The impact of climate oscillations on the surface energy budget over the Greenland Ice Sheet in a changing climate

Silva, Tiago; Abermann, Jakob; Noël, Brice; Shahi, Sonika; van de Berg, Willem Jan; Schöner, Wolfgang

doi:https://doi.org/10.5194/tc-2021-388

Preprints

https://doi.org/10.5194/tc-2021-388

Preprints

14 Jan 2022

Status: a revised version of this preprint is currently under review for the journal TC.

The impact of climate oscillations on the surface energy budget over the Greenland Ice Sheet in a changing climate

Tiago Silva¹, Jakob Abermann^1,3, Brice Noël², Sonika Shahi¹, Willem Jan van de Berg², and Wolfgang Schöner^1,3 Tiago Silva et al.

Show author details

¹Institute of Geography and Regional Science, Graz University, Austria
²Institute for Marine and Atmospheric Research, Utrecht University, Netherlands
³Austrian Polar Research Institute, Vienna, Austria

Received: 17 Dec 2021 – Discussion started: 14 Jan 2022

Abstract. Climate change is particularly strong in Greenland primarily as a result of changes in advection of heat and moisture fluxes from lower latitudes. The atmospheric structures involved influence the surface mass balance and their pattern are largely explained by climate oscillations which describe the internal climate variability. Based on a clustering method, we combine the Greenland Blocking Index and the North Atlantic Oscillation index with the vertically integrated water vapor to analyze inter-seasonal and regional impacts of the North Atlantic influence on the surface energy components over the Greenland Ice Sheet. In comparison to the reference period (1959–1990), the atmosphere has become warmer and moister during recent decades (1991–2020) for contrasting atmospheric circulation patterns. Particularly in the northern regions, increases in tropospheric water vapor enhance incoming longwave radiation and thus contribute to surface warming. Surface warming is most evident in winter, although its magnitude and spatial extent depend on the prevailing atmospheric configuration. Relative to the reference period, increases in sensible heat flux in the summer ablation zone are found irrespective of the atmospheric circulation pattern. Especially in the northern ablation zone, these are explained by the stronger katabatic winds which are partly driven by the larger surface pressure gradients between the ice/snow-covered surface and adjacent seas, and by the larger temperature gradient between near-surface air and the air above. Increases in net shortwave radiation are mainly connected to high-pressure systems. Whereas in the southern part of Greenland the atmosphere has gotten optical thinner, thus allowing more incoming shortwave radiation to reach the surface, in the northern part the incoming shortwave radiation flux has changed little with respect to the reference period, but the surface albedo decreased due to the expansion of the bare ice area.

Download & links

Preprint (PDF, 6010 KB)

Supplement (17965 KB)

Tiago Silva et al.

Status: final response (author comments only)

RC1:
'Comment on tc-2021-388', Anonymous Referee #1, 01 Feb 2022

Silva et al. examines the influence of the North Atlantic Oscillation (NAO), Greenland Blocking Index (GBI), and a cluster-aggregation of the aforementioned indices along with integrated water vapor (IWV) over the Greenland Ice Sheet (GrIS) on regional surface energy budget (SEB) changes, derived from the polar-adapted Regional Atmospheric Climate Model (i.e., RACMO2), between the 1959-1990 and 1991-2020 periods. In addition to deconstructing the GrIS-wide and regional SEB and thermodynamic variables (e.g., skin temperature, IWV, and near-surface specific humidity) by phase of these raw and clustered climate indices, the authors also correlate the accumulation and ablation zone rates of change associated with the indices’ phases to each of these variables individually for winter and summer seasons. Mesoscale processes forcing SEB changes (e.g., loss of local sea ice increases in wind speeds due to strengthening surface pressure gradients) are also discussed in the context of results. Main conclusions include that GrIS surface warming is most pronounced during winter (following the strongest period of Arctic Amplification), but the associated magnitude and spatial pattern of temperature changes depend on the prevailing atmospheric pattern and presumably its frequency. Meanwhile, sensible heat flux increases in the summer ablation zone are found regardless of the atmospheric circulation pattern, further signaling the importance of mesoscale controls (e.g. katabatic wind strengthening due to land-sea temperature and pressure changes) on low-elevation melt.

