the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 26 Apr 2022
Surface circulation properties in the Eastern Mediterranean emphasized using machine learning methods
Georges Baaklini
Roy El Hourany
Milad Fakhri
Julien Brajard
Leila Issa
Gina Fifani
Laurent Moriter
Abstract. The Eastern Mediterranean surface circulation is highly energetic, composed of structures interacting stochastically. However, some main features are still debated, and the behavior of some fine-scale dynamics and their role in shaping the general circulation is yet unknown. In the following paper, we use an unsupervised neural network clustering method to analyze the long-term variability of the different mesoscale structures. We decompose 26 years of altimetric data into clusters reflecting different circulation patterns of weak and strong flows with either strain or vortex-dominated velocities. The vortex-dominated cluster is more persistent in the western part of the basin, which is more active than the eastern part due to the strong flow along the coast, interacting with the extended bathymetry and engendering continuous instabilities. The cluster that reflects a weak flow dominated the middle of the basin, including the Mid-Mediterranean Jet (MMJ) pathway. However, the temporal analysis shows a frequent and intermittent occurrence of a strong flow in the middle of the basin, which could explain the previous contradictory assessment of MMJ existence using in-situ observations. Moreover, we prove that the Levantine Sea is becoming more and more energetic as the activity of the main mesoscale features is showing a positive trend.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(5057 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(5057 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Georges Baaklini et al.
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2022-202', Anonymous Referee #1, 31 May 2022
Dear Authors,
Please fid below my review of your paper entitled "Surface circulation properties in the Eastern Mediterranean emphasized using machine learning methods". I am not especially an expert in machine learning methods, but the manuscript is well written and relatively easy to read. I did not detect any flaw and the science seems to me sound. Overall I enjoyed the study and I only have a few minor comments. Overall I recommend a minor revision. See below.-- Main comment --
1. My main comment is that the sections of the paper could be re-aranged in a more traditonal way while keeping the same content.
For example, Section 3 - Results, is actually a "Results and Discussion" section since the results are discussed with it ( and at the moment, there is not a real Discussion section)2. Section 4 is a bit floating... I wonder if it is really needed (it is interesting, but personally I would remove since it is only based on a single deployment of 3 drifters).
3. There are also a lot of figures. They are all discussed and potentially interesting, but it is a lot...
-- Other Specific comments --
- Introduction and Figure 1: An improved larger map with the political boundaries is needed. This will help the reader not familiar with the region.
- Paragraph starting L.35: This seems a key aspect and could be introduced better.
- L.35: "debatable" (is there any reference to this?)
- L. 50: 25-yr of altimetry: maybe cite "International Altimetry Team (2021), Altimetry for the future: Building on 25 years of progress. Advances in Space Research. doi: 10.1016/j.asr.2021.01.022."
- The acronym SOM is spelled out several times (L.53, 79, 97, 98, ...)
- L.53: The sentence starting by "It is use to" is badly contructed.
- L.69: replace "till" by "and", and remove semi-colon ";"
- L.86: Mention how the altimetry product was accessed (e.g. Copernicus? accessed date? doi? citation?)
- Figure 3: Should we see a spatial map here? Are the pixels organized as in Fig 2?
- L.134: Would be nice to have a better description here of the clusters. Later you say "high EKE" (clusters 1-2) or "high vorticity" (clusters 4-5) and this is useful. I would mention it here too.
- Section 2.4: I wonder if the study area should appear of the beginning of the method...
- L.136: wrong construction with the sentence.
- Figure 4, 6, etc.: I would choose a different color scheme for the clusters. The 2 blues and 2 reds are too similar (I printed the manuscript and you can't tell the difference). The colormap from Figure 11 would be best, for example....
- L.138: "guided by the isobath" -> say which one.
- L.160 and L.162: What you call "variation" is actually "standard deviation" and should be call this way.
- L.161: Replace "Depending" by "Based"
- Section 3: It is really Results and Discussion...
- L.204 (whole paragraph): Maybe discuss a little bit more how the changes in satellite and sensors may have impacted (e.g. any abrupt changes corresponding to satellite changes?). Maybe also a good place to put your current Section 4 if you decide to keep it...
- L.221: Not sure how to interpret the sentence starting with "The along-slope..."
- L.222: "C1 and C2" -> If I interpret correctly, it is mostly C2...
- L.223: Here a map with political boundaries would help too.
- L.230: "along its potential path" -> along or accross?
- L.254: You mean their distance to the "closest" isobath between 1000, 2000, or 3000m? (same comment for Fig 13 caption).
