SUMMARY
We report a mode of ionospheric resonance induced by the 2022 January 15, Hunga Tonga–Hunga–Ha'apai volcanic eruption, captured globally through ground-based Global Navigation Satellite Systems networks. In the near-field region, we observe a strong ionospheric disturbance in the total electron content, including contributions from acoustic and gravity waves. In contrast, propagation of gravity waves, seismic airwaves diffracted by topographic barriers and/or atmospheric Lamb waves can be documented for far-field ionospheric disturbances. Specifically, the atmospheric wave propagation towards the Hawaii Islands involved three distinct modes of ionospheric perturbation through Rayleigh, gravity and atmospheric Lamb waves. Further, we also noted that the peak-to-peak amplitude of the disturbance reached ∼ 8.7 per cent of the background electrons, which is significantly higher than the recent volcanic explosions in the Japanese Islands and the 2020 chemical explosion in Beirut. The ionospheric total electron content perturbations can provide important information about the volcanic source, hence they could be useful as a measure of eruption intensity and for real-time hazard monitoring.
1 INTRODUCTION
The Hunga Tonga––Hunga-Ha'apai volcano (HTHH) is located in the Kingdom of Tonga (home of ∼105 000 people) in the Southwest Pacific along the Tonga–Kermadec intra-oceanic arc. It is a large underwater edifice that has produced a series of eruptions for at least ∼900 yr (Brenna et al. 2022). The latest spectacular eruption that occurred on 2022 January 15 (∼VEI6, Dalal et al. 2022; Sharma & Scarr 2022) can be considered a ‘once-in-a-century’ event (Fig. 1a). Two small, ∼2-km-long, uninhabited islands (i.e. Hunga Tonga and Hunga Ha'apai), rising to 114 m above sea level, currently represent the summit of the dominantly submarine volcano. These two islands were connected by a new island (central cone) produced during the late-2014 to early-2015 eruption sequence (Brenna et al. 2022) and again separated after the recent explosive eruption (Fig. 1b). The massive plume broke the ocean surface at about 4:16:20 UT, on 2022 January 15 (https://earthquake.usgs.gov) and within the next 2 hr, it had expanded into a mushroom-shaped, ∼3 mile-wide plume rising into the air to ∼34 miles, as witnessed by several satellites (Perry 2022) (Fig. 1a). Numerous barometric arrays and ocean bottom pressure sensors around the globe captured Lamb and shock waves that rippled through the earth's atmosphere (https://nctr.pmel.noaa.gov/tonga20220115/) (Figs S1 and S2, Supporting Information, Dalal et al. 2022). This powerful explosion was heard in New Zealand, Alaska, Yuko and Canada (Carroll 2022). Lamb waves propagate horizontally with a velocity of ∼310–340 m s−1, are concentrated in the troposphere, and can get excited by large eruptions, explosions, impacts and earthquakes (Nishida et al. 2014).
The 2022 HTHH volcanic eruption and ionospheric disturbance. (a) Real-time satellite imagery of 2022 January 15 (04:14:45 UTC) HTHH Volcanic eruption, acquired by Himawari-8 (https://himawari8.nict.go.jp.) (b) The location of the volcanic eruption and GNSS sites are marked by the red star and yellow circles, respectively. The white circles represent the radial distance of 1000, 2000 and 2500 km from the eruption. The bottom panel represents Google Earth imagery before (2022 January 13) and after (2022 February 2) the eruption.
The Pacific Tsunami Warning Centre (PTWC) reported tsunami waves of 1.15 m at Nuku'alofa (Tonga), but local wave heights might have reached ∼15 m, and three people perished in Tonga (https://www.spc.int/). Affected shores along the Pacific coastlines of the United States, including Alaska and Hawaii, Chile, Japan and New Zealand declared a state of emergency after this violent eruption, but the eventual far-field tsunami turned out to be moderate, with two casualties reported in Peru (https://www.spc.int/). Surprisingly, early and higher-than-expected tsunamis also occurred in the distant Caribbean and Mediterranean, suggesting that the tsunami was in part due to atmosphere-ocean interactions (Sommerville et al. 2022).
