Looking From Space, Researchers Find Pollution Spiking Near E-Commerce Hubs
Joe Berberich
Link to Paper: https://www.nature.com/articles/s41467-024-50000-0
Link to Article: https://www.nytimes.com/2024/07/24/climate/e-commerce-truck-pollution.html?searchResultPosition=3
Globalization and the rise of e-commerce have changed the way consumers purchase goods. Corporations like Amazon have constructed warehouse receiving, sorting, and fulfillment facilities and enlarged delivery vans and semi-trailer fleets (Amazon alone has over 200,000 vehicles of these types) to address consumerism in the digital age. These facilities increasingly densify into warehouse hubs filled with thousands of loading docks and parking spots. The increase in warehousing and goods movement increases traffic-related pollutants like NO2 and fine Particulate Matter (PM) associated with pediatric asthma development, cardiovascular disease, and premature death. To make matters worse, the health effects disproportionately affect low-income and minority communities. Prior concerns of environmental racism emboldened policymakers, like Kamala Harris in 2011, to establish NOx and diesel PM reductions in Southern California; New York and Colorado are considering similar legislation.
A recent article by Hiroko Tabuchi, published in The New York Times on July 24, 2024, discusses a recent scientific paper that examines the spatial relationship between warehouse hubs and heightened NO2 emissions across the contiguous United States. While previous studies allude to an overrepresentation of minority communities within 1 km of warehouses, they often are narrowly focused (one warehouse) or limited to a handful of states. In the paper by Kerr et al. (2024), they conducted a nationwide study on air pollution and environmental inequities using geographic and remote sensing data to generalize their results to nearly all people in the United States.
The paper used numerous datasets to investigate the relationship between NO2, warehouses, and demographics. CoStar, a commercial real estate database, identified ~ 470,000 properties that were then constrained to 149,075 warehouses due to size requirements (> 20,000 sq. ft). CoStar provided warehouse square footage, parking spaces (delivery vans), loading dock numbers (semi-trailers), and the year of construction. The authors used remote sensing NO2 data using the TROPOspheric Monitoring Instrument (TROPOMI), that was validated by the EPA's NEMO NOx dataset of anthropogenic emissions. TROPOMI is a satellite that measured tropospheric columns of NO2 at 1:30 pm local time as it overpassed locations. The native grid-size resolution (3.5 km x 5 km) was improved (1km x 1km) through oversampling. A 7 km arbitrary grid buffer surrounding the warehouses was created to catch downwind warehouses. Finally, the authors used census tract information from the American Community Survey (U.S. Census Bureau) to collect racial and ethnicity information statistically quantified by least squares regression and the nonparametric Spearman’s rank correlation coefficient (r).
The paper found, over the past decade, increases in warehouse construction (117%), building area (147%), loading dock area (400%), parking space area (158%), and clustering (66%). Clustering is evident as 0.3% of counties situate 20% of all warehouses in the United States. There is more local traffic that impacts air pollution with greater and larger warehouses. NO2 levels were associated with the increase in warehousing, showing a 17.9% increase on average across the United States. Importantly, single warehouses showed less of an increase compared to the largest clusters (11.5% vs. 23.1%). Clustering explains nearly 40% of NO2 variance near warehouses in low-population-density locations (less evident in high-population-density locations that correspondingly have higher anthropogenic emissions). The increase in NO2 emissions is experienced differently regarding demographics. Near the largest warehouses clustering, racial-ethnic minorities make up a greater fraction of the local population. For example, as a fraction of those who live near warehouses, the Hispanic population is over-represented by 2.4x near the largest warehouses (28.1% from 8.3%). As a result, the Hispanic population experiences higher NO2 levels (29.1% increase on average). This result is similar for Pacific Islanders (2.9x with 6.2% from 1.6%) and more than one race (2.6x). The authors found that the African American population encounters slightly less NO2 level increases than the average across the United States (17.9% increase). This initially surprising result is attributed to the high African American population living in the southeastern United States where NO2 concentrations are lower. To close their paper, Kerr et al. (2024) discuss the numerous policies and stakeholders that can address NO2 emissions. They discussed emission standards for new vehicles (federal), the expedition of EV charging stations (state), low-emission/environmental zones (local), and replacement of older trucks, and the commitment to electrify truck fleets (corporate).
