Source emission contributions to particulate matter and ozone, and their health impacts in Southeast Asia
By: Lily Press
Research paper: "Source emission contributions to particulate matter and ozone, and their health impacts in Southeast Asia"
Background
Southeast Asia (Brunei, Myanmar, Cambodia, Timor-Leste, Indonesia, Laos, Malaysia, Philippines, Singapore, Thailand, and Vietnam) is currently in a period of industrialization. As a result of this anthropogenic activity, the region has seen an increase in emissions and air pollution (1). While historically biomass burning has contributed to emissions in the region, industrialization has resulted in a rise of NOx, CO, VOCs, and SO2 emissions from vehicles and shipping (1). Of particular note has been increases in PM2.5 (particulate matter with diameters of 2.5 microns and smaller; fine inhalable particles) and ozone (2). This is owed to their documented deleterious health effects when in the troposphere, such as respiratory problems, cardiovascular impacts, and premature mortality (2).
However, the sources of these pollutants, their concentrations, and resultant impacts on health in Southeast Asia have not been thoroughly investigated, complicating mitigation strategies. As the region is diverse in sectors, environment, and population, different regions have different emission profiles (1). Transboundary air pollution (TAP) is quite prevalent in the region, complicating knowledge on where pollution originated (3). TAP is defined as pollution that travels out of the borders of the country in which it was generated, with regional TAP being TAP that remains within the geographic region, and super-regional TAP extending across long-ranges (1). As such, the peer-reviewed article aimed to determine sources and health effects of these pollutants.
News Article
The article begins by introducing the study, saying that the study suggests regulation on air pollution is a route that could majorly reduce air pollution regulation by 2050. The authors then go on to introduce what ozone is, and why we don't want it in our troposphere. In particular they mention the major sources of local ozone emissions that the paper found: biogenic and road transport emissions. Biogenic emissions are defined for the reader as emissions from natural sources, and the confounding factors of TAP on air quality are also mentioned. The author in the paper then mentions a call-to-action that the paper had, which called for greater collaboration on air pollution mitigation. The article concludes with support for continued research on this subject, highlighting the direct connection between in depth studies like these and our ability to mitigate and prevent the impacts that are observed.
Research Paper
The authors' goal here was to determine where PM2.5 and ozone observed in SE Asia in 2018 came from, and to determine the health impacts as a result of their presence in the region.
To model the distribution and origin of these pollutants, the authors used a known chemistry/meteorology transport model system, WRF-CMAQ. This model system allowed the authors to look into the transport and dynamics of PM2.5 and ozone emissions/formation. It involves both weather modeling and chemical transport. In addition to this model system, the authors formulated concentration-response models (CRMs) to evaluate how different emission sources impact local PM2.5 and ozone concentrations, and extrapolate that towards health effects.
After confirming the validity of their model in section 3.1, the authors in section 3.2 evaluated regional distribution of target compounds during a given time period-- specifically, the four monsoon seasons-- and averaged concentrations across these four periods. This was in an effort to elucidate spatial distributions of pollutants, decoupled from temporal changes as best they could. It was found that the highest PM2.5 concentrations were in industrialized areas, namely, cities. Local industry was a major contributor to PM2.5 concentrations. With respect to ozone, the highest concentration was seen in the suburbs of Jakarta, Indonesia. The main ozone sources were from road transport and biogenic activity. Interestingly, major cities saw depressed ozone concentrations. This resulted from VOC-depletion regimes and elevated NOx concentrations facilitating reactions with ozone.
After evaluating regional distributions, seasonal effects on target pollutant concentrations were evaluated so that mechanisms of formation could be determined. PM2.5 concentrations were found to be highest in the first two monsoon seasons, likely from fire-emissions during this time period. In coastal regions, such as the Phillipines, sea spray emissions also contributed massively to PM2.5 concentrations, with concentrations increasing during windier seasons. Notably, despite the different geographic locations, Singapore had seasonal variation in PM2.5 concentrations similar to Indonesia and Malaysia, and it was found that 51.5% of PM2.5 were from TAP. Ozone was also observed to be higher during the first two monsoon seasons for mainland SE Asia. Biogenic emissions provided a major contribution to these elevated concentrations, supplying 40-50% of observed ozone. In Singapore, an urban area, ozone was depressed during the latter two monsoon seasons. During this time, VOC-limited regimes and higher NOx concentrations resulted in ozone concentration depression and PM2.5 increases. For some coastal SE Asian countries, ozone concentrations were relatively independent of season.