The 2022 eruption had an explosive force more than 500 times that of the atom bomb dropped on Hiroshima, Japan, in 1945 (Mallapaty 2022). This event offers us a rare and unique opportunity to study the phenomena of ‘ionospheric resonance’, that a signature of perturbation in the ionosphere through the short- and long-path propagation of several types of atmospheric waves initiated by the blast (e.g. acoustic, gravity and Lamb waves, Fig. 1c). The ionospheric disturbance caused by various waves can be distinguished by their propagation speed and frequency (Jin et al. 2019). These include seismic air waves (∼300 m s−1), acoustic waves (300–1500 m s−1), atmospheric gravity waves (100–1000 m s−1), seismic Rayleigh waves (2000–3000 m s−1) and Lamb waves (310–340 m s−1, Nishida et al. 2014; Astafyeva et al. 2019; Jin et al. 2019), etc.
In the young field of ‘ionospheric seismology,’ Global Navigation Satellite Systems (GNSS), such as the American GPS, Russian Glonass, European Galileo, Japanese QZSS and Chinese BeiDou, have become the principal instruments and enable us to observe variations in ionospheric total electron content (TEC) from the phase differences of dual-frequency microwave signals from the satellites (Komjathy et al. 2016; Astafyeva 2019). The atmospheric waves generated from various natural hazards and explosions cause ionosphere disturbances (Hines 1972; Komjathy et al. 2012; Dautermann et al. 2009). These include volcanic explosions (Heki 2006), anthropogenic surface explosions (Kundu et al. 2021) and underground nuclear tests (Park et al. 2014). In this letter, we report on diverse modes of ionospheric oscillations triggered by atmospheric waves caused by the 2022 HTHH eruption.
2 DATA AND METHODS
2.1 Data sets
To estimate TEC variation associated with the 2022 January 15 HTHH volcanic eruption, we have collected RINEX GNSS data from New Zealand, Hawaii, Fiji, Samoa and other islands situated near the volcanic eruption. These GNSS stations are operated and maintained by Geological and Nuclear Sciences Limited (GNS), New Zealand, Pacific GPS Facility, University of Hawaii, and Geoscience Australia, and the data are archived at the Scripps Orbit and Permanent Array Centre (SOPAC, http://sopac-old.ucsd.edu/dataBrowser.shtml). We use the standard sampling rate of the 30 s in the daily GNSS file to quantify the TEC changes in the ionosphere during the HTHH volcanic eruption. Further, we used the Planetary Index (Kp) and Disturbance Storm Time Index (DST) to evaluate geomagnetic activity during the studied day, which can be obtained from NASA/GSFC's Space Physics Data Facility's OMNIWeb service (https://cdaweb.gsfc.nasa.gov/). The barometric pressure and water level data used in Fig. 4(a), and Figs S2 and S6 (Supporting Information) are archived at the National Oceanic and Atmospheric Administration (NOAA, https://tidesandcurrents.noaa.gov/map/index.html).
2.2 GNSS-TEC processing strategy
3 RESULTS
The phase difference between the two microwave frequencies, L1 and L2 can be subsequently converted to TEC (1 TECU = 1016 electrons m−2) along the ray path, and we explore the time-series of TEC from global GNSS stations induced by ionospheric disturbances in the near- and far-field regions of the 2022 HTHH volcanic eruption. From Kp and DST index analysis, we observe that 2022 January 15 started and ended with high Kp index values (> 4) and low DST index values (< −50 nT). However, low Kp index (< 4) and high DST index (> −50 nT) were observed at the time of the HTHH eruption (Fig. S3, Supporting Information). It has been suggested that, when the Kp index < 4 and DST index > −50 nT, the geomagnetic conditions are considered as quiet (Chakraborty & Morley 2020; Luo et al. 2020; Senapati et al. 2022;Pedatella et al. 2010). Therefore, we considered that the geomagnetic activity is quiet during the time of HTHH eruption (Fig. S3, Supporting Information).
