The fire was in Siberia. Its smoke had crossed thousands of miles, drifting over the Arctic, across Alaska and then over northern Canada. For scientists working on the first Injected Smoke and Pyrocumulonimbus Experiment field campaign, that distance meant an opportunity.
For David Peterson, a research meteorologist at the U.S. Naval Research Laboratory in Monterey, California, and principal investigator for INSPYRE, the campaign represented the culmination of years spent studying how wildfire smoke enters and moves through the upper atmosphere.
"I've been fascinated by weather since I was a kid, and I never really grew out of it. I went on to study meteorology in college, and for the past decade, I've been studying these extraordinary storms that are generated by wildfires," he said. "There's still so much we don't understand about them, and I think that's what makes the work so exciting. The more we discover, the more questions we find."
Using aircraft furnished with scientific equipment, an NRL-led team intercepted the Siberia-generated plume over three consecutive days, collecting direct measurements of smoke carried high into the atmosphere. The observations will help researchers address a gap in atmospheric forecasting: predicting where the smoke plume travels, at what altitude, and how it changes the air surrounding it.
The sampling capped the 2026 field deployment of the NASA-funded experiment, which lasted three months. The campaign studied pyrocumulonimbus clouds, or fire-generated thunderstorms. These can produce lightning that ignites additional fires, winds that spread flames, and hail and violent downdrafts that threaten low-flying aircraft.
The mission relied on a NASA ER-2 aircraft, which flew above active weather systems with instruments that observed storms from high altitude, and a U.S. National Science Foundation Gulfstream V aircraft, which took samples directly from inside the smoke plumes. The aircraft logged more than 100 flight hours each, completing 16 and 17 flights, respectively.
They observed fires from Arizona to Canada and followed transported smoke as far as Alaska. A ground team also deployed mobile radar and lidar systems near active fires.
The mission operations center was based at Rocky Mountain Metropolitan Airport in Broomfield, Colorado. The ER-2 aircraft operated from the Montana Air National Guard base at Great Falls International Airport.
On the ER-2, equipment measured terrestrial gamma-ray flashes associated with lightning activity inside fire-generated storms. Other instruments measured the size and intensity of the fire fronts, smoke plume dynamics and the surrounding atmospheric conditions.
"We know that large and intense wildfires tend to generate their own weather, and now we finally have direct measurements of it," Peterson said.
The measurements address a central challenge for INSPYRE. Researchers have long observed smoke from satellites, but satellite imagery cannot reveal everything inside a plume. Scientists must make assumptions about particle size and how particles absorb or reflect sunlight. Those properties help determine the plume's effects on the atmosphere and on light passing through it.
For Peterson, replacing some of those assumptions with comprehensive measurements was one of the campaign's most important achievements.
During the Siberian plume sampling, both aircraft observed the smoke, with some flights coinciding with satellite overpasses. The ER-2 also descended to about 38,000 feet to sample the plume directly. Researchers can now compare measurements taken inside the smoke with observations from above and use them to improve atmospheric models.
Together, the observations gave Peterson and his team a rare view of the same smoke event from multiple perspectives.
"We were able to follow the smoke as it moved across the [region]," Peterson said. "That gives us a chance to connect what happened at the fire with what we eventually measured thousands of miles away."
For the Navy, locating a large smoke plume is only part of the challenge. Predictions of its altitude and particle properties matter for visibility, aviation and optical systems. The NRL supplied additional ER-2 flight-hour funding to support measurements relevant to those needs.
"From a Navy perspective, lasers, satellite communication and targeting cameras will be affected by these smoke plumes," Peterson said. "If you're shooting a laser through a smoke plume or a dust plume, it degrades how the systems work."
Peterson often compares fire-generated clouds to giant chimneys that can propel smoke into the stratosphere, much like a volcanic eruption.
At a fire in Utah, researchers observed a cloud pushing smoke to about 36,000 feet without meeting the team's working definition of a pyrocumulonimbus. The case raised a practical question about whether other types of fire-driven clouds can produce large smoke plumes high in the atmosphere that matter for forecasting.
"You can see it literally growing in front of you as you're moving toward it," Peterson said, describing the fire-driven convection he saw firsthand aboard a smoke-plume flight.
That moment captures the challenges fire behavior can bring. A cloud that appears ordinary from a distance can rapidly become an efficient pathway for smoke to reach the upper atmosphere.
Instrument teams are now checking and processing the data. Research groups will examine fire behavior, cloud development and transported smoke, while preparing for a second deployment in 2027.
For Peterson, the next phase is as important as the flights themselves. The team must determine which observations can improve models, identify where current forecasts diverge from reality and translate the campaign's measurements into tools that can support operational decisions.