the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Novel methods to study sea ice deformation, linear kinematic features and coherent dynamic elements from imaging remote sensing data
Abstract. Satellite Synthetic Aperture Radar (SAR) data are commonly utilized for calculating sea ice displacements and, consequently, sea ice deformation strain rates. However, strain rate calculations often suffer from a poor signal-to-noise ratio, especially for products with a spatial resolution higher than 1 km. In this study, we applied a new filtering method to strain rate calculations derived from Sentinel-1 SAR image pairs with a spatial resolution of 800 m. Subsequently, we employed a power law to evaluate the deformation rates at decreasing spatial resolutions, assessing the quality of the filtered data. Upon positive evaluation of the filtered data, we introduced two innovative methods for sea ice deformation assessment. The first method, named 'damage parcels' (DP) tracking, involved the combined analysis of displacements and deformation strain rates to monitor divergence and convergence within the sea ice cover. Additionally, we proposed a new term to describe the behavior of the winter pack: 'Coherent Dynamic Elements' (CDE). CDEs are cohesive clusters of ice plates within the pack ice that move coherently along Linear Kinematic Features (LKFs). The second novel method developed in this study focused on exploring the geometrical properties of these CDEs. Both methods were applied to the winter collection of Sentinel-1 SAR imagery available during the N-ICE2015 campaign. Our results revealed a cyclically changing winter sea ice cover, marked by synoptic events and transitions from pack ice to the marginal ice zone. The DP were continuously tracked over a period of three weeks, including a major storm, revealing a remarkably slow healing process of existing LKFs. Furthermore, the CDE analysis demonstrated the presence of elongated CDEs with a density ranging from 5 to 20 per 100 km by 100 km, and the shortest distance between LKFs was found to be 5–10 km.
- Preprint
(4598 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2023-2626', Anonymous Referee #1, 24 Nov 2023
Novel methods to study sea ice deformation, linear kinematic features and coherent dynamic elements from imaging remote sensing data by Polona Itkin
Summary
Automatically identifying sea ice dynamic features is challenging. In this paper the author presents several new methods to estimate several dynamic features from SAR imagery using the N-ICE2015 study period as a test case. I like the ideas and methods presented in this paper and they certainly add the understanding of sea ice dynamics. This paper has new elements however, I found some sections and items challenging to fully grasp. I think this paper can be published it just requires some revisions to improve its presentation (readability, clarity, etc.). I hope my comments help the author improve this work.General Comments (major)
1. I really like the idea of CDE’s but their definition is a bit confusing. It seems to me CDE’s are an architecture or framework or terms (not term singular) that certain variables can be used to collectively describe winter pack ice. Am I right? However, you first define Coherent Dynamic Elements (CDE) as the boundary of rigid ice plates (Line 58 and 59). OK. In the Abstract you say CDE describes the behaviour of the winter pack but nothing in the paper including your Conclusion relates the winter pack behaviour during N-ICE2015 in that context. I thought I was missing something. Further, if a new term is introduced, then the definition must be consistent. Your definition and usage of CDE needs revision throughout the text otherwise readers will be scratching their heads as its meaning and usage. I suggest defining the CDE framework (with associated variables) earlier in the paper and explicitly describe how these terms can be used collectively to understand winter pack behaviour with evidence from N-ICE2015.2. I understand why the power law was employed for accuracy/quality assessment but it is not the easiest section to comprehend. Perhaps it is my ignorance. Nevertheless, I think this section needs to be revised as casual readers will struggle – I did. I see is no reason why a simple buoy to SAR deformation comparison cannot be performed. The buoy data is available from the lead author (Itkin et al., 2015). Further, the two-way comparison is far more useful anyways and what casual readers will be looking for. I think the power law quality check metrics can still be included but the author needs to add some additional “bread and butter” comparison statistics for casual readers.
3. On Line 10 you state, “Our results revealed a cyclically changing winter sea ice cover, marked by synoptic events and transitions from pack ice to the marginal ice zone.” However, this really was not investigated in the paper. There is no synoptic data in the paper. Again, casual statements like these will leave readers confused because this analysis is nowhere to be found in the paper. Why not add some supporting synoptic data (spatially) to make the manuscript more comprehensive?
4. There are so many acronyms and notation that the reader often forgets or has to refer back to what the definition is. There is nothing wrong with spelling things out in full and in fact it makes your paper more accessible to casual readers. Considering removing some of the notation for text.
Specific Comments (minor)
Line 19
What implications? A good to idea to state what they are i.e. For example, …Lines 22-25
How can increased deformation erode the long-term memory of ice thickness? As I read Mitch’s paper he and co-authors state predictability is lost with the onset of melt. Or are you suggesting winter-time deformation will complicate winter ice thickness retrievals? You need to be explicit about the link between deformation and seasonal prediction.Line 49
Those are not really references related to RADARSAT-1 and RADARSAT-2. I suggest the following:Mahmood, A., Crawford, J.P., Michaud, R., and Jezek, K.C. 1998. “Mapping the world with remote sensing.” Eos, Transactions, American Geophysical Union, Vol. 79(No. 2): pp. 17, 23
Z. Ali, I. Barnard, P. Fox, P. Duggan, R. Gray, Peter Allan, Andre Brand & R. Ste-Mari (2004) Description of RADARSAT-2 synthetic aperture radar design, Canadian Journal of Remote Sensing, 30:3, 246-257, DOI: 10.5589/m03-078
Line 53
I think the RGPS has some done a lot more than derive scaling laws and intersection angles with respect to understanding sea ice dynamics.Line 54-55
The spatial resolution of “deformation estimates from SAR” has been…Line 62-65
Redundant. You just stated most of this information in the previous paragraph.Line 70
You already defined SAR.Line 78
As with previous commentLine 99
How where the SAR images pre-processed? Were they calibrated? I think some details on this is required.Line 189
See General Comment #1.Line 378
The Conclusions do not really match (are missing) some of the items presented in the Introduction.Line 400
Can something be said as to the applicability of these techniques to summertime conditions? Or are these strictly limited to the winter time?Figure 2 and 3:
Probably a good idea to note in the Figure Caption the artifacts or bad data presented (Line 145)Citation: https://doi.org/10.5194/egusphere-2023-2626-RC1 -
AC2: 'Reply on RC1', Polona Itkin, 14 Feb 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-2626/egusphere-2023-2626-AC2-supplement.pdf
-
AC2: 'Reply on RC1', Polona Itkin, 14 Feb 2024
-
RC2: 'Comment on egusphere-2023-2626', Andrew Mahoney, 28 Dec 2023
Please see attached PDF containing my review
-
AC1: 'Reply on RC2', Polona Itkin, 14 Feb 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-2626/egusphere-2023-2626-AC1-supplement.pdf
-
AC1: 'Reply on RC2', Polona Itkin, 14 Feb 2024
Viewed
HTML | XML | Total | BibTeX | EndNote | |
---|---|---|---|---|---|
361 | 141 | 34 | 536 | 20 | 18 |
- HTML: 361
- PDF: 141
- XML: 34
- Total: 536
- BibTeX: 20
- EndNote: 18
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1