the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Enhanced resolution capability of SWOT sea surface height measurements and its application in monitoring ocean dynamics variability
Abstract. The wavenumber spectrum of sea surface height along ground profiles is commonly determined to quantify the magnitude of detectable ocean dynamic features by altimetry missions. In this paper, wavenumber spectral were calculated and compared for HY2B, Saral/AltiKa, Sentinel-3A, and SWOT. The wavenumber power spectral density(PSD) of sea surface height (SSH) was averaged using weighted methods across multiple along tracks within defined boxes. The deduced resolution capabilities were also compared and analyzed, evaluated using the relevant definition of one-dimensional mesoscale resolution capability. We verified that the latest wide-swath SWOT mission offers significantly improved measurements. For example, in the vicinity of Kuroshio, the one-dimensional mesoscale resolution of SWOT is about 25 kilometers, twice the resolution capability of conventional satellites. In addition, the quality of measurements declined obviously over regions where the eddy kinetic energy gets larger. Finally, a global analysis of ocean dynamics variability scales was conducted based on two cycles of SWOT data using reciprocal power spectral analysis. The results showed significant geographic and temporal variations in the ocean dynamics variability scales, which are mainly relative to sea state variability. The regions with large scales of ocean dynamics variability are concentrated in oceans with strong currents and unstable sea states, such as the Kuroshio Current, the Gulf Stream, and the Antarctic Circumpolar Current. In addition, the scale of ocean dynamics variability is not necessarily large where eddy kinetic energy is large, such as the equator and the northwest Indian Ocean current area. Ocean dynamics variability also varies across seasons.
- Preprint
(1927 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2024-3005', Anonymous Referee #1, 02 Nov 2024
The article employs wavenumber spectral analysis to conduct experiments and introduces a novel method for global data statistics. It compares data derived from four distinct satellite types, thereby highlighting the advantages of the SWOT satellites. Additionally, the authors analyze global ocean-scale changes by defining a new parameter and utilizing SWOT satellite data. While the proposed methodology and the new parameter are intriguing, they require further clarification and validation. The English writing should be further polished. I recommend that this manuscript be considered for publication, contingent upon addressing the following modifications.
- The paper does not sufficiently demonstrate the significant advantages of the authors' improvements to the algorithm compared to existing methods. To better emphasize the necessity and effectiveness of the proposed enhancements, it is recommended that a comparison with the traditional averaging method be included.
- The averaging method presented in Appendix A employs distance-weighted averaging across a range of orbital data. However, it is crucial to assess whether this method is scientifically sound. I request a detailed explanation supported by appropriate examples and a discussion of the statistical weight distribution involved in this approach.
- Footprints of SWOT, HY2B, Sara and S3A are different. The resolutions of SSHs for these 4 missions are also different. How about the effects of different resolutions on the results?
- line 55: Xu Y et al. -> Xu et al.
- line 68: Vergara O et al. -> Vergara et al.
- section 2.1: Time spans for different satellite altimetry missions are different.
- line 103: What is an uncorrected data product?
- line 104: What is an uncorrected data product?
- line 105: What is an uncorrected data product?
- line 106: What is an full corrected data product?
- line 113: What about the frequency of SSHs used? 40Hz or 20Hz?
- line 128: The sentence may repeat.
- line 131: What is the nadir-stellar point?
- line 140: along-orbit SSH -> along-track SSH.
- line 154: What is the SSH anomalous wavenumber power spectral density?
- Line 165: What are these three parameters in detail?
17: fig. 2: How to define and determine the slop? This is not global.
18: fig. 3: This is not global.
19: fig. 4: This is not global.
- fig. 5: This is not global. In line 275, the colorbar units in Figure 5c are absent. Please ensure that the units are indicated for clarity and proper interpretation of the data.
- Section 4 introduces a new parameter; however, the experiments are conducted exclusively using data from the SWOT 21-day repeat cycles. Is it feasible to conduct experiments using data from the SWOT one-day repetition cycles for specific locations? This approach would more effectively illustrate SWOT's contributions to understanding ocean sub-mesoscale dynamics.
- How to show the dynamic mechanism?
- fig.6: This is not global.
- The reference cited at the end of line 72 is incorrect, and the reference formatting at the end of line 55 is inconsistent. Please review the reference formatting throughout the manuscript to ensure uniformity.
