ArcGIS API for JavaScript

Air Quality Disparities Are Growing Among Ethnic Groups and Income Levels

A new study by Harvard and Esri researchers has determined that air quality has generally improved across the US since 2000. But it has not improved equally for everyone – namely Black, Asian, Latinx, and low-income populations.

Led by Dr. Francesca Dominici and her colleagues at the Harvard Data Science Initiative, in collaboration with Esri’s spatial statistics team members, Xiaodan Zhou, Jie Liu, and Ting Lee, the study highlights racial disparities in air quality using visualizations made with the ArcGIS API for JavaScript, a powerful mapping tool. The study examines demographic and air quality data from the years 2000 to 2016 to investigate trends for fine particulate levels- air particles smaller than 2.5 micrometers wide (PM2.5)- in more than 32,000 ZIP codes.

The results show that diminished air quality is common in low-income and ethnically diverse ZIP codes suggesting environmental injustice could be a key reason for air quality disparities. Areas where people have fewer tools and money to fight injustice result in a lack of enforcement, regulation, and remediation. 

These maps show majority Black ZIP codes (left) and majority white ZIP codes (right) in colors ranging from dark red to green. Red indicates high levels of PM2.5 and green indicates low levels of PM2.5 for the years 2000 and 2016.

The maps above show air quality levels for primarily Black ZIP codes (right) and primarily white ZIP codes (left). The top maps show PM2.5 levels in the year 2000, while the bottom maps show PM2.5 levels in the year 2016. Notice how, between the years 2000 and 2016, majority-Black ZIP codes saw air quality improve from less healthy red levels to acceptable lime green levels. Meanwhile, majority-white ZIP codes saw air quality improve from red to the optimal dark green levels, marking the most significant air quality improvement. In 2016, the average PM2.5 concentration for the Black population was 13.7% higher than that of the white population. These maps show that while air quality is improving overall across the US, the rate of improvement is not equal and varies based on the racial makeup of an area.

These maps show low-income ZIP codes (left) and high-income ZIP codes (right) in colors ranging from dark red to green. Red indicates high levels of PM2.5 and green indicates low levels of PM2.5.

Similar disparities were observed when comparing income-level. The maps above show PM2.5 levels for low-income communities (right) and high-income communities (left). The tops maps show PM 2.5 levels for the year 2000 and the bottom maps show PM2.5 levels for the year 2016. Notice that, between the years 2000 and 2016, air quality in low-income communities improved from being in the dangerous red range to the acceptable lime green range. Meanwhile, air quality in high-income communities improved from the dangerous red range to the optimal dark green range, making the amount of improvement much greater for high-income communities. While air quality did improve for both communities, these maps suggest that the income of a ZIP code plays a strong role in the amount of air-quality remediation an area receives.

Dr. Dominici smiles at the camera

Even as air quality is improving, it is not improving equally for all people.

Francesca Dominici PhD Harvard University

Most notably, the results also show that as the Black population increased within a ZIP code, the amount of PM 2.5 particles also increased. This was especially true in ZIP codes where more than 85 percent of the population was Black, with a similar trend observed for Latinx populations. However, the opposite trend was observed for predominantly white ZIP codes. As the white population increased in a ZIP code, the amount of PM2.5 in the air decreased.

This figure shows what happens as the percentage of an ethnic population increases within a ZIP code. Notice how as a ZIP code becomes more white (green line) PM2.5 levels decrease, while as a ZIP code becomes more Black (red line), Latinx (light green line), or Asian (blue line) PM2.5 levels increase.

Dominici said seeing the disparities on maps was key to helping readers of the study understand the dire situation, pointing to the partnership with Esri as being particularly helpful.

“By working with Esri, not only have we kept this paper at a very high level of scientific rigor, which is important, we have also taken many steps forward by being able to communicate and visualize our results,” she said.

The team harnessed a variety of Esri tools in the study. ArcGIS Pro was used to explore the disparities of the data. The ArcGIS Pro Project feature allowed the team to handle missing data, resolve spatially misaligned data, and conduct exploratory mapping. The team also utilized ArcGIS Online to share data for cross-team collaboration. Lastly, ArcGIS API for JavaScript was used to create the maps in an informative, and aesthetically pleasing way that was computer memory efficient.

“The visualizations shine a light on relative disparities in a very intuitive and straightforward way,” said Zhou. “Anyone can capture the disparity pattern easily from our maps and animations. We did these with ArcGIS API for JavaScript. It’s a powerful tool for building stunning 2D and 3D visualizations.”

 

Anyone can capture the disparity pattern easily from our maps and animations.

Xiaodan Zhou Product Engineer, Esri

The study was created to be easy to understand for all people, rather than being statistically complicated, Dominici explained.

“The more accessible we make our results, the higher the impact they’re going to have. It’s not about pointing fingers or agreeing and disagreeing. It’s about being able to visualize and see. Look at the data,” she said, throwing up her hands in emphasis.

Dominici urged people with a GIS and statistics background to download the data and look at the code, which will allow them to see step-by-step what was done and potentially apply the team’s methodology to other situations, “not only in the context of air pollution but also in the context of weather, pesticides, climate change, and more.” In addition to open access to the team’s code, supplemental videos are available that show how the team created each map. The study’s code and data can be found here.

“People that live in both clean-air areas and dirty-air areas, now have the visualization of maps combined with a peer-reviewed paper in Nature to advocate for change,” said Dominici. “By keeping the study simple, everyday people can engage with the figures and use them to push for targeted legislation in their area.”

The data can also be downloaded, and individuals can look up their ZIP code to see the disparities in their own community.

“It’s really about mobilizing the entire community to advocate for change,” said Dominici.

To learn more about ArcGIS for air quality analysis explore air quality layers in the ArcGIS Living Atlas of the World, access stories, and discover resources.

About the author

Victoria is an Industry Marketing Manager for Corporate Science Communications supporting Chief Scientist of Esri, Dr. Dawn Wright. She has a background in Earth and Environmental Science from Vanderbilt University. When she isn't marketing, she's passionate about science communications, sustainable fashion, photography, and wildlife.

Connect:
0 Comments
Inline Feedbacks
View all comments

Next Article

Accessibility Sessions and Presentations at Esri User Conference 2022

Read this article