The new CNES-CLS18 global mean dynamic topography

Mulet, Sandrine; Rio, Marie-Hélène; Etienne, Hélène; Artana, Camilia; Cancet, Mathilde; Dibarboure, Gérald; Feng, Hui; Husson, Romain; Picot, Nicolas; Provost, Christine; Strub, P. Ted

doi:https://doi.org/10.5194/os-17-789-2021

Articles | Volume 17, issue 3

https://doi.org/10.5194/os-17-789-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/os-17-789-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 17, issue 3

Research article

|

17 Jun 2021

Research article |

| 17 Jun 2021

The new CNES-CLS18 global mean dynamic topography

Sandrine Mulet, Marie-Hélène Rio, Hélène Etienne, Camilia Artana, Mathilde Cancet, Gérald Dibarboure, Hui Feng, Romain Husson, Nicolas Picot, Christine Provost, and P. Ted Strub

Download

Final revised paper (published on 17 Jun 2021)
Preprint (discussion started on 07 Jan 2021)

Interactive discussion

Status: closed

RC1:
'Comment on os-2020-117', Anonymous Referee #1, 22 Jan 2021

Overview

This very well written manuscript describes how the newest version of the CNES mean dynamic topography product is calculated, and is a straightforward read for those familiar with the 2009 and 2013 versions. As a user of the 2018 product already, I am happy to see this published in the peer-reviewed literature. The authors have produced a product of great value to the community, and I particularly enjoyed the case studies in Section 8 to validate the results and demonstrate the improved performance. Many of the results and figures in this section, such as Fig. 12, are very striking. As the authors note in their conclusions, many different factors went into these improvements; while one certainly wonders about the relative impact of improved techniques vs. higher data density/higher spatial resolution, in the end a user wants all of these improvements.

I have only very minor comments, and recommend that the manuscript be published with only minor reivions.

Specific comments

94: there should be a citation or acknowledgement for the SD-DAC.

113: what does “section 0” mean? Should this say section 5?

168: is there a reason why the undrogued drifters were not used for z=0m? Is this due to slippage (noted later in the manuscript)?

170-172: no lowpass is applied to the Argo data, because the floats aren’t at the surface long enough to allow for this. The authors should note that explicitly, and that these data thus include much more noise from high frequency motion. This again makes me wonder why the ~hourly undrogued drifter data wasn’t used for z=0m. [NOTE: the authors address this on lines 265-268. I left this comment as a notice that readers may be wondering about this earlier.]

201-205: this is a great result! Very well presented.

263: how does this work at very low latitudes? Wouldn’t max(Pi, 24h) go to infinity?

Technical issues

65, 178: misplaced ().

226: font size change.

Citation: https://doi.org/10.5194/os-2020-117-RC1
- RC2:
  'Reply on RC1', Anonymous Referee #2, 03 Feb 2021
  
  I agree on the comment of referee #1 that this is a well written manuscript that nicely describes the updated methodology to produce the MDT and presents striking improvements of the new MDT compared to older versions.
  
  What I would like to see are comparisons with geodetic MDTs that apply new combined geoid models using GOCE data (XGM, SGG-UGM, GOCO05c, GECO). In l48 it is stated, that for spatial scales shorter than 100km other information than those provided by geodetic MDTs alone is needed. But actually, when applying combined geoid models clearly signal is found below 100 km scale. This has been shown at least for the Gulf Stream and the Kuroshio (Siegismund 2020). I'm pretty sure the CNES-CLS18 MDT resolves shorter scales than any geodetic MDT, but to show this would even underpin the additional value of the new CNES-MDT compared to any pure geodetic MDT.
  
  Minor points:
  
  225, 227, 267, 272 : I guess, Rio (2012) is the right reference.
  
  A lot of 'section 0' is found (101,113,123,258).
  
  Citation: https://doi.org/10.5194/os-2020-117-RC2
  - AC2: 'Reply on RC2', Sandrine Mulet, 26 Feb 2021
    
    Thanks a lot for your comment. I agree that geodetic MDTs computed from combined geoid model could resolve scale smaller than 100km, I should precise that I talk about “satellite only” geoid model. However, to be independent from any apriori MDT, we do not want to use combined geoid models that use a priori MDT to convert altimetric information to gravimetric signal.
    I agree that using combined geoid model, geodetic MDTs have valuable signal at scale shorter than 100km especially in main currents but in less energetic area residual noise could be important and dealing with this noise is not trivial.
    I also agree that some comparisons with other geodetic MDTs could be useful and this will be considered in the revised manuscript.
    
    Citation: https://doi.org/10.5194/os-2020-117-AC2
- AC1: 'Reply on RC1', Sandrine Mulet, 04 Feb 2021
  
  Thanks a lot for your positive comment and for your specific questions that will help us to clarify the manuscript. Below you will find answers to your comments that will be taken into account in the revised manuscript.
  94: there should be a citation or acknowledgement for the SD-DAC. Lumpki et al, 2013: Lumpkin, R., S. Grodsky, M.-H. Rio, L. Centurioni, J. Carton and D. Lee, 2013: Removing spurious low-frequency variability in surface drifter velocities. J. Atmos. Oceanic Techn., 30 (2), 353—360, http://dx.doi.org/10.1175/JTECH-D-12-00139.1.
  113: what does “section 0” mean? Should this say section 5? Yes indeed
  168: is there a reason why the undrogued drifters were not used for z=0m? Is this due to slippage (noted later in the manuscript)? Exactly, we have noted in previous studies (Rio et al., 2014) that Argo floats are much less affected by wind slippage compare with undrogued drifting buoys certainly thanks to their design. Consequently, we do not need to correct Argo float drift from wind slippage and thus we can use them to estimate Ekman model at the surface (z=0). We will clarify that in the revised manuscript.
  Rio, M.-H., S. Mulet, and N. Picot (2014),Beyond GOCE for the ocean circulation estimate: Synergetic use of altimetry,gravimetry, and in situ data provides new insight into geostrophic andEkman currents,Geophys. Res. Lett., 41, doi:10.1002/2014GL061773.
  170-172: no lowpass is applied to the Argo data, because the floats aren’t at the surface long enough to allow for this. The authors should note that explicitly, and that these data thus include much more noise from high frequency motion. This again makes me wonder why the ~hourly undrogued drifter data wasn’t used for z=0m. [NOTE: the authors address this on lines 265-268. I left this comment as a notice that readers may be wondering about this earlier.] We will then mention this point earlier in the manuscript.
  263: how does this work at very low latitudes? Wouldn’t max(Pi, 24h) go to infinity? We’ve forgot to say that in practice an upper bound is set at 6 days for the cutoff wavelength. We have to add this information.
  
  Citation: https://doi.org/10.5194/os-2020-117-AC1

Peer review completion

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

AR by Sandrine Mulet on behalf of the Authors (02 Apr 2021) Author's response Author's tracked changes Manuscript

ED: Publish as is (23 Apr 2021) by Anne Marie Treguier

Short summary

Satellite altimetry has revolutionized ocean observation by allowing the sea level to be monitored with very good spatiotemporal coverage. However, only the sea level anomalies are retrieved; to monitor the whole oceanic signal a temporal mean (called mean dynamic topography, MDT) must be added to these anomalies. In this study we present the newly updated CNES-CLS18 MDT. An evaluation of this new solution shows significant improvements in both strong currents and coastal areas.