Numerical simulation and observed TEC change in the near field. (a) Maps of TEC change at three time (i.e. 11.0, 12.5 and 15.0 min after eruption) at 300 km altitude due to the upward propagating acoustic pulse (with a period 80 s) in the atmosphere, considering the geomagnetic field orientation near the HTHH volcano (i.e. declination: 13.70° and inclination: −41.85°). Here, we assumed the sound velocity structure of the US Standard Atmosphere 1976 and the Chapman distribution of the electron density as a function of altitude. (b) Relative electron density (blue) and velocity of acoustic wave (red line, U.S. Standard Atmosphere 1976) plotted as a function of height. (c) Ray tracing of the acoustic wave for zenith angles 0°–22° performed assuming the velocity profile shown in (b). The geomagnetic field (B) represented by black arrows. (d) The trajectory of subionospheric points (SIPs) of the GNSS sites (marked by cyan squares) for the period of 3–10 UT. The yellow star indicates the location of the volcanic eruption. Black dots along the SIPs indicate the time of the eruption. Grey dots along the circle indicate the different satellite azimuths at 250 km distance from the eruption. (e) The synthetic TEC changes observed at different satellite azimuths (0°–360° at 30° interval), assuming satellites with elevation angles 60°. (f) The observed (grey) and synthetic (red colour) STEC change in satellite numbers 10, 18, 23 and 24 for different GNSS stations (cyan square in Fig. 2d) situated near the eruption. QZSS and BeiDou satellites tracked by these GNSS stations are shown in Fig. S4 (Supporting Information). The vertical line represents the time of the volcanic explosion (04:14:45 UT).
To investigate the ionospheric resonance induced by the atmospheric waves, we examined TEC time-series data from several continuous GNSS stations in the near-field (< 1000 km) and far-field regions (1000–6000 km) from the epicentre of the Hunga–Tonga eruption (Fig. 1b). Irrespective of the very limited number of GNSS stations close to the location of the eruption, we are able to observe a combined mode of ionospheric oscillation (∼0.5–3.5 TECU) caused by the acoustic and gravity waves, immediately after the eruption in the near-field region. The strong ionospheric disturbance signals are clearly captured in the slant ionospheric TEC (STEC) time-series at ground-based GNSS stations located ∼700 km north of Tonga (e.g. FTNA, PTVL, LAUT and SAMO) due to the interaction with geomagnetic fields (Fig. 2 and Fig. S4, Supporting Information). However, the GNSS stations situated in the south of Tonga (e.g. CKIS) do not show such strong disturbance in the STEC, as the southward propagating wavefront gets parallel with the earth's magnetic field and attenuates southward propagation (Fig. 2f). We have also noted that the ionospheric disturbances continued up to 5–6 hr after the eruption, which might be due to the post-volcanic evening equatorial plasma bubbles (EPB, Aa et al. 2022). Moreover, we have reproduced the arrival times, waveforms and relative intensities of the observed ionospheric disturbances by following Heki's (2006) approach (for detail, see Supporting Information). From this analysis, we conclude that the first N-shaped disturbances in the observed TEC time-series have a good fit with synthesized curves, suggesting that the model has a good agreement with the observed data (red curve in Fig. 2f). However, using this model approach, it is only possible to model the ionospheric disturbances induced from the acoustic waves, 12–15 min after the eruption, which is one of the limitations of this model. Therefore, in the present study, we are unable to model the entire TEC time-series, as the ionospheric disturbances appear to be caused by a combination of acoustic waves, gravity waves, and post-volcanic evening EPB. Further, these GNSS stations also tracked the QZSS and BeiDou satellites, and we found similar signals of the ionospheric oscillations to those found in data from GPS satellites (Fig. S4, Supporting Information). These disturbances in the ionosphere may be caused by eruption-induced perturbations at frequencies of acoustic and gravity waves. However, it is difficult to isolate each component (acoustic or gravity wave) from the superimposed STEC signature.
