View from the summit

Hunga Tonga-Hunga Ha'apai major eruptions

13 January 2022 15:00-23:50 UTC and 14 January 04:51 UTC-17 January 04:30 UTC

Photo credit: Damien Grouille

View from the summit
View from the summit

In January 2022 the underwater volcano Hunga Tonga-Hunga Ha'apai in the Pacific Ocean erupted twice causing major gravity and shockwaves to reverberate around the world.

Last Updated

03 October 2023

Published on

17 January 2022

By Ivan Smiljanic (CGI), Jan Kanak (SHMU), HansPeter Roesli (Switzerland),  Federico Fierli, Remko Scharroo, Thomas August and Marc Crapeau (EUMETSAT), and Simon Proud (National Centre for Earth Observation)

13 January eruption

NOAA’s GOES-17 satellite captured the eruption of the Hunga Tonga-Hunga Ha'apai volcano on 13 January 2022 (Figure 1). The extent of the ash plume can be seen, as well as multiple rippling gravity waves emanating outward. Gravity waves are generated, similarly to convective events, by an energetic density fluctuation; wavelength, as visible in the animation, can range from just a few kilometres to a few hundred kilometres. This type of imagery has the finest spatial resolution of all the bands.

Figure 1: GOES-17 ABI Band 2, 13 January  15:10 UTC. Source: NESDIS

Himawari-8 also captured imagery of both the eruption and resulting gravity waves (Figure 2).

Himawari-8 TCOL and IR 13 Jan 2022
Figure 2: Himawari-8 AHI True Color (left) and IR10.8, 13 January 18:25 UTC

For the close-look area the time differences for the IR10.4 band (AHI B13) has been done (Figure 3), which revealed a surprising pattern of outward-spreading circular waves, probably due to a serendipitous coincidence of wave length and delta-time in contrast (the explosion on 15 January had a more chaotic pattern).

Figure 3: Difference between two scans of Himawari-8 ABI IR10.4 channel, 13 January, 15:15 – 23:55 UTC.

From the temperature field in motion (Figure 4), aside from the aforementioned gravity wave signature, westward moving ‘puffs’ above the cloud anvil can be seen (in green temperature shades around 200 K). These puffs, or ejections, are embedded in the stratospheric air, which is warmer than the main cloud anvil (by roughly 20 K) and also embedded in the westerly flow at around 20,000 m. Confirmation on the vertical temperature profiles comes from a nearby radiosounding at Pago Pago airport.

Figure 4: Himawari-8 IR10.4, 13 January 15:00-23:50 UTC

On the comparison of the Himawari-8 Cloud Phase RGB and Dust RGB, and wind field at two different heights (200 hPa and 70 hPa), on 13 January, 21:00 UTC (Figure 5), the volcanic cloud appears as yellow in the Cloud Phase RGB, while the higher volcanic puffs in stratosphere appear in ochre shades in the Dust RGB.

Himawari-8 Cloud and Dust RGB comparison

Dust RGB compare1

Figure 5: Himawari-8 Cloud Phase RGB and ECMWF wind field at 200 hPa (left) and Dust RGB and ECMWF wind field at 70 hPa (right), 13 January 21:00 UTC.

GOME-2 (from both Metop-B and -C missions) provides sulphur dioxide total column (SO2). SO2 is emitted in volcanic eruptions and the effect of the Hunga Tonga-Hunga Ha'apai explosion can be seen as a large red area on the bottom right in Figure 6. SO2 may convert to sulphuric acid when reacting with water vapour, thus, remaining in the atmosphere for longer.

GOME-2 sulphur dioxide total column, 13 Jan 2022
Figure 6: GOME-2 SO2 vertical column (from AC-SAF data), 13 January. Source: IASB-BIRA, Belgium

15 January eruption

The eruption on 13 January was only a precursor of what was to come. Two days later, on 15 January, the underwater volcano's explosive eruption was so massive, the impacts were felt across the world.

A first shockwave travelled around the globe, with a second recorded days later. The shockwaves were so widely experienced that they showed up on pressure sensors at ground level and were also visible from space.

