Summary

“Coastal climate variability in Northeast Greenland and the role of changing sea ice and fjord ice” by Shahi et al presents a process-level analysis of the local and large-scale drivers of slope temperature gradients (STG) measured at weather stations around the Zackenberg station array. The authors additionally use passive microwave sea ice observations, atmospheric data from UAV retrievals, and ERA5 reanalysis and the Regional Atmospheric Climate Model (RACMO) data fields to evaluate complex associations between nearby Greenland Sea and fjord ice cover conditions, upper-air circulation patterns, and STGs and their impacts on surface mass balance of the adjacent A.P. Olsen (APO) ice cap. Toward understanding these interactions, key findings include high offshore fjord and Greenland sea ice fraction (SIF) is linked with near-surface inversion-like conditions (shallow STG), less snowfall/ablation at APO under anticyclonic upper-air conditions aloft. In the case of low SIF, the opposite associations are found with steep STGs, more snowfall and ablation under a lower pressure/cyclonic regime in the mid-troposphere.

Changes to the regional cryosphere do not typically occur in isolation (i.e., such as local glaciers melt during a warm spell while nearby sea ice does not and vice versa), and the authors put forth a commendable effort to link changes in glacier surface mass balance (SMB) and associated surface-atmosphere interactions within an observation-rich region at Zackenberg. That said, I offer a few suggestions to draw attention more clearly to the paper’s themes. In particular, the authors might re-consider if the title reflects the narrative. The authors may also consider what role a changing sea ice/fjord ice cover through time plays in their results. More specific recommendations and minor comments are listed below by line (L) number referencing the submitted version of the manuscript.

General Comment on the Analysis

1) Much like other Arctic marginal seas, summers of the last decade have seen a decline in Greenland Sea ice conditions, which contrast earlier analysis years of the late 1990s and early 2000s when sea ice was more prevalent, for example. Does this change in the ice-coverage and its seasonality (timing of melt and freeze onset) across time impact the interpretation of high versus low sea ice fraction years with regards to the atmospheric processes examined?

Specific Comments within Sections

L11-25: In the abstract, a concluding sentence on implications of this work in a changing climate is needed, especially with regards to nearby Greenland/fjord SIF and local APO SMB changes. To this end, within this section the authors might consider more clearly emphasizing regional cryosphere coupling in line with the key findings of the paper.

L163-164: Was there a threshold established (>+/- x sigma) for establishing these unrealistic data points? Please clarify.

L185: Please be more specific about this “information” - what local camera satellite products and space-time resolutions are used?

L297-299: Was the composites sensitivity to these STG thresholds tested? It would be a good idea to comment on how the results change, if at all, by using different STG strength criteria.

L321: Are the SIF and STG also non-normally distributed with respect to time also? In other words, is there a year within the respective (seasonal) time series where the mean and/or variance change, and how do composites before/after these potential changepoint(s) affect interpretation of results presented here? Greenland Sea ice cover has changed in terms of freeze/melt onset (see Stroeve et al., 2017 cited in manuscript) and extent through much of the annual cycle (see Peng and Meier, 2018, Annals of Glaciology, doi:10.17/aog.2017.32), so some indication on how the SIF evolution with time affects the frequency of low/high SIF with respect to STG, SMB and related processes should be discussed.

L454-461: The u10m direction could also be emphasized in this section (I could be wrong, but my understanding is +u = westerly & offshore (in this case), -u = easterly & onshore).

L477/Figure 8: Might it make sense to flip the color ramp such that more ice (colder conditions) are blue hues and less ice (warmer conditions) are red hues? Figure 11 involving APO SMB applies this rationale to its anomaly SMB color ramp.

Technical Corrections

L15: near-fjord sea or land ice conditions?

L19: Would suggest removing “Evidently” and start sentence with “A positive…”

L21: Instead of “change” I’d recommend adding a descriptor (i.e., reduction)

L67: In the atmospheric science community, zonal references east-west and meridional north-south, so would consider switching terms here for the sake of clarity.