The New York Times article efficiently summarizes the scientific paper in its tagline, "Research showed truck-related releases of nitrogen dioxide, ..., concentrated around some 150,000 warehouses nationwide." From the tagline, readers quickly understand that truck emissions near warehouses are responsible for air pollution. Still, the article has limitations. The author only mentions the overall increase in NO2 level; it does not distinguish between singular warehouses and highly clustered warehouses. It also rounds the NO2 increase from 17.9% to 20%, slightly dramatizing results. In addition, although the article mentions the 2.5x proportion of Hispanic individuals living near the largest clusters, they ignore the 2.9x increase for pacific islanders, the 2.6x for more than one race, and the slight decrease for African Americans. Finally, the article does not discuss the numerous policies that can combat NO2 emissions — it only mentions one piece of legislation coming out of California in 2011.
I believe the article faithfully reproduces the scientific paper despite some discrepancies. It does a great job linking to the source material and being readily digestible. In less than 600 words, it summarizes the main points of a scientific paper ten times its size. It is approachable and consistent with the scientific findings. The issue presented is its limited scope. Tabuchi's article does not describe the nuances between warehouses and clustered warehouses, the percentage increase in size and scale of warehouse operations, or the higher (or lower) levels of NO2 for racial/ethnic groups other than the Hispanic population. Some information must be lost to condense a paper of this scale. With that in mind, I rate it an 8.5/10.
I completely agree with your statement that the article faithfully reproduced the scientific paper with minor subtractions from the supporting information. It is a surprisingly short article but does a nice job of summarizing the paper and relating it to current(past) events. It makes sense that they choose to not include some data and dramatize some numbers due to their targeted audiences, and the more shocking conclusion they have, the more views they will get. I like how they link not just to the paper but also to examples they bring up so that the reader can easily access some background reading or if the readers want to learn more about it.
ReplyDeleteReally good point Richard! Online sources are incentivized to dramatize somewhat to improve engagement. I agree that the author struck a balance between slight exaggeration and the published science in such few words. You bring up a good example in which the author linked the background readings regarding the lawsuit initiated by Kamala Harris and the California Law on e-commerce pollution. Audiences interested in the topic can easily continue their learning beyond the original article.
DeleteThe publication by Kerr et al is a highly nuanced study, and I think that Tabuchi was able to summarize the key points in a concise way. She was able to provide the context behind warehouse production as it was needed throughout her writing, instead of isolating the background information in the beginning as in the Nature paper. She was able to effectively address some of the limitations of the study, such as the lack of data on other pollutants produced around warehouses in the United States. However, Tabuchi only mentioned that Kerr et al performed their study with data from the Federal Highway Administration and the Census Bureau. A majority of the data on warehouses came from a commercial real estate company called CoStar. I believe this should have been discussed as it has the potential to introduce bias into the study. Kerr et al even stated:
ReplyDelete"CoStar does not provide information on quality control or assurance, and it is difficult to verify the spatial distribution and characteristics of warehouses from CoStar as we know of no other comprehensive dataset–either publicly available or proprietary–that provides information on warehousing."
In my opinion, this does not detract from the quality of the research by Kerr et al, but if I had skimmed the New York Times article, I would have appreciated a truthful discussion of the credibility of the data used in the study.
What a thorough dive into the data, Genevieve! You are right in addressing the potentially problematic source with CoStar. The CoStar data has severe limitations, mentioned in the paper:
Delete"Some of the warehouses are missing data on these characteristics, such as the year built (9.9%), the number of loading docks (40.7%), and the number of parking spaces (31.1%)."
There is some inherent bias in using a source that lacks some of the key data points used in the study. But as you mention, it should not detract from the quality of the research. Kerr et al. defended their research by saying:
"We note, however, that CoStar is the most extensive commercial real estate database in the U.S., undergoes vetting and updates by their team of researchers, and is commonly used in academic research"
Sometimes, the best and most appropriate data is not available. In this case, I believe that Kerr et al. used the most with what they had. Still, to your point, the NYT article never mentioned CoStar, so the audience may have assumed the data came from the Federal Highway Administration, the Census Bureau, or satellite data (all of which would be false). The audience may have never understood the biases that existed in the dataset. Why Tabuchi intentionally described the article this way confuses me as well.
I agree with everything you said. The article did a god job generalizing the data for the target audience. The scientific paper was much more in depth, but I feel like it was a good idea to not include all of the advanced data. When general articles try to do this, the data may become misinterpreted by the audience which can lead to consequences that were not intended. In conclusion, I believe that your rating of 8.5/10 is accurate because the article did a good job at generalizing the data.
