What is “Good” Air?

As covered in my previous blog post, air quality can have severe repercussions on human health. How then should authorities determine if air quality is suitable for outdoor sports? At what threshold would the ambient air go from being good to not-so-good? Authorities have attempted to develop various air quality indexes (AQIs) in order to ensure that citizens have access to reliable and accurate evaluations of current air conditions. Far from being absolute and without fault, however, the accuracy of such AQIs shall be evaluated in this blog post.

Priti and Kumar (2022) emphasise that AQIs are not only information tools used to communicate air quality condition to the public, but also useful tool for authorities to track and predict short and long-term air quality trends. This thus presents a conundrum for developers: the model has to be clear-cut and easy enough to understand such that the general public are able immediately take the necessary precautions, while at the same time remaining comprehensive enough for authorities to have a solid grasp over air quality. AQIs typically include various components like scale and category (e.g. good, hazardous). The similarities between various AQI models, however, stop here. Since 1960, authorities have elected to develop AQIs in various ways. Some models churn out calculations based on the concentrations of selected pollutants in a localised area (e.g. PSI), while others may evaluate air quality based on the relative risk of daily mortality (e.g. RRAPI) (see Figure 1).

Upon the evaluation of these AQI models, Priti and Kumar (2022) conclude that most models suffer from a mix of ambiguity, eclipsing, rigidity, spatial aggregation, or pollution aggregation shortfalls (see Figure 2).

Shortfall	Description	Example
Ambiguity	Overestimating air quality i.e. reporting air quality as worse than it actually is	Ontario Air Quality Index (OAQI) Measures suspended particulate matter and SO2 As the calculation of the overall index combines individual indices, it is prone to ambiguity
Eclipsing	Underestimating air quality i.e. reporting air quality as better than it actually is	Canadian Air Quality Indices (CAQI) Proposed in 1975, 8 parameters (including NO2, suspended particulate matter, and visibility) are used to determine air quality After applying these 8 parameters to air quality in urban, around urban, and outside urban areas, a national index is generated using population as the weighting criteria Though this allowed the CAQI a large spatial coverage, it also obscures the unevenness of air quality across and outside urban areas
Rigidity	Aggregate functions are unable to accommodate new pollutants	PSI (max operator) Only takes into account one criteria pollutant at a time
Spatial and/or Pollution Aggregation	Failure to take into consideration the synergistic and antagonistic effects of different pollutants over a wide area	Multipollutant air quality index (MAQI) Takes into account the relative risk of exposure to PM2.5, PM10, NO2, SO2, O3 and CO individually, and does not consider how these pollutants may act on one another

Figure 2: A summary of the shortfalls that has plagued various AQIs since 1960. Note that due to these shortcomings, several of these AQIs may no longer be used today.

Furthermore, Priti and Kumar (2022) are of the view that AQIs are simply too limited in their coverage of pollutants. The indexes mainly consist of PM_2.5, PM₁₀, NO₂, SO₂, O₃ and CO, and while these are no doubt key pollutants with harmful effects on individual health (Priti & Kumar, 2022; World Health Organization, n.d.), there is insufficient scrutiny on Hazardous Air Pollutants (HAPs). HAPs consists of 188 pollutants that have been found or suspected to be carcinogenic, such as benzene (found in gasoline), and asbestos. Heavy metals are also capable of adsorbing to PM2.5 particles, which goes undetected in most AQIs (Li et al., 2021). In addition to Priti and Kumar’s (2022) suggestion that the choice of parameters should be targeted to the locality and purpose of different AQIs (e.g. including meteorological factors to develop an AQI for alert-purposes), I also believe that these parameters should be continuously updated and informed by sound research and advancements in technology (e.g. new findings regarding previously undetected pollutants). This is so that AQIs can remain relevant and useful for both local authorities and the general public.

The reliance on breakpoint concentrations also raises another question: how exactly are these breakpoint concentrations of each pollutant determined? Authorities frequently gloss over the determination of breakpoint concentration of pollutants, with little justification. Priti and Kumar (2022) are quick to point out that the breakpoint concentrations of various pollutants often vary between developed and developing countries despite such breakpoints being supposedly based on epidemiological studies. This is corroborated by a comparison of the breakpoint concentrations under the US, China, and Singapore’s AQI models against the World Health Organisation’s (WHO) air quality guidelines (AQG) (Figure 3). Indeed, the determination of breakpoint concentrations do not only involve the potential impacts on human health, but is also a representation of authorities’ attempts (or even struggles) to balance economic development and public health (Priti & Kumar, 2022). In the process, air pollution is normalised.

