Ongoing Development of the Bass Strait GNSS/INS Buoy System for Altimetry Validation in Preparation for SWOT

Round 1

Reviewer 1 Report

In this paper, the authors used GNSS (GPS + GLONASS) + INS instead of GPS-only + INS, to improve the SSH observation precision. PPK and PPP methods were considered to obtain high-precision SSH results. And the results are evaluated with the in situ mooring by 4 deployments. Finally, the authors tried to analyze the influence of sea conditions (wind, current, wave) on the accuracy of buoy SSH measurement. The authors have done a lot of work, but there are still some problems as follows,

1. Line 60-61, “While the study offered some insight into the precision of GPS/INS equipped buoys, the integration of non-GPS constellations was not considered, and the basic dynamics of the buoy were not investigated.” Not an accurate presentation.  Although Zhou did not consider other constellations (i.e., Galileo, BDS, and GLONASS) besides GPS, Huang et al., (Monitoring High-frequency Ocean signals using low-cost GNSS/IMU buoys) used the combination of GNSS and IMU to monitor the sea surface. Please check.

2. How does the buoy supply power during the experiment? Which modules consume power? What is the GNSS/INS power consumption?

3. Why not use GPS + BDS + Galileo instead of using GPS + GLONASS to avoid the inter-frequency bias?

4. Line 103-104, "The pressure gauge observes the integrated pressure field and is not sensitive to the wave field given the depth of deployment". If the pressure information is not sensitive to the wave information, will there be a large error in SSH observation? What is the accuracy of the bottom-mounted pressure gauge? I think the authors need to give its parameters because we don't know whether GNSS+INS is more accurate or the pressure gauge is more accurate. Analysis of buoy motion parameters may be meaningless in the following manuscript if GNSS+INS accuracy is higher.

5. The sampling rate of the pressure gauge is about 20min, and it is not sensitive to high-frequency (such as wave) data. Can the influence of waves on GNSS+INS SSH measurement residuals be effectively analyzed?

6. What type of antenna was used in the experiment? Does the Marine environment have a significant impact on GNSS observation, such as a large multipath effect, or a high sea state will cause signal loss?

7. Line 169-170, “a differenced series between the SSH and the reconstructed tide series based on a decade long mooring SSH series at JAS CP was calculated.”Mooring data measures water depth (or pressure), while GNSS measures sea surface height (or ENU coordinates with respect to the base station). How to compare the values of two different reference coordinate systems? Please indicate in the manuscript.

8. Single-frequency or dual-frequency observation used for DD? In the case of dual frequency observations, is the IF model or uncombined models used in the calculation?

9. What products and models are used in the PPP? Is the PPP fixed solution performed? If the fixed solution is adopted, what about the bias products?

10. Due to the influence of wind and waves, the buoy will tilt, resulting in the height change from the antenna reference point to the sea surface. Whether this error is considered in this article, it is recommended to refer to and cite the paper “Di, M., et al. "Evaluation of Real-Time PPP-Based Tide Measurement Using IGS Real-Time Service." Sensors 20.10(2020).”.

11. Line 237, “the wave direction from INS can be calculated,” The buoy will rotate due to the combined force of waves and the tether. Can the direction measured by INS represent the direction of the wave at this time?

12. Line 271-273, ”With GLONASS introduced, up to 15 total satellites were tracked, forming a better geometry for estimation leading to solutions with noticeably reduced deviations from a smooth trajectory.” Why not add the analysis of DOP value? 

13. After adding the GLONASS data, does it affect the fixed rate of GPS? What are the fixed rates before and after adding GLONASS?

14. How to set the weight between GNSS(GPS) and INS in the combination?

15. In Figure 5, we can see that the wind stress in the final part of Window 1 is large, even larger than that in Window 3. But why the magnitude of the wind wave in Window 1 is smaller than that in Window 3?

16. Line 412-413, “Figure 6a shows that the trend of the instantaneous unsmoothed SSH is in generally good agreement with the titling and the decomposed roll and pitch” Why the interval of unsmoothed SSH data is 30s?

17. Line 445-456, “given the in-situ mooring is unlikely to be sensitive to current and waves.” It is recommended to add tank test to verify whether it is sensitive to current or waves.

18. Line 457-458, “Under this scenario of external forcing setup, a quadratic fit between the buoy tilt and wave energy sensed by the buoy is shown in Figure 7”, How to calculate the wave energy?

19. Line 480, “… the wind stress projected to the tether direction and the B-M residual…”, How to project the wind stress to tether direction?

20. In figure 8, the authors try to find the relation between the current and B-M res. How does the current affect the results of the vertical direction? The current is horizontal and will not affect the vertical direction. The accuracy of GNSS/INS positioning is related to its own data quality (GNSS may be subject to certain interference, but it is generally random). 

21. Line 537, “… in the high frequency domain below the 2-hour band”. I think 2 hours is not the high-frequency part,  the main problem is that the sampling rate of the mooring is too low.

22. Line 541-542, “The sidereal multipath on the carrier phase is likely to affect the base stations onshore more than the offshore buoys.”. What is the environment of the base station? In general, the multipath effect in the marine environment is higher than that on land, but the multipath effect mainly affects the pseudo-range observation and has little impact on the carrier observation. 

