Evaluation of NAM Meteorological Predictions for Driving the North American Air Quality Forecasting Capability Predictions

The exposure to ozone pollution can lead to health effects such as cardiovascular diseases, which is why predicting ozone is important for the public’s knowledge. NOAA’s National Air Quality Forecasting Capability (NAQFC) is designed to provide numerical model guidance of ozone and particulate matter for the United States to aid state forecasters and the public for reducing health risks. The Community Multiscale Air Quality Model (CMAQ) is an integral part of NAQFC towards producing ozone and PM model predictions through 48 hours to the public for over ten years.

NAQFC predictions are generated twice per day (at 06 and 12 UTC cycles) at 12 km resolution with outputs at 1 hour time intervals through 48 hours and distributed at http://airquality.weather.gov. The CMAQ chemical transport model is driven by the NOAA National Centers for Environmental Prediction (NCEP) North American Model (NAM) weather predictions on a 12 km horizontal grid. The NAM was upgraded in 2017 (V4) in part to reduce a warm 2 m temperature bias in summer. CMAQ was also upgraded to V5.0.2 in June 2017. For several months, the models were run in parallel with the old and current version of the NAM along with the two versions of CMAQ. The upgraded CMAQ and NAM both had equally reduced ozone prediction biases, however, there was a smaller effect on PM2.5. By increasing the cloud’s opaqueness, which also caused a reduction in incoming shortwave radiation, the NAM 2 m temperature improved, which often time led to improvement in CMAQ ozone predictions. In this study, the NAM’s ability for predicting meteorological fields such as temperature, cloud cover, near surface winds, and capturing boundary layer processes were evaluated for an episode in Maryland and New York during June 2017. Ceilometer data at two sites were used to evaluate model planetary boundary layer height predictions. In addition, the NAM CONUS Nest (3 km horizontal resolution), higher resolution operational NWP model will be evaluated for its utility to driving air quality model predictions.