Now we know that the fire featured a rare phenomenon that demonstrates just how extreme it was, with the National Weather Service’s announcement Wednesday afternoon that two fire tornadoes were associated with the blaze.
Unlike firewhirls, which can commonly be seen forming from the ground up over campfires and even small brushfires, fire tornadoes are actual tornadoes. They are born from rotating clouds or smoke plumes in environments where winds change speed and/or direction with height. The funnels themselves aren’t filled with fire, but rather a superheated column of smoke, ash and debris.
Two tornadoes have been confirmed thus far from the blaze on Sept. 5 — one near Mammoth Pool Reservoir in the Sierra National Forest well southeast of Yosemite, and another near Bass Lake, about five miles east of Oakhurst. Mammoth Pool Reservoir is the location where more than 200 people were safely evacuated by military helicopters because of the lack of an escape route as the fire rapidly advanced.
The National Weather Service in Hanford, Calif., which was responsible for rating the tornadoes, determined their strength based on observed tree damage. Ordinarily, the surveys would have been conducted sooner, but visibility limitations due to remnant smoke made getting to the damaged regions challenging until recently.
Jerald Meadows, the warning coordination meteorologist at the Hanford office, said both tornadoes shared common features. “The main contributing factor was the debarking of all the pine trees up with the Mammoth Pool tornado,” Meadows said. “They both uprooted trees to the root balls and snapped large pines. But the [EF1 tornado] did not have any signs of true debarking. We’re probably talking the difference between 100 and 110 miles per hour.”
The Mammoth Pool tornado, which touched down inside the Wagner Campground, snapped several two-foot-diameter trees about 20 to 30 feet above the ground; it was rated as having winds of 115 to 125 mph. The Huntington Lake fire tornado had winds of 90 to 107 mph, and the NWS noted that it was “the result of unprecedented fire activity.”
The surveys were conducted by an incident meteorologist from the National Weather Service already dispatched to the fire. It’s common for meteorologists to be sent to the field to assist fire personnel in predicting fire behavior and keeping firefighters aware of evolving fire behavior.
That particular meteorologist tweeted that he believed the EF1 tornado was a “strong, short-lived event,” since the felled vegetation did not appear to be sandblasted. During stronger or longer-duration tornadoes, dirt from the ground can become swept into the vortex and buffet areas in the twister’s path.
The EF2 tornado near Mammoth Pool was associated with an impressive radar signature at 3:17 p.m. local time Sept. 5. The smoke plume associated with the Creek Fire had towered to 41,000 feet — the height of a strong thunderstorm — and the entire plume was rotating clockwise at the time.
That’s opposite to how most tornadic thunderstorms in the Northern Hemisphere spin, but it appears that weather conditions have repeatedly favored clockwise-spinning rotations in California’s smoke plumes this year.
On Doppler radar, one can also discern the rotational couplet that produced the tornado. Pixels plotted in red mark winds moving away from the radar, while green denotes wind blowing toward the radar.
It’s easy to spot the two bright splotches of color juxtaposed against each other where the tornado was ongoing.
Shortly after the tornado occurred, the plume billowed to what one fire researcher referred to as an “insane” 55,000 feet. That means that the plume was probably undergoing quick vertical development with a strengthening updraft at the time of the tornado, which probably aided in its formation.
“A tornado warning was considered but not issued,” said Meadows, who feared that disseminating such an alert might leave people unnecessarily conflicted about deciding whether to shelter or evacuate.
“A tornado warning for a fire opens up a can of worms,” he said. “We want to make sure we’re messaging properly, and we were talking to fire crews letting them know of the circulations we were seeing.”
Before the onslaught of fire tornadoes that has been a hallmark of 2020′s blazes, some National Weather Service offices have had internal discussions and concluded that they would not issue tornado warnings for wildfire-related twisters. While the National Weather Service hasn’t issued specific policy guidance to its 122 forecast offices on how to handle fire tornadoes, Meadows suspects considerable research will be needed to reach a resolution.
“Being a government agency, it takes time for us to develop policy,” he said. “We have to make sure we have science behind us 100 percent. [Fire tornadoes] in general are different dynamically than a normal tornado. We need to get the science down before we start building policy about when we alert the public about such things. We don’t want to overburden emergency services or the public with limited science.”
Fire tornadoes have happened before, but they’ve been observed with an unusual frequency this year. A deadly EF3 fire tornado occurred with the Carr Fire in Redding, Calif., in July 2018, and marked the first instance of an Enhanced Fujita scale rating being assigned to fire tornado damage in the United States.
Previously, fire tornadoes came to prominence after a high-end EF3 fire tornado that traveled 12 miles through the Canberra, Australia, bush fires of 2003. Fire tornadoes and other extreme fire behavior were also observed during the devastating Australian bush fires of 2019-2020.
In the case of the 2018 Redding fire tornado, damage to high-voltage electrical towers led meteorologists to the conclusion that winds up to 143 mph were probably present in the vortex. That places it among the top three strongest tornadoes ever to have occurred in California.