Sometimes you run across things by simply looking out the window of a commercial airliner that are large, unexpected, and surprising.
Such was the case on July 31, 2012 when Maggie de Grauw, flying from Samoa to Auckland, NZ found a kilometers wide raft on the surface of the Pacific Ocean. She took a few photos, some of which were forwarded to Dr. Scott Bryan at Queensland University of Technology in Oz. Dr. Bryan was an author of an article on pumice rafts.
Once the photo was received, the detective story started: What was this? Was it erupted? If so, when? From where? And under what circumstances? The date time stamp on the photo combined with the flight path of the jet allowed scientists to start the process of locating the rafts and the source.
The expanding raft was photographed before the first on site sampling, August 10, by a Royal NZ Naval vessel, which found a portion of the raft that was 0.6 m thick, a kilometer wide and extended as far as the eye could see. It was made of golf ball to soccer ball sized pieces of pumice, some smooth, and much of irregular, rough shape.
During this time, investigators also started taking a look as satellite imagery and seismic data for the previous month. They located a short seismic swarm 17 – 18 July in the region. They found MODIS imagery of a hot spot on July 18. They found MODIS imagery of a vapor plume on July 18. MODIS is primarily thermal imagery. Finally, they first found visible satellite imagery of a pumice raft dated July 19.