A data fusion approach to assessing the contribution of wildland fire smoke to fine particulate matter in California

The escalating frequency and severity of global wildfires necessitate an in-depth understanding and monitoring of wildfire smoke impacts, specifically its contribution to fine particulate matter (PM2.52.5). We propose a data-fusion method to study wildfire contribution to PM2.52.5 using satellite-derived smoke plume indicators and PM2.52.5 monitoring data. Our study incorporates two types of monitoring data, the high-quality but sparse Air Quality System (AQS) stations and the abundant but less accurate PurpleAir (PA) sensors that are gaining popularity among citizen scientists. We propose a multi-resolution spatiotemporal model specified in the spectral domain to calibrate the PA sensors against accurate AQS measurements, and leverage the two networks to estimate wildfire contribution to PM2.52.5 in California in 2020 and 2021. A Bayesian approach is taken to incorporate all uncertainties and our prior intuition that the dependence between networks, as well as the accuracy of PA network, vary by frequency. We find that 1% to 3% increase in PM2.52.5 concentration due to wildfire smoke, and that leveraging PA sensors improves accuracy.

Impact/Purpose

Develop time and space Bayesian algorithm for calibration of Purple Air sensors. Though Purple Air sensors are continuously improving, they do not have equal performance over time. Additionally, bias of these sensors depends on weather patterns, elevation, as well as distance from the regulatory monitor. To provide the most accurate information that reflects user experience of ambient air quality we will develop a spatial and temporal Bayesian Hierarchical model which will dynamically calibrate each monitor's observation. To dynamically deliver the information to the user in real time we will adopt a multiscale approach.   The escalating frequency and severity of global wildfires necessitate an in-depth under- standing and monitoring of wildfire smoke impacts, specifically its contribution to fine particulate matter (PM2.5). We propose a data-fusion method to study wildfire contribution to PM2.5 using satellite-derived smoke plume indicators and PM2.5 monitoring data. Our study incorporates two types of monitoring data, the high-quality but sparse Air Quality System (AQS) stations and the abundant but less accurate PurpleAir (PA) sensors that are gaining popularity among citizen scientists. We propose a multi-resolution spatiotemporal model specified in the spectral domain to calibrate the PA sensors against accurate AQS measurements, and leverage the two networks to estimate wildfire contribution to PM2.5 in California in 2020 and 2021. A Bayesian approach is taken to incorporate all uncertainties and our prior intuition that the dependence between networks, as well as the accuracy of PA network, vary by frequency. We find that 1% to 3% increase in PM2.5 concentration due to wildfire smoke, and that leveraging PA sensors improves accuracy.

Citation

Yang, H., S. Ruiz-Suarez, B. Reich, Y. Guan, AND A. Rappold. A data fusion approach to assessing the contribution of wildland fire smoke to fine particulate matter in California. MDPI, Basel, SWITZERLAND, 15(17):1, (2023). [DOI: 10.3390/rs15174246]

DOI: A data fusion approach to assessing the contribution of wildland fire smoke to fine particulate matter in California