L114-116: I’d suggest combining the two sentences, such as “…the present study examines the statistical relationships between STG, SIF, and other atmospheric variables and their physical connections.”

L119: The year range of ZR temperature monitoring could be listed here

L126: Established by whom (ClimateBasis?) and when in 1995? The following sentences mention 1996 hence clarifying the starting year and month.

L136-137: Suggest change “for the entire” to “across Greenland”

L152: Suggest modification to “These datasets provide…”

L207: Instead of cross do you mean “x” marker?

L281-283: This is a bit ambiguous; are there numerical values/ranges associated with steep and shallow STGs?

L347: “During the UAV measurement period?...”

L353-354: If possible, would recommend an estimate of this “same elevation” value.

L357: Do you mean “screen-level” on the temperature instrument? Please clarify.

L378: Does “early” need to be in parenthesis here?

L380: Do you mean total spring days? Please clarify here and similarly in the sentence that follows.

L395/Figure 5: Might a colorbar be added to the figure to depict the SIC gradient (faint to dark blue) that evolves during the freeze-melt periods?

L432: Remove comma after “snow” and suggest removing “and as expected,” also

L443-444: Would suggest removing this sentence as it does not add substantive summary of the previous results described.

L450: GrSea sea ice edge averaged for the full period? Please clarify what the dashed black line shows in each panel.

L512 and 562-563: “~27 change in SIF” – is this % change? There are a few instances in the paper where sea ice fraction (SIF) is not associated with units.

L564: “…greater impact on the air close to the surface than higher above” such as at what atmospheric layer(s) according to your analyses?

L675: Synoptic-scale time lag of the processes? It is good to acknowledge this timescale as it is relevant to the Stroeve et al. (2017) process interpretation and likely the same timescale of processes that interact in your analysis.

L679: Change to “According to the authors,…”

# General comments

Shahi and coauthors analyze the factors controlling the vertical temperature structure of the atmosphere over the Zackenberg region of northeast Greenland. They utilize a dense network of weather stations arrayed throughout varying elevations in this topographically complex area, along with UAV observations, atmospheric reanalysis, and regional climate model output. Their main findings are: (1) atmospheric inversions - characterized by a more "shallow" slope temperature gradient between the low- and high-elevation stations - are more common during the cold season and less frequent during the summer; (2) inversions are associated with anomalously high 500 hPa geopotential heights and regional sea ice cover; and (3) the influence of these environmental conditions on the surface mass balance of the A. P. Olsen Ice Cap are complex, particularly in the summer when lower sea ice fraction is associated with more melt at lower elevations but more mass gain through accumulation at higher elevations.

This is a interesting study overall and the authors have done a nice job of integrating their rich observational dataset with other contextual data sources to understand the environmental controls on atmospheric temperature profiles in the region. The UAV observations are novel in this part of Greenland, to my knowledge, and are a good complement to the station data. The references are thorough and relevant. However, I do have comments on several aspects of the paper that should be improved to make it suitable for final publication. The writing style is often difficult to read with extensive use of acronyms and parenthetical asides, and I think the results could be communicated more effectively by using more plain-language descriptions where possible. I am also not completely convinced that the authors' assertions about the causative influence of sea ice conditions on the regional atmospheric conditions are robustly supported by their analyses. More detail on these comments and some additional comments are provided in the specific comments and technical corrections.

# Specific comments

In my opinion the writing style in the manuscript should be more clear. The writing is burdened by excessive use of acronyms, parenthetical phrases, and repeating statements of the same fact. These factors disrupt the flow of the paper. One example is the sentence from L16-18 sentence in the abstract: "...we find that shallow, i.e. more positive (inversions) or less negative than the mean condition, STGs are associated with a positive anomaly in geopotential height at 500 hPa and surface pressure over East Greenland" - it seems to me this could be simply written as "we find that temperature inversions are associated with positive 500 hPa geopotential height and surface pressure anomalies".

