Our team members have modeled NO2 using land-use regression (LUR) for the past ten years. We plan to standardize and document the models, and temporally adjust them to monthly and annual periods (2000 to present). Looking forward, a new suite of temporally-adjusted LURs incorporating a chemical transport model (a project from Health Canada and Environment and Climate Change Canada) is in the works.
Looking ahead – hourly air pollution from North America’s geostationary satellite
In 2019 the world’s first geostationary satellite (TEMPO) for measuring air pollution will produce high resolution (~5km) maps every daytime hour of tropospheric NO2, SO2, and aerosols. These maps can be used to estimate pollutant concentrations for all Canadian cities. CANUE members Randall Martin and Chris McLinden are part of the TEMPO science team. Prior to the launch, we will develop algorithms to access these data in real-time to map surface concentrations.
A new index of air pollution
By combining the PM2.5, NO2 and SO2 satellite estimates, we are exploring methods to identify two primary PM2.5 components: combustion-related PM (likely dominated by traffic, indicated by high PM2.5 and NO2); and certain industry-related PM or coal/oil power plants (more-likely associated with high PM2.5 and SO2).
Estimating gas and diesel emissions
In close collaboration with the Transportation Data Team, we will – for the first time – produce independent exposure surfaces for gasoline and diesel traffic-related air pollution at the urban to suburban scale. Isolating health effects of these two sources will have significant policy implications.
Historical air pollution estimates
Our team members recently developed a nationwide, satellite-derived estimate of PM2.5 concentrations for multiple time windows, mapped to six digit postal codes. This resource will be available on our portal. To support analyses of multiple pollutants, we will release estimates for NO2 (recently improved), O3, and for the first time, SO2. Different exposure time windows (monthly and annual estimates, back to at least 1990) will also be available.