The level of detail provided in linking the climate indices/oscillations to the SEB and thermodynamic variables is commendable and presents a more detailed picture of atmosphere-GrIS surface forcing than is typically presented in comparable studies. However, this amount of detail also presents challenges with regards to clearly distilling key results. As such, the main findings could be more clearly stated in the abstract and especially in the conclusions. If key takeaways and related points could be more clearly stated through the manuscript, this study could be a valuable addition to the literature.

My comments are provided by line number (LN) or specific figure below. While most are minor in nature, the total number of comments may tilt the paper toward the category of major revision.

LN14: Do you mean clouds have become optically thinner? Please clarify.

LN43: Can you clarify what is intended and ultimately hypothesized by “tilt within large-scale structures may have an impact at different locations”? Work by Woollings et al. (2008) J. Climate and Hanna et al. (2018) Int. J. Climatol. has shown that that the setup of Greenland blocks tends to precede by a couple of days downstream positive North Atlantic SLP anomalies in the vicinity of the Icelandic Low (i.e., -NAO conditions). Perhaps referencing this work may help clarify the large-scale structural reference?

LN76: It would be good to emphasize around this point in the introduction the explicit goal and primary research questions of the study. These would help build upon some previously mentioned hypotheses (e.g., LN 74-75) by adding more structure and thus guidance for the reader toward analyses that lie ahead.

LN 117: List the flux terms units, W/m^2?

LN126-127: I think you could move this sentence (beginning with “Using the 62 years…”) to L131 and explicitly list the sub-periods that were the result of equally dividing the total years in the dataset.

L155: Both the NAO and GBI indices should be defined here (e.g., domains, methods, papers defining such, etc). Moreover, both atmospheric indices are derived at z500, did you look at the surface NAO (i.e., Hurrell PCA or weather station-based NAO)? In this context, I recommend in the paper that you address why only z500 indices are used or why NAO from SLP data is not used. This discussion would seem appropriate since you are exploring through cluster analysis how co-varying characteristics of these atmospheric patterns (along with IWV) may impact GrIS surface conditions.

LN167-168: To clarify the sentence, I recommend substituting “predominant” with “prevailing” then remove “prevailing” in LN168.

LN172: The clustering approach could use more description. Why did you pick 3 clusters? Did you select these based on subjective or objective criteria? Are the results sensitive to the number of clusters selected and analyzed?

LN179: Be more specific on what data is shown in the Figure S3 scatterplots. This is very vague as currently written.

LN183: “The positive phase of NAG is connected…” Is this “connection” illustrated somewhere either graphically or statistically? This would be helpful to show the reader to see what +NAG entails.

Figure 2 caption: The last sentence is unclear. I suggest mentioning that data from 1991 onward is found to the right of the gray vertical line.

L192: Why show the 925 hPa height anomaly rather than SLP, a field typically used in defining surface pressure characteristics of the NAO.

L193: “vertical tilting structure” meaning what? Please clarify.

L199: Do you mean “winter” instead of “spring” is when the equator-to-pole air temperature contrast is maximized?

Figure 3: Should the colorbar label r_s(Seasonal ablation…) be r^2 as it is in the caption?

Figure 4: “The percentage of each NAG phase used...” can you please clarify what this means? As I interpret it, it sounds like some +/-NAG days were composited and some were not without explanation as to why since f0 and f- percentages do not sum to 100% (as in Fig 5).

LN253: Remove “configuration.” Also, since these surface temperature and radiative fields (i.e., Fig 5) increase regardless of atmospheric pattern, does that suggest that warming climate is the main culprit in driving these fluxes that impact SMB? What link is being made with adjacent marginal seas; they respond similarly to these fields as the GrIS?

LN 270-271: In comparing 1991-2020 against the reference period, do you mean “increased surface-based inversions…”

LN272-273: This sentence is confusing; the IWV increase over the northern GrIS is not related to local cloud water content? Please clarify. Is this shown in the analysis and if so, then where?

LN293: “high summer GBI values…” – where is this analysis shown?

LN 297: “crucial role of NAO advecting heat and moisture…” through storms/the storm track migrating poleward toward Greenland?

LN311-313: This sentence is a bit hard to follow. I recommend splitting it into two sentences.

