Table 2 Summary of model details.
Model | WRF-Chem v3.6.1 | HRRR-Smoke | 1D Freitas |
|---|---|---|---|
Resolution | 4 km horizontal, 53 vertical levels | 3 km horizontal, 40 vertical levels | 70 vertical levels, 100 m spacing |
Initializaiton/Spin-Up | 00UTC initialization, 12 h spin up, 76 h analysis, 84 h forecast, model meteorology passed to Freitas model34 | 00, 06, 12, 18 UTC initializations, 48 h forecasts, model meteorology passed to Freitas model34 | Uses model profles of T, P, geopotential, RH, wind, potential temperature, specific humidity. |
FRE Flux Assumptions | 2.69–13.86 kWm−2 for forest, 0.4–3.10 kWm−2 for woody savanna/cerrado, and 0.57 kWm−2 for grassland/pasture/ cropland. | 3.4 kWm−2 for grassland to 4.8 kWm−2 for forest for 14 vegetation categories as defined by the IGBP28,43,44 | Choose the Freitas maximum heat flux for the dominant fuel category |
Fire Size Assumptions | Converts MODIS/VIIRS FRP into area using fuel-varying heat fluxes | ||
Flaming Fraction of Emissions Assumption | 50% flaming, undergo plume rise | 75% flaming, undergo plume rise | N/A |
Key Parameterizations | Rapid Radiative Transfer Model for GCMs45, Noah Land Surface Model46, aerosol-aware microphysics47, Grell-Freitas cumulus34, Mellor–Yamada-Janjić PBL48, MEE = 3.549, MOSAIC Aerosols50 | Aerosol-aware microphysics47,51 Mellor–Yamada–Nakanishi–Niino PBL scheme52 | N/A |
Data Assimilation | Near-real-time inversion of MAIAC AOD V653,54 based on 6 largest fires in the region19 | None | N/A |
Emissions Source | QFED v2.4, 0.1 degrees resolution, regridded38 | Top-down from MODIS/VIIRS FRP39 | N/A |
Variables Extracted | PM2.5 curtains, model PBL and theta-derived mixed layer heights19, land use fractions, Richardson number and Brunt-Vaisala frequency via centered-difference at twice and half the PBLH. | PM2.5 curtains, model PBL height, Richardson number and Brunt-Vaisala frequency calculated via centered-difference at twice and half the PBLH. | Maximum plume top height |