We observed far-field ionospheric disturbances (∼0.4–0.8 TECU) induced by the southward atmospheric wave propagation in New Zealand (Fig. 3). Fig. 3(b) represents a high-bandpass filtered (3–10 min) STEC time-series obtained from GPS satellite 10 at ten ground-based GNSS stations ∼2000–3000 km south of the eruption, along with the trajectory of SIPs calculated assuming the thin ionosphere at altitude 300 km (Fig. 3a). In Fig. 3(c), we present a diagram indicating the distance to the SIPs as a function of time with colours showing the STEC anomaly. We observe two distinct modes of ionospheric disturbances in the TEC time-series in New Zealand (Figs 3b and c). The first mode of disturbance, observed at 5–6 UT with the anomaly propagating at an apparent speed of ∼540 m s−1, is related to the atmospheric gravity waves. The second mode of disturbance observed at 6–7 UT propagated at an apparent speed of ∼300 m s−1. We suggest that the TEC perturbations induced by the acoustic waves should not be observable in this region because of the geomagnetic directivity (Fig. 2). However, TEC perturbations induced by gravity waves, diffracted seismic airwaves, or Lamb waves can all be considered as suitable candidates for the second mode of observed signals. One possibility is that gravity waves have travelled obliquely upward in the southward direction at a velocity of ∼100–1000 m s−1. It is also possible that the seismic airwaves propagated horizontally with a velocity of ∼300 m s−1 over the ocean and were diffracted into the upper atmosphere due to the topographic barrier in New Zealand (Lacanna et al. 2014). Finally, the upward propagation of the atmospheric Lamb waves, which travel at the speed of sound, also oscillates in the ionosphere (Fig. 3d). Although the Lamb wave-induced oscillations are confined to the troposphere, the Lamb wave energy can leak into the thermosphere through the atmospheric resonance at acoustic and gravity wave frequencies and carrying substantial wave amplitudes at high altitudes (Ogawa et al. 1982; Heki 2022).
Long-path ionospheric disturbances. (a) The trajectory of subionospheric points (SIPs) of GPS satellite 10 at different GNSS stations situated in New Zealand. The green squares indicate the locations of GNSS sites. (b) The STEC change observed for satellite 10 at different GNSS sites (shown in part a). The distance between the GNSS station and HTHH volcanic eruption represent in the right-hand side of each sip plot. The vertical line represents the explosion time (04:16:20 UT). (c) The observed STEC anomalies as a function of time (horizontal axis) and distance from the eruption (vertical axis), are calculated along the surface assuming an eruption at 175.39° W, 20.546° S. Note two modes of ionosphere disturbance observed in New Zealand. The first mode of STEC anomalies propagate at a velocity of ∼540 m s−1 and reflect atmospheric gravity waves (AGW), whereas the second mode of STEC anomalies propagate at a velocity of ∼300 m s−1 and may be triggered by either the AGW, diffracted seismic or Lamb waves. (d) Schematic representation of ionospheric disturbance caused by the different waves (i.e. acoustic, atmospheric gravity and Lamb waves) generated from a volcanic eruption.
We suggest that the upward propagation of the Lamb waves and subsequent ionospheric disturbances can be documented more clearly at faraway GNSS stations located in the mid-Pacific in the Hawaii Islands (∼5000 km) and to the northwest in the Japanese Islands (∼8000 km), as other waves would either have been attenuated at these distances or are expected to arrive at different times. Although the radiation characteristics of Lamb waves have already been described theoretically and also in laboratory-based experiments using waveguide sensors (Park et al. 2020), these waves have never been reported as ionosphere disturbances after a volcanic eruption or surface explosion captured by GNSS observations (see also Zhang et al. 2022).
