Extension of WoFs-Smoke to Ingest GOES-R Fire Radiative Power Retrievals

The Warn-on-Forecast System (WoFS) was recently extended to include the ability to forecast smoke aerosols generated by wildfires resulting in a new system known as WoFS-Smoke. This system generates 0-6 hour probabilistic forecasts of smoke and related parameters in a regional domain with a horizontal grid spacing of 3 km. Smoke aerosols are injected into the atmosphere using a 1-D plume height simulation model that relates the satellite retrieved fire properties to the potential plume height. Smoke is then allowed to be transported by the model downstream and interact with its thermodynamic environment. The current system is based on methods employed by HRRR-Smoke, which primarily uses fire properties retrieved from polar orbiting sensors. Using only these data represents a significant limitation since the temporal sampling of an individual fire may be on the order of hours. Thus, smoke from rapidly developing and evolving wildfires is not accurately forecast. To address this limitation, WoFS-Smoke has been modified to utilize retrieved fire properties from GOES-R series of satellites, which are available at a much higher temporal frequency. The new version of WoFS-Smoke ingests wildfire properties at 15 minute intervals corresponding to each data assimilation cycle. Comparing WoFS-Smoke experiments with and without GOES-R fire properties for several extreme wildfire events showed that including these data improved smoke forecasts associated with the developing phase of the fires. Additionally, improvements in the forecast evolution of smoke concentrations was observed due to greater sampling of changes in wildfire characteristics. In particular, using these data allowed for quicker shutoff of smoke in the system for fires that became inactive. This version of WoFS-Smoke is currently being tested in a realtime configuration.