Evaluation of Dust and Smoke Forecasts from NOAA’s High Resolution Rapid Refresh Forecast System (RRFSSD)

NOAA in collaboration with Global Systems Laboratory (GSL) has developed a next generation convection-allowing, rapidly-updated prediction system named as Rapid Refresh Forecast System (RRFS). Smoke and dust model (RRFS-SD) is added in the system to predict more accurate smoke and dust and improve visibility and weather forecasts. RRFS-SD is based on the Unified Forecast System (UFS) using the Common Community Physics Package (CCPP) framework. The model uses high spatial resolution VIIRS I-band and high-frequency GOES-16/17 fire radiative power (FRP) data are ingested into the model to estimate both biomass burning emissions and fire heat fluxes every hour. Inline turbulent mixing of smoke and dust within the boundary layer scheme, hourly wildfire potential to predict the evolution of the biomass burning emissions, smoke interactions with the double-moment microphysics scheme are implemented in the system. The FENGSHA dust scheme developed by ARL is included in RRFS-SD to forecast dust emission. The operational model configuration will feature a 3 km grid covering North America and include deterministic dust and smoke forecasts every hour out to at least 18 hours, and forecasts out to 60 hours four times per day at 00, 06, 12, and 18 UTC.

In this study we have used satellite (VIIRS and GOES) and ground based observations (AIRNOW and AERONET sites) to evaluate both surface concentration and vertically integrated total column extinctions. The capabilities of RRFS-SD are evaluated based on some of the big wildfire and dust events between 2019 and 2023.