LN315: To clarify, is the suggestion that the summer wind speed increase over northern GrIS is due to the near-complete summer melt of Baffin (particularly) ice cover, a typical feature of its annual cycle? Atmospheric circulation patterns can accelerate the melt, but their intensity and orientation could also presumably affect the onset of such summer wind increases.

LN 322-324: It would be a good idea to direct the reader to this figure or analysis within the paper.

LN382: Change “has” to “have”

Supplemental Material:

Figure S8: Label seasons at the top of the graphic as with Figure S7, etc.

Figures S12-14: I am confused what these graphics actually show and what the units on each concentric circle represent. Please clarify.

Citation: https://doi.org/10.5194/tc-2021-388-RC1
- AC1: 'Reply on RC1', Tiago Silva, 15 Apr 2022
  
  Dear RC1, please find our answers to the reviews in the Supplement.
  
  Citation: https://doi.org/10.5194/tc-2021-388-AC1
RC2:
'Comment on tc-2021-388', Anonymous Referee #2, 01 Feb 2022

This is an interesting novel study of the effects of the NAO and GBI and their combined influence on the Greenland Ice Sheet surface energy balance. The paper is fairly insightful and is reasonably well presented overall, although it would be useful to add an explanation of exactly how the NAG (influence of the North Atlantic over Greenland) time series was derived. It should also be clarified somewhere whether the reported correlation coefficients are based on detrended datasets. The analysis is based on an interesting and worthwhile hypothesis that the tilt within North Atlantic jet stream structures may have differing spatial (and temporal) effects on the near-surface impacts of jet-stream changes, and is best quantified by combining NAO and GBI rather than taking one of these measures in isolation.

Specific comments

Line 92 (P4): why not use the most recent (and best) ERA5 ECMWF reanalysis (which is available back to 1950) to force the RACMO for the whole time period?

Line 97 (P4) “based on the lowest 5% albedo values between 2000 and 2015” – how many values/how frequent?

L162 (P7) etc. – are the reported correlation coefficients based on de-trended data?

L183 (P8): how exactly is the NAG time series calculated?

L187 (P8) “the influences exerted by NAO and GBI may differ” – this is an interesting result.

L190 (P8) “the 95^th percentile of IWV is mainly connected to positive NAG phases in summer and winter”: Fig. S3 suggests (dark circles) that high IWV is mainly associated with the neutral (grey) cluster then.

L198 (P8) “in winter +NAG frequently contributes the most to surface accumulation”. What is the reason for this? If +NAG means more Greenland Blocking, this should be associated with fewer storms in south-east Greenland.

Fig. 3 (and elsewhere): are the reported correlation coefficients based on de-trended data?

L238 (P12) “-NA in winter promotes more IWP at the Northeast” – this seems unclear from Fig. 4(a).

L240 (P12) “The RH2m…” – where is this shown? Should this refer to the q2m plots?

Citation: https://doi.org/10.5194/tc-2021-388-RC2
- AC2: 'Reply on RC2', Tiago Silva, 15 Apr 2022
  
  Dear RC2, please find our answers to the reviews in the Supplement.
  
  Citation: https://doi.org/10.5194/tc-2021-388-AC2
RC3:
'Comment on tc-2021-388', Anonymous Referee #3, 11 Feb 2022

General comments

In this study, the authors use a cluster analysis of NAO, GBI, and column water vapor to derive a “North Atlantic influence on Greenland” (NAG) index. RACMO2 output is then used to investigate atmospheric and cryospheric conditions across different NAG phases and their changes across a 1991 break point in summer surface mass loss. Results describe a large array of seasonal anomalies in atmospheric conditions and surface energy balance components across the NAG phases for each season during the pre- and post-1991 periods.

I found this paper difficult to follow due to the large number of figures and sub-panels within figures and the organization of the paper, as it lacks a clear statement of the research questions or summary of what important new information was learned in this study. I also agree with the editor that there is insufficient originality, at least with how the results are presented in current form. However, there do not appear to be any technical flaws in the methods employed, and I do think there is potential for some of the results to form a nice study if they are better organized. I encourage the authors to think about what they consider to be the most important and novel findings contained within their many analyses and distill these findings into a focused message for readers to take from the paper. As an example, the contrast between moistening in northern Greenland and drying / clearing in southern Greenland under +NAG conditions is an interesting finding. In addition to the specific comments and technical corrections below, I would recommend that the authors simplify the figures, and restructure the discussion so that a large part of the findings in the main paper are not describing figures found in the supplement.

