the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 28 Oct 2025
Locating the Coseismic Ionospheric Disturbance from the 2010 Mw7.8 Mentawai, Indonesia, Tsunami-Earthquake
Abstract. Forecasting tsunamis using seismological and geodetic observations remains challenging. However, tsunamis generate acoustic waves that travel into the atmosphere and affect the ionosphere, creating disturbances in total electron content (TEC). These disturbances, known as Coseismic Ionospheric Disturbances (CIDs), can be observed using GNSS data. Identifying the source location of the CID could provide valuable information about whether a tsunami has been triggered. In this study, we attempt to determine the source location of the CID following the 2010 Mw7.8 Mentawai tsunami-earthquake, using onset time measurements of the CID on vTEC waveforms. Our approach combines acoustic wave propagation within a realistic atmosphere alongside a Bayesian inversion framework that considers uncertainties in the observations and the forward problem itself. Our approach search for the effective height of the ionosphere rather than fixing its altitude, as is commonly done. Our results suggest an effective height of the ionospheric of around 250 km, which is below the F-layer peak. Furthermore, we show that the inferred source location of the CID does not coincide with the zone of maximum seafloor uplift or the area where the tsunami initiates. An alternative model based on the assumption of a homogeneous atmosphere locates the source of the CID within the area where the tsunami initiates. While this might imply that the realistic atmosphere model does not allow acoustic waves to propagate at the correct speed, our findings demonstrate that both models have the same apparent acoustic wave speed. Therefore, we argue that the approach using a realistic atmosphere provides an accurate location of the CID source because it takes the complexity of propagation paths into account. We offer several explanations as to why the source location the CID does not match the area where the tsunami initiates, such as the effect of the spatial extent of the actual source, more complex tsunami initiation dynamics due to bathymetry effects, or potential errors in the earthquake source models that could shift the area where the tsunami initiates. This work demonstrates the potential of using ionospheric observations to complement existing tsunami monitoring strategies.
- Preprint
(5315 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed (peer review stopped)
- RC1: 'Comment on egusphere-2025-5088', Kosuke Heki, 30 Nov 2025
-
RC2: 'Comment on egusphere-2025-5088', Anonymous Referee #2, 28 Feb 2026
Report on ‘Locating the Coseismic Ionospheric Disturbance from the 2010 Mw7.8 Mentawai, Indonesia, Tsunami-Earthquake’ by Twardzik, Rolland, Munaibari, and Mikesell
The manuscript investigates the source location of the coseismic ionospheric disturbance (CID) generated by the 2010 Mw7.8 Mentawai tsunami earthquake using onset times measured on GNSS-derived vTEC waveforms. The authors implement a Bayesian inversion framework that incorporates both data and prediction uncertainties and compute acoustic travel times in a realistic atmosphere based on NRLMSISE-00 and horizontal wind models. Unlike many previous studies, the effective ionospheric height is treated as an unknown parameter rather than fixed a priori. The results suggest an effective ionospheric height of ~260 km and a CID source location that does not coincide with the maximum uplift or the tsunami initiation area when a realistic atmospheric model is used. The study is methodologically interesting and potentially valuable for tsunami early warning applications. However, several aspects require clarification and additional robustness tests before the conclusions can be considered fully supported.
General Comments
The manuscript does not report the geomagnetic conditions at the time of the event. TEC perturbations in the 2–10 mHz band may also be influenced by background TIDs or geomagnetic activity. Please report relevant geomagnetic indices (e.g., Kp, Dst, AE) for the event day and for at least 3 days before the event and discuss whether the ionosphere was geomagnetically quiet. This would strengthen the interpretation that the detected perturbations are indeed coseismic in origin.
In addition, I suggest adding a control day within the same month (October 2010) and at similar local time, under geomagnetically quiet conditions and without significant seismic events capable of producing ionospheric disturbances. Comparing representative 2–10 mHz filtered vTEC time series (and simple RMS/PSD metrics in the same band) between the Mentawai day and the control day would help demonstrate that the detected oscillations are not typical background variability.
Given the low-latitude setting of the Mentawai event, where electrodynamic processes (E×B drifts, evening instabilities, equatorial plasma irregularities) can significantly affect TEC variability, please discuss how such effects might influence onset-time identification and travel-time interpretation. In particular, clarify whether geomagnetic/electrodynamic conditions were quiet and whether background equatorial variability could bias source localization.
NRLMSISE-00 provides a climatological neutral atmosphere constrained by solar and geomagnetic indices but does not represent transient mesoscale perturbations (e.g., gravity waves or background TIDs) that may have been present at the time of the event. Please discuss how unmodelled atmospheric variability could influence acoustic travel-time predictions and whether this uncertainty is fully captured by the prediction covariance matrix Cp.
Furthermore, NRLMSISE-00 is known to perform less reliably during geomagnetically disturbed conditions, as it represents a climatological neutral atmosphere and does not capture storm-time dynamics. Please report the geomagnetic conditions at the time of the event and discuss whether potential storm-time deviations could affect acoustic travel-time predictions beyond the uncertainties included in Cp.
If available, satellite-based thermospheric composition data (e.g., TIMED/GUVI) near the event time could help confirm that no significant storm-time composition anomalies were present. At minimum, the manuscript should document geomagnetic and solar conditions to justify the use of a climatological background atmosphere.
