The massive 2022 eruption of Hunga – Tonga Hunga -Ha’apai focused worldwide attention on volcanoes along the Tonga – Kermadec volcanic arc.  While most popular post-eruption videos covered the eruption and destruction of the aerial portion of the erupting volcano, they also covered impact of the eruption on surrounding islands, which was substantial.  We ran across one of these while streaming entertainment video a few months ago.  It ended with speculation about the next Tongan volcano to produce a similar massive eruption and selected Tofua.  With that in mind, lets take a look at Tofua.

 Tofua is an oval shaped island, 10 x 8 km in diameter, with an area of 80 km2.  Its 5 km diameter caldera hosts a cold, freshwater lake.  Rims top out at 515 m above the lake which is 500 m deep.   It is located 6 km S from neighboring Kao, some 29 km NW from the closest neighboring Tongan Island, Kotu, 151 km N of Tonga, and 89 km NNE from the remaining aerial portion of Hunga Tonga Hunga Ha’apai. 

There are three post-caldera cones, one of them, Lofia still active, in eruption 2022 – 2023.  Eruptive products are basaltic andesites, andesites and some dacite flow from one of the older cones.  The three cones and their lava flows occupy the N end of the caldera lake. 

The island was designated as a national park in 2001.  It is overgrown by the largest undisturbed Tongan tropical rain forest.  It is also an important bird area and has been uninhabited for the last 30 years, though it was settled on and off before then.  Smithsonian GVP lists 1,556 people within 5 – 30km.  I believe most of these are on the inhabited islands to the E. 

The island is densely vegetated, though somewhat thinner in the N part.  Vegetation varies from bushes to palms and trees.  There are many coconut and fruit (banana) trees.  There is a small plantation on the N part of the island.  The island does not have a classic sandy beach.  Its shores are mostly rock, pebbles and cliffs.  Some areas of the island gets high waves during strong storms.  There is not much remaining human infrastructure on the island. 

The western world first discovered Tofua in 1774 when Captain Cook sailed between Tofua and Kao.  He did not land and observed smoke rising from the island.  Tofua played a part in the Apr 28, 1798, Mutiny on the Bounty, which took place some 56 km from the island.  After being cast off the ship, Captain Bligh navigated his crowded launch on a 41-day voyage to Tofua.  They reprovisioned and sailed to Timor, reaching it Jun 14.  There was one casualty, stoned to death on Tofua by natives. 

There is an oral tradition of the formation of the caldera some 1,000 years ago and subsequent creation of neighboring Kao. 

The island is not actively monitored, though visited by volcanologists from time to time.  We previously covered volcanoes along the Tonga – Kermadec Ridge in our Niuafo’ou Caldera, Exceptionally Violent Volcanic Eruptions, Havre Seamount and Tafu – Maka posts. 

Region

Volcanoes in the region are part of the Tongan volcanic arc, a subset of the larger Tonga – Kermadec Ridge that stretches all the way S to New Zealand.  We will restrict our discussion of neighboring volcanoes to those listed by Volcano Discovery within 100 km N and S of Tofua.  Distances will be measured from Tofua.  Seafloors in this part of the arc are generally 1,500 – 1,200 m below sea level. 

Fonuafo’ou

 Fonuafo’ou (Falcon) Island is an ephemeral island located some 70 km SSE from Tofua.  It was initially reported by the HMS Falcon as a low shoal in 1865.  Islands up to 6 km long were formed in 1885 and 1927 eruptions.  By 1933, the island was over 145 m high.  Passing ships often reported smoke issuing from the location.  The latest reported activity was 1936.  By 1949, the island had eroded beneath sea level.  Subsequent eruptions took place in 1970 and 1993.  The summit is some 17 m below the waves. 

Feinga Seamount

Feinga Seamount is located some 43 km SSW from Tofua and 32 km NE of Fonuafo’ou.  It is topped with a 4 km diameter caldera open to the N.  There are no reports of recent volcanic activity from this seamount. 

Kao

Kao is the highest and one of the most spectacular Tongan volcanic islands.  It rises steeply from the ocean some 6 km N of Tofua Island.  It is connected to Tofua by a submarine ridge.  The island is elongated NNE-SSW, with slopes greater than 35°.  It is topped with a series of small coalescing craters.  The lower flanks are densely covered by jungle.  Upper slopes are almost entirely unvegetated.  There are no fresh lava flows, though the island is uneroded, suggesting it is relatively new.  There is an 1847 report of smoke pouring forth from Kao. 

