Developing an algorithm to determine woodsmoke events

Woodsmoke can have a significant effect on air pollution, with significant contributions to PM10 and PM2.5 concentrations (Hellén et al. 2008). There is a considerable evidence that woodsmoke adversely affects health (Naeher et al. 2007). Especially in densely populated areas, such as the Netherlands, people are likely significantly exposed to woodsmoke. However, distinguishing woodsmoke from other air pollution sources is challenging. A good understanding of the specific contribution of woodsmoke is essential for implementing effective strategies to improve air quality.Levoglucosan, a very specific marker of woodsmoke, is usually sampled from filters (24-hour averaged) and analyzed off-line in a laboratory. This approach is relativity expensive and makes it difficult to attribute short-lived peaks in air quality parameters to woodsmoke. In this study, we investigate the possibility of identifying a specific fingerprint of woodsmoke by analyzing multiple species simultaneously.An elaborate dataset was gathered in the Netherlands (in the province of North-Holland) for one winter (2023­-2024). Our dataset consists of filter measurements (PM2.5), PM fractions (Pallas FIDAS), ultrafine particles (UFP, total particle count), black carbon (BC, Aethalometer AE33), carbon monoxide (CO), and nitrogen oxides (NOx). Additionally, an ACSM (Aerosol Chemical Speciation monitor) was used to determine the chemical composition of the aerosols.In a previous campaign, the ACSM showed promising results for measuring levoglucosan from m/Q channel 60 on a 10-minute timescale. Therefore, we examined whether other compounds correlate with this channel. We found that BC showed a very promising relation with the m/Q channel 60 of the ACSM. Additionally, PM measured by the FIDAS peaked at similar times but also exhibit peaks when the m/Q channel 60 did not detect anything. We developed an algorithm to automatically identify woodsmoke events from measured BC and PM concentrations. This algorithm was verified dusing the ACSM levoglucosan results. For all the woodsmoke events determined by the algorithm the ACSM also measured peaks in the m/Q channel 60. Conversely, not all peaks found in the m/Q channel 60 were distinguished as woodsmoke events by the algorithm. The discrepancy occurred for 56 hours out of the 4654 hours sampled. These discrepancies predominantly occurred in May; we hypothesize that this was due barbequing.A new campaign is underway at the time of writing this abstract (winter 2024–2025). In this campaign, we will evaluate the capabilities of the ACSM to measure levoglucosan more accurate by using a standard vaporizer instead of a capture vaporizer. Furthermore, we assess whether woodsmoke events can be distinguished by combining BC/ PM sensor data from citizens. By integrating spatially resolved measurements, we aim to provide actionable insights into localized pollution patterns, enabling targeted interventions to reduce exposure in high-risk areas. Hellén, H., Hakola, H., Haaparanta, S., Pietarila, H., & Kauhaniemi, M. (2008). Influence of residential wood combustion on local air quality. Science of the total environment, 393(2-3), 283-290. Naeher, L. P., Brauer, M., Lipsett, M., Zelikoff, J. T., Simpson, C. D., Koenig, J. Q., & Smith, K. R. (2007). Woodsmoke health effects: a review. Inhalation toxicology, 19(1), 67-106.