ReplyDeleteI agree that the article portraits the main concepts of the scientific paper. The article is to some degree biased because it keeps out certain information. However, I think some of the information was not included to keep the ideas centralized. For example, if the article had talked about a decrease NO2 exposure in African American communities, the argument that minority are disproportionately affected would be confusing. While this is again a bias to support the conclusion, I think include all the excess information would distract average Americans and cause them to not believe conclusions made from the scientific paper. I did enjoy that the article contained the extra example from Kamala Harris that was not in the scientific paper.
ReplyDeleteFair point Nico. The exclusion of a decrease NO2 exposure in the American American community was done intentionally to avoid confusing a general audience. When presented with damning evidence and one surprising exception (which can be explained away by removing the southern states as Kerr et al. explain), people predisposed to rejecting the argument would cling to the exception to "prove" their case. This is a very valid argument that you make and one that I did not consider in my rating. In my rating, I focused on the need for science to accept nuance rather than ignoring data points that authors do not like. Still, your point is valid — maybe it is best that the NYT article avoided mudding their conclusions with an exception.
DeleteThe NYT article does do a great job summarizing considering how detailed the scientific article is and there isn't really an expectation for it to cover every single detail, rather gather the main points for its target audience (The average American) which it does. I think also by design, Tabuchi draws you in to reading the Nature article which, I would argue, should be the ultimate goal with this type of article. Many would probably agree that the Nature article might not be digestible by the average American, but I think even just reading the abstract, introduction, or discussion gives a ton more information that otherwise would've gone unnoticed.
ReplyDeleteGenevieve also makes a really great point that the NYT article should include a disclaimer about the dataset in use and that it may not be completely accurate. This is very important because, especially in our current technological age, transparency is extremely important and Tabuchi could have easily included a short discussion of the dataset without damaging the impact of the study.
I see where you're coming from Theo. I agree with you wholeheartedly that the NYT article summarizes the piece nicely and avoids getting into the nitty-gritty details for intentional and reasonable reasons. However, I disagree with your assessment that Tabuchi draws the reader into the Nature article. For a scientifically literate audience, maybe they would read the actual paper. However, I believe that it is the role and perhaps the responsibility of news authors (like NYT) to summarize scientific work to a different (non-academic) audience. The target audience of the NYT is, in my opinion, a more general audience (like teachers, politicians, people in finance, or even high school students). It is most likely that these individuals will accept the article at face value and share the (limited) information with others. In that lens, the article slightly errs since it did not include the nuance or the CoStar disclaimer, as Genevieve noted.
DeleteI agree that the article did a good job translating a long scientific paper into an easy to consume newspaper article. I also agree with your criticism on not focusing on the increased pollution around clustered warehouses. I disagree that It is an issue that so much finer details were smoothed over in the news article. Finer details such as the exact effect on each minority group is not necessary information that needs to be conveyed to a layperson consuming news.
ReplyDeleteSome criticism about the paper and the news article is putting all of this pollution solely at the foot of E-Commerce. I do not doubt that E-Commerce has had a significant effect on air polution and increased warehouses, but I feel it is inaccurate to say that every warehouse is an E-Commerce warehouse. As well, I agree that the paper could have talked about country wide initiatives to regulate warehouse, or the lack of such initiatives, instead of just focusing on California.
I see where you're coming from acknowledging that not all warehouses are e-commerce warehouses. To the authors' credit, the original scientific paper includes only warehouses above 20,000 ft in size. Other than perhaps military warehouses (which may have been a category in the CoStar data), I do believe that it is accurate that these would be e-commerce. A lot of buildings codified as "warehouses" necessarily store goods. Even if they are a warehouse for a brick-and-mortar store, the items must be moved via truck or semi-trailers. As a result, nearby pollution at these locations is an intriguing scientific motivation. Further, with the rise of Amazon and e-commerce in general, I do not think it is as misleading to connect warehouses with e-commerce. Maybe it would have been better to say that consumerism is a more relevant driver of higher NO2 levels, and maybe if the study was conducted in the early 2000s the same result would have been found (as globalization would have already occurred). In short, you are right; specifying e-commerce may be a bit misleading. What I think is undeniable is the evidence that consumerism has increased the construction of warehouses and has increased localized pollution for socio-economically disadvantaged groups.
DeleteThis comment has been removed by the author.
ReplyDeleteI think you did a great job summarizing the main takeaways here, and I agree that the NYT article did a great job of conveying the findings from the Kerr et al. study. I am very impressed that Tabuchi was able to write such an effective and concise article, but I do think the discussion of socio-economic impacts was a bit lacking and that the rounding of the NO2 increase from 17.9% to 20% was unnecessary as 18% would be just as digestible to the reader, making this seem like an exaggeration. With that being said, I do think this article was able to convey the key points in layman's terms and was able to emphasize why a study like this is important by discussing policy issues.
