Two Looks at One Comet, a Month Apart, Only Half Agree
On Feb. 9, 2023, astronomers pointed an infrared spectrograph at a comet and measured the temperature of the water gas pouring off it. The warmest water was not at the comet. It was a little way out, in the cloud of gas and dust around it, and on the side facing away from the Sun it stayed warm for longer. That reading is now peer-reviewed and published, and it is still an open question, because a month after that night the same collaboration looked again with JWST and ALMA and could not find it.
The comet was C/2022 E3 (ZTF), which for a few nights around its closest approach to Earth was just visible without binoculars from a properly dark site, reaching a visible magnitude of about 5. It passed within about 0.28 astronomical units of us on Feb. 1, roughly a quarter of the Earth's distance from the Sun, and that is what made the measurement possible: from that range the iSHELL spectrograph at NASA's Infrared Telescope Facility in Hawaii could map how conditions changed across the inner coma instead of averaging the whole cloud into one number. The results were published Sept. 11 in The Planetary Science Journal, by Nathan X. Roth and colleagues.
Most of what that night produced is not in dispute. iSHELL securely detected eight molecules in the coma, among them water, carbon monoxide, methane, hydrogen cyanide and carbonyl sulfide, and set most of them below the average for comets from the Oort cloud, the distant reservoir that supplies long-period visitors. Carbonyl sulfide was the exception. The comet's water output also swung over the course of the night, which the authors tie, cautiously, to the spin of the nucleus. Both of those have been seen at other wavelengths: Biver and colleagues, observing with the IRAM 30-meter telescope and the Odin space telescope, found the same pattern of low abundances, though not the same numbers molecule by molecule, and the same short-term swing in output, possibly linked to the comet's roughly 8-hour rotation. Several of their co-authors are also on the iSHELL paper.
What iSHELL reads is a rotational temperature, the way energy is shared among tumbling water molecules, and it tracks the temperature of the gas only where the molecules are packed closely enough to keep colliding. Close to the nucleus they are. Along the line from the Sun to the comet, the water registered 84 K at the nucleus, roughly 190 °C below freezing, and rose to 88 K about 135 km out on the shadowed side. The uncertainties are a couple of degrees, so the rise is real and small: a few degrees, not a hot spot.
The firmer result is farther out. On the sunward side the water cooled to about 65 K within 320 km of the nucleus; on the shadowed side it did not reach that temperature until nearly 550 km. A coma cools as it expands, and this one did. It just cooled at different rates in different directions, and that asymmetry, unlike the small off-nucleus peak, survived the second look.
For both patterns the authors have a preferred explanation. Grains of dust carrying water ice are pushed away from the Sun by sunlight itself; carried out into the coma, they turn their ice straight to gas some distance from the nucleus. Water released that way arrives warm, and it arrives where the grains are rather than at the surface. The paper says its profiles are consistent with that kind of icy grain sublimation and may indicate it.
What makes the preference more than a guess is the alternative they set aside. Electrons in the charged gas around a comet can heat its coma too, but that happens near a particular boundary, and modeling puts it about 750 km from this nucleus, against temperature rises seen within a couple of hundred kilometers. The same double signature has been reported before, in comet 103P/Hartley 2, where infrared mapping found two modes of water release, one at the surface and one out in the coma.
Then there is the other paper. On March 1 to 4, 2023, about a month after the iSHELL night, JWST and ALMA were turned on the same comet, and the results were published in February 2026 in The Astronomical Journal, with Kiernan D. Foster of the University of Virginia as first author. On the asymmetry the two agree: the shadowed hemisphere was warmer, and Foster's team reports that enhancement as statistically significant. On how temperature changed with distance they part company. Foster and colleagues saw it fall steadily outward on both sides, with no evidence at all of a rise off the nucleus.
This is not two camps. Several of the same people are authors on both papers, Roth and Foster among them: one collaboration, reporting two sets of observations that do not match. The two looks were also not the same measurement. The February peak is in water alone, while the March data cover methanol as well, seen by two instruments that sample the coma differently, and the two papers reach for different physics to explain why temperature falls off with distance at all.
The explanation the February team offers for the mismatch is the most interesting line in the paper, and it is a proposal rather than a settlement. The thermal physics of this coma, they write, did not necessarily stay constant with time, so icy grain sublimation may not have been a constant effect. On that reading neither measurement is wrong. The comet was shedding icy grains near its closest approach and had largely stopped a month later, farther from the Sun. That would mean a single night's observation, however sharp the picture, captures the comet only as it was that night.