Himawari-8 imagery of shockwaves

The most impressive feature in the Himawari-8 IR10.4 loop (Figure 7) are the shock waves coming from the volcano on 15 January. These even initiated convective events in a wide area around the volcano. In the black and white scale of temperatures (180–320 K) it is easy to see the warmer layer of stratospheric cloud that moved westwards. This cloud was initially at roughly 235 K, which according to nearby Pago Pago radiosounding would correspond to a height closer to 30 000 m. The brighter anvil below this cloud was at the tropopause level, i.e. at the stable atmospheric layer where gravity waves could oscillate, with temperature changes from roughly 195 K to 200 K (crest to trough).

Figure 7: Himawari-8 IR10.4, 15 January 03:00-18:00 UTC

Some rather interesting preliminary calculations emerging from the looping imagery, i.e. changes in time and space (compared to Pago Pago radiosounding), are:

  • Horizontal cloud expansion at the speed of 360 km/h (120 km from 04:20 to 04:40 UTC — E-B line in Figure 8)
  • Vertical cloud expansion at speed of ca. 80 km/h /(from 0 to -85°C from 04:00 to 04:10 UTC)

Himawari-8 channel comparison

04:40 UTC compare1

Figure 8: Himawari-8 VIS0.64 channel comparison at 04:20 and 04:40 UTC, 15 January

Figure 9 follows the explosive eruption through advanced 500 m resolution of VIS0.64 channel during the local evening hours. Notice the speed of vertical and horizontal expansion of the cloud that was a mixture of mostly volcanic ash, SO2 and water vapour gases.

Figure 9: Himawari-8 VIS0.64 channel loop from 03:00 and 06:00 UTC, 15 January

The Himawari-8 Airmass RGB image (Figure 10) captured this high stratospheric cloud in unusual yellow shades, with white underlying cloud at the tropopause level. This cloud was skirted by the pale cyan ring of subliming water vapour at the high levels, due to an abrupt pressure change.

Himawari-8 Airmass RGB 15 January 05:10 UTC
Figure 10: Himawari-8 Airmass RGB, 15 January 05:10 UTC. Credit: CIRA

Meteosat imagery of shockwaves

On the comparison of Meteosat and Himawari imagery the first shockwave can be seen converging on Algeria (Figure 11). This video shows fluctuations caused by the shockwave in temperature measurements of around 2 K from the 6.2 micron bands on sensors aboard the Himawari-8 and Meteosat-11 satellites.

Figure 11: Himarwari-8 (left) and Meteosat-11 (right) imagery of the shockwave. Credit: Simon Proud/National Centre for Earth Observation.

The Meteosat-11 animation (Figure 12) captures the movement of shockwaves from 15 January 13:30 UTC to 16 January 02:30 UTC, in 15-minutes time-steps. Image is centred to 0˚ longitude. It shows how the circular shaped wave was oscillating between the volcano located in Pacific Ocean and a north Africa inflexion point.

Figure 12: Meteosat-11 IR6.2 brightness temperature differences between successive images*, 15 January 13:30 UTC to 16 January 02:30 UTC

In the Meteosat-8 animation, from 15 January 12:00 UTC to 16 January 11:30 UTC (15-minutes time-steps) (Figure 13), thanks to the eastern satellite position, the wave can be very clearly seen travelling from west Australia over the Indian Ocean and from the south after passing over Antarctica.

Figure 13: Meteosat-8 IODC (centred to 41.5˚ E), IR6.2 brightness temperature differences between successive images*, 15 January 12:00 UTC to 16 January 11:30 UTC

*Algorithm used to calculate the images uses SEVIRI IR6.2 data. Counts are converted to calibrated radiances and radiances to brightness temperatures. Then the image differences are calculated from two consecutive images in time. Changes resulting from wave movement are quite low but they are amplified by the increased contrast of the image. Shockwaves are highly visible over cloud-free areas and when the displacement of the wave per time step is equal to half the wavelength.

Gravity waves seen by IASI

The IASI infrared sensors flying 50 minutes apart on Metop satellites B and C were able to detect the gravity waves. The animation in Figure 14 shows the temperature sounding in the upper stratosphere from the operational IASI products, and the temperature anomaly compared to the previous days.

Figure 14: Temperature field around 3 hPa from IASI-B and IASI-C and the temperature anomaly

Because shockwaves are formed when a pressure front moves at supersonic speeds and pushes on the surrounding air, they can register on pressure sensors at weather stations. HansPeter's 15 January reading from his weather station in the Ticino, Switzerland saw clear spikes (Figure 15).

Shockwave measured in the Po Valley
Figure 15: Air pressure reading in Ticino on 15 January, the shock wave is clearly seen as high-low kinks before 20:00 UTC.

