Imagine you're watching a pot of water boil so violently that the steam column punches upward into a sheet of plastic wrap stretched across the top of the pot. The wrap bulges upward, forming a smooth dome — and then the boiling column rips through it and keeps going. That dome is a pileus cloud. This NASA observation caught one forming atop a fire-generated thunderstorm column, and for the first time connected the full chain: ground fire behavior, convective plume dynamics, and cloud microphysics, all from a single airborne pass. On August 11, 2026, NASA's ER-2 high-altitude research aircraft flew over Montana's Sand Creek fire as part of the INSPYRE (Injected Smoke and Pyrocumulonimbus Experiment) campaign. The aircraft carried MASTER, a 50-band airborne imaging spectrometer covering wavelengths from 0.66 to 11.33 micrometers, along with downward-pointing radars. The timing was fortunate: the fire was making a rapid run up the western side of Mount Comet, winds aligned with drainage channels, and the resulting updraft was extreme enough to produce a pyrocumulus cloud capped by a pileus. Pileus clouds form when a fast-rising column of air — in this case heated by wildfire — slams into a moist horizontal layer above it, displacing that layer upward like air forced over a mountain. As the displaced moist air rises and cools, water vapor condenses into a thin, smooth cap. The mechanism is well understood from thunderstorms and volcanic eruptions, but observing it over a wildfire with simultaneous multi-spectral ground-truth of the fire perimeter is genuinely rare. The pileus appeared in only one of the roughly 30-minute imaging passes, consistent with their typical lifespan of minutes. The MASTER instrument's shortwave-infrared channels revealed actively burning spots along the fire perimeter (appearing pink and yellow in false color), while burned areas showed as brown and smoke as blue-purple. The natural-color channels captured the pyrocumulus tower and its pileus cap. Critically, the downward-pointing radars simultaneously probed the internal structure of the cloud. Deputy project investigator Neil Lareau described this as the 'core of what we're after with INSPYRE' — connecting surface fire dynamics to convective plume to cloud processes in a single coordinated observation. The scientific value here is methodological rather than theoretical. Pileus formation physics isn't new — Garrett et al. (2006) documented pileus-like cirrus near the tropopause above thunderstorms during NASA's CRYSTAL-FACE campaign in 2002. What's new is the observational completeness: 50 spectral bands, radar cloud structure, and surface fire mapping captured simultaneously over a single pyrocumulus event. This matters because the injection height and cloud dynamics of pyrocumulus events determine how high wildfire smoke reaches into the atmosphere, which in turn affects regional and hemispheric air quality and radiative forcing. The INSPYRE campaign is part of a broader push to understand pyrocumulonimbus (pyroCb) events — the most extreme fire-weather phenomena, capable of injecting smoke into the stratosphere. These events are increasing in frequency as fire seasons intensify. The Sand Creek pileus observation is a proof-of-concept for the kind of multi-sensor airborne observation needed to build predictive models of when fires generate their own weather. The limiting factor isn't sensor technology but intercept timing: getting the aircraft overhead during the minutes-long window when these dynamics unfold. For the atmospheric science community, this is a clean demonstration that INSPYRE's observational architecture works as designed. The next question is whether the campaign can catch a full pyroCb event — where the fire cloud punches through the tropopause — with the same sensor suite. That would be the observation that moves the field from documenting pyrocumulus to predicting pyroCb blowups.