ReplyDeleteGreat work summarizing both the NYT article and the Kerr et al. study, as well as discussing the strengths and weaknesses of the NYT reporting of the findings. I agree that the Tabuchi article was generally successful in conveying the main findings of the research paper, but left out a few key details. I was particularly surprised that they chose to only talk about the impacts of NO2 for just one racial/ethnic group. As a reader, it leaves me questioning whether other racial/ethnic groups are also impacted by NO2 near E-commerce and why they chose to not report on it. You did a good job concisely reporting a more holistic view of these statistics, which I feel would have benefitted the NYT article. I am similarly disappointed that the only piece of legislation the Tabuchi article mentioned was over 10 years ago. I have to imagine there have been more recent policies related to combatting NO2 emissions - I would curious to learn more about those.
ReplyDeleteI agree with your point that the NYT article overall does a good job with presenting a summary of the main ideas reported in the scientific paper! I want to point out (since i have yet to see anyone mention this) that although the rounding of 17.9% to 20% was generally unnecessary, I believe that the author of the NYT article rounded it to 20% due to the abstract of the Nature article stating that the increase was “nearly 20%”. After I noticed that in the abstract, it made a lot more sense why Tabuchi decided to go with 20% over 18% when rounding up. However, I think it would have been much better for Tabuchi to have included some indicator that the percentage was not actually 20% just like how the abstract used the terms “nearly”, because it definitely comes off as a slight exaggeration of the actual findings and puts into question other findings mentioned in the article. In addition, I was also disappointed that the wider impact of NO2 emissions were not touched on besides the one piece of legislation mentioned. I feel like it is helpful to include the motivation behind the study, significance of the results, and wider impacts in order to better contextualize them, especially if you are unfamiliar with the field. I felt that the article did not do an amazing job of letting the readers know why this study is worth reporting on and what is or could be done knowing these findings.
ReplyDeleteExcellent blog post. I agree that this article was able to efficiently summarize the main ideas from the scientific article to the general public. An aspect you brought up that I also would liked to have seen a little more from the article is a nod to additional policies/methods for reducing pollution levels. To me it seems that many times, especially with respect to the climate, the news only focuses on how terrible everything is (which is obviously good to know...) but not on highlighting the possibility for change or empowering people to make change/have optimism. I think reminding readers that there is hope to make things better is a good thing to include.
ReplyDeleteGreat job on the post and great selection. I really appreciated learning about these clusters of warehouses that bring increased pollution from truck traffic. I agree that the article did a good job summarizing the paper in a concise but overall accurate way. I really appreciate the accessibility of the information in the article. Because the article is so short and concise it allows the reader to take away the biggest picture items from the paper. I do think the article could be improved by adding information about TROPOMI. I appreciated how you commented on this satellite instrument but I think the general public would benefit from learning about it too. I just learned about it in another class and I was so excited that there is daily data available on air pollution around the world. I think this is an amazing resource that everyone can benefit from learning about.
ReplyDeleteFantastic work on the blog entry. I think the article provided a valuable overview and is slanted toward highlighting the necessity of a call to action to reduce global emissions, I believe it offers a useful summary. On the other hand, the paper goes into great depth and provides all the material required for readers to understand the complexity of the research. The essay goes into great detail about how crucial it is to comprehend and manage air quality. And given the way they presented the problem, I totally agree with the 8.5 rating.
ReplyDeleteI think this article is quite interesting. The paper itself does a great job at collecting data from points around the globe and shows the true impact that these warehouses can create and for what communities they affect. I would say the article does a decent job at describing these facts but does fail to mention some pretty important facts such as the specific demographics that are effected, the percentage of air pollution, as well as the ability for change. While it does make attention to the cause I could see it approached slightly different.
ReplyDeleteGreat work on the blog post! You did an excellent job summarizing the key finding of the paper and the article, as well as effectively highlighting the strengths and weaknesses of the article. I think this was an excellent choice of article in particular. Although I agree that it omitted some significant findings such as not distinguishing between single warehouses or clusters as well as the the slight decrease in African Americans living in near clusters, I think it overall did a great job of summarizing and condensing the paper into a decently digestible read for someone that is not in specialized in this topic. The topic of environmental justice is immensely important, so the more reach that information such as this can get, I think the better. Overall, I agreed with your rating and commend you on a job well done.
ReplyDelete