Probabilistic forecasts of ozone and PM 2.5 from the Community Multiscale Air Quality (CMAQ) model.

A bias-correction Kalman filter - Analog (KFAN) method was developed previously for post-processing surface ozone and PM2.5 predictions from the NOAA Community Multiscale Air Quality (CMAQ) forecasting system used in the National Air Quality Forecasting Capability (NAQFC). This method has been used operationally at the National Centers for Environmental Prediction (NCEP) for PM 2.5 since January, 2016, and for ozone since December, 2018, with forecasts available at airquality.weather.gov. Forecasts for both species are shown as deterministic predictions.

The next step is the generation of gridded probabilistic forecast information from the ensemble of forecast analogs. Several approaches for calculating probabilistic forecasts will be discussed. The dataset used in this study includes the CMAQ model based driven by the FV3 meteorological forecasting system, with results analyzed over several months in both 2018 and 2019, including episodes with wild fires. The final method selected employs a statistical interpolation procedure using surface ozone or PM2.5 analogs and a diurnally and seasonally dependent gridded climatology (or prior) of those variables from the CMAQ forecasts. Reliability diagrams, Brier skill scores, and spread-skill relations are used to assess the accuracy of the probabilistic forecasts.