23. Line 575-576, “We note typical percentages of outliers from PPP was ~10% compared to 6% for DD”, Will the adoption of multi-GNSS in PPP improve the results? Maybe there are too few GPS satellites.

24. Line ”594-596”, “but it is not impacted as much as the DD method by the sidereal harmonics in the lower frequency bands, likely driven by multipath effects at land-based reference sites.” It is recommended to use two static stations to verify the assumption that the error is caused by multipath.

Author Response

Thanks for reviewer #1's valuable comments. Please see the attachment for the response.

Author Response File: Author Response.docx

Reviewer 2 Report

The authors worked on the GNSS/INS buoy system for altimetry validation in preparation for SWOT. The authors said that a comparison between GPS and GNSS solutions suggested up to 3 cm reduction in the root mean square of the buoy minus in situ mooring SSH residual over the selected sidereal periods. Then, comparison between double differencing and precise point positioning solutions suggested a possible common mode error external to GNSS processing. Moreover, GNSS and INS observations were used during periods to investigate the possible influence of buoy dynamics. However, there are some comments and questions about their work:

- In the abstract, the word in situ is repeated three times. Using synonyms to make your writing more flexible.

- The acronyms have to be defined before using, and only defined one time at the beginning: e.g. RMS, PPP, ...

- The Methods Section should come before the Data Section.

- You should distinguish between Inertial Navigation System (INS) and Inertial Measurement Unit (IMU).

- Line 84: Why do you choose GNSS data at 2 Hz and INS data at 100 Hz?

- Line 149: Why do you use only GLONASS and GPS satellites from many GNSS satellites?

Author Response

Thanks for reviewer #2's valuable comments. Please see the attachment for the response

Author Response File: Author Response.docx

Reviewer 3 Report

Review of the article : « 

 Ongoing Development of the Bass Strait GNSS/INS Buoy Sys-2 tem for Altimetry Validation in Preparation for SWOT »  by Zhou et al.

The paper focuses on GNSS equipped buoys which is important for the study of the accuracy of Sea surface Height altimetry product. The work investigates the use of GNSS vs GPS only positioning accuracy. Here GNSS is mostly GPS+GLONASS constellation. They use DD and PPP techniques. Their analysis of the position with GPS and GNSS shows an improvement with more satellites and less transients signals (due to dilution of multipath). They further investigate the buoy dynamics (wave height, current velocity) using GNSS +INS. The test site is the Bass Strait near Tasmania in Australia.

General comments:

The paper reads well.

Technical comments:

There is a lot of processing work behind the results. TRACK is modified to use GPS and GLONASS, then INS observations. GIPSYX use only GPS. I have not used GIPSYX, but I wonder if it can also be modified to include other constellations? It would have been nice to see a DD vs PPP study (not only GPS) in order to look further into the sideral spurious signals. Except that, I can understand well the strategy used to include GLONASS in section 3. All the steps are well documented (outliers removal, ambiguity estimation and update). I would like to see a block diagram to have a good overview. It will help the unexperimented readers to grasp quickly the methodology. I would also suggest that 3.1 and 3.2 are under the same section  such as Processing of the GNSS observations

Keep the name of section 3.3. However, Section 3.3.1 is not very clear. You may need to put the derivations in the appendix, if not necessary for the results. Same for Section 3.3.2.

Results:

The analysis is sound and well explained.

The first part shows well the addition of GLONASS (GNSS+GPS) on the GPS/INS and GNSS/INS positioning solution (Figure 3). The sideral frequencies are smoothed out with the GNSS/INSS solution. The RMS decreases slightly.

The second part is dedicated to PPP vs DD solution with GPS only. The sideral harmonics are not shown at the same frequency. This is expected due to the difference in the methodology.

The third analysis on the Buoy dynamics when including INS is very interesting (in particular Figure 5). I am not working with Buoy, but the Figure shows the time series of the different parameters estimated from the Buoy position and correlated with the various in situ sensors. I believe the figure caption should be more explicative, especially the upper panel and the three windows. The reader needs to find a lot of details in the text.

Overall, the paper is very interesting. I believe that the paper can be accepted.

Author Response

Thanks for reviewer #3's valuable comments. Please see the attachment for the response.

Author Response File: Author Response.docx

Reviewer 4 Report

This paper enhanced an existing buoy aproach and make contribution for better understanding of the buoy dynamics. The contents of this manuscript is well organised, the figures is well presented and clear.

General comments and minor points:

1. For Fig.3b, it should be better to give ecological distribution fitting for the current tow histograms. The fitting parameters can present some conclusions.

2. For section 4.2, I am not sure whether the DD solutions is achieved by GPS or GNSS?

3. The points’s color of No.63 and 64 deployment is too close,especially in Fig.8 a and b. It may be better to make the difference more significant.

4. In the section 5.1, would you refer as perspectives how other GNSS, such as BDS, Galileo, can help improve the positioning performance?

Author Response

Thanks for reviewer #4's valuable comments. Please see the attachment for the response.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Thank you for the revisions made. Now it improve much and is ready to be accepted.