	24-hour PM2.5 (μg/m3)	8-hour O3 (μg/m3)	24-hour NO2 (μg/m3)
WHO AQG	15	100	25
US AQI	12.0	54	–
China IAQI	35	100*	40
Singapore PSI	13	118	–

*Peak O₃ concentration over an 8-hour period are used in calculations, as opposed to average.

Figure 3: Comparison of the maximum concentration limits of selected pollutants in what is considered to be “healthy air” by the WHO and national governments. Dashes indicate that the AQI model does not take into account that pollutant over the specified period of time. The Singapore PSI, for example, only takes into account 1-hour NO2 concentrations when it is equal to or above 1130 μg/m³ . (Sources: WHO, US AirNow, Chen et al. (2016), Singapore National Environmental Agency)

Lastly, Priti and Kumar (2022) briefly acknowledge the danger of arbitrary breakpoint concentrations. I would like to expand on this point by linking it to sports and athletes using a study done by Nowak et al. (2022) on the effects of air pollution on athletes. Determining and reflecting the accurate relative risks of the air quality is particularly important in sports, as athletes rely on AQIs to determine if it is safe to exercise. The health impacts of misleading AQIs are accentuated in athletes owing to the strenuous nature of certain sports. Nowak et al. (2022) writes that during such strenuous physical activities (such as running a marathon), athletes find themselves needing to breathe through both their nose and mouth in order to increase their oxygen intake. Needless to say this greatly increases the amount of particulate matter that they breathe in, while mouth-breathing also bypasses the body’s “natural defenses against microbial or particulate matter” (Nowak et al., 2022, p.3).

In addition to diminishing athlete performance (Guo & Fu, 2019), the health effects of exercising in poor air conditions are already being seen in distance runners and endurance athletes. These negative impacts are not hypothetical. Already, there is a higher prevalence of asthma or asthma-like symptoms amongst such athletes, and this phenomenon is hypothesised to be directly caused by hyperventilation amidst exposure to particulate matter (Nowak et al., 2022). While Guo and Fu (2019) had written on the risk of running amidst air conditions that are known to be poor, it is not enough for these athletes to simply stop training or competing during periods of “poor” air quality. The cumulative effects of low doses of pollutants (i.e. air quality marked as “good” by certain AQI models) are enough to impair the lung function of athletes over just a single year (Lee et al., 2023). Thus, it would seem that AQI models are not reflecting the real threat of pollutants in the air to a sufficient degree.

What, then, might this mean for the future of outdoor sports, when it is becoming apparent that athletes are being directly harmed by the environment that they necessarily need to be in? My next blog post shall cover some of the strategies that have been used by authorities to tackle air pollution within their territory.

 

References:

Chen, W., Tang, H., & Zhao, H. (2016). Urban air quality evaluations under two versions of the national ambient air quality standards of China. Atmospheric Pollution Research, 7(1), 49-57.

Guo, M. & Fu, S. (2019). Running with a mask? The effect of air pollution on marathon runners’ performance. Journal of Sports Economics, 20(7), 903-928.

Lee, H. Y., Kim, H. J., Kim, H. J., Na, G., Jang, Y., Kim, S. H., Kim, N. H., Kim H. C., Park, Y.-J., Kim, H. C., Yun, Y.-K., & Lee, S. W. (2023). The impact of ambient air pollution on lung function and respiratory symptoms in elite athletes. Science of The Total Environment, 855, 158862.

Li, F., Yan, J., Wei, Y., Zeng, J., Wang, X., Chen, X., Zhang, C., Li, W., & Lü, G. (2021). PM2. 5-bound heavy metals from the major cities in China: spatiotemporal distribution, fuzzy exposure assessment and health risk management. Journal of Cleaner Production, 286, 124967.

Nowak, A. S., Kennelley, G. E., Krabak, B. J., Roberts, W. O., Tenforde, K. M., & Tenforde, A. S. (2022). Endurance athletes and climate change. The Journal of Climate Change and Health, 100118.

Priti, K., Kumar, P. (2022). A critical evaluation of air quality index models (1960–2021). Environmental Monitoring and Assessment, 194(5), 324.

World Health Organisation. (n.d.). Ambient air pollution data. https://www.who.int/data/gho/data/themes/air-pollution/ambient-air-pollution