After evaluating sources and concentrations of emissions, premature fatalities induced by PM2.5 and ozone pollution were evaluated for each country in the region with respect to local emission source (Figure 1). Of the 899,000 fatalities induced in 2018 in SE Asia, 77% were attributed to local emissions. Urban centers saw higher rates of premature fatalities, largely from PM2.5 emitted via industrial or residential activities, though some countries did see notable contributions from fire as well. Elevated PM2.5 proved more fatal than ozone in this study, with PM2.5 and ozone being attributed to ~66% and 10% of total fatalities, respectively. Every single country had more PM2.5 induced fatalities compared to ozone. This disparity was a result of ozone having less risk per concentration unit than PM2.5 (this was determined from the concentration-response curves). Additionally, ozone fatalities depended on precursor emissions, specifically, NOx and VOCs. As a result, biogenic emissions were a major contributor to ozone deaths (30.1-64.4%, depending on the country, and excepting Singapore), as they supplied VOCs for ozone formation. Road transport emissions were another major contributor to ozone-related deaths, as these emissions were both high concentration and the public were more frequently exposed to them. Road transport accounted for 5.9-40.1% of ozone deaths, depending on the country.
The contributions of transboundary air pollution to premature fatalities were evaluated in a similar manner, but sources of the TAP were not identified (Figure 2). 206,000 of the 899,000 pollutant-induced premature fatalities (22.9%) were a result of TAP. Over half of ozone-related mortalities in SE asia were from TAP, rather than from local emissions. Super-regional TAP showed larger impacts than regional TAP for ozone. This was explained by the long lifetime of ozone and of the precursors that form ozone. An additional confounding factor was that the the most populous SE Asian countries are bordered by sea, resulting in greater opportunity for super-regional TAP in these areas (2x greater TAP occurs here compared to global average), and thus exposing larger populations to ozone. For PM2.5, TAP generally did not contribute as much to premature fatalities as local emissions did, though there were exceptions observed with East Timor and Singapore, both of which had ~40% of premature fatalities from regional TAP.
Analysis and Criticisms of the News Article
The news article does a great job of taking the information they provide and making it digestible for the average person. They define unknown terms and concepts, and clearly illustrate why what they're pointing out matters. They mention one of the major findings of the paper: locally-emitted ozone that results in premature fatalities is largely emitted from biogenic and road transport sources. The article does a great job of presenting a positive perspective on the paper as well. Despite these positives to the article, I had many many issues with it. I am not convinced that the author of this article did more than briefly skim the research paper, and I would not be surprised if they did not even read it. The news article describes a study and result from this paper that never occurred and was not a finding: that ozone related deaths could be significantly reduced by 2050, and that multiple different scenarios were looked at to determine this. I think there is a chance that the author did not link the correct paper, and that this may have been what happened, but it is unclear. The most important issue of the research article is the determination of sources of PM2.5 and ozone emissions, and the resultant impacts on human health, but the author of the news article does not touch on the work done to determine this.
While I do understand having some trouble communicating the article, considering it had an enormous amount of data to sort through, and the major findings were hard to determine as a result, the news article author completely missed the mark on this one. The major findings of this paper were the following:
1) PM2.5 is the leading cause of air pollution exposure-related premature fatalities
2) local PM2.5 emissions resulting in premature fatalities were largely from industrial and residential emissions, while for ozone it was largely from biogenic and road transport emissions
3) fatalities related to local emissions were largely a result of PM2.5 exposure, while fatalities related to TAP were largely a result of ozone exposure
As the paper did not communicate this information, only focused on ozone, and barely mentioned any of the main takeaways, I give this article a 3/10 (points retained for its effectiveness at defining foreign terms for non-scientific readers, and highlighting the importance of research for policy making and impact mitigation).
Citations
(1) Gu, Y.; Fang, T.; Hung, S. Source Emission Contributions to Particulate Matter and Ozone, and Their Health Impacts in Southeast Asia. Environment International 2024, 108578–108578. https://doi.org/10.1016/j.envint.2024.108578
(2) United States Environmental Protection Agency. Particulate matter (PM) basics. United States Environmental Protection Agency. https://www.epa.gov/pm-pollution/particulate-matter-pm-basics.
(3) US EPA, O. Transboundary Air Pollution. US EPA. https://www.epa.gov/international-cooperation/transboundary-air-pollution.
I'm glad you pointed out that the news article reported a finding that is not a true conclusion of the journal article. Actions like this by this journalist and others can be very dangerous, as many people are not going to click on the scientific article and fact-check the journalist. Even if they tried to read it, scientists often write with so much jargon that their papers become inaccessible to a general audience with little to no scientific background. A news website--which most people would consider trustworthy--presenting incorrect information is how misinformation spreads. This is why journalists and science communicators are so important and should always accurately represent the work they are sharing with the public. Great summary and analysis!