We report far-field ionospheric disturbances (∼0.2–0.5 TECU) induced by the long-path atmospheric wave propagation in the Hawaii Islands that reveal three distinct modes of oscillation, capturing a signature of Rayleigh, atmospheric gravity and the Lamb waves (Fig. 4 and Fig. S5, Supporting Information). We consider filtered STEC time-series obtained by GPS satellites 32 and 31 at ten representatives ground-based GNSS stations from Hawaii Island (Fig. 4a and Fig. S3, Supporting Information). The L1 phase of Lamb waves travelled from SE to NE along the shortest path with a speed of ∼0.3 km s−1 and passed over the Hawaii region at 8–9 hr UTC. The second passage of Lamb wave (L2) propagated along the long path occurred ∼12 UTC on 2022 January 16. These waves kept rotating around the globe several times and were subsequently captured in the barometric and GNSS stations (Figs S1 and S2, Supporting Information, and Figs 4 and 5). We observe that the time of TEC change in the ionosphere is slightly earlier than the barometric pressure change induced by the L1, L4 and L5 phases of the Lamb wave. This may be due to the satellite observing geometry, which led to the synchronous arrival of the Lamb waves and the TEC perturbations (Fig. 5). The TEC change caused by the L2 and L4 phases of the Lamb wave is later than the arrival of the Lamb wave in the barometric pressure station (Figs 4 and 5). This may be possible due to AWG excited by the Lamb waves takes 20–30 min to reach and disturb the TEC in the ionosphere. It has also been observed that, unlike the barometric pressure change, waveforms and amplitudes of ionospheric disturbances exhibit large diversity, which reflects their turbulent nature (Figs 4 and 5). We also note a much earlier arrival of atmospheric gravity waves in the STEC time-series at GPS satellite 32 about 1.3 hr after the eruption, followed by a prominent signature of the Lamb waves at GPS satellite 31, about 3 hr later, and the arrival of tsunamic waves in Hawaii occurred much later than the atmospheric gravity and Lamb waves (Figs S6 and S7, Supporting Information). Although we expect spatio-temporal variation in Lamb wave propagation, due to the coupling of regional wind-flow or shallow-water-wave-induced tractions, the idealized Lamb waves assume the surface to be traction-free (Lowe 2001). Irrespective of that complexity, GNSS stations at Hawaii Island have prominently captured this new mode of ionospheric disturbance (Figs 4b and 5).
Long-path atmospheric Lamb and gravity wave (AGW) induced ionospheric disturbances. (a) The STEC change observed in satellites 32 (3–6 hr) and 31(7–10 hr) of different GNSS sites located in the Hawaii Islands at ∼5000 km from the eruption (green squares in the top inset). The red curves represent the barometric pressure changes observed at the Hilo Bay and Kawaihae stations (red dots in the top inset). Note that the ionospheric disturbances in satellite 31 is slightly earlier than the L1 phase of Lamb waves in the barometric pressure time-series. Different phases of Lamb waves (i.e. L1, L2, L3, L4 and L5) and corresponding TEC changes are shown in Fig. 5. The first tsunami arrivals observed at tide gauges at Hilo Bay and Kawaihae occurred hours after the AGW (Fig. S6, Supporting Information). (b) Schematic representation of ionospheric disturbance induced by the atmospheric Lamb and gravity waves.
Barometric pressure change by atmospheric Lamb waves and corresponding TEC change in the ionosphere. Top panel represents atmospheric pressure change in Hilo bay and Kawaihape station at Hawaii Island. The boxes indicate the different phases of atmospheric Lamb waves (i.e. L1, L2, L3, L4 and L5). Bottom panel represents TEC change in the ionosphere caused by the different phases of atmospheric Lamb waves, observed in the different GNSS stations shown in Fig. 4(a). Numbers within the bracket are the satellites number of the corresponding GNSS station.