Unnamed

The Unnamed seamount 19.44°S 174.95°W is located some 33 km NNE from Tofua.  It is topped with a 6 km caldera.  Google Maps shows its flanks rippled on three sides which may be eruption debris, flank collapses or lava flows.  Volcano Discovery lists this seamount as inactive over the last 10 ka, likely extinct.

Metis Shoal (Lateiki)

Metis Shoal (Lateiki) is a submarine volcano located 63 km NNE from Tofua.  It produced a series of ephemeral islands since first confirmed in the mid-19^th Century.  It was initially observed by western explorers in 1781 and subsequently eroded away.  During the 20^th Century, waves were observed breaking on rocky reefs or sandy banks to a depth of 10 m or less.  Dacite tuff cones formed during eruptions 1967 and 1979 were quickly eroded below the ocean surface.  A 1995 eruption built a dome forming an island 280 m in diameter and 43 m high.  It persisted until Dec 2006. 

An Aug 2006 eruption created pumice rafts between Fiji and Tonga.  These dissipated by Oct.  An eruption Oct 2019 created an island that disappeared by mid-Jan 2020.  This eruption created an ash plume that triggered aviation warning. 

Home Reef

Home Reef is a submarine volcano midway between Metis Shoal and Late Island located some 86 km NNE from Tofua.  It has built ephemeral islands repeatedly eroded back below the waves in multiple eruptions 1852, 1857, 1984, 2006 and 2022.  Two of those produced pumice rafts.  Eruptive products here are dacites.  Most eruptions here in the VEI 2-3 range. 

The 1984 eruption built a small island 500 x 1,500 m.  It produced a plume at least 12 km high, and pumice rafts that traveled as far as Fiji and Australia.  The island was eroded below sea level within a few months.

The 2006 eruption also built a temporary island.  The pumice raft from this eruption also made it to Fiji and was about the same size as the 1984 raft.  The island was about the same size as the one produced by the 1984 eruption.  It was first reported by a passing yacht.  Rafts were imaged by satellite.  MODIS thermal imagery documented a small, hot crater.

The most recent eruption was Sept 2022.  Discolored water around the eruption site was imaged starting in May.  The new island emerged 10 Sept along with a gas and steam plume rising as much as 2 km above the volcano.  By 16 Sept, the island was 180 m in diameter and 170 m high.  Ash emissions continued with the plume measured to 8 km.  Discolored water surrounded the island and drifted mainly S.  Activity tailed off by the end of the month.  By 29 Sept, there was no ash in the plume.  There was persistent fumarole and lava flow activity.  Explosions on 5 and 6 Oct put ash plumes 3 km above the new island.  The island grew to 280 m in diameter, 15 – 18 m high.  The last thermal anomaly was recorded on 17 Oct.

Late Island

Late Island is a 6 km wide circular volcanic island located some 108 km NNE of Tofua.  It tops out at 540 m.  It is topped with a 400 m wide, w50 m wide summit crater with an occasional lake.  The andesite / basaltic andesite volcano rises some 1,500 m from the surrounding sea floor.  There are cinder cones on the flank N of the summit crater, W and N of a plateau 100 – 150 m below the summit and on the NW coast.  There is a graben-like structure on the NE flank with two large pit craters.  The lower one is partly filled with a saltwater lake.  Explosive eruptions from these in 1790 and 1854 may have produced lava flows. 

A Nov 2022 visit to Late observed warm ground surfaces, steam rising from cracks and new vents at the summit.  The new features were too small to show up in satellite imagery.  Volcanologists concluded that the hydrothermal system was causing the new features.  The cracks were attributed to past volcanic activity and there was no new risks to local communities.  Aviators and mariners were asked to report any observable changes on the island. 

The western world first discovered Late Island in 1781 with a Spanish expedition.  It was later visited by French and British explorers in 1787 and 1791.  There was a large pumice raft observed NW of Late Island in Aug 2019.  This raft was not produced by Late. 