ReplyDeleteI think it's shocking how the news article claimed results and further steps that were simply not mentioned in the study they linked. I agree that it's likely they linked the wrong source, or that they had an additional source that spoke on regulation reducing ozone-related deaths by 2050 that was not properly sourced. This study, while data-dense, is very clear that its focus was on finding the source of ozone and PM2.5 emissions, and linking those to deaths in the region, and not as focused on "next steps".
ReplyDeleteThe model simulations used to calculate and understand the likely health impacts of the ozone and PM2.5 exposure rates in SE Asia are fascinating and also confusing in exactly where they drew certain values from, most likely because I'm personally unfamiliar with how the algorithms work. In these circumstances, effective science communication becomes really essential to distill it to a general person, which is where the author's poor article becomes that much more disappointing.
Wonderful analysis and summary!
Thank you for the thoughtful comment! I was also fascinated/confused on where they drew values from. I did my best to understand, and read some of the cited papers in the process, and will try to communicate how I understand it. The authors began by developing a regional chemistry transport model. They did this by combining known datasets and models, specifically: “FNL reanalysis” (which represents the day-by-day history of weather in the region, derived from combining forecasts and observational data), land cover data, a known global-chemistry model, air quality data, emission data, and meteorological observations. This established a regional chemistry transport model informed by weather, pollution, and potential reactivity. So cool, right? However, its the ability of equations to govern these models that eludes me– it seems like there are so many confounding variables that modeling on such a large scale must be nearly impossible. Though, it clearly is not! Once the authors have their model, they are able to “tag” their pollutants of interest using a tool called “ISAM,” which is specifically for apportioning sources. How this works, again, is another point of confusion for me. These algorithms allow for the monitoring of transport and reactivity of both the pollutants and their precursors. Once the emission sources, amounts, and transport are understood, then information on fatality incident rates and populations could be factored in. This allowed for the development of their concentration-response model. To the best of my knowledge, it is the understanding of the “baseline” spatiotemporal variability in these pollutants in conjunction with knowledge of emission sources and transport dynamics that allow for attribution of fatalities to particular emission sources.
DeleteGreat job, Lily. I'm very glad you caught the article's inconsistences. I find it interesting that all I see is ads when I open up the news site. This leads me to believe it is not the most credible source. The paper is very interesting. I did not know that natural events contribute as heavily as they do to pollution. How do sea spray emissions increase the PM 2.5 concentrations? Also, what caused the fire emissions that led to the PM2.5 concentrations to be highest in the first two monsoon seasons?
ReplyDeleteGreat questions! Sea spray emissions include the release of both sea salt ions (sodium, chloride, potassium) and ions originating from the crust of the earth (calcium, magnesium). Since these ions are in solid form, aka particulate matter, and they are often <2.5 microns large, they fit the definition of PM2.5. This was surprising to me as I had not previously thought much about what sea spray emissions would consist of. The fire emissions being raised during the first two monsoon reasons is a result of multiple factors. The first being “burning season,” which varies from country to country, but could generally be attributed to farmers burning old crops to clear land for the next seasons planting (1). Since this often occurs in March/April (Spring), this coincides with the 2nd monsoon season. Additionally, the 1st monsoon season, in January, occurs at a similar time when cooler and dryer air is in the region (1). This facilitates spontaneous fire development as well.
DeleteCitations: (1) It’s Fire Season in Southeast Asia. earthobservatory.nasa.gov. https://earthobservatory.nasa.gov/images/91771/its-fire-season-in-southeast-asia.
This analysis was great! It was interesting to see how far off the popular science article could be. The news article was very short and not very in depth.
ReplyDeleteI noticed that the only health impact the study looked at was premature mortalities. These were moralities related to respiratory and cardiovascular diseases. I wonder if there is also data on people still living with these respiratory and cardiovascular diseases how much wider the scope of damage would be these people were considered as well. I also would be interested in seeing how the deaths are determined to be related to PM2.5 or ozone.
I can only imagine that the scope of damage is significantly wider than is represented by the mortality data. The amount of math that goes into modeling these scenarios though is mind-boggling. While I wouldn't be surprised to learn of a study modeling the impacts on living people, it seems to me that tracking this would be harder, making modeling it even more challenging.
DeleteI was able to figure out some more information on how they determined mortalities though. Ozone fatalities were determined by assuming that a linear increase in ozone concentration results in an exponential increase in RR ("relative risk"). This means that an increase in ozone concentration increases the risk of health effects exponentially, according to the model. Then, the calculated RR is plugged into an equation called a CRM (concentration-response model). This equation uses baseline ozone concentrations, the population count, and the determined RR to calculate the number of premature fatalities.
Determining this for PM2.5 was more complex as you cannot assume that a linear increase in PM2.5 concentration results in an exponential increase in RR. This is because
below a PM2.5 concentration of 2.4 micrograms / m^3, one can assume an RR roughly equivalent to 0. For ozone, any concentration is a risk-inducing concentration-- this is not the case for PM2.5, and that difference has to be factored into the equations. Luckily, a 2016 paper developed the math for modeling PM2.5 mortalities, and that is the math this paper used.