Regarding acronyms, I understand it is challenging to strike a balance between excessive acronym use on the one hand and not using the same full phrase repeatedly on the other. However, I think the authors should consider which acronyms are truly necessary and substitute concise plain-language descriptions where possible. For example, the sentence from L629-630 would be more readable if written as something like "Figure 12 shows the comprehensive picture of the potential relationship between the vertical temperature structure, mid-tropospheric circulation, sea ice cover, and ice cap surface mass balance."

As a related comment, I found the language used to describe slope temperature gradients confusing, especially when trying to think about the slope temperature gradient in the context of inversions. For example, in L455-456 and at many other points, the authors use the terms "steeper (shallower)" to refer to the slope temperature gradient. This is confusing because the slope temperature gradient is directly related to atmospheric inversions, but inversions are themselves also often referred to as "shallow" (with a completely different meaning to the word in this context). And at first glance, one would think the term "steeper" would refer to a stronger inversion, but it is actually the opposite, as "steeper" STGs mean there is a greater decrease in temperature from the surface to the higher-elevation stations. The authors also throw in the term "positive STG" (e.g. L460) to refer to the stronger inversion / shallower STG conditions. I suggest substituting plain-language physical descriptions for acronyms in at least some places to remind the reader of the physical meaning of the STG variable - i.e. simply state if there is an inversion present or, alternatively, if the temperature decreases with height.

I am not sure the authors have convincingly proven that the correlation between regional sea ice and atmospheric conditions in the Zackenberg region represents a direct causative influence of sea ice on the atmosphere. For example, the authors find a seasonal relationship between Greenland Sea sea ice fraction and Zackenberg temperature and precipitation (L644-646). I agree that it is very likely that reduced sea ice contributes to atmospheric warming and moistening. However, could it not also be the case that the sea ice and atmospheric conditions are both influenced by the same large-scale circulation patterns, leading to a correlation between sea ice and atmospheric conditions that is not a one-directional causal influence of sea ice on the atmosphere?

Similarly, do the authors account for the climatological evolution of sea ice conditions within seasons in their analyses? For example, the authors find that, on a daily time scale during the summer, reduced sea ice is correlated with reduced SMB in the ablation area. Again, I agree that the reduced sea ice conditions likely influence atmospheric warming, but could it not also be the case that both sea ice cover and SMB are correlated with the climatological pattern of warming temperatures throughout the course of the summer? I apologize if the authors have already addressed these questions in their methodology, however I could not find any reference to these issues in Section 2.9.

The Introduction does a nice job of explaining why the vertical structure of temperature (and not just temperature in general) is important to study. I think some of this language should be adopted in the Abstract to explain the overall objectives and importance of the study, instead of simply stating that the purpose of the study is "to capture this complexity" (L11-12).

As a related comment, I don't think the title accurately captures the contents of the study. The study primarily concerns the controls of large-scale circulation and regional sea ice cover on the *vertical temperature structure* of the atmosphere in the Zackenberg region, but the vertical temperature profile aspect is not contained in the title.

Fig. 1 is a nice overview map and helps the reader understand the spatial setting of the study.

L84: Should "equatorward" be "poleward" here? I.e. the warm air mass is located in an anomalously poleward area due to the blocking?

L96-97 and elsewhere: Is it necessary to use the "SIF" abbreviation so frequently? At least in some places the text could just name sea ice as "sea ice" rather than SIF.

L99: Is the abbreviation "GrSea" really necessary?

L113-114: I think the A. P. Olsen Ice Cap SMB analysis should be considered objective 3. It is a substantial part of the analysis in the paper. Correspondingly, there should be a separate "section 3.4" starting with L371, instead of grouping the SMB analysis into section 3.3.