4 DISCUSSION
The electron density disturbances in the F region of the ionosphere induced by the volcanic activity could serve as a measure for the intensity of volcanic explosions (Manta et al. 2020; Cahyadi et al. 2021; Kundu et al. 2021). Therefore, we have compared the amplitude of the STEC disturbance caused by the HTHH volcanic eruption with the recent volcanic explosions in Japan (i.e. from 2004 to 2015) and the 2020 chemical explosion in Beirut. Here, we consider the amplitude of the STEC disturbance that is normalized by the background VTEC value. The background VTEC values at the time and location of eruption are obtained from the GIM, (Fig. S8, Supporting Information). From the GIM, the background vertical TEC during the 2022 HTHH volcanic eruption is ∼36.4 TECU (Fig. S8, Supporting Information), and the amplitude of STEC (calculated from FTNA, PTVL, LAUT and SAMO stations) is ∼ 3.2 ± 0.04 (Fig. S4, Supporting Information), which makes the relative amplitude of TEC (i.e. the ratio of observed STEC and the background VTEC) ∼8.7 per cent (Fig. S9, Supporting Information). This is significantly higher than the ionospheric changes caused by recent volcanic explosions in Japan and the chemical explosion in Beirut (Cahyadi et al. 2021; Kundu et al. 2021; Fig. S9, Supporting Information).
Moreover, the 2022 January 15 Hunga Tonga volcanic eruption involved five explosions between 4 and 5 UT (Astafyeva et al. 2022). The main explosion occurred at 4:16:20 UT, which caused larger ionospheric disturbances as compared to the other four explosions. In this study, we have analysed ionospheric TEC disturbances caused by the main explosion. if we consider the later explosions, it will not affect our results, because, in the near-field region, we have shown the ionospheric TEC time-series from 3 to 10 UT, in which all the multiple peaks can be observed. However, it is not our aim to identify these individual peaks in the present draft.
The 2022 HTHH volcanic eruption had such explosive power that it could be detected high up in the sky in the ionosphere, even several thousand kilometres from the source. The gravity and Lamb waves are observed well ahead of the tsunami waves (Figs S6 and S7, Supporting Information) and can potentially play a role in future early tsunami warning systems. Moreover, Lamb waves can only originate from extremely large events (Donn & Shaw 1967; Mikumo 1968; Nishida et al. 2014), and this HTHH volcanic eruption is probably the first event, in which the atmospheric Lamb wave-induced ionospheric disturbances are captured by the global GNSS network (see Zhang et al. 2022; Themens et al. 2022; Astafyeva et al. 2022; Ghent & Crowell 2022; Heki 2022; Matoza et al. 2022). The observed changes in the ionosphere contain information about both the source processes and propagation of atmospheric waves associated with major natural hazards on the earth's surface, and have potential for hazard monitoring, early warning and mitigation in the wider field of ‘ionospheric seismology’.
5 CONCLUSIONS
From the above analysis, we conclude the following salient points:
The Lamb waves induced by the 2022 HTHH volcanic eruption travelled around the Earth several times and disturbed the electron density of the ionosphere.
We observe a combined mode of ionospheric oscillations (∼0.5–3.5 TECU) caused by acoustic waves and gravity waves in the near-field region (< 1000 km from eruption). This ionospheric anomaly propagates northward from the eruption.
Two modes of ionospheric disturbances are observed in the New Zealand region. The first mode of disturbance propagates at an apparent speed of ∼540 m s−1 and is related to the atmospheric gravity waves. The second mode of disturbances advances at an apparent speed of ∼300 m s−1 and may be related to diffracted seismic airwaves, gravity waves, or Lamb waves.
The peak-to-peak ionospheric disturbance is ∼0.2–0.5 TECU in the Hawaii Islands, involving three distinct modes of ionospheric perturbation: Rayleigh, gravity and atmospheric Lamb waves.
The atmospheric gravity waves in the STEC time-series and Lamb waves in the barometric pressure arrive earlier than the tsunami waves in the Hawaii Islands and may play a potential role for tsunami warning systems.
The peak-to-peak amplitude of the disturbance reached ∼ 8.7 per cent of the background electrons, which is significantly higher than the recently recorded volcanic explosions in the Japanese Islands and the chemical explosion in Beirut.
SUPPORTING INFORMATION
Figure S1. Location of the barometric station (yellow circles). The red star represents the location of the Hunga–Tonga volcanic explosion (Dalal et al. 2022).