Citation: https://doi.org/10.5194/egusphere-2024-3005-RC1 -
AC1: 'Reply on RC1', Yong Wang, 03 Jan 2025
Dear reviewer:
The author's team would like to thank you for reviewing the paper and providing useful feedback and suggestions. We have carefully read and responded to your comments. Your comments are in black font, our explanatory response is in blue font, and the corresponding revision in the manuscript is in red font.
-
RC2: 'Reply on AC1', Anonymous Referee #1, 06 Jan 2025
The authors have addressed all comments and proposals.
Citation: https://doi.org/10.5194/egusphere-2024-3005-RC2
-
RC2: 'Reply on AC1', Anonymous Referee #1, 06 Jan 2025
-
RC3: 'Comment on egusphere-2024-3005', Anonymous Referee #2, 05 Feb 2025
The paper applies wavenumber spectral analysis to compare SSH data from four altimetry missions—HY2B, Saral/AltiKa, Sentinel-3A, and SWOT. It evaluates their spectral characteristics and introduces a new method for global statistical analysis. A key focus is on SWOT’s ability to resolve small-scale ocean variability, leveraging its finer spatial resolution. The study highlights the differences in spectral content among these missions and attempts to quantify global-scale variations using SWOT data by introducing a new parameter.
While the study presents a well-motivated analysis, the manuscript would benefit from clearer phrasing in several sections to improve readability and eliminate ambiguities. Additionally, certain methodological aspects require further justification, particularly regarding data preprocessing and resolution effects. I recommend that this manuscript be considered for publication, provided the authors address the following major concerns:
Major Comments
- The averaging method used to compute spectral slopes (Appendix A), which relies on a distance-weighted scheme, raises concerns about its statistical validity. A detailed justification of this approach, including examples and a discussion of the weight distributions, is necessary to ensure the robustness of the analysis. Additionnaly, A direct comparison with conventional approaches (e.g., standard averaging methods) is needed to show the advantages of its proposed methodology over standard spectral analysis techniques.
- The section"Global Analyses of Ocean Scale Changes" lacks clarity, primarily due to insufficient explanation of key methodological steps. The authors refer to Appendix B, which introduces spectral coherence methods without clearly linking them to the main analysis. The section does not explicitly explain how the proposed global-scale parameter (derived from mutual power spectra) captures ocean variability. The transition from Appendix B’s coherence-based method to a global variability assessment is particularly vague.
Specific Comments
l.28: “…sub-mesoscale activity can also reverse the cascade of energy from…”
l. 55: Xu et al. (2012) instead of Xu Y et al. (2012)
l. 71: “Another altimeter ”Which one?
l. 84: “reciprocal power spectral analysis” The term “reciprocal power spectral density” isn't a standard term, better referring to cross spectral density.
l. 90-91: The explanation of how SWOT data is split into along-track components lacks details on the method used. Does this involve simple subsampling, or is interpolation applied?
l. 92: From October to November 2023: Does it mean October AND November?
l. 93-94: “To compare […] by SWOT’s KaRIn”: The sentence structure could be improved to convey the intended meaning more effectively.
l. 103: “orbit data” rather than “orbit fix data”.
l. 104-108: There is some confusion in this section due to the repetition of sentences. Is the GDR corrected or uncorrected? What is the difference between a corrected and an uncorrected product? Additionally, is the SGDR corrected or uncorrected?
l. 116-119: The advantages of Ka-band in SARAL/AltiKa (lines 116–119) should be supported by relevant literature.
l. 118: There seems to be a repetition of the word “water”. A final period is missing between “data” and “The final...”. Altair band: This seems to be a typo. It likely refers to the AltiKa Ka-band, which, if I’m not mistaken, operates exclusively in the Ka-band. Therefore, the phrase 'The final choice was...' may be unnecessary.
l. 119: Why were only cycles 175 and 176 selected for SARAL/AltiKa? Do these correspond to the October-November 2023 period?
l. 120: “The Sentinel-3A (S3A) satellite carries the SRAL altimeter…” : I would specify “…carries Synthetic Aperture Radar Altimeter (SRAL)…”
l. 127: As mentioned earlier, please specify why cycles 104 and 105 were selected.
l. 131: What are sub-stellar points?
l. 132: “SWOT carries s Ka-band radar” : Typo: 's' should be changed to 'a'.
l. 137: “Selected data from SWOT's ocean […] are selected for this paper.” Remove the first “selected”
l. 140, 142, 144: replace “along-orbit” with “along-track”
l. 140-142: What kind of corrections are applied?? Clarifying this step is essential for reproducibility.