L121-122: The geographical description of "west" vs. "east" is confusing here. So the ZR is ~40km west of the outer coast and ~70km east of the GrIS?

L134: It would be helpful to go ahead and state the elevation of ZAC and M3 stations in this sentence. As is, the reader must reference Table 1 and Figure 2 on the subsequent pages to figure this out.

L133-137: Is there spatial variability in the characteristics of inversions in the ZR? Or is the assumption that the stations are sufficiently near to each other that they are vertically sampling the same air mass?

Figure 3: The gray shadings of the UAV vertical profiles in the JJA panel for 2017 and 2018 are difficult to distinguish from one another. Perhaps a color other than gray could be used for one of the years?

L218: Psurf is not an upper air variable.

L226-230: Have the authors considered using sea ice data from a higher-resolution source when available, for example AMSR data (Meier et al. 2018) from 2012-present?

L252-262: Nice idea to evaluate RACMO over ice-free surfaces and compare to ice-covered, this is a novel contribution of the study.

L343-357: UAV profiles

- Where were the UAV profiles taken in relation to stations M3 and ZAC? Can the locations be marked on the map in Figure 1?

- What were the temporal sampling characteristics of the UAV profiles? Were the flights in each of 2017 and 2018 performed on the same day? Were the 2017 and 2018 flights performed in similar times of the year (more specific than summer)?

Figure 5: The plot lines and confidence interval shading are difficult to distinguish from the sea ice background shading. Consider different color choices or perhaps plotting sea ice as a line rather than shading. If keeping the shading, the color shading should probably be mapped to specific values rather than a non-quantitative "transitional period" - there is no way for the reader to discern what sea ice values this phrase refers to.

Figures 6-7: These maps are visually busy and difficult to read. I suggest using a color other than white for the coastlines so they can be more easily seen. Are the dots and the black mesh both necessary?

Figures 6-11: What is the climatological reference period from which the anomalies are calculated? Is it 2004-2019 as mentioned in the Figure 6 caption, or 1996-2020 as mentioned in section 2.9.2?

L522: It would be nice to provide a little more detail on what mesoscale forcing drives the northerly wind during the winter, like the explanation the authors provide for the onshore winds during summer.

L527-534: The result that areas near the coast are warmer than inland during high SIF days is an interesting and unexpected finding.

Figure 12: I am struggling to wrap my head around the result that higher 500 hPa heights are associated with a *decrease* in ablation, and vice versa. This is counter intuitive because one would expect higher temperatures and greater melt with a 500 hPa ridge overhead. Is the explanation that the Z500 increases are local rather than large-scale in extent (e.g. L630-632)?

L671-677: This paragraph about the influence of SIF on SMB is specific to the summer, correct? This should be mentioned here if so.

L686-689: This study also used UAV data.

In the discussion and/or conclusion, it would be nice to see some discussion about implications of this study for climate change. How will projected future decreases in sea ice extent and potential changes in large-scale atmospheric circulation patterns affect the vertical thermal structure of the atmosphere (i.e. the STG) in the Zackenberg region? What does this mean for the future SMB evolution of Greenland's peripheral ice caps and glaciers (as well as the main ice sheet)?

# Technical corrections

L81: "anticyclone... which is" --> "anticyclones... which are"

L411 and elsewhere: "moistures" --> "moisture"

L432: "excepted" --> "expected"

L512-517 and elsewhere: include units (%) for SIF values.<mark style="background: #FFF3A3A6;"></mark>

# References

Meier, W. N., T. Markus, and J. C. Comiso. (2018). AMSR-E/AMSR2 Unified L3 Daily 12.5 km Brightness Temperatures, Sea Ice Concentration, Motion & Snow Depth Polar Grids, Version 1 [Data Set]. Boulder, Colorado USA. NASA National Snow and Ice Data Center Distributed Active Archive Center. https://doi.org/10.5067/RA1MIJOYPK3P. Date Accessed 03-14-2023.