the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 21 Aug 2025
First 2D record of a tsunami by SWOT satellite: observation data and preliminary numerical simulation of the 19 May 2023 tsunami near the Loyalty Islands
Abstract. The NASA-CNES altimetry mission SWOT (Surface Water and Ocean Topography) deployed in December 2022 embarks a Ka-band Radar Interferometer (KaRIn), providing a 120 km-wide swath sea level measurement. On 19 May 2023, SWOT was able to record a 2D signature of the tsunami generated by the Mw 7.7 earthquake southeast of the Loyalty Islands (southwest Pacific Ocean), about 1 hour after the earthquake, on a straight SSW-NNE path. Comparison between numerical models and real measurements was performed to assess SWOT’s ability to monitor tsunami waves. A uniform coseismic slip rupture model allows to satisfactorily fit the regional observations. Testing models against a dynamic representation of the tsunami wavefield (instead of static) show a good phase agreement, but simulated amplitudes and energy spectra are lower than the measurements. However, this SWOT unprecedented 2D observation critically inform on tsunami propagation and modeling, and offer a breakthrough perspective for better predictions.
- Preprint
(2628 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-3926', Anonymous Referee #1, 10 Oct 2025
-
AC1: 'Reply on RC1', Jean Roger, 14 Nov 2025
The work entitled “First 2D record of a tsunami by SWOT satellite: observation data and preliminary numerical simulation of the 19 May 2023 tsunami near the Loyalty Islands” focuses on satellite radar observation of a tsunami event on 19 May 2023. It is the first dynamic observation of a two-dimensional tsunami wave field. The satellite’s orbit allowed the observation of the wavefront propagating in both directions from the epicenter, about 1 h after the earthquake. I wonder why the authors call it 2D observation, when in fact it is a 3D observation because it includes the sea surface elevations (Sea Level Anomaly) within a two-dimensional domain. These observations are compared with numerical simulations performed with COMCOT model, showing generally good agreement with observations, although some differences reveal weak points of the simulations and indicate that there is room for future model improvements. Simulations show a small delay in the main peak of the tsunami wavefront and underestimated amplitudes, especially for the secondary waves in their propagation to the north. The results presented are relevant and highlight the usefulness of SWOT data to improve tsunami predictions. Some paragraphs and figures show inconsistencies that must be corrected (see attached file). The manuscript presents new data, and I recommend publication once the minor changes have been addressed. I hope my comments are useful to the authors and help improve the quality of the paper.
- We thank the reviewer for his positive feedback and constructive comments, and we will consider the comments one by one in the revision process.
Specific comments:
Title and abstract: Why 2D record? It is an image of a 2D domain, but it includes the sea level anomaly, which is the third dimension, therefore, perhaps it could be called a 3D record?
- Indeed, this a 2D record from a 2D altimeter, compared to the usual 1D record done with 1D altimeters. In fact, traditional nadir altimeters survey data along a single track providing 1D profiles of the sea surface height. SWOT is equipped with a KaRIn which surveys data across a wide swath providing 2D (bidimensional) “snapshots” of the oceanic surface (e.g., Peral et al., 2024). For clarity we propose to rewrite the title this way: “First mapping of a tsunami wavefield by SWOT satellite: observation data and preliminary numerical simulation of the 19 May 2023 tsunami near the Loyalty Islands.”
Lines 116 to 119: This paragraph starts “On 18 May 2023”…. “This corresponds to pass number 147 of cycle 525”. The tsunami occurred on 19 May, so I think this date is mistaken. And then, in line 118 it says “measurements from 19 May 2023 (pass 147 of cycle 524 … and 525 …)” this part must be clarified, as cycle 524 and 525 should correspond to different dates (are those 18 May and 19 May?).
- We are going to correct the numbers and cycles accordingly. First, you’re right, it should start with “On 19 May 2023…”. Then, the cycle numbers are right, but the date is wrong: it should be “measurements from the 18 May 2023 (pass 17 of cycle 524 before the tsunami…”. Note that we identify a typo in the document with pass 147 being pass 17 in reality. In fact, there are only 28 pass per day in the considered period of time. This typo will be corrected everywhere in the manuscript.
Caption of Figure 3: “(the optimal interpolation OI is the MIOST)” Please include this “optimal interpolation” in figure 3b and explain it if necessary, as it is not clear in its current form.
The optimal interpolation can be added to figure 3b and related details added to the caption, however, we are not sure it is relevant to show.
