the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Impacts of Arctic warming on ice nucleating particles over recent decades: Distributions and contributions of dust, marine organic aerosols, and bioaerosols
Abstract. Aerosols serve as ice nucleating particles (INPs) and play a critical role in the formation of mixed-phase clouds. These clouds are prevalent in the lower and middle troposphere of the Arctic and exert a strong influence on both regional and global climate. However, limited understanding of INP sources and their temperature-dependent behavior has hindered accurate predictions of aerosol-cloud interactions in the Arctic. In this study, we investigate the sources, spatial distributions, seasonal variations, and long-term changes of INPs in the Arctic using a global climate-aerosol model that explicitly represents INPs from three Arctic aerosol species: mineral dust, marine organic aerosols (MOA), and bioaerosols. Simulations covering the period 1981–2020 show that Arctic-sourced INPs account for more than 70 % of total INPs in the Arctic lower troposphere. Dust is the largest contributor (36 %), followed by bioaerosols (28 %) and MOA (9 %). They exhibit distinct spatial and seasonal patterns, underscoring the importance of representing multiple INP species and applying appropriate parameterizations for each when modeling INPs and mixed-phase clouds in the Arctic. Over the past four decades, Arctic warming increases local emissions of all three aerosol species by 4.7–18 % because of the retreat of snow and sea ice. Nevertheless, INP concentrations in the Arctic lower troposphere decline by 19–29 %, primarily because the INPs per unit aerosol mass decrease with increasing temperature. This indicates that the temperature-driven reduction of ice nucleating efficiency outweighs the emission-driven increase of INP abundance, except in regions with substantial local increases of emissions.
- Preprint
(5258 KB) - Metadata XML
-
Supplement
(1319 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-6214', Anonymous Referee #1, 05 Mar 2026
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2026/egusphere-2025-6214/egusphere-2025-6214-RC1-supplement.pdfCitation: https://doi.org/
10.5194/egusphere-2025-6214-RC1 -
RC2: 'Comment on egusphere-2025-6214', Anonymous Referee #2, 15 Apr 2026
The paper "Impacts of Arctic warming on ice nucleating particles over recent decades: Distributions and contributions of dust, marine organic aerosols, and bioaerosols" by Ren et al. presents a study on present and future aerosol emissions in the Arctic and their ice nucleating efficiency compared to transported/non-Arctic aerosol sources. They found that the major part of the ice nucleating particles (INPs) in the Arctic come from local sources. They show that for future simulations the Arctic aerosol emissions increase, but the fraction of INPs decrease because of the overall change in temperature. The study is interesting and the aerosol setup of the model with 11 tracers impressive. The findings are very helpful for future research and model development in the Arctic and can contribute to help to understand Arctic feedback processes. However, more explanation on the methodology, how the results were derived and some additional suport figures are needed. I therefore suggest that the paper should undergo major revisions (I hope all points can be adressed and the quality of the paper be improved so that it can be published and its results made available for the research community).
Major comments:
- More explanation on the methodology is needed: how are the INP exactly calculated? It states in the study that it is calculated instantaneous based on T and aerosol conc. and somewhere else that the claculation happens offline. Does that mean it is calculated on the 30 min (timestep of the model) output values for T and aerosol conc.? Is that output saved on this high time resolution? What information is used for the aerosol size, aerosol area, aerosol mass, aerosol volume (depending on the freezing parameterization)?
- The aerosol setup of CAM-ATRAS misses some information that makes it hard to understand the study setup. Can you please spell out all 11 aerosol species used in the model (dust, sea salt, MOA, pollen, bacteria, fungal spores, anthropogenic burning (=BC?), biomass burning (=BC?) are mentiond but that is not 11 species all together)? And then how many tracers does the model use in total - the 11 species + extra tracers (4?) for the distinction between Arctic/non-Arctic aerosols?
- I find the formulation of one of the major hypothesis (INA is reduced in future at ambient temperature) a bit confusing and would rewrite it. INA is a function of temperature and at the same temperature the INA value of one species is the same (all simulated decades). What is changing is the ambient temperature and that the INP efficiency is lower at warmer temperature. I would phrase this as an ambient temperature effect rather than a decrease in INA.
- The Arctic INPs are calculated based on different parameterization schemes as the non-Arctic INPs. That makes sense, but given that you yourself show the high sensitivity of the results on the freezing parameterization scheme did you make a sensitivity test (this could be for a shorter time span, e.g. one year) how the result looks like if the same schemes are used? For example how does the ratio of Arctic dust INP/dust INP change if for both sources the same freezing parameterization scheme is used?