l. 149: “…Fourier transforming it…” I would suggest to write instead: “performing a Fourier transform on it”
l.154: “We calculated the SSH anomalous wavenumber power spectral density (PSD) for each mission…” This is unclear. Are the authors referring to the power spectrum of SSH anomalies?
l. 157: The phrase "departed from the previous method of averaging" is vague. What method did the authors adopt instead? The preprocessing steps for calculating the PSD in each 10°x10° box follow methods similar to Dufau et al. (2016). However, instead of averaging all individual PSDs within a box (as done in previous methods), the authors suggest a different way to compute the average PSD for each box. This section requires further clarification
l. 166: “... is the 2 km sampling rate…” : “resolution” instead of “rate”.
l. 169-170: “For wavelengths below 25 km for the first three missions, the 1 Hz SSH error level was estimated by fitting a level to the spectrally flat noise levels present in the PSD maps (Figure.1).” This sentence is confusing. I assume Figure 1 presents the unbiased spectra (i.e., without the constant noise level), is that correct? If so, the sentence needs to be rephrased for clarity. I would also suggest adding the noise level to the plot, if possible, to visualize the intersection with the slope. Additionally, including references to Xu and Fu (2012) and Dufau et al. (2016) would be helpful.
l. 175: “…by removing the esFmated constant error level below ...” Remove “below”.
l. 176-178: “Diverse methods for calculating the Power Spectral Density (PSD) […] in the estimated PSD slope range” : I agree but should not be here.
l. 179: “Hence, for the first three conventional missions, we chose wavelengths in the range of 70-250 km”. Missing references: Dufau et al., 2016; Le Traon et al., 2008; Xu & Fu, 2011.
l. 184-185: “Due to the presence of […] etc (Boas et al., 2022)” : I believe this should not be a sentence on its own, but rather linked to the previous one for better coherence.
l. 185: “wavelength range” instead of “wavelength”
l. 186-188: The one-dimensional mesoscale resolution capability essentially represents the shortest wavelength detectable in along-track altimeter observations where the signal exceeds the noise, correct? Perhaps this sentence could be rephrased for clarity.
Global resolution capability of altimetry satellites
l. 207: “[...] theoretical predictions from SQG and GG theories.” Add a reference please.
l. 208-209: The authors should specify the figure number referenced in the cited paper.
l. 212: “[…] oblique pressure tide […]” : Do the authors mean baroclinic instead of oblique?
l. 261: “satellite altimetry observes” instead of “satellites altimetry observed”
l. 226: “for one month” instead of “for the one month”.
l. 227: “[…] results of previous studies.” Add references please.
l. 230: “northwest Pacific” instead of “Pacific northwest”.
l. 239-241: “However, for the data from the period […] potential error in one of the satellite's corrections” Could the authors clarify this statement?
l. 243-244: How the authors explain the high-level noise pattern in the mid-north Atlantic for SWOT (Figure 3c)?
l. 256: Dufau et al. (216) instead of Dufau C et al. (2016)”.
l. 269: “[…] where the peak slope occurs.” Would it be possible to include a plot to visually represent this statement?
l. 272: “Experiments have demonstrated…” Add references
Global analyses of ocean-scale changes
l. 284: As I mentioned earlier, it would be better to refer to the cross-spectral density.
l. 288: Marks et al. (2016) instead of Marks K M et al (2016).
l. 290: Why 0.5? Add some details please.
l. 294: Western Boundary Currents
l. 297: northwest Indian Ocean current
l. 306: “world’s ocean current regions” Remove “world’s”
Appendix B
l. 403-407: How are equations 5 and 6 linked? Is R the cross-correlation function and S the cross-spectral density (Fourier transform of the cross-correlation)? Also, in equation 6, is P equivalent to S? Please be careful with the notation. Additionally, could you define all the variables?
l. 408-409: “The coherence function can judge the […] cycle signals to infer resolution capability.” Already mentioned earlier, see l. 397-398. Marks et al. (2016) instead of Marks K M et al (2016).
l. 417: Which kind of interpolation is performed? Linear?
Figures
Figure 1: The authors should specify whether these spectra are biased or unbiased.
Figures 2, 3 and 5c: The units are missing.
Figure A1: The legend mentions some orange points, but I cannot see them in the plot.
References
l. 481 and 499: The reference formatting is inconsistent.
Viewed
HTML | XML | Total | BibTeX | EndNote | |
---|---|---|---|---|---|
209 | 74 | 182 | 465 | 9 | 12 |
- HTML: 209
- PDF: 74
- XML: 182
- Total: 465
- BibTeX: 9
- EndNote: 12
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1