Updated caption : Figure 3: SWOT SLA and Power Spectral Density. (a) Zoom on SWOT SLA pass 17 on 19 May in 1D along track of the Nadir and KaRIn instruments for several across track distance, at 51, 31 and 11km either side of the nadir position. The dashed lines represent the optimal interpolation removed for each profile, with respect to colors.
- It is important to remove mesoscale ocean variability (but also gyres and other potential large structures) to avoid confusing their signatures with tsunami signature. Mesoscale signature could be higher than tsunami amplitude (not the case for the signal at latitude of 18S°). Difference of signal with Oi removing or not is visible at Figure 2 between 2b & 2c.
Section 4: Please include an introduction paragraph between lines 176 and 177 indicating the contents of section 4.
- Section 4 will now be introduced with the following sentence to describe its content: “Simulation results and measurement data are compared (§4.1) and the misfit between them is further discussed (§4.2) as well as the residual errors from SWOT signal (§4.3).”
Line 188: “tsunami propagation wavefield 62’ after the earthquake”. It indicates that figure 5 shows the simulation 62’ after the Te but in line 207 it says that the results are shown “at t=1h”. Clarify which time was used.
- Indeed, the time on Figure 5 is 62 min after the earthquake, and we will correct the time in the text.
Lines 206 and 207: “In five minutes, the tsunami would have travelled ~50 km”. Include a discussion about the lack of phase agreement in figure 4c. If the north wavefront was observed at 1h 2 min, the difference in phase of around 20 km (for the first wavefront) may be easily explained.
- That’s true, we are going to add a short sentence to reflect this estimation.
Section 4.1: In general, I think this section would be better structured if the results of figure 5 were shown first followed by the zoomed-in view of figure 4 on the north wavefront. Additionally, the results and analysis performed in figure 4 could be improved by showing the results of the COMCOT model at time 1h 2min which will likely achieve better agreement with the north wavefront observation.
- We thank the reviewer for his suggestion. We are going to swap figure 4 and 5 as it makes more sense effectively. The simulation snapshot at one hour will be replaced with the one at 1h 2 min as suggested.
Technical corrections:
Figure 2 and lines 153 to 157: The caption of figure 2 refers to the different panels as left, center, and right, while the text in the next paragraph refers to 2a (line 153), 2b (line 154), b (line 156) and 2c (line 157). Please include the letters in figure 2, correct the caption, and line 156 accordingly.
- The letters will be included in figure 2 as suggested, and the caption and line 156 will be modified accordingly.
Lines 170 and 173: References to figure 3b and 3c. It seems that they should refer to 3a and 3b, respectively. 3c does not exist.
- The reviewer spotted an issue coming from a previous version of figure 3. We are going to update the text, accordingly, keeping only mentions of figure 3a and 3b.
Figure 3: Please include the title of the X-axis in panel b.
- The title will be added to the X-axis.
Line 180: It explains the results shown in figure 4c without introducing the contents of this panel c before. Please include a description of what the profiles of panel care.
- The following description of panel c will be added before line 180: “Figure 4c shows a comparison between SWOT data and the simulation model along the nadir.”
Also, we noticed that the caption of Figure 4c does not explain what the vertical blue lines are: they symbolize the dashed circles visible on Figure 2. This will be added to the caption.
-
AC1: 'Reply on RC1', Jean Roger, 14 Nov 2025
-
RC2: 'Comment on egusphere-2025-3926', Anonymous Referee #2, 15 Oct 2025
This paper presents a very interesting study on the use of SWOT satellite data for detecting a tsunami event generated by the Mw 7.7 earthquake southeast of the Loyalty Islands (Southwest Pacific Ocean). The authors used the Level-3 Low Rate Sea Surface Height SWOT product, which is easily accessible to researchers, and compared it with numerical model simulations. Their analysis shows a strong consistency between the SWOT observations and the model results.
Tsunami is a type of event for which observations by satellites are rare and depends on “luck”, but when captured, they provide unique opportunities to study the event and its dynamics in ways that models alone might not allow (especially with a satellite like SWOT with high spatial resolution and advanced capabilities). This article therefore opens the door for more in-depth analyses and demonstrates the potential of SWOT studying such oceanic phenomena and events.
Although I am not a tsunami expert, I found that the literature review about tsunami detection from satellite altimeters is well presented in the introduction. The presentation of the SWOT altimeter and its capabilities is also concise and very clear.