- L.280: Resonable to conclude that is could be underestimated while only comparing a single event with 3 drifters?Citation: https://doi.org/10.5194/egusphere-2022-202-RC1 -
AC1: 'Reply on RC1', Georges Baaklini, 16 Jul 2022
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2022/egusphere-2022-202/egusphere-2022-202-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Georges Baaklini, 16 Jul 2022
-
RC2: 'Comment on egusphere-2022-202', Anonymous Referee #2, 01 Jun 2022
The paper studies the long-term variability of the mesoscale structures in the Eastern Mediterranean Sea (EMS) using geostrophic velocities derived from altimetry. Authors found an interment occurrence of the mid-Mediterranean jet and an increase of the mesoscale activity in the Levantine basin.
I think that the topic is of interest and the approach that the authors try to follow is correct but the paper in its present form is still far from being publishable. Sections are mixed and the Ms is difficult to follow needing some editorial work. Figures do not provide the information that the authors want to present being in some cases repetitive or understandable.
Main concerns.
*The work is based on the application of SOM with a huge number of neurons and then grouping them using the HAC. The input layers correspond to the zonal and meridional velocities as well as the OW parameter. Then, at different sub regions statistics for the different parameters are evaluated including seasonal variations of clusters. If the objective of the Ms is to understand the mesoscale dynamics of the EMS the approach would be first to perform a temporal SOM analysis to the (ug,vg) velocities (or alternatively the MKE) to obtain the zones of co-variability. This would also provide the time series of the velocities in each of the patterns. This has to be done in conjunction with a spatial SOM that will give the main mesoscale structures in the basin. The BMUs of these spatial patterns decomposition will give the seasonality that the authors want to explain. However if the objective is to analyse the eddy activity in the area I suggest to change for the input data the EKE, MEKE and the OW parameter. In the paper no mention is given to which SOM they are applying nor the 5 clusters that they finally ended.
Minor concerns.
* I assume that the data corresponds to daily velocities but this is never stated in the Manuscript. Why using daily data and not weekly or monthly if the objective is to analyse mesoscale structures?.
Figure 1 and 2 can be merged.
Figure 3. What are the units in the colorbar?
Figure 4A. What is the message in this figure?
Figure 5. I suggest defining the areas directly with the SOM (see main concern)
Figures 6-10. I don't understand the message behind these graphics.
Ln 90. Why using OW and velocities as input?
Page 6. Why 1400 neurons?
Lines around 120. What do you mean with ''SOM is well organized’’?
Line 132, C$ instead of C5
Line 156 ''iso-MDT'' . There are no isolines in this plot
Section 3.1. See Main concerns.
Citation: https://doi.org/10.5194/egusphere-2022-202-RC2 -
AC2: 'Reply on RC2', Georges Baaklini, 16 Jul 2022
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2022/egusphere-2022-202/egusphere-2022-202-AC2-supplement.pdf
-
AC2: 'Reply on RC2', Georges Baaklini, 16 Jul 2022
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2022-202', Anonymous Referee #1, 31 May 2022
Dear Authors,
Please fid below my review of your paper entitled "Surface circulation properties in the Eastern Mediterranean emphasized using machine learning methods". I am not especially an expert in machine learning methods, but the manuscript is well written and relatively easy to read. I did not detect any flaw and the science seems to me sound. Overall I enjoyed the study and I only have a few minor comments. Overall I recommend a minor revision. See below.-- Main comment --
1. My main comment is that the sections of the paper could be re-aranged in a more traditonal way while keeping the same content.
For example, Section 3 - Results, is actually a "Results and Discussion" section since the results are discussed with it ( and at the moment, there is not a real Discussion section)2. Section 4 is a bit floating... I wonder if it is really needed (it is interesting, but personally I would remove since it is only based on a single deployment of 3 drifters).
3. There are also a lot of figures. They are all discussed and potentially interesting, but it is a lot...
-- Other Specific comments --
- Introduction and Figure 1: An improved larger map with the political boundaries is needed. This will help the reader not familiar with the region.
- Paragraph starting L.35: This seems a key aspect and could be introduced better.
- L.35: "debatable" (is there any reference to this?)
- L. 50: 25-yr of altimetry: maybe cite "International Altimetry Team (2021), Altimetry for the future: Building on 25 years of progress. Advances in Space Research. doi: 10.1016/j.asr.2021.01.022."
- The acronym SOM is spelled out several times (L.53, 79, 97, 98, ...)
- L.53: The sentence starting by "It is use to" is badly contructed.
- L.69: replace "till" by "and", and remove semi-colon ";"
- L.86: Mention how the altimetry product was accessed (e.g. Copernicus? accessed date? doi? citation?)
- Figure 3: Should we see a spatial map here? Are the pixels organized as in Fig 2?
