Evaluation of Offline-Coupled FV3GFS-CMAQ over the U.S. in Support of the Next Generation of the National Air Quality Forecast Capability

Since the development of the U.S. NOAA’s National Air Quality Forecast Capability (NAQFC) for real-time air quality forecasting over the continental U.S. (CONUS), significant progress has been made to improve the forecasting skill of NAQFC. However, moderate-to-large biases remain in forecasting surface O₃ and PM_2.5. The current version of the NAQFC consists of the Community Multiscale Air Quality (CMAQ) model version 5.0.2 offline coupled with the North American Mesoscale Forecast System (NAM) model.

The NAM will be replaced by a new Finite Volume Cube-Sphere Global Forecast System (FV3GFS) in the next generation of NAQFC. In support of the development of this new generation of NAQFC, the offline-coupled FV3GFS-CMAQ v5.0.2 has been applied for real-time air quality forecasting over the U.S. since July 2017. In this work, the offline-coupled FV3GFS-CMAQ v5.0.2 will be comprehensively evaluated in terms of both meteorological and air quality forecasts. The observations to be used for model evaluation will include meteorological measurements from NCDC and PRISM and chemical measurement data from AIRNOW, AQS, CASTNET, IMPROVE, CSN, and PAMS. The evaluation will include relevant meteorological variables such as temperature, relative humidity, wind fields, precipitation, and planetary boundary layer heights and chemical species such as O₃, PM_2.5, and their gaseous precursors (e.g., SO₂, NO_x, and VOCs). The analysis of forecasting products will be performed in terms of performance statistics and spatio-temporal agreement against available observations at various time scales (e.g., hourly, diurnally, daily, monthly, seasonally, and annually) over various geographical domains (e.g., CONUS, western and eastern U.S.). The main objective is to identify major model biases in forecasting meteorology and air quality associated with the offline-coupled FV3GFS-CMAQ v5.0.2 and suggest potential model improvement to further enhance the NAQFC’s forecasting skill.