Figure S2. Time-series of the barometric pressure at stations shown in Fig. S1. The different arrival phases of the Lamb wave (i.e. L1, L2, L3, L4, L5 and L6) are represented by the red and blue lines (Dalal et al. 2022).
Figure S3. Space weather indices around the HTHH eruption. Here we show two geomagnetic activity indices (Kp and DST indices), from 2022 January 10 to 20. Note during eruption time on 2022 January 15, the Kp < 5 and DST > −50 nT, which indicates that it was a geomagnetic quiet day.
Figure S4. The STEC change observed at different GNSS sites. The vertical dashed line represents the explosion time (04:14:45 UT). Note that ionospheric disturbances are seen shortly after the eruption. These stations also tracked QZSS and BeiDou satellites, showing similar disturbance signals in the ionosphere.
Figure S5. The STEC change observed in the signal from satellites 31 and 32 at different GNSS sites on Hawaii Island. Note that the left-hand panel represents the ionospheric disturbance induced by the AGW and Rayleigh wave. The middle and right-hand panels represent the ionospheric disturbance caused by leaky Lamb waves.
Figure S6. Time-series of water level variation, STEC, and pressure change in Hawaii Island. The dashed line represents the time of Hunga–Tonga volcanic eruption. The inset image shows the location of the meteorological/tide gauge stations. Note that arrival time of tsunami waves at Hawaii is much later than the Lamb wave and AGW.
Figure S7. Variation of water level in Hawaii due to 2022 January 15 Honga–Tonga volcanic eruption. The red and black curves are modelled and observed wave amplitudes respectively (taken from NOAA, https://nctr.pmel.noaa.gov/). Some tsunami arrivals appear well before that predicted by the NOAA model. That may reflect the component of the tsunami induced by the atmospheric waves. This is not so obvious at Hilo.
Figure S8. VTEC distribution from GIM. The global ionospheric data on 2022 January 15 is obtained from the Goddard Space Flight Center NASA (https://cdaweb.gsfc.nasa.gov/index.html/).
Figure S9. Disturbance amplitude of STEC and its comparison with volcanic explosions in Japan and Beirut chemical explosion. Comparison of the maximum observed STEC amplitudes (blue squares) to background VTEC (green squares) for Hunga–Tonga volcanic explosion with five volcanic eruptions in Japan and the Beirut chemical explosion. Note that the relative amplitude of the Hunga–Tonga volcanic explosion near five times higher than the Beirut and five volcanic explosions in Japan (Modified after Kundu et al. 2021).
Please note: Oxford University Press is not responsible for the con- tent or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be di- rected to the corresponding author for the paper.
ACKNOWLEDGEMENTS
We thank Kosuke Heki for discussions of the results, which significantly improved the quality of this work. BS and SR have been supported by the NITR research fellowship. RB acknowledges support by the Miller Institute for Basic Research in Science at UC Berkeley. We thank the editor, Prof Diego Melgar and Prof Dylan Mikesell and two anonymous reviewers for their constructive comments, which improved the quality of the manuscript.
AUTHOR CONTRIBUTIONS
BK and RB provided research idea. BK, SR and BS performed GNSS data analysis and estimated the TEC changes. BK wrote the original manuscript and BS performed numerical simulations for the observed TEC changes. All authors took part in finalizing the manuscript.
DATA AVAILABILITY
RINEX GNSS data are available on the SOPAC (http://sopac-old.ucsd.edu/dataBrowser.shtml). The barometric pressure data are archived at the NOAA (https://tidesandcurrents.noaa.gov/map/index.html). The Kp index can be obtained from the German Research Centre for Geosciences (GFZ, https://www-app3.gfz-potsdam.de/kp_index/). The GNSS, barometric pressure data and space weather data used in this paper can be obtained online (https://doi.org/10.6084/m9.figshare.19189091.v1). Fortran source codes used in this work are given in http://www.ep.sci.hokudai.ac.jp/∼heki/software.htm.
COMPETING INTERESTS
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
