The South Fork fire complex had been burning since being started by lightning on 31 July 2014 (InciWeb), and a pyroCb was seen to develop late in the day on 09 August 2014. A 4-panel comparison of GOES-15 (GOES-West) and GOES-13 (GOES-East) 0.63 µm visible channel and 3.9 µm shortwave IR channel images (above; click image to play animation; also available as an MP4 movie file) showed that there was a fire hot spot flare-up around 22:00 UTC (red pixels on the shortwave IR images), with the rapid development of the pyroCb cloud beginning around 23:45 UTC.
The corresponding GOES-15 and GOES-13 10.7 µm IR channel images (below; click image to play animation; also available as an MP4 movie file) revealed that cloud-top IR brightness temperatures of the pyroCb cloud became colder than -40º C (brighter green color enhancement) after 00:15 UTC; the coldest pyroCb cloud-top IR brightness temperature was -43.4º C on the 01:30 UTC GOES-15 R image.
Using the 00 UTC 10 August rawinsonde data from Salem, Oregon (below), the -40 to -43º C cloud-top IR brightness temperatures roughly corresponded to altitudes around 10-10.5 km (pressure levels of 280 to 254 hPa). The presence of northeasterly winds in the upper troposphere explained the southwestward drift of the pyroCb cloud. The tropopause was at a height of 13 km (178 hPa), at a temperature of -61.7º C.
NOAA ARL HYSPLIT model forecast trajectories initialized at the location of the first -40º C cloud-top IR brightness temperatures (below) suggested an anticyclonic transport path that would bring the high-altitude smoke back northward over Oregon and Washington within 72 hours.
During the following overnight hours, a comparison of Suomi NPP VIIRS 3.74 µm shortwave IR and 0.7 µm Day/Night Band images at 09:49 UTC or 2:49 am local time (below) showed the hot spots (yellow to red color enhancement on the shortwave IR image) and the bright glow of the fires on the “visible image at night” Day/Night Band.