After checking the comments of the first reviewer, I find myself in agreement with their comments and I don’t have further technical corrections to provide. I therefore propose a minor revision, and two additional comment/suggestions that could further improve the manuscript:
- To better quantify the time lag between the simulated main peak and the SWOT peak, I suggest providing a comparison spatially restricted between −19° and −18° latitude and using a simulation output closer in time to the SWOT overpass above this restricted area. This would make the comparison more precise.
- I also propose that the authors change the title of section 4 to something more generic like “Comparisons and discussion”, and change the title of section 4.1 to something like “Comparison between observations and model output” to better reflect the content and structure of the section.
Citation: https://doi.org/10.5194/egusphere-2025-3926-RC2 -
AC2: 'Reply on RC2', Jean Roger, 14 Nov 2025
This paper presents a very interesting study on the use of SWOT satellite data for detecting a tsunami event generated by the Mw 7.7 earthquake southeast of the Loyalty Islands (Southwest Pacific Ocean). The authors used the Level-3 Low Rate Sea Surface Height SWOT product, which is easily accessible to researchers, and compared it with numerical model simulations. Their analysis shows a strong consistency between the SWOT observations and the model results.
Tsunami is a type of event for which observations by satellites are rare and depends on “luck”, but when captured, they provide unique opportunities to study the event and its dynamics in ways that models alone might not allow (especially with a satellite like SWOT with high spatial resolution and advanced capabilities). This article therefore opens the door for more in-depth analyses and demonstrates the potential of SWOT studying such oceanic phenomena and events.
Although I am not a tsunami expert, I found that the literature review about tsunami detection from satellite altimeters is well presented in the introduction. The presentation of the SWOT altimeter and its capabilities is also concise and very clear.
- The authors thank the referee for these encouraging remarks.
After checking the comments of the first reviewer, I find myself in agreement with their comments and I don’t have further technical corrections to provide. I therefore propose a minor revision, and two additional comment/suggestions that could further improve the manuscript:
- To better quantify the time lag between the simulated main peak and the SWOT peak, I suggest providing a comparison spatially restricted between −19° and−18° latitude and using a simulation output closer in time to the SWOT overpass above this restricted area. This would make the comparison more precise.
- The objective of the paper is to highlight SWOT’s ability to record tsunami. The numerical simulation was done only to confirm that SWOT recorded the tsunami wavefield, and we were expecting large discrepancies in amplitude and phase, which is finally not so bad. In fact, the tsunami generation model we are showing here is not the best and due to the complexity of the seismic source, being strongly non double-coupled, more work needs to be done on the source itself before going further in comparing numerical results and SWOT records (or example at DART locations). So, we could increase the sampling rate of the simulated snapshots to be able to choose a closer snapshot in time, but due to the source quality, it may not provide better comparison.
- I also propose that the authors change the title of section 4 to something more generic like “Comparisons and discussion”, and change the title of section 4.1to something like “Comparison between observations and model output” to better reflect the content and structure of the section.
- We agree with the referee suggestions to change the section titles and will update them accordingly.
Citation: https://doi.org/10.5194/egusphere-2025-3926-AC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 2,069 | 129 | 17 | 2,215 | 36 | 33 |
- HTML: 2,069
- PDF: 129
- XML: 17
- Total: 2,215
- BibTeX: 36
- EndNote: 33
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The work entitled “First 2D record of a tsunami by SWOT satellite: observation data and preliminary numerical simulation of the 19 May 2023 tsunami near the Loyalty Islands” focuses on satellite radar observation of a tsunami event on 19 May 2023. It is the first dynamic observation of a two-dimensional tsunami wave field. The satellite’s orbit allowed the observation of the wavefront propagating in both directions from the epicenter, about 1 h after the earthquake. I wonder why the authors call it 2D observation, when in fact it is a 3D observation because it includes the sea surface elevations (Sea Level Anomaly) within a two-dimensional domain. These observations are compared with numerical simulations performed with COMCOT model, showing generally good agreement with observations, although some differences reveal weak points of the simulations and indicate that there is room for future model improvements. Simulations show a small delay in the main peak of the tsunami wavefront and underestimated amplitudes, especially for the secondary waves in their propagation to the north. The results presented are relevant and highlight the usefulness of SWOT data to improve tsunami predictions. Some paragraphs and figures show inconsistencies that must be corrected (see attached file). The manuscript presents new data, and I recommend publication once the minor changes have been addressed. I hope my comments are useful to the authors and help improve the quality of the paper.