Tofua Caldera

The 10 x 8 km Tofua Island is the aerial portion of a single stratovolcano.  This stratovolcano collapsed in a pair of climactic eruptions around 1030 AD, forming a 5 km diameter caldera.  This caldera contains a freshwater lake, up to 500 m deep in some places.  The N portion of the lake has been filled with three post-collapse cones, lavas, spatter and ashfall.  One of them, Lofia, is the active cone of the system. 

The best physical description of Tofua is Bauer’s 1970 Geology of Tofua Island, Tonga.  One of the interesting observations is that he believes the caldera collapse is continuing on the N, E and SE sides.  This is visible along fault displacements along the rim creating hummocky topography, small grabens, steps with scarps 15 – 30 m high. 

Large stream valleys do not exist on Tofua.  All are narrow, though some are deep.  They are mostly on the SE – SW windward flanks, extending perhaps 500 m inland.  They are mainly cut into post-caldera tuffs.  Gullies along the SE – SW caldera walls cut into pre-caldera rocks.  Most of the island is circled by a cliff averaging 50 m high.  The cliff is steepest where lavas are exposed by undercutting of massive lava layers by wave action and their collapse. 

Four main units mapped on Tofua:  Hamatua Formation (precaldera basaltic andesites, andesites and dacites); the Hokula Froth Lava; Kolo Formation; and the Lofia Formation (most recent). 

The Hamatua Formation is mainly lava flows erupted before caldera collapse.  It is only exposed in scattered patches on the flanks of the main cone.  It is best exposed in the walls of the caldera, nearly 500 m thick Individual flows are around 20 m thick.  There is little vertical joining in the central mass.  Most of these lavas are basaltic andesites and andesites.

The Hokula Froth Lava overlies the Hamatua Formation.  It is separated by an erosional layer and only exposed only in the sea cliff.  Its extent is unknown and thought to have come from the upper part of the mountain.  It does not exist in the caldera wall.  Total thickness of several layers of this lava approach 10 m.  There is some columnar joining, and the layer is easily cut by wave action.  The 1970 Bauer description of this mentions froth lavas first identified in Italy.  An internet search describes froth lavas as welded pumice. 

The Kolo Formation is mainly lapilli breccias, tuffs, with interbedded basaltic andesite lava flows, unconsolidated ash, cinders and one thick andesite lava flow.  The Tuff breccia is a poorly sorted airfall.  It is found on the S, E and N flanks of the island, forming a layer over 30 m thick.  Near Hamatua village, it fills a valley with layers up to 60 m thick.  Source vents for these tuffs are on the NE, E and S caldera rims.  They resemble maars in other parts of the world.  Explosions that formed these craters took place close to the surface and were probably phreatic.  The unconsolidated tephra is younger than the tuff breccia.  It erupted from a series of cinder cones on the caldera rim.  Concentric faults formed during the collapse allowed rising magma to reach the surface and erupt. 

The Lofia Formation is mainly confined to the caldera, exposed on the N flank of the volcano.  The part outside the caldera is only a thin veneer due to spatter from cones on the N caldera floor.  Lofia Formation lavas, spatter and pyroclastics cover rocks of the Hamatua Formation that collapsed into the caldera.  Contact with Kolo Formation rocks on the rim indicate that both erupted simultaneously.  Pyroclastic layers are thin and poorly sorted.  There is evidence of a directed blast from Lofia toward the caldera wall.  All Lofia lavas stayed on the caldera floor.

The largest structural features on Tofua are three types of caldera collapse faults.  The main faults bound the sunken block of the caldera floor.  There are faults parallel to the rim fault that magma erupted through and fault along the rim which eruptions did not take place. 

The maximum displacement of the main ring faults is over 460 m.  The base of the rim wall has been buried by later volcanics and more recent collapses from the rim wall itself.  The secondary faults that magma erupted through are the second most numerous.  They are parallel to the rim and found on the N, E and SE rims.  They seem to represent an outward spreading of the zone of caldera collapse.  Tofua is unique in that no eruptions took place from radial fissures.  

The main cone of Tofua was built by Hamatua lavas.  They have not yet been dated.  There was period of quiet during which erosion cut deep valleys into the cone.  Caldera collapse took place 1030 AD.  Timing of eruption of the Hokula Froth Lava before or after the collapse is unknown.  There is some speculation that the volatiles in the froth lava may have triggered the collapse.  An unknown amount of the froth lava flowed into the sea.  The top of the froth lava was then eroded.  Tension cracks from the expanding caldera provided conduits for rising magma to erupt from multiple locations around the rim creating the Kolo Formation.  Finally, activity within the caldera erupted the Lofia Formation.  This may have been concurrent with the Kolo eruptions, though Kolo stopped and Lofia is ongoing. 

