Orly Stampfer

Project title: Use of Low-Cost Sensors in Community-Engaged Research on Wood Smoke in the Lower Yakima Valley

Degree: MPH | Program: Environmental and Occupational Health (EOH) | Project type: Thesis/Dissertation
Completed in: 2018 | Faculty advisor: Catherine Karr


Rural lower Yakima Valley, Washington is home to communities including Latinx farmworker families, the Confederated Tribes and Bands of the Yakama Nation, and Native Americans of other tribes. Episodic poor air quality impacts this region, reflecting sources of particulate matter with a diameter of less than 2.5 micrometers (PM2.5) that include residential wood smoke and agricultural biomass burning. University of Washington (UW) partnered with the Yakama Nation Environmental Management Program and local institutions, Heritage University and White Swan High School, to develop community-engaged research on wood smoke. Both a strong partnership and reliable methods are necessary for community-engaged research. The overall aim of this thesis was to address the feasibility of forming a community-academic partnership and using low-cost sensors to examine wood smoke emissions. This study assessed the following: (1) what are community-academic partnership perspectives on trust, equitable and culturally informed processes, and community involvement in research? and (2) can the low-cost sensor be easily calibrated, and can low-cost sensor particulate matter size distributions be used to predict wood smoke emissions? Aim (1) used semi-structured interviews with each member of the UW team (n=6) and most of the community partners (n=6). Responses were coded and analyzed using a grounded theory approach. Funding, Dialogue, and Formal roles and processes were structural components that provided foundations for the following themes as practices: Community partner capacity building, Academic presence in community, Recognition of community strengths, Understanding the significance of historical and current community dynamics and culture, and Transparency in the research process. Each of the practices contributed to research question outcomes: trust, equitable and culturally informed processes, and community involvement in research. Based on these findings, recommendations were developed for rural, multicultural community-academic partnerships involved in air quality research. Aim (2) used data collected at a tribal air monitoring site in Winter 2018 with a laser based, low-cost, 5-bin particle counter and a 5-wavelength aethalometer (MA200 Aethlabs). Delta-C, the absorbance difference at 375nm-880nm, was used as an indicator of biomass burning. Low-cost sensor PM2.5 calibration used regression parameters from an 8-day co-location with a tribal beta-attenuation monitor. The 80th percentile of the hourly Delta-C:PM2.5 ratio was used to signify wood smoke-enriched hours. The low-cost sensor particle size distributions during these wood smoke-enriched hours were compared to those below 50th percentile of the hourly Delta-C:PM2.5 ratio. Hourly PM2.5 mean was 6.7 μg/m3 (standard deviation 6.7 μg/m3, range 1.3-31.8 μg/m3). Measured particle size distribution did not differ between wood smoke-enriched and low wood smoke hours. The correlation between Delta-C and PM2.5 was higher during wood smoke-enriched hours (0.84) vs. low wood smoke hours (0.75). This suggests that while the low-cost sensor captures wood smoke emissions, further analysis exploring application of low-cost methods to isolate wood smoke episodes are needed. Through evaluation of community engagement and assessment of the performance of low-cost sensors, this thesis contributes to knowledge of best practices in conducting community-engaged research on air quality and provides information on the potential use of low-cost sensors in identifying wood smoke in a rural US setting. These results may facilitate future studies of rural air quality in multicultural communities historically underserved in air quality research. URI