The study is based on a single event. While appropriate for a methodological proof-of-concept, the manuscript would benefit from additional robustness tests, such as jackknife or bootstrap analysis over onset picks and sensitivity tests to filtering choices, ICCS thresholds, and prior bounds. These analyses would help demonstrate that the inferred source location and effective ionospheric height are stable with respect to reasonable perturbations of the input data and methodological choices.
Ideally, inclusion of a second case study or control event would further strengthen the generality of the conclusions.
The manuscript presents an interesting and technically advanced framework for CID source localization, and the Bayesian treatment of uncertainties is commendable. However, given the small amplitude of the TEC perturbations, the low-latitude setting, and the reliance on a climatological atmospheric model, additional contextual analysis and robustness tests are required. I therefore recommend major revision, with the expectation that the authors can address the points above to significantly strengthen the manuscript.
Citation: https://doi.org/10.5194/egusphere-2025-5088-RC2
Status: closed (peer review stopped)
- RC1: 'Comment on egusphere-2025-5088', Kosuke Heki, 30 Nov 2025
-
RC2: 'Comment on egusphere-2025-5088', Anonymous Referee #2, 28 Feb 2026
Report on ‘Locating the Coseismic Ionospheric Disturbance from the 2010 Mw7.8 Mentawai, Indonesia, Tsunami-Earthquake’ by Twardzik, Rolland, Munaibari, and Mikesell
The manuscript investigates the source location of the coseismic ionospheric disturbance (CID) generated by the 2010 Mw7.8 Mentawai tsunami earthquake using onset times measured on GNSS-derived vTEC waveforms. The authors implement a Bayesian inversion framework that incorporates both data and prediction uncertainties and compute acoustic travel times in a realistic atmosphere based on NRLMSISE-00 and horizontal wind models. Unlike many previous studies, the effective ionospheric height is treated as an unknown parameter rather than fixed a priori. The results suggest an effective ionospheric height of ~260 km and a CID source location that does not coincide with the maximum uplift or the tsunami initiation area when a realistic atmospheric model is used. The study is methodologically interesting and potentially valuable for tsunami early warning applications. However, several aspects require clarification and additional robustness tests before the conclusions can be considered fully supported.
General Comments
The manuscript does not report the geomagnetic conditions at the time of the event. TEC perturbations in the 2–10 mHz band may also be influenced by background TIDs or geomagnetic activity. Please report relevant geomagnetic indices (e.g., Kp, Dst, AE) for the event day and for at least 3 days before the event and discuss whether the ionosphere was geomagnetically quiet. This would strengthen the interpretation that the detected perturbations are indeed coseismic in origin.
In addition, I suggest adding a control day within the same month (October 2010) and at similar local time, under geomagnetically quiet conditions and without significant seismic events capable of producing ionospheric disturbances. Comparing representative 2–10 mHz filtered vTEC time series (and simple RMS/PSD metrics in the same band) between the Mentawai day and the control day would help demonstrate that the detected oscillations are not typical background variability.
Given the low-latitude setting of the Mentawai event, where electrodynamic processes (E×B drifts, evening instabilities, equatorial plasma irregularities) can significantly affect TEC variability, please discuss how such effects might influence onset-time identification and travel-time interpretation. In particular, clarify whether geomagnetic/electrodynamic conditions were quiet and whether background equatorial variability could bias source localization.
NRLMSISE-00 provides a climatological neutral atmosphere constrained by solar and geomagnetic indices but does not represent transient mesoscale perturbations (e.g., gravity waves or background TIDs) that may have been present at the time of the event. Please discuss how unmodelled atmospheric variability could influence acoustic travel-time predictions and whether this uncertainty is fully captured by the prediction covariance matrix Cp.
Furthermore, NRLMSISE-00 is known to perform less reliably during geomagnetically disturbed conditions, as it represents a climatological neutral atmosphere and does not capture storm-time dynamics. Please report the geomagnetic conditions at the time of the event and discuss whether potential storm-time deviations could affect acoustic travel-time predictions beyond the uncertainties included in Cp.
If available, satellite-based thermospheric composition data (e.g., TIMED/GUVI) near the event time could help confirm that no significant storm-time composition anomalies were present. At minimum, the manuscript should document geomagnetic and solar conditions to justify the use of a climatological background atmosphere.
The study is based on a single event. While appropriate for a methodological proof-of-concept, the manuscript would benefit from additional robustness tests, such as jackknife or bootstrap analysis over onset picks and sensitivity tests to filtering choices, ICCS thresholds, and prior bounds. These analyses would help demonstrate that the inferred source location and effective ionospheric height are stable with respect to reasonable perturbations of the input data and methodological choices.
Ideally, inclusion of a second case study or control event would further strengthen the generality of the conclusions.
The manuscript presents an interesting and technically advanced framework for CID source localization, and the Bayesian treatment of uncertainties is commendable. However, given the small amplitude of the TEC perturbations, the low-latitude setting, and the reliance on a climatological atmospheric model, additional contextual analysis and robustness tests are required. I therefore recommend major revision, with the expectation that the authors can address the points above to significantly strengthen the manuscript.
Citation: https://doi.org/10.5194/egusphere-2025-5088-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 967 | 776 | 116 | 1,859 | 96 | 93 |
- HTML: 967
- PDF: 776
- XML: 116
- Total: 1,859
- BibTeX: 96
- EndNote: 93
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Please see the attached WORD document. My identity could be disclosed.