- L.134: Would be nice to have a better description here of the clusters. Later you say "high EKE" (clusters 1-2) or "high vorticity" (clusters 4-5) and this is useful. I would mention it here too.
- Section 2.4: I wonder if the study area should appear of the beginning of the method...
- L.136: wrong construction with the sentence.
- Figure 4, 6, etc.: I would choose a different color scheme for the clusters. The 2 blues and 2 reds are too similar (I printed the manuscript and you can't tell the difference). The colormap from Figure 11 would be best, for example....
- L.138: "guided by the isobath" -> say which one.
- L.160 and L.162: What you call "variation" is actually "standard deviation" and should be call this way.
- L.161: Replace "Depending" by "Based"
- Section 3: It is really Results and Discussion...
- L.204 (whole paragraph): Maybe discuss a little bit more how the changes in satellite and sensors may have impacted (e.g. any abrupt changes corresponding to satellite changes?). Maybe also a good place to put your current Section 4 if you decide to keep it...
- L.221: Not sure how to interpret the sentence starting with "The along-slope..."
- L.222: "C1 and C2" -> If I interpret correctly, it is mostly C2...
- L.223: Here a map with political boundaries would help too.
- L.230: "along its potential path" -> along or accross?
- L.254: You mean their distance to the "closest" isobath between 1000, 2000, or 3000m? (same comment for Fig 13 caption).
- L.280: Resonable to conclude that is could be underestimated while only comparing a single event with 3 drifters?Citation: https://doi.org/10.5194/egusphere-2022-202-RC1 -
AC1: 'Reply on RC1', Georges Baaklini, 16 Jul 2022
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2022/egusphere-2022-202/egusphere-2022-202-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Georges Baaklini, 16 Jul 2022
-
RC2: 'Comment on egusphere-2022-202', Anonymous Referee #2, 01 Jun 2022
The paper studies the long-term variability of the mesoscale structures in the Eastern Mediterranean Sea (EMS) using geostrophic velocities derived from altimetry. Authors found an interment occurrence of the mid-Mediterranean jet and an increase of the mesoscale activity in the Levantine basin.
I think that the topic is of interest and the approach that the authors try to follow is correct but the paper in its present form is still far from being publishable. Sections are mixed and the Ms is difficult to follow needing some editorial work. Figures do not provide the information that the authors want to present being in some cases repetitive or understandable.
Main concerns.
*The work is based on the application of SOM with a huge number of neurons and then grouping them using the HAC. The input layers correspond to the zonal and meridional velocities as well as the OW parameter. Then, at different sub regions statistics for the different parameters are evaluated including seasonal variations of clusters. If the objective of the Ms is to understand the mesoscale dynamics of the EMS the approach would be first to perform a temporal SOM analysis to the (ug,vg) velocities (or alternatively the MKE) to obtain the zones of co-variability. This would also provide the time series of the velocities in each of the patterns. This has to be done in conjunction with a spatial SOM that will give the main mesoscale structures in the basin. The BMUs of these spatial patterns decomposition will give the seasonality that the authors want to explain. However if the objective is to analyse the eddy activity in the area I suggest to change for the input data the EKE, MEKE and the OW parameter. In the paper no mention is given to which SOM they are applying nor the 5 clusters that they finally ended.
Minor concerns.
* I assume that the data corresponds to daily velocities but this is never stated in the Manuscript. Why using daily data and not weekly or monthly if the objective is to analyse mesoscale structures?.
Figure 1 and 2 can be merged.
Figure 3. What are the units in the colorbar?
Figure 4A. What is the message in this figure?
Figure 5. I suggest defining the areas directly with the SOM (see main concern)
Figures 6-10. I don't understand the message behind these graphics.
Ln 90. Why using OW and velocities as input?
Page 6. Why 1400 neurons?
Lines around 120. What do you mean with ''SOM is well organized’’?
Line 132, C$ instead of C5
Line 156 ''iso-MDT'' . There are no isolines in this plot
Section 3.1. See Main concerns.
Citation: https://doi.org/10.5194/egusphere-2022-202-RC2 -
AC2: 'Reply on RC2', Georges Baaklini, 16 Jul 2022
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2022/egusphere-2022-202/egusphere-2022-202-AC2-supplement.pdf
-
AC2: 'Reply on RC2', Georges Baaklini, 16 Jul 2022
Peer review completion
Journal article(s) based on this preprint
Georges Baaklini et al.
Georges Baaklini et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 281 | 57 | 8 | 346 | 6 | 6 |
- HTML: 281
- PDF: 57
- XML: 8
- Total: 346
- BibTeX: 6
- EndNote: 6
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(5057 KB) - Metadata XML