Review of Jann Schrod and Heinz G. Bingemer, A view on recent ice-nucleating particle intercomparison studies: Why the uncertainty of the activation conditions matters

This manuscript, and the associated supplement, describe in detail a cause of deviation between commonly used instruments which determine the abundance of ice nucleating particles in laboratory and field studies. It has been widely noted, and specifically in recent papers in AMT and ACP, that ice nucleation measurements are critical for understanding climate, cloud formation, precipitation and a host of other atmospherically relevant issues.

The authors’ focus is on uncertainty in temperature, which is one commonly noted cause for divergence of measurements. Although this paper is very narrowly focused, it is none the less an important addition to the literature. Several suggestions for improvement are listed below. After revisions, this paper could be suitable for publication in AMT.

This paper would benefit from a thorough re-write to emphasize what it is and what has been done previously.

This starts with the title. The paper deals with the impact of temperature uncertainty on INP measurements. It is not a comprehensive look at all uncertainties that impact INP measurements. The title seems to imply the latter, since ‘temperature’ – the sole focus – isn’t mentioned. I suggest “A view on recent ice-nucleating particle intercomparison studies: Why temperature uncertainty of the activation conditions matters” is actually what is done in this paper. I further suggest the authors go through and make the wording in the paper (namely abstract) consistent with their temperature uncertainty theme and focus.

The paper has a decided tone that temperature uncertainty haven’t been addressed previously. This is absolutely untrue. Essentially all, starting with even the earliest papers on INP measurement, addressed temperature uncertainty. The line in the abstract, “A regularly overlooked reason for these discrepancies may be related to uncertainties and errors in the temperature measurement.” is therefore not true – temperature uncertainty isn’t “overlooked”, it can be found in essentially all papers on this topic. To illustrate this point, a recent intercomparison paper referenced here (DeMott et al., 2024) uses the term ‘temperature’ no less than 100 times and details temperature uncertainty and other factors, e.g., aerosol spreading in instrument lamina – mentioned in this paper only briefly - among a more comprehensive list of reasons for experimental divergence.

I suggest the authors go through and make the wording in the paper consistent with this being an extension of what has been previously discussed in INP papers on temperature uncertainty and not as if this is the only, or one of a very few, doing so. In fact, on page 4, the authors note they couldn’t find sufficient temperature data in their 10 listed intercomparisons in only 21% of cases. That implies that there is significant treatment in the vast majority of cases.

To offer a compromise, I believe what the authors are driving at is (1) they have conducted a more thorough treatment of how temperature uncertainty can impact INP concentration - the value of this paper - and (2) that intercomparisons should provide more information, consistent with the quantities they list towards the bottom of page 3 and elsewhere in the text. This is a suggestion that can be emphasized in the abstract and conclusions (i.e., that more data should be provided). The statements that temperature uncertainty aren’t addressed is, by their own references, simply not true.

In conclusion, where this paper shines is in its treatment of how much deviation can be expected from a given level of temperature uncertainty. This is an important addition to the literature and is worthy of publication. The paper suffers from is the incorrect and overly strong statements that temperature is not treated in previous papers. If the latter can be eliminated this work will be much stronger.

Review of Shrod et al. A view on recent ice-nucleating particle intercomparison studies: Why the uncertainty of the activation conditions matters

This manuscript addresses the uncertainties associated with in situ ice nucleation measurements based on a review of a number of intercomparison studies made over the last ten years. It illustrates how temperature uncertainties can be the result of the large spread in the range of ice nucleation measurements. They calculate these estimated uncertainties as a function of temperature and determine an error Factor (EF). This EF is then evaluated using a number of commonly used parametrizations from the literature. This paper is well written and pleasant to read. It is a very important topic to address and this work will likely contribute to the motivation to generating new measurement guidelines and approaches when comparing instruments.

However at the end of the manuscript, the impression is that, the current uncertainty in ice nucleation measurements is so great that we cannot rely on these measurements when interpreting particle ice-nucleating properties. It would be a useful addition to this manuscript to include a list of recommendations that can be brought forward into future measurements. The community is already striving to reduce the uncertainties in the measurements.  

