10th International Aerosol Conference September 2 - September 7, 2018 America's Center Convention Complex St. Louis, Missouri, USA
Abstract View
AQ & U: A Layered Framework for Integrating Sensor Data of Variable Quality and for Engaging Citizens about PM2.5 Exposure
KERRY KELLY, Pierre-Emanuel Gaillardon, Miriah Meyer, Ross Whitaker, Anthony Butterfield, Pascal Goffin, Tom Becnel, Amir Biglari, Tofigh Sayahi, University of Utah
Abstract Number: 1337 Working Group: Low-Cost and Portable Sensors
Abstract The emergence of commodity sensors is changing the way we think about and estimate our personal exposures to potentially harmful air-quality events. AQ & U is building a layered framework for integrating sensor data of variable quality using state-of-the-art data modeling and visualization coupled with a citizen-science effort to engage residents to host and maintain sensors across the city. The goal of AQ&U is to provide real-time, localized estimates of PM2.5 levels across the Salt Lake Valley - a region that periodically experiences some of the worst short-term PM2.5 pollution episodes in the country. During the winter of 2017 the Salt Lake Valley had over 100 sensing nodes in the AQ&U network, and the network included high-quality data from state monitors and research-grade instrumentation as well as lower quality information from community networks of low-cost, PM2.5 sensors. We performed laboratory and field calibration on a portion of the low-cost sensors, but the quality of the measurements from the existing network of community sensors was less certain. Sensor readings and uncertainty estimates for each sensor were used in a regression model to estimate PM2.5 concentration at the neighborhood scale. Complexities, such as elevation and roadways, are being added to the model. Citizens can view the sensor data, estimates of PM2.5 and uncertainty through engaging visualizations. AQ&U takes a citizen-centric approach both in the way that we deploy and maintain our sensor network, as well as in the way we design tools for public access. We rely on individual and school volunteers to host sensors and help identify poor-quality data. For students, we developed a hands-on teaching module that incorporates building blocks and simple electronics to build a light-scattering PM detector. This allows students to understand the operating principal of their sensor, making it less of a “black box”.