Data Fusion with Uncertainty Quantification for Sub-City-Scale Assessments and Forecasting of PM2.5 and Trace Gases

NATHAN PAVLOVIC, Carl Malings, K. Emma Knowland, Christoph Keller, Stephen Cohn, Callum Wayman, Alan Chan, Sean Wihera, Sean Khan, John White, Daniel Westervelt, Randall Martin, Sonoma Technology, Inc.

     Abstract Number: 569
     Working Group: Aerosols Spanning Spatial Scales: Measurement Networks to Models and Satellites

Abstract
Sub-city-scale air quality estimates and forecasts are a critical need for areas that are disproportionately impacted by poor air quality or intermittent air quality events such as wildfires. Many information sources can support air quality assessments and forecasting, including atmospheric chemistry model outputs, satellite retrievals of column aerosols and gases, and surface-based air quality monitoring data from both regulatory and low-cost instruments. These sources of data span a broad range of spatial and temporal scales with respect to resolution and coverage. Our team is developing a data fusion approach with data from low-cost sensors, reference-grade monitors, models (CAMS, GEOS-CF), and satellites to develop near-real-time (NRT) forecasting of PM_2.5 and gases (NO₂ and O₃). The NRT forecasting will generate an hourly forecast of sub-city-scale (e.g., 1 km resolution) air quality estimates that will support decision-making needs for end users. Furthermore, the data fusion framework includes quantification of uncertainties in the resulting fused estimates based on the variability of each input data source and among the data sources. These capabilities will allow air quality managers to better understand their local air quality situations, including relative confidence in the fused estimates for different atmospheric constituents, locations, and times, leading to better informed air quality management decisions. This presentation (1) covers the underlying methodology of the data fusion and uncertainty quantification approaches, with a focus on the PM_2.5 component; (2) provides an update on the status of its implementation; and, (3) presents early qualitative and quantitative results and case studies focused on aerosol events, such as wildland fire smoke and dust storms.