• Are there some methods that have shown to have smaller uncertainties and more reliable measurements?
• Should the community compare similar instruments (same make/model) and avoid comparing different types of ice nucleating measurements?
• How can these temperature measurements be improved?

The submitted manuscript, "A view on recent ice-nucleating particle intercomparison studies: Why the uncertainty of the activation conditions matters," from Schrod and Bingemer, is a detailed investigation into the potential consequences of temperature uncertainties associated with various measurement techniques, targeted at determining ice nucleating particle (INP) concentrations.  I find the manuscript well written, fairly easy to digest, and also note that it raises an important point of discussion for the community participating in these measurements.  Although the title is somewhat general, the submitted manuscript really only deals in detail with temperature uncertainty.  As the authors point out, even though many investigators discuss and consider temperature uncertainty, the common practice is that results like activation curves are not reported with ΔT error bars.  I believe that the author's main point is that, especially when activation curves are steep, small temperature changes mean pronounced changes in INP activity parameters.  Although there are several technical corrections needed throughout the manuscript, I would suggest that with those corrections this is a suitable manuscript for publication and a valuable contribution to the field.  I suggest points for technical correction in (page number, line number) form below.

Itemized technical corrections: 

Abstract -- The authors should consider (for the atmospheric community) if they would like to make a more clear distinction between there choice of EF to mean "error function", given this is also a common notation for "emission factor".

(page 2, line 36) -- strike "out", should simply read, " ....ice particles precipitate earlier..."

(3,77) "Per" is strange to begin a sentence with, I would suggest, "For every 5..."

(3,80)  replace "highest" with "most" and "aerosol" with "particles" 

(3,85) strike "according to Eq. (1) and Eq. (2):" and replace with "as"

(3,88) should be "methods sections"

(4,95) I suggest replacing "of" with "corresponding to the"

(4,103) Rephrase: We considered studies for our investigation when the following criteria are all met:

(4,104) specify "published since 2015" as in the last 10 years will not age well.  

(4, 109) rephrase, "We identified..."

(4, 111) strike "a few"

(4,124) The sentence beginning, "A number of uncertainty estimates...." needs to be clarified and/or expanded.  What kind of interpretation?  

(5, 140) -- When discussing the Castarède et al., 2023, paper I would suggest the authors also highlight that in this manuscript the authors make some effort to argue that in CFDCs (in particular, PINCii there) it might be that the important activation condition is in fact the strongest thermodynamic forcing condition present within the chamber at a given time.  The details appear to be discussed more in depth in the first author's PhD thesis.  But this raises an important issue with online type instruments. At times activation conditions are the important reported parameter, not simply INP counting.  The two types of measurements will not be impacted in the same way by temperature uncertainty of such chambers.

(6, 175) The range of temperatures is strangely presented.  Mixing digits and text, and "tenth" should at a minimum be "tenths" I believe.  

(8, 198) suggest: ....different instruments usually agree to within 1 order....

Table 2: n_m is introduced in table without first being defined in text.  Also in the first bullet related to the DeMott et al., 2017 paper, it is not clear weather differences get smaller or larger as concentrations go above or below 1 INP/L. Please rephrase so intent is clear.

(10, 247) perhaps: ...density per unit surface...

(11, 260) "upper grey line" and in fact all of the "grey isolines" referred to in the caption are extremely difficult to distinguish.  

(11, 262) This claim that "a hypothetical instrument mistakenly assuming to measure at -25 would actually report nINP of the true temperature of -23.5 is difficult to visualize with the presented figures.  Can the reader be coached through how to understand this?  

(12, 282) suggest: δT is larger

(12, 293) suggest: ...is acceptable, increasing only to an EF...

Conclusions:  I think for the offline droplet/assay freezing methods the fact that time dependence is largely ignored needs to be mentioned again (as I believe it is in the introduction) in the paragraph spanning pages 13 and 14.

(15, 369) suggest "possibly" should be replaced by "possible"

(15, 380)  This is a great question, and I applaud the authors trying to take one step to solving this underlying problem.

Appendix B: suggest: The following figures provide further details for (or perhaps from) the analysis presented in....