- Related to the point before: it would be great if you could add an overview plot of all INA(T) parameterizations for one example aerosol (=size, mass, area, volume).
- Fig. 1: I guess it is limited by the temperature at which INP were observed (INP observations were available for all locations for the temperature plotted here?) what can be plotted here. However, a comparison can look very different at warm vs. cold temperatures.
- Fig. 2: This figure could be biased by the choice of the height (here lower troposphere) since T is coupled to that. Is it vertically integrated for all pressure levels >730 hPa right now? How does the result differ for a range of different heights or more interestingly for different temperature (but to show that exolicitely instantaneous temperature binned output would be needed)?
- Fig. 4: To support this result or add additional value to it and the corresponding discussion, please add (can also be in the Appendix):
- The same figure but for different seasons. That would help to see if there is a connection between INP values and snow/sea ice cover.
- Add the same figures for dust, MOA, bioaerosols and non-Arctic INPs - it would be great to see a map where transported aerosols matter for Arctic INP.
- Fig. 4: Same as for Fig. 2: How much does the choice of vertical level influence the results? The level should differ quite a bit between non-Arctic transpored aerosol, Arctic transported aerosol and Arctic local aerosol.
- Fig. 4: I guess you don't have saved instantaneous temperature binned output, so it is difficult to request this afterwards, but if there is this kind of output: it would be super interesting to see this results for different temperature bins.
- Section 3.5: This section disturbed the flow of the last part of the paper. I think the conclusion of this section and the findings are really important. However, the authors should consider shifting this section to the appendix and just tie the conclusions into existing (other) text to keep the flow.
- Data availability: Nowadays data should be made available, so please consider sharing at least the processed data used for the figures (more data/not averaged etc. is of course also appreciated).
Minor and technical comments:
- line 41: You write originating from plants here, is it not rather algae och marine species (plant might be a confusing word)?
- line 45: You can remove the sentence "For example...".
- line 58: The sentence on the Matsui et al. study does not really make sense in the current form (or I don't understand it).
- line 145: You write that the dust emissions are likely overestimated - can that be checked?
- line 145-147: In my opinion your argumentation here does not make sense. The dust emissions are likely overestimated which in turn leads to an overestimation of dust INP for the Svalbard location. That sounds reasonable. However, I don't understand why the INP parameterization is important here - that should not be affected by an overestimation of dust emissions (it just propagates the error).
- Fig. 1: Why is there one location (Northern Norway) with a lot smaller/no variability?
- Fig. 1: The data points are a mean over 9 years? Could be specified in the caption.
- Fig. 1: How big were the temperature bins for the simulated data that were used to derive the simulated INPC at one specific T?
- Sect. 3.2: How were the vertically mean INPC calculated?
- Line 197ff: I am not sure if you can make this statement.
- Fig. 5: I am surprised that there is never areas with 0 Arctic INPs? Is that because it is a average value over the whole time span?
- Fig. 5: The monthly mean T can confuse. The INP values are not caculated based on this temperature.
- Fig. 5: Here temperature binned would also be great. This figure in the current form is a bit hard to read because what is seen is based on a mixture of aerosol profile and temperature at the level where Arctic aerosols exist.
- Line 230f: I don't agree with that these results show a strong correlation... You would need to calculate this and I don't see something like that in the paper?
- Fig. 6: Can you add here the spatial distribution from one period (1981-1990 is probably a good choice) or even both as well? That would be a nice extra information in addition to the change plot.
- line 247: Write "change of mean values" (as used later) instead of "rate of change".
- line 249: Agreed that if there is a correlation it would show in your figure. However, you did not do a correlation analysis and you do not show spatial correlation here, so please rewrite.
- line 258-264: Add some explanation already here?
- line 269-270: This is very confusing written: it is not driven by a reduction of INP (which I phrased like that would intuitively read as reduction of aerosols), but a change in temperature that affects the INP efficiency.
- Fig. 8: I would move the LowBac part here to the Appendix.
- Fig. 9: This is a mean over the whole time period? How variable is the result?
- Code availability: I would remove "reasonable" request (upon request is enough).
- Line 367-369: You could explain this more/elaborate on this more.Citation: https://doi.org/10.5194/egusphere-2025-6214-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 178 | 78 | 18 | 274 | 42 | 20 | 40 |
- HTML: 178
- PDF: 78
- XML: 18
- Total: 274
- Supplement: 42
- BibTeX: 20
- EndNote: 40
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1