Status Update: Is smoke on your mind? Using social media to
determine smoke exposure
Bonne Ford1, Moira Burke2, William Lassman1, Gabriele Pfister3, and Jeffrey R. Pierce11Department of Atmospheric Science, Colorado State University, 1371 Campus Delivery, Fort Collins, CO 80523 2Facebook, Menlo Park, CA 94025 3National Center for Atmospheric Research, 3450 Mitchell Lane, Boulder, CO 80301
Received: 10 Jan 2017 – Accepted for review: 17 Jan 2017 – Discussion started: 19 Jan 2017
Abstract. Exposure to wildland-fire smoke is associated with negative effects on human health. However, these effects are poorly quantified. Accurately attributing health endpoints to wildland-fire smoke requires determining the locations, concentrations, and durations of smoke events. Most current methods for determining these smoke-event properties (ground-based measurements, satellite observations, and chemical-transport modeling) are limited temporally, spatially, and/or by their level of accuracy. In this work, we explore using social-media posts regarding smoke, haze, and air quality from Facebook to determine population-level exposure for the summer of 2015 in the western US. We compare this de-identified, aggregated Facebook data to several other datasets that are commonly used for estimating exposure, such as satellite observations (MODIS aerosol optical depth and Hazard Mapping System smoke plumes), surface particulate-matter measurements, and model (WRF-Chem) simulated surface concentrations. After adding population-weighted spatial smoothing to the Facebook data, this dataset is well-correlated (R2 generally above 0.5) with these other methods in smoke-impacted regions. Removing days with considerable cloud coverage further improves correlations of Facebook data to traditional exposure datasets, which implies that the population is less aware of smoke on cloudy days relative to sunny days. The Facebook dataset is better correlated with surface measurements of PM2.5 at a majority of monitoring sites (163 of 293 sites) than the satellite observations and our model simulation are. We also present an example case for Washington state in 2015, where we combine this Facebook dataset with MODIS observations and WRF-Chem simulated PM2.5 in a regression model. We show that the addition of the Facebook data improves the regression model's ability to predict surface concentrations. This high correlation of the Facebook data with surface monitors and our Washington state example suggests that this social-media-based proxy can be used to estimate smoke exposure in locations without direct ground-based particulate-matter measurements.
Ford, B., Burke, M., Lassman, W., Pfister, G., and Pierce, J. R.: Status Update: Is smoke on your mind? Using social media to
determine smoke exposure, Atmos. Chem. Phys. Discuss., doi:10.5194/acp-2017-26, in review, 2017.