Great analysis! I am shocked that the news article's author made claims so far off from the results which were reported in the paper. Additionally, I thought that the article was even titled in a way that makes it clear to me that the author was most concerned with the article's engagement rather than actually accurately reporting the paper's findings. I really liked the paper and thought it was interesting that biogenic emissions play such a significant role in ozone-related health impacts. Like someone else mentioned, I would also be interested to see data on ozone health impacts past only premature deaths.
ReplyDeleteHi Lily, great analysis! It's interesting to me that the news article claimed results that the article doesn't say. I feel like this could be an example of how fake news begins because people could take this article and cite that when they talk about this topic. Im curious if the authors of the article are aware of this and if they could do anything to make sure that a false narrative isn't spread citing their paper.
ReplyDeleteWith any article nowadays I feel as though the author of most news relating to something don't generally fully understand or read the thing it is talking about, a scientific paper in this case. Obviously this severly detracts from its use as a news article as you showed in your evaluation but it also can lead to sort of telephone game with the information as many other news outlets tend to copy the first ones to talk about something. (except for institutions with higher integrity.) This may be what happened here and that this article copied another similar piece and took the ozone reduction conclusion from that or may be simply made something up to fill space who knows. In response to that I like how you focus on the PM2.5 aspect of the research paper to make up for the articles shortcoming. Although you also succinctly mention the ozone aspect as well. It is also interesting to see the Southeast Asian front for this issue as the European study someone else presented, seem to find that developing Europe seemed to have more of a problem with ozone from other countries ie industrialized ones. Granted a large part of that study was that the ozone came from hemispheric sources or unknown determinance.
ReplyDeleteWhen I read the news article first I thought that the research paper must be very focused or just quite short but it was quite long and looked at several different pollutants and emissions. I definitely the news article did a bad job, leaving so much out and even making a detailed claim not stated in the research paper. One thing I found confusing that I never really cleared up was how you state they found super-regional TAP to have a larger ozone impact than regional TAP, but they also never identified the sources of the TAP, so how did they differentiate between regional and super-regional TAP?
ReplyDeleteThis comment has been removed by the author.
DeleteHi Zackary, great questions. I would say your questions arise from a mixture of the paper being unclear about determining super-regional TAP affects, and me not being precise enough in my language. Sorry for the part I played in your confusion! Regional TAP is easier to determine since it is modeled by their algorithms. Thus, for regional TAP, sources are able to be determined and tracked. Super-regional TAP provides greater difficulty since it extends beyond the region covered by their models. As the models the authors use account for local seasonal biogenic and anthropogenic emissions, they assumed that any PM2.5 or ozone not be attributed to a particular source or region (their “baseline” PM2.5 and ozone levels with respect to time and location) were the result of super-regional TAP.
DeleteI really like that you did not shy away from giving the article a bad review, and I like how after your criticism of the article, you went back and re-stated exactly what the research paper concluded so that there was no confusion between the article and the paper. I find it really interesting how PM2.5 causes such a high percentage of local premature fatalities, but a much lower percentage of TAP fatalities. Do you have insight into why PM2.5 doesn't travel beyond local regions? Is it because PM2.5 are generally smaller molecules or potentially more/less reactive than other molecules that cause ozone formation in regions beyond where they originated? Thank you for your very in-depth analysis!
ReplyDeleteHi Kevin, thanks for the questions. I don't have a definite answer to this one, but based on my understanding of the paper and the patterns seen within it, I do have an educated guess. I cannot say that this effect is a result of PM2.5 instability-- especially since PM2.5 is a category and not one specific compound. What I think it has more to do with is the correlation between population density and pollution rates. In a higher population density area, like a city, there will be significantly more emissions, resulting in greater PM2.5 concentrations. Since there are many more people in the area, there is thus a greater % of the population locally exposed to PM2.5. To phrase this differently, I think the PM2.5 has the capability to travel beyond local regions, but it does so much damage within those local regions that the regional and super-regional effects are smaller in comparison. Additionally, in cities, there is oftentimes ozone depression as a result of minimal trees bringing the region into a VOC-limited regime. With the major population centers at minimized risk for ozone effects, but more rural areas facing elevated risk, this contributes to the % PM2.5 fatalities being skewed towards local emissions.
DeleteHi, this is a good write up. I like that you caught the inconsistencies between the article and the research paper and were not afraid to point them out to people. I am interested in how this sort of thing might happen or how something like that would get published. I also thought that you did a good job summarizing the paper but I am curious about a few things. One thing I am curious about is what kinds of pollutants make up the particulate matter that the paper is talking about.
ReplyDelete