Real-time Modeling of Air Quality Estimates due to Traffic Emissions at Hyperlocal Scales

We present an approach for predicting near-road air quality in urban areas at hyperlocal scales in near real-time. We introduce C-REAL - a new web-based modelling system to study air pollution exposures due to traffic-related sources at a community or city scale. The dispersion algorithms are based on the C-LINE modeling system augmented with near real-time traffic activity and meteorological inputs. Specifically, we use meteorology from the National Oceanic and Atmospheric Administration (NOAA’s) High Resolution Rapid Refresh (HRRR) model, traffic activity from Google’s Waze, and emissions factors for NO_x and primary PM_2.5 from MOVES-2014 as a function of road type, vehicle type, temperature and speed. The results are available continuously for every hour (current, two hours prior and two hours in the future) through the web-based interface for four cities in the U.S. – Kansas City, KS Houston, TX, New York City, NY and Oakland, CA. In this presentation, we will present illustrative examples of this tool being applied in Kansas City, KS and Oakland CA, and comparisons of model predictions to Google Street View (GSV)-based measurements. The presentation will summarize the capabilities of C-REAL and its potential applications for providing air quality estimates in near real-time at high spatial resolutions in support of epidemiological and public health studies.