Specific comments

Did the authors examine trends in the frequency of NAG phases, or did they only look at changes in atmospheric conditions over time during each NAG phase?

L18–45: The opening paragraph of the Introduction is quite long and does not provide a compelling introduction to the research topic that the authors investigate. I think it would make more sense to first introduce the problem of Greenland surface melt and its atmospheric drivers (the second paragraph), before moving on to the indices that are used to help quantify these atmospheric drivers (first paragraph).

L21: Liu and Barnes (2015) is a good reference on the relationship between Rossby wave breaking and poleward moisture transport in the vicinity of Greenland.

L29: I’m not sure it’s correct to say that the NAO phase “explains most of the heat and moisture transported poleward”. It’s more accurate to say that the NAO phase affects the location and magnitude of poleward heat and moisture transport, and provide a reference on this.

L32: GBI simply quantifies the mean 500 hPa geopotential height over a Greenland-centered domain, as the authors state in the previous sentence. It does not directly quantify the strength and moisture transported over the Greenland domain although it is correlated with these quantities (see the Barrett et al. 2020 paper the authors already cite). See Wachowicz et al. 2020 for a more nuanced discussion of the GBI and comparison with other blocking metrics.

L120 and Figs. 5, S9–S11: It is not clear how the method of dividing the adjacent seas into four areas is actually used to assess potential sources of moisture. I am having trouble understanding what the numbers in the corners of Figs. 5 and S9–S11 (the “differences in composites between adjacent seas”) represent.

L123: It should be stated explicitly at the beginning of section 2.3 that the reason for the break point detection is to form the basis for subsequent analyses of atmospheric and glaciological conditions before and after the break point. As it stands now, this section reads like it is reporting research findings, rather than describing a method that will be used to produce the results of the study.

L172–180: State up front that you are using a k-means clustering method (rather than first describing the method and naming it as k-means clustering at the end of the description).

L181: I don’t think the “influence of the North Atlantic over Greenland” is an accurate description of what the NAG index produced by the cluster classification provides. Maybe describe as the “influence of regional climate” on Greenland instead. (The AMV index, which specifically quantifies oceanic conditions, is discussed in the Introduction and in L159 in the Data and Methods, but doesn’t appear to be included as an input to the NAG index.)

L226, 232–236, 311–315, and 368–370: The authors should consider that the stronger wind speeds during the +NAG phase are not strictly katabatic but are enhanced by the interaction of a strengthened synoptic-scale pressure gradient with the Greenland ice sheet’s orography. I would suspect this is especially true for the winter cases where the authors find that increased wind speeds and SHF occur during +NAG. Previous studies have described this synoptically-driven wind enhancement as the Greenland “barrier jet” or “plateau jet” – see e.g. Meesters 1994, van den Broeke and Gallée 1996, Moore et al. 2013, Mattingly et al. 2020.

L255, Figs. 4–6: I assume all the results in Figs. 4–6 (e.g. the increasing trend in TCWV in northern Greenland described in L255) are produced from RACMO2 data? If so this should be explicitly stated in the figure captions and the text.

L282–284: This statement about the seasonal preconditioning effect of skin temperature warming appears to contract the finding in L166–168 that there is no relevant time-lag response between seasonal GrIS surface mass fluxes and the predominant atmospheric circulation pattern prevailing in the preceding seasons.

L314: How would decreasing ice in neighboring seas contribute to an increase in summer wind speed? Please explain in more detail.

Technical corrections

L2: The word “fluxes” is not needed since “advection” already describes the horizontal flow of heat and moisture.

L2: surface mass balance of what? (state definitively that it’s the SMB of the Greenland Ice Sheet)

L2: “pattern” --> “patterns”

L14: “optical” --> “optically”

L14–16: Run-on sentence. Consider splitting into two sentences.

L15: “shortwave radiation flux” should be “shortwave radiation” or “shortwave radiative flux”

L18: north of the *climatological location of* the jet stream

L63: “largest” --> “most intense”?