Readings from MeteoSwiss stations also clearly show the shockwaves (Figure 16 and 17).

Readings of the first shock wave at 6 MeteoSwiss stations
Figure 16: Mean sea level pressure readings of the first shock wave at 6 MeteoSwiss stations, all south of the Alps. You see the slight time shift between the orange (most northern) and the blue (most southern) trace. Credit: MeteoSwiss
Readings of  two shock waves at 4 MeteoSwiss stations
Figure 17: Mean sea level pressure readings of two shock waves at four MeteoSwiss stations, two north of the Alps (blue-red) and two south of the Alps (green-black). They show the passage of the both shock waves. The difference in arrival time between north and south station shows that the first wave was coming from north (across the North Pole) and the second from south. Credit: MeteoSwiss.

SO2 plume

Volcano Discovery reported that NASA's Ozone Mapping and Profiler Suite (OMPS) measured the eruption column height reaching up to 30 km.

In the animation of Himawari ASH RGB images from 15 January 06:30 UTC to 17 January 04:30 UTC (Figure 18), the spread of SO2 cloud from the volcano can be followed all the way to the Australian continent (bright green or blue-green shades).

Figure 18: Himawari ASH RGB, 15 January 06:30 UTC to 17 January 04:30 UTC

Figures 19 and 20 show the evolution of the SO2 plume as seen by IASI. The animations start just before the biggest eruption that took place the 13 January. The maps have been created using the operational Metop-B and C IASI L2 SO2 products (BRESCIA algorithm from ULB/LATMOS and AC SAF).

Figure 19: IASI SO2 13-19 January
Figure 19: IASI SO2, 13-19 January
Figure 20: IASI SO2, without the IASI PDUs pattern in background, 13-19 January
Figure 20: IASI Dust contamination index, 13-19 January

Observing the tsunami with Copernicus altimeter missions

It is rare to capture tsunami waves in satellite altimetry, the satellite need to be just there in the right place at the right time. Tsunami waves travel fast, are shallow (around 10 cm), and have long wavelengths (tens to hundreds of kilometers) when in deeper waters. It is when they arrive near the coast that they grow to a formidable size, as the wave bunches up to much smaller wavelengths.

Around 4 hours 45 minutes after the start of the major eruption and earthquake, the Copernicus altimeter missions Sentinel-6A Michael Freihlich and Jason-3 passed within 100 km of the Hunga Tonga-Hunga Ha'apai volcano. As the sea level anomaly (height of the sea level compared to a long-term mean) image shows (Figure 21), both satellites observed many trains of waves with wavelengths of tens to hundreds of kilometers and height variations of about 50 cm, peak to peak. Similar features were seen by Sentinel-3B passing some 15 minutes later and about 250 km west of the volcano. For comparison two other passes of Sentinel-3B and Sentinel-3A are shown from the previous day, that do not show such features — evidence that what is seen on 15 January are tsunami waves reflecting from islands and bottom topography, as well as newly-generated tsunamis from the ongoing eruptions. Note that these are really long-wavelengths waves and should not be confused with the wind-driven waves that we are normally accustomed to. As shown in the significant waves heights data (Figure 22), the wind-driven waves were totally nominal.

Figure 21: Sea level anomalies from Sentinel-3A and -B, Sentinel-6 and Jason-3, 15 January 04:15 UTC.
Figure 21: Sea level anomalies from Sentinel-3A and -B, Sentinel-6 and Jason-3, 14 and 15 January 04:15 UTC.
Figure 22: Significant waves heights from Sentinel-3A and -B, Sentinel-6 and Jason-3, 15 January 04:15 UTC.
Figure 22: Significant waves heights from Sentinel-3A and -B, Sentinel-6 and Jason-3, 14 and 15 January 04:15 UTC.

Additional content

AC SAF SO2 estimates for 15 January (Twitter)
Explosive eruption of the Hunga Tonga volcano (CIMSS Satellite Blog)
Volcano Discovery's Hunga Tonga-Hunga Ha'apai volcano pages
Tonga eruption: How its impact spread so widely and violently (BBC News)
Himawari Ash RGB Quick Guide
Atmospheric pressure reading from Elm weather station, 15 January (Credit: Jörg Asmus)
Volcanic eruption in Tonga and the Pacific - Charter Activation (Disasters Charter)