities with the worst environmental effects because of their proximity to toxic sites, only two were in California, whereas 14 were in Maryland and seven were in Florida. Mothes will present this research at AGU’s Annual Meeting 2023 on 13 December.

Creating a Nontoxic Carceral Landscape

“I am delighted to see scientists developing and applying these tools to advance our knowledge of the range of environmental threats facing incarcerated populations,” said David Pellow, an environmental scientist at University of California, Santa Barbara and head of the Global Environmental Justice Project.

“These studies provide powerful visualization tools that can facilitate a deeper understanding of the challenges of environmental injustices facing incarcerated persons, corrections officers, and others working and living in and around carceral institutions,” said Pellow, who was not involved with these projects.

Mothes’s group has made these data sets public and has also released the code they used to calculate their environmental vulnerability index. Researchers can generate new maps of environmental vulnerability as they gather updated environmental data or data covering other carceral facilities like county jails, Mothes said.

Ovienmhada’s team is working to codesign a geographic information system (GIS) that integrates environmental and socioeconomic data to aid prison ecology researchers, policymakers, and activist organizations.

Pellow said he was particularly encouraged by the scientists’ partnerships with groups such as Campaign to Fight Toxic Prisons, which works with and advocates for incarcerated persons and their families. “The reason why these projects have the potential to create real social change is precisely because of the collaborative nature of this work between scientists and laypersons,” Pellow said.