Evaluation of AIRPACT Air Quality Forecasts Against Ground-based Remote Sensing Observations

Washington State University runs an operational air quality forecast model over the Pacific Northwest region, which is called AIRPACT (Air Indicator Report for Public Awareness and Community Tracking). AIRPACT is routinely evaluated against satellite-based aerosol optical depth (AOD), an indicator of the extinction of solar radiation by particulate matters (or aerosols) in an atmospheric column. Unfortunately, the satellite measurements have poor temporal and spatial coverage because of the polar orbit of satellites and the frequent cloud coverage in our area. To overcome these limitations and gain an additional dimension of data for model validation, we evaluate AIRPACT against ground-based AOD measurements from NASA’s AERONET (AErosol RObotic NETwork) remote sensing network as well as surface visibility data from IMPROVE (Interagency Monitoring of Protected Visual Environment) nephelometer network: 7 AERONET sites and 3 IMPROVE sites within the AIRPACT model domain. We chose August 2015 for the evaluation due to the prevalence of wildfires and therefore higher atmospheric aerosol levels during that month. We find that the predicted AOD is consistently lower than observations, but otherwise captures diurnal and synoptic variability. Using AERONET (total volume AOD) and IMPROVE (surface-level AOD) datasets, we find that the under-predicted AIRPACT AOD is not caused by vertical distribution of aerosols: we find similar errors and biases in the model AOD between total column and surface layer. To examine any systematic problems in aerosol optical assumptions used in the AIRPACT, we also compare the evaluation results from surface visibility with the surface aerosol mass concentrations evaluation.