New PM2.5 index captures both how high and how often air pollution strikes

Fine particulate matter smaller than 2.5 micrometers, known as PM2.5, is among the most closely watched pollutants in the world. Regulators report annual averages, public health agencies track short-term spikes, and news outlets broadcast color-coded air quality alerts. Yet a new study argues that the standard toolkit leaves a gap: none of the familiar numbers fully captures both how severe pollution gets and how often it happens in a single, intuitive figure. In a paper published in Environmental Monitoring and Assessment, Ren Zhang of the Center for Molecular Medicine and Genetics at Wayne State University School of Medicine introduces the PM2.5 Magnitude–Rank Index, or PMRI, a single number designed to summarize the entire concentration–frequency structure of a year of daily air quality data.

The definition is disarmingly simple. PMRI is the largest integer k such that at least k days in a year have daily mean PM2.5 concentrations of k micrograms per cubic meter or higher. A city with a PMRI of 23, for example, experienced at least 23 days on which the daily average concentration reached 23 micrograms per cubic meter or more. The metric borrows its logic from the h-index, the well-known bibliometric measure introduced in 2005 that captures both the productivity and citation impact of a researcher. Just as an h-index of 30 means a scientist has published at least 30 papers cited at least 30 times each, a PMRI of 30 means a metropolitan area endured at least 30 days with concentrations of at least 30 micrograms per cubic meter. The parallel is more than cosmetic: both indices formalize a magnitude–frequency tradeoff without relying on any externally chosen threshold.

That threshold-free character is central to the argument. Conventional regulatory metrics depend on policy decisions. Exceedance counts, for instance, ask how many days crossed a specific limit such as the 35 micrograms per cubic meter standard for short-term PM2.5 exposure in the United States. Change the threshold and the count changes, sometimes dramatically, which makes comparisons across jurisdictions and eras awkward. Annual means, meanwhile, smooth away the episodic character of pollution: two cities can share nearly identical yearly averages while one suffers a handful of severe smog episodes and the other endures a chronic, moderate haze. PMRI sidesteps both problems by letting the data itself define the cutoff, and it responds to both the height and the width of the pollution distribution.

To test the index, Zhang applied it to daily PM2.5 records from the U.S. Environmental Protection Agency’s Air Quality System AirData archive, covering 96 U.S. metropolitan statistical areas over the ten-year period from 2013 to 2022. After screening for data completeness, 918 metropolitan-area years qualified for analysis. Across this decade-long, nationwide sample, PMRI values ranged from 6 to 38, with a median of 16. The range tells a story in itself: the cleanest qualifying city-year in the dataset had at most six days reaching six micrograms per cubic meter, while the worst had nearly forty days at or above roughly forty micrograms per cubic meter, a level several times higher than the World Health Organization’s 2021 guideline values for annual and daily exposure.

The statistical performance of the new index is a key part of the paper. PMRI correlated strongly with established metrics: the Spearman rank correlation with the annual mean concentration was 0.83, with the annual 98th percentile it reached 0.95, and with the number of days at or above 15 micrograms per cubic meter it climbed to 0.96. These are high correlations, which means PMRI is not measuring something alien; it tracks the same underlying pollution burden that regulators already monitor. But correlation is not identity. The author shows that PMRI can distinguish metropolitan-area years with similar annual means but different concentration–frequency profiles, the exact situation in which a single average conceals meaningful differences in exposure patterns that matter for health.

The distinction matters because the health effects of PM2.5 depend on more than the yearly average. These particles, small enough to penetrate deep into the lungs and enter the bloodstream, are linked in the epidemiological literature to cardiovascular disease, respiratory illness, and premature mortality. A population exposed to a steady moderate background faces a different physiological challenge than one battered by repeated intense episodes, even if the annual averages converge. By encoding both magnitude and frequency in one number, PMRI offers a way to flag such differences that annual means alone cannot. The index also avoids the policy dependence of exceedance counts, which shift whenever a standard is revised, complicating long-term trend analysis.

The mathematical structure of PMRI also gives it a pleasing robustness. Because it is defined as a rank over the sorted distribution of daily concentrations, it is insensitive to extreme outliers in the way that means and percentiles can be, and it requires no distributional assumptions about the data. Its interpretation requires no unit conversion, no category mapping, and no reference table: the number is directly readable as days matched to concentration levels. The author provides Python code and a sample input dataset in the supplementary materials, and the full set of PMRI values and selected annual metrics for all 918 metropolitan-area years is published as a supplementary table, making the analysis reproducible and extensible to other regions and pollutants.

The study situates PMRI within a long tradition of air quality indices, from the Pollutant Standards Index to the revised Air Quality Index frameworks compared in earlier literature. Those indices serve vital communication functions, translating complex monitoring data into categories the public can act on, but they compress information into bins and depend on defined breakpoints. PMRI is not proposed as a replacement. Rather, the paper positions it as a transparent complement: a summary statistic that coexists with annual means, percentiles, and exceedance counts, each answering a different question about the same pollutant record. For researchers comparing cities, for regulators evaluating multi-year trends, and for historians of air quality, a threshold-free magnitude–rank summary adds a dimension that existing metrics do not provide.

There are natural extensions. The same magnitude–rank construction could in principle be applied to other pollutants monitored on daily cycles, such as ozone or PM10, and to other spatial units, from individual monitoring stations to entire countries. The author notes that the index is defined for annual summaries, and its behavior under different averaging windows or in regions with different pollution regimes remains to be explored. What the current study demonstrates is the feasibility and interpretability of the approach on a large, high-quality national dataset spanning a decade of U.S. urban air quality.

For a field drowning in dashboards, thresholds, and color codes, the appeal of a single self-explanatory number is obvious. A PMRI of 16, the median value in this dataset, means that a typical American metropolitan area in the study period endured at least sixteen days a year with concentrations of at least sixteen micrograms per cubic meter, a level well above the WHO annual guideline. As fine particulate pollution continues to shape public health policy worldwide, tools that make the structure of exposure legible at a glance may prove as influential as the averages they complement. The full paper, including the data and code, is openly described in Environmental Monitoring and Assessment, and the underlying daily measurements remain publicly available from the EPA archive for anyone wishing to compute the index for their own city.

Subject of Research: A new summary statistic, the PM2.5 Magnitude–Rank Index, characterizes the magnitude and frequency of fine particulate pollution in U.S. metropolitan areas

Article Title: A PM2.5 magnitude–rank index for summarizing annual concentration–frequency structure

Article References: Zhang, R. (2026). A PM2.5 magnitude–rank index for summarizing annual concentration–frequency structure. Environmental Monitoring and Assessment, 198(9), Article 996. https://doi.org/10.1007/s10661-026-15858-1

Image Credits: AI Generated

DOI: 10.1007/s10661-026-15858-1

Keywords: air quality, PM2.5, environmental monitoring, public health, air pollution index, epidemiology, Environmental Monitoring and Assessment, magnitude, rank, index, summarizing, annual