Greenland Ice Mapping Project: ice flow velocity variation at sub-monthly to decadal timescales

Joughin, Ian; Smith, Ben E.; Howat, Ian

doi:https://doi.org/10.5194/tc-12-2211-2018

Articles | Volume 12, issue 7

https://doi.org/10.5194/tc-12-2211-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/tc-12-2211-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 12, issue 7

Research article

|

11 Jul 2018

Research article |

| 11 Jul 2018

Greenland Ice Mapping Project: ice flow velocity variation at sub-monthly to decadal timescales

Ian Joughin, Ben E. Smith, and Ian Howat

Download

Final revised paper (published on 11 Jul 2018)
Supplement to the final revised paper
Preprint (discussion started on 27 Feb 2018)
Supplement to the preprint

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

SC1: 'Comment on computed velocity trends in south-west Greenland', Andrew Tedstone, 06 Apr 2018
AC1: 'Response to Comment by Tedstone et al', Ian Joughin, 10 Apr 2018
- SC2: 'Reply to AC1', Andrew Tedstone, 23 Apr 2018
RC1: 'review of Joughin et al.', Anonymous Referee #1, 11 Apr 2018
RC2: 'Review of: Greenland Ice Mapping Project: Ice Flow Velocity Variation at submonthly to decadal time scales', Anonymous Referee #2, 23 Apr 2018
- AC2: 'Response to Reviewers 1 and 2 and to second comment by Tedstone et al', Ian Joughin, 01 May 2018

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Ian Joughin on behalf of the Authors (01 May 2018) Author's response

ED: Referee Nomination & Report Request started (07 May 2018) by Marco Tedesco

RR by Anonymous Referee #2 (18 Jun 2018)

Suggestions for revision or reasons for rejection

Greenland Ice Mapping Project: Ice Flow Velocity Variation at submonthly to decadal time scales

I find this paper rather challenging to review. One major issue is that the authors present detailed methods for the derivation of velocities from Sentinel-1a/b data but there is relatively little attention given to the corresponding methods and uncertainties for the analysis and discussion of observed changes in velocity and likely causes. To the author’s credit, several of their arguments are partially or fully supported by previously published work. Nevertheless, in several cases, it is difficult for the reader to get a sense of the likelihood that the authors’ conclusions are valid. Here I provide a few suggestions for further improvement:

- In the abstract the authors state “contrary to earlier results we do not find a slowdown over much of the southwest Greenland”. The authors state in the manuscript that their results are likely not statistically different from those of Tedstone et al. and P11, L12 states that the authors do find a slowdown. I would therefore recommend removing/changing this statement from the abstract.

- I continue to recommend a thorough analysis of the agreement between TSX and Sentinel-1a/b velocities. What is the distribution of velocity differences for all TSX and S1 data separated in time by +/- 10 days? Can the authors plot the relationship between local velocity gradient and velocity difference? Can authors resample the TSX data to show the relationship between velocity gradients, image resolution and error? It would be very helpful to provide the results of the analysis as a table so that the users of the data have quantitative metrics to assess the new dataset.

- P10, L13-15: KB9 appears to have the smallest horizontal gradients in velocity but the maximum disagreement. This does not seem to support the conclusion that differences are a function of surface gradients and sensor resolution.

- Section 3.5 and 3.6. As with Reviewer 1 I do not understand the stacked velocity plots or the % change numbers presented in the paper. What points are the velocities taken from? What does it mean if the cumulative velocity changes by X%... Would it not make more sense to normalize by glacier velocity? This way the authors could make a statement along the lines of “on average glaciers sped up by X%.

- Where do these errors come from? They seem overly optimistic given the paucity of data before 2015. Do they agree well with the standard deviation to the residuals to the linear fits?

- General grievance: All velocity axes are scaled to very large ranges which looks good but makes it difficult compare products. Since comparison of products is, to a large degree, the motivation for the figures, could the authors also provide a plot for slow moving glaciers (100 – 300 m/y) so that the reader can better assess the agreement between products?

- Speculations/conclusions of Section 4.2 rely heavily on estimates of basal/fjord topography. For increased transparency, can the authors kindly include the latest bed profile from bedmachine and discuss how using this profile would impact their conclusions.

- The authors include a paragraph (P15, L26 to P16, L4) on why the Koge Bugt glacier may have stayed near its current position and why other glaciers will/have likely done the same. While not necessarily wrong, this paragraph has little to do with the rest of the analysis presented in the manuscript. Is this not a chicken or egg argument? Does the trough form where the ocean is or does the ocean go where the trough is? I don’t think that anyone will dispute that abrasion is more prevalent where the ice enters the ocean and glaciers flow is at a maximum. But the location and rate of erosion will also depend on the geology and thermal regime. Since this study does not address geomorphologic processes by which glacier troughs are formed I continue to recommend that this paragraph be removed.

Minor Comments:
- P5, L18: insert “is” between “it: and “best”

- P13, L10: Where does the 1% value come from and is this not also the areas identified as most important and thus targeted by TSX? A more quantitative analysis of the disagreements between sensors of different resolutions would help the reader to determine the impact of the observed difference.

- P14, L18 Why did the glacier advance in 2017 and not 2015 if it was colder in 2015?

- P15, L5 Is the distance that a perturbation propagates not also dependent on bed slope?

- P16, S4.3 and 4.4. Why was Jakobshaven not included? The y-axis scales all the way to 80,000 m/yr… Jakobshaven would only increase this by another 14,000 m/y.

- P17, L17: Wouldn’t a measurement that integrates over a much longer period be less sensitive to temporal sampling errors and also result in lower m/yr. errors? It seems a difference of 48 day would be inconsequential when looking at the total displacement over a full year. If velocities where 20% higher during the maximum period of sampling difference (48 days) then this would only result in a 2.6% error that should be random with time.

- P8, L31: Does the S1 processing use updated DEMs as done for the TSX processing?

- P18, L2: elevation should be in units of “m”

Hide

ED: Publish subject to minor revisions (review by editor) (18 Jun 2018) by Marco Tedesco

AR by Ian Joughin on behalf of the Authors (19 Jun 2018) Author's response Manuscript

ED: Publish as is (20 Jun 2018) by Marco Tedesco

Short summary

We describe several new ice velocity maps produced using Landsat 8 and Copernicus Sentinel 1A/B data. We focus on several sites where we analyse these data in conjunction with earlier data from this project, which extend back to the year 2000. In particular, we find that Jakobshavn Isbræ began slowing substantially in 2017. The growing duration of these records will allow more robust analyses of the processes controlling fast flow and how they are affected by climate and other forcings.