Near-roadway air pollution: evaluation of fine particulate matter (PM₂.₅) and ultrafine particulate matter (PM ₀.₁) in Interior Alaska

Abstract

Particulate air pollution in the form of fine (PM₂.₅) and ultrafine (PM₀.₁) particles has become a global concern, especially in urban areas with high population and vehicular traffic. Considerable research has been carried out to understand the underlying processes that impact particulate pollution, but most studies have been conducted in warmer and urban regions such as in California. The Fairbanks North Star Borough (FNSB), in Interior Alaska, provides an interesting example of a relatively small- to mid-sized northern locality (population ~100,000) with persistent air quality issues and extremely cold climatic conditions for a major part of the year. Since December 2009, the FNSB has been designated a nonattainment region by the U.S. Environmental Protection Agency for the federal PM₂.₅ standard. As part of their remediation efforts, the borough and state have undertaken increased monitoring by using an on-roadway monitoring vehicle (sniffer vehicle) and stationary near-roadway sites for air quality measurements, beyond what is required for regulatory compliance. One of the goals of this project was to develop a novel data investigation and analyses methodology for the geospatial air quality data collected by the borough's mobile monitoring vehicle (years 2012-15), to shed light on the PM₂.₅ issues faced by the FNSB. In addition, this research also undertook measurements of ultrafine particle (UFP) concentration levels at four road weather information system (RWIS) sites in the FNSB region. UFPs, though unregulated, are considered to have significant human health impacts and no known studies have investigated UFPs in FNSB. In addition to UFPs, other parameters such as PM₂.₅, traffic, and weather data were measured at the same locations to investigate any underlying trends/correlations with UFPs. In the first part of the research with mobile monitoring, data were categorized in nine different groups based on their mean and standard deviation values to determine the spatiotemporal distribution of PM₂.₅. This novel way of grouping data allows identification of locations with consistently poor and consistently better air quality, by going beyond the simple analyses of means and accounting for variability and standard deviation in the data. In addition to hotspot identification, analyses found that average on-roadway PM₂.₅ concentrations were higher in North Pole (27.2 μg/m³) than in Fairbanks (12.9 μg/m³), and that average concentrations were higher in the background stationary monitoring data (29.4 μg/m³) than in the mobile monitoring data (20.0 μg/m³) for the study period. Not surprisingly, significant negative correlations (R² = 0.49 for Fairbanks, and R² = 0.31 for North Pole) were found between temperature and PM₂.₅. Temporal distribution of the data suggests that PM₂.₅ levels increase gradually in the months of October and November, peak during the months of December, January, and February, and quickly plummet beginning March. In the latter part of the study, data on UFP measurements were collected at four RWIS sites in the FNSB for four days between March 1 and 18, 2017, for five continuous hours each day. Among other parameters, PM₂.₅ concentrations, temperature, relative humidity, wind speed, and traffic volume data were collected. Data were analyzed to develop correlations between UFPs and other parameters, to compare data from this study with other studies, and to determine current roadside UFP concentration levels in interior Alaska. Fairbanks roadside locations showed higher mean UFP counts (41,700 particles/cm³) than the North Pole (22,100 particles/cm³) locations. Similarly, for the period of study, Fairbanks roadside locations showed higher PM₂.₅ concentrations and traffic counts (6.3 μg/m³; 15 vehicles/min) than the North Pole (4.6 μg/m³; 10 vehicles/min) locations, both being well below the on-roadway and background PM₂.₅ concentrations estimated in the first part of this report. Multilinear predictive models were developed for estimation of UFPs and PM₂.₅ based on weather and traffic parameters. This first study of UFPs in Alaska improves our understanding of near-roadway UFPs in cold regions.