Nov 2000 Geologic Mapping

A geologic mapping team spent 8 days on on Tofua itself in Nov 2000. They reported the following in the Smithsonian GVP Dec 2001 Bulletin Report (BGVN 26:12):

Flanks rise steeply to a well-defined rim 515 above sea level.  The inner caldera walls are steep into the large, cold caldera lake at 30 m above sea level.  The most recent volcanic activity took place from three cinder cone complexes in the N half of the caldera. 

The W cinder cone is densely forested and eroded.  The E complex has four distinct overlapping smaller cinder cones with small craters.  Young basaltic andesite lavas erupted from these cones and flowed into the lake.  There are varying levels of vegetation on the flows.  The youngest may have been erupted during the 1958 – 1960 series of eruptions. 

The N cinder cone (Lofia) is large and vigorously degassing.   Its cone is 500 m in diameter, topping out at 380 m.  The summit crater is 70 m in diameter with vertical inner walls.  It was completely filled with SO2-rich steam during the visit.  There were intermittent noises from the crater floor.  Occasional orange reflections from clouds above the caldera rim were seen from the sea on two nights.  No evidence of recent spatter was found around the crater rim.  No indication of significant volcanic activity since the 1958 – 1960 eruption sequence.  Fire fountaining from Lofia created a 20 m thick lava flow on the NW caldera rim.  Breadcrust bombs were ejected and welded in place during eruptions before and during 1958 – 1960. 

Three craters were observed during May – Jun 2006 fieldwork at the bottom of the vent.  One was quiet, one had active fumaroles, one had an orange glow.  Three days of thermal alerts were observed in 2004, 9 days in 2008.  They all came from the same area in the N half of the caldera.  At least a couple of the 2008 thermal alerts were sun glint off the ocean surface.

Caldera Formation

Caldera formation took place some 1,000 years ago via a VEI – 5-6 eruption.  There were two phases of this eruption, each producing an ignimbrite.  The first and largest eruption caused the inward collapse of the stratovolcano producing the Tofua ignimbrite and creating the caldera.  This was deposited in a radial fashion over the entire island.  It is also associated with Plinian fall deposits up to 0.5 m thick on islands more than 40 km away.  Common scoria bombs are distributed throughout the ignimbrite, meaning a small degree of water – magma interaction during the eruption.  The eruption was thought to be due to a water-saturated shallow (less than 6 km) dacite magma.  The chamber roof collapsed causing a Plinian eruption column.

After a short hiatus, large-scale faulting in the SE part of the island produced a second explosive phase of the eruption.  This was due to recharge of the magma changer with a chemically distinct magma.  This eruption was similar to the first one, though smaller.  It emplaced the Hokula ignimbrite in the NW part of the island. 

The events deposited at least 8 km3 of juvenile material.  The eruptive column is thought to have reached at least 12 km, depositing fall on islands 40 – 80 km to the E.  These eruptions are tentatively assigned to the 1030 AD volcanic sulphate spike in Antarctic ice sheets. 

A 1999 geologic team mapped the thick compound ash layer on several neighboring islands 30 – 50 km E from Tofua.  This ash layer has three distinct units and abundant pumice clasts.  The ash is thought to represent part of the pyroclastic flow sequence associated with caldera forming eruption on Tofua.  Erosion of these layers left pumice stones on the beaches of the islands.  The stones are thought to be small blocks of andesitic pumice from the ash layers.

Eruptions

There were at least 12 eruptions reported from Tofua since 1774.  Half of these were VEI 2 eruptions.  There were two in the 18^th Century, 1774 and 1792.  Four of these were reported in the 19^th Century 1845 – 1885.  The 20^th century was similarly busy with eruptions 1906, 1958 and 1993.  The most recent two were 2004 and an ongoing series of eruptions starting 2015.  Note that the island has a vigorous hydrothermal system and an active intracaldera cone, so “smoke” and fumarole reports are common.  The 1958 – 1960 series of eruptions forced resident islanders to evacuate for a year or more. 

