An all encompassing Fire RGB allows a forecaster to monitor three important components to an ongoing wildfire in a single RGB: the wildfire hot spot, the smoke plume, and the burn scar. In order to do so, the RGB combines the 3.9 um shortwave IR channel for hot spot detection (red component), the 0.87 um component for land change (green component), and 0.64 um for smoke (blue component). This RGB is fairly similar to the “Day Land Cloud Fires RGB”, already in AWIPS, but allows for detection of smaller/cooler wildfires given the inclusion of 3.9 um vs 2.25 um, and easier detection of smoke plumes and burn scars given tweaked ranges. This RGB was first discussed here.

One-minute imagery from GOES-East was available over the southwest US on the 16th courtesy of WFO Tuscon with a reasoning of “Several IMETs deployed”. Viewing the Mangum fire in north-central Arizona during the afternoon of 6/16/2020 using the Fire RGB, the large hot spot is obvious as several red pixels (large red contribution with small green and blue; Fig 1). The thick smoke plume emanates from the hot spot to the northeast, and is characterized by medium cyan colors (similar green and blue contributions, small red). Finally, areas that had burned the previous days to the southwest of the hot spot are almost black with little contributions from all three components. Elsewhere, highly vegetated areas (such as the region around/in which the fire is burning) appear deep green, the Colorado River and other waterways are dark blue to almost black, while non-vegetated areas are dull green. The 4-panel in Fig 2 includes the RGB during the same period along with the three components.

The AWIPS menu entry for this “Fire RGB” is available upon request.

VIIRS DNB NCC imagery captured the light from the wildfire during the overnight hours early on the 16th in three passes (two from SNPP, one from NOAA-20; Fig 3). The active areas of the wildfire are obvious as a ring of light around previous burned area. The most active region of the wildfire, on the northern front, is also diagnosed from the imagery. Finally, a relatively dim glow adjacent to the wildfire is suspected to be scattering of light off of the smoke/haze. Changing position of the wildfire and suspected haze glow is due to parallax given different viewing angles with each of the three frames.

The HRRR-Smoke model, which utilizes information from polar-orbiting satellites (VIIRS, MODIS), provides a prediction as to how smoke will evolve from the fire in the near future. In this case, the model (early morning cycle) accurately predicted smoke associated with the Mangum fire to spread northeast across southeast Utah, northwest Colorado, and into southern Wyoming by the evening (Fig 4). This information can be used by local officials to anticipate smoke reports from the public, and associate them with a particular fire upstream.

Bill Line, NESDIS and CIRA