L91: ERA5 is the most recent reanalysis product from ECMWF (it’s not an “earlier product”)

L211: The abbreviation “0NAG” is used repeatedly from this point forward without previously being defined in the text. It appears to be defined in the caption for Figure 4, but its meaning should be explicitly stated in the text at first use.

L331: Delete the word “or” at the end of this line.

L335: Accumulation zone has been decreasing *in area*?

L337: Insert the word “zone” after “accumulation”

L366-367: “vertically distributed changes” --> “vertical distribution of changes”

References

Liu, C. and Barnes, E. A.: Extreme moisture transport into the Arctic linked to Rossby wave breaking, J. Geophys. Res. Atmos., 120, 3774–3788, https://doi.org/10.1002/2014JD022796, 2015.

Mattingly, K. S., Mote, T. L., Fettweis, X., van As, D., Van Tricht, K., Lhermitte, S., Pettersen, C., and Fausto, R. S.: Strong Summer Atmospheric Trigger Greenland Ice Sheet Melt through Spatially Varying Surface Energy Balance and Cloud Regimes, J. Climate, 33, 6809–6832, https://doi.org/10.1175/JCLI-D-19-0835.1, 2020.

Meesters, A.: Dependence of the energy balance of the Greenland ice sheet on climate change: Influence of katabatic wind and tundra, Q.J Royal Met. Soc., 120, 491–517, https://doi.org/10.1002/qj.49712051702, 1994.

Moore, G. W. K., Renfrew, I. A., and Cassano, J. J.: Greenland plateau jets, Tellus A: Dynamic Meteorology and Oceanography, 65, 17468, https://doi.org/10.3402/tellusa.v65i0.17468, 2013.

van den Broeke, M. R. and Gallée, H.: Observation and simulation of barrier winds at the western margin of the Greenland ice sheet, Q.J Royal Met. Soc., 122, 1365–1383, https://doi.org/10.1002/qj.49712253407, 1996.

Wachowicz, L. J., Preece, J. R., Mote, T. L., Barrett, B. S., and Henderson, G. R.: Historical Trends of Seasonal Greenland Blocking Under Different Blocking Metrics, Int J Climatol, 41, E3263–E3278, https://doi.org/10.1002/joc.6923, 2020.

Citation: https://doi.org/10.5194/tc-2021-388-RC3
- AC3: 'Reply on RC3', Tiago Silva, 15 Apr 2022
  
  Dear RC3, please find our answers to the reviews in the Supplement.
  
  Citation: https://doi.org/10.5194/tc-2021-388-AC3

Tiago Silva et al.

Supplement

https://doi.org/10.5194/tc-2021-388-supplement

Tiago Silva et al.

Viewed

Total article views: 771 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
587	159	25	771	42	13	12

HTML: 587
PDF: 159
XML: 25
Total: 771
Supplement: 42
BibTeX: 13
EndNote: 12

Views and downloads (calculated since 14 Jan 2022)

Month	HTML	PDF	XML	Total
Jan 2022	221	70	4	295
Feb 2022	208	33	9	250
Mar 2022	48	21	3	72
Apr 2022	55	12	4	71
May 2022	33	18	4	55
Jun 2022	19	5	1	25
Jul 2022	3	0	3

Cumulative views and downloads (calculated since 14 Jan 2022)

Month	HTML	PDF	XML	Total
Jan 2022	221	70	4	295
Feb 2022	208	33	9	250
Mar 2022	48	21	3	72
Apr 2022	55	12	4	71
May 2022	33	18	4	55
Jun 2022	19	5	1	25
Jul 2022	3	0	3

Viewed (geographical distribution)

Total article views: 723 (including HTML, PDF, and XML) Thereof 723 with geography defined and 0 with unknown origin.

Country	#	Views	%

Discussed

Latest update: 06 Jul 2022

Short summary

In order to overcome internal climate variability, this study investigates spatio-temporal changes of atmospheric drivers within key atmospheric circulation patterns (ACP) over the Greenland Ice Sheet (GrIS). We present the extent of the recent tropospheric warming and increase in water vapor as dependent on season and on the prevailing ACP along with regional impacts of atmospheric drivers on the GrIS surface energy components.


Total:	0
HTML:	0
PDF:	0
XML:	0