Smoke was reported by a passing ship in Jun 1979.  Most of the reported eruption sites were investigated 11 – 18 July.  They observed large fluctuations in the intensity of fumarole activity.

Fumarole activity in 2000 put almost continuous steam above the island.  At the time, it was considered to be normal state of the island.  A geologic mapping and observation team visited the island for 8 days in late Nov. 

There was near continuous activity 2010 – 2011, with a 1.3 km plume reported in Jul 2011.  The Wellington VAAC issued aviation warnings through 19 Jul, though the cloud was not visible on satellite imagery.  Glow from the active crater was visible up to 50 km from the volcano during the night.  The crater was littered with lava stone.  There was active spatter in the crater.  Low level eruptions produced larger ash plumes.  11 satellite thermal alerts were issued 2011 – 2013.  Four more were issued in 2014, though no ash advisories were issued in 2014.  The anomalies were more frequent beginning Feb 2018 – Jul 2020.  They were all centered on the Lofia crater.  Activity continued into 2021 with intermittent hotspots.  The most active cluster was Feb 2021 which included SO2 plumes and consecutive days with thermal anomalies. 

Activity increased noticeably Feb 2021 with stronger thermal anomalies and SO2 plumes.  A total of 114 days of thermal hotspots were reported Aug 2021 – Jul 2022.  The volcano is considered in 2023 to be in eruption. 

Tectonics

Tonga – Kermadec subduction zone extends 2,000 – 2,500 km from New Zealand to Fiji.  Activity is due to the W subduction of the Pacific Plate beneath the Indo – Australian Plate.  Collision of the Louisville Seamount Chain with the trench segmented the zone, dividing it into the Tonga Trench in the N and the Kermadec Trench in the S.  The collision may have influenced when and where back arc rifting took place.  Rifting creating the Lau Basin began 6 – 5.5 Ma.  This evolved into mature seafloor spreading with multiple spreading centers by 4 Ma.  Arc volcanism shifted from the remnant Lau Ridge to create the new Togua Arc 3.5 Ma.  In the S, rifting of the Kermadec Arc less than 2 Ma created the Havre Trough. 

The sharp bend of the Tonga Trench from NE to W at its far N changed action along the trench from subduction in the S to strike – slip in the N.  This fault has been described as a STEP fault, associated with vertical tearing that causes a piece of the subducting plate to remain at the surface. 

Behind the arc, the modern Lau Basin has a “V” shape, 500 km wide at the N to 200 km in the S.  In the S portion, there is a single, segmented spreading center.  The number of spreading centers increases to the N, indicating increasing tectonic complexity.  The Niuafo’ou Microplate to the NW is suggested to exist between the Indo-Australian and Tonga Plates.  It is bounded by and contains numerous spreading centers, and propagating rifts.  It might be more structurally complex (multiple microplates) than currently thought.  There are likely several other microplates in the Northern Lau Basin with poorly defined boundaries. 

Conclusions

While Tofua is a vigorously active volcanic system, there is nothing I have seen writing this post that suggests it is the Next Big Thing out of Tonga.  On the other hand, there is nothing to suggest it won’t be.  But extraordinary expectations require extraordinary evidence, and none as yet exists.  The submarine volcano is certainly large enough.  Caldera formation is certainly recent enough.  And the resurgent volcanism in the caldera is certainly active enough to suggest that this system will continue to be active.  Subduction continues as will magma generation along the volcanic arc.  It is enough, I think to simply mark Tofua as a dangerous, active volcano without tabbing it as the source of the next eruption that changes the world. 

Additional information

Magma evolution in the primitive, intra-oceanic Tonga Arc:  petrogenesis of basaltic andesites at Tofua volcano, Caufield, et al, Mar 2012

Mafic Plinian volcanism and ignimbrite emplacement at Tofua volcano, Tonga, Caulfield, et al, Apr 2011

Tofua volcano, Tonga Islands, South Pacific Ocean, NASA Visible Earth

Report on Tofua (Tonga) – August 2021, Relief Web, Oct 2021

Volcanoes on land and in the ocean around Tonga, Pacific Ocean, Active volcanes, Vulkaner.no

The genesis of silicic arc magmas in shallow crustal cold zones, Adam, et al, Aug 2016