A novel approach to the development of 1-hour threshold concentrations for exposure to particulate matter during the response to episodic air pollution events

Background
Episodic air pollution events, eg wildfires, result in significant increases in pollutant concentrations over a short duration. These events are an increasing phenomenon because of climate warming. Moreover, despite evidence showing that inflammatory response in humans to particulate (PM) exposure occurs in only a few hours, the averaging period for PM exposure guideline values is 24-h. Thus, short-duration (1-h averaging period) exposure guidelines are needed to inform population exposure advice.

Methods
Two approaches were used to generate 1-h guideline values for PM exposure: (1) data from UK major air pollution incident monitoring was used to determine a probabilistic relationship between 1-h PM10 or PM2.5 measurements and exceedance data for a range of 24-h guideline values (GVs); (2) we applied Receiving Operating Characteristic statistical analysis to an extensive US PM monitoring data set (38 million hours of data), allowing the determination of 1-h threshold concentrations that predict exceedances of US EPA air quality guideline (AQI) category boundaries.

Results
We found that the maximum-observed 1-h PM concentration in any rolling 24-h averaging period is an excellent predictor of exceedances of 24-h guideline values. This has allowed us to develop de-facto 1-h GVs that correspond to a range of air quality guidance. E.g. we propose a 1-hr PM10 GV of 550μgm-3 for evacuation of exposed areas. Our 1-h GV for evacuation is similar to the (withdrawn) 1–3-h average ‘Recommended Action Level’ for closing public buildings, which was set of 526μgm-3 for PM10/PM2.5 during wildfires.

Conclusions
This straightforward approach allows for the development of 1-hr GVs that can be used to make much more timely public health protection decisions during the response phase to an episodic air pollution incident.