Spectroscopic identification of water emission from a main-belt comet

Main-belt comets are small Solar System bodies located in the asteroid belt that repeatedly exhibit comet-like activity (that is, dust comae or tails) during their perihelion passages, strongly indicating ice sublimation1,2. Although the existence of main-belt comets implies the presence of extant water ice in the asteroid belt, no gas has been detected around these objects despite intense scrutiny with the world’s largest telescopes3. Here we present James Webb Space Telescope observations that clearly show that main-belt comet 238P/Read has a coma of water vapour, but lacks a significant CO2 gas coma. Our findings demonstrate that the activity of comet Read is driven by water–ice sublimation, and implies that main-belt comets are fundamentally different from the general cometary population. Whether or not comet Read experienced different formation circumstances or evolutionary history, it is unlikely to be a recent asteroid belt interloper from the outer Solar System. On the basis of these results, main-belt comets appear to represent a sample of volatile material that is currently unrepresented in observations of classical comets and the meteoritic record, making them important for understanding the early Solar System’s volatile inventory and its subsequent evolution.


Fig 1:
The ice features in the 103P spectrum shown here are very hard to see on this scale, and I don't think that these are relevant.It is perhaps confusing to show two different 103P spectraone would be sufficient to make the comparisons with 238P.Pg 6.It is perhaps worth noting that MBCs being different to other comets, and therefore the limits on water production derived from CN limits not necessarily being all that strong, was discussed/predicted previously (e.g. in the review by Snodgrass et al that you cite earlier in the paragraph).It is good to have an observational confirmation of this.
Pg 7. The authors give radii for active 'patches' corresponding to the active areas required (in a table in the methods section).The authors could consider giving these radii in the text here, and discussing whether or not these patch sizes support (or not) popular MBC activation theories, such as a recently uncovered area of ice in a crater.At 100-200m in radius the areas are perhaps larger than have been predicted in some models, and would be a large crater, but perhaps not unexpected on a body this size.Can anything be said about expected impact rates / crater sizes / the population of currently active MBCs that would be expected to correspond to this? Pg 9/10.The only point in this letter that I have some (small) doubts about is the section on comparing dust-to-gas ratios with other comets.I think that the opening sentence of this paragraph (that dust-to-gas infers 'composition') promises a bit much.There are many unknowns.I think that it is reasonable to say that dust-to-gas can be compared with other comets to say whether or not this MBC is typical or unusual, and that the arguments given show that it could have a range of values that overlap with other comets, but care must be taken not to read too much into it.It would also be worth adding a caveat that the comparison with Rosetta results on 67P is really comparing very different types of measurement (to an already highly debated number).
Pg 22.I'm not sure how helpful the activity timescale calculation is, given the many uncertainties involved, and because this ends with a slightly hand-waving 'few orbits' to end up with a number that is consistent with what we already know observationally -that this (and other) MBCs are active for at least a few orbits.Perhaps this section would be better with a comparison with previous thermal model predictions for active lifetimes for MBCs, and some discussion as to why the initial ~1 year timescale is very short.Given that the 2nd model doesn't appear to be a very good fit, and it isn't described anywhere, I don't think it needs to be shown.Comments from the reviewers are marked with ">".
We have realized that the sublimation temperatures of ammonia ice and carbon dioxide ice are similar, which adds interesting complexity to the lack of CO2 coma and the presence of the rounded 3-μm band, possibly due to ammonium salts.We have added a comment to this effect: "However, ammonia and CO2 ice have similar sublimation temperatures, therefore it may be that Read had these volatiles in the past, but they have since been lost." > Fig 1: The ice features in the 103P spectrum shown here are very hard to see on this scale, and I don't think that these are relevant.It is perhaps confusing to show two different 103P spectra -one would be sufficient to make the comparisons with 238P.This is a good point, and we have removed one, specifically leaving the "icy" spectrum since we do compare to it in Fig. 3.The caption is edited accordingly.> Pg 6.It is perhaps worth noting that MBCs being different to other comets, and therefore the limits on water production derived from CN limits not necessarily being all that strong, was discussed/predicted previously (e.g. in the review by Snodgrass et al that you cite earlier in the paragraph).It is good to have an observational confirmation of this.
We have added a sentence to explicitly cite this point: "This conclusion is in agreement with previous predictions that the CN/H2O ratio of the general comet population may not be representative of main-belt comets.^[Snodgrass et al. 2017]" > Pg 7. The authors give radii for active 'patches' corresponding to the active areas required (in a table in the methods section).The authors could consider giving these radii in the text here, and discussing whether or not these patch sizes support (or not) popular MBC activation theories, such as a recently uncovered area of ice in a crater.At 100-200m in radius the areas are perhaps larger than have been predicted in some models, and would be a large crater, but perhaps not unexpected on a body this size.Can anything be said about expected impact rates / crater sizes / the population of currently active MBCs that would be expected to correspond to this?
We have investigated this somewhat, and find that 10-m impactor is needed to account for the active area.However, the energy of such an impact is enough to disrupt the nucleus, given nominal assumptions.Therefore, it initially seems unlikely, but perhaps the parameters of sub-catastrophic impacts may be tuned to produce the needed active area.We state these conclusions in a new paragraph in the main text and a new section in the Methods (pages 3 and 11), but further investigation should be the topic of a separate work.> Pg 9/10.The only point in this letter that I have some (small) doubts about is the section on comparing dust-to-gas ratios with other comets.I think that the opening sentence of this paragraph (that dust-to-gas infers 'composition') promises a bit much.There are many unknowns.I think that it is reasonable to say that dust-to-gas can be compared with other comets to say whether or not this MBC is typical or unusual, and that the arguments given show that it could have a range of values that overlap with other comets, but care must be taken not to read too much into it.It would also be worth adding a caveat that the comparison with Rosetta results on 67P is really comparing very different types of measurement (to an already highly debated number).
Certainly the method by which one would infer the interior properties based on the coma demands assumptions, which we largely left unstated.Regardless, we agree with the reviewer that there are many unknowns.This paragraph, now in the Methods, as been revised.The intro paragraph to the sub-section "Dust-to-gas ratio" now reads: "The coma dust-to-gas ratio may be measured from our data and compared to other comets."We quote ~1 instead of 0.85 for the dust-to-volatiles mass ratio of 67P, and refer the reader to Choukroun et al. ( 2020) for more discussion on the topic.Finally, our concluding sentence has been revised accordingly: "Therefore, our conclusions are that comet Read has a [coma] dust-to-gas ratio [broadly] consistent with the general comet population, which [suggests it may have] formed in a region of the protoplanetary disk with abundant water ice."> Pg 22.I'm not sure how helpful the activity timescale calculation is, given the many uncertainties involved, and because this ends with a slightly hand-waving 'few orbits' to end up with a number that is consistent with what we already know observationally -that this (and other) MBCs are active for at least a few orbits.Perhaps this section would be better with a comparison with previous thermal model predictions for active lifetimes for MBCs, and some discussion as to why the initial ~1 year timescale is very short.
We now reference the more physical analyses of Prialnik andRosenberg (2009) andSchorghofer (2016).But since this is the first water production rate measurement, we think that some attempt to use the observed value is useful.To avoid the hand waviness, we have also revised our calculation to instead estimate the total water mass lost considering the rh**-3 activity dependence from the model of Hsieh et al. (2009).In our revised text, the dust-to-gas mass loss rate ratio leads us to suggest that the ice layer retreats faster than the surface.Therefore, the surface devolatilizes with time, which would ultimately quench activity.> Fig 4 .I assume that the apparent separate bright spot that is just separated from the central condensation in the F227W image is some artefact?It would perhaps be worth commenting on in the caption at least.Indeed, it appears to be an artifact in a single frame, perhaps due to 1/f noise on the detector.We have commented on this fact in the caption.> Fig 5 .The two models shown in this figure are not described in the text.Given that the 2nd model doesn't appear to be a very good fit, and it isn't described anywhere, I don't think it needs to be shown.
The second model was a fit to a smaller wavelength range and we have removed it from the figure.In the "Methods: Reflectance spectrum sub-section" we clarify that two models with different wavelength ranges were fit to the data and analyzed to assess the thermal emission.> Fig 8 .Please add an error bar for 238P in this plot (or explain in the caption if this would be too small or too large to plot).
The error bar was about the same size as the marker.However, we have revised our uncertainty calculation on this value.Because this ratio is based on a comparison of NIRSpec and NIRCam data, the absolute calibration uncertainty needs to be accounted for.The log10 uncertainty increased from 0.04 to 0.06.We have also made the marker fill color transparent so that the error bar can be more clearly seen.

Fig 4 .
Fig 4. I assume that the apparent separate bright spot that is just separated from the central condensation in the F227W image is some artefact?It would perhaps be worth commenting on in the caption at least.

Fig 8 .
Fig 8. Please add an error bar for 238P in this plot (or explain in the caption if this would be too small or too large to plot).