[GaryNumanVevo]

I've found the HRRR experimental smoke models to be quite good. It's a bit cumbersome to use, but good enough for 18H forecasts

https://rapidrefresh.noaa.gov/hrrr/HRRRsmoke/

[cmason]

In my experience these have been completely wrong recently for my neck of the woods (PDX) even at 6-12 hours. Cliff explains in one of his posts that this is because of not modeling the inversion. I’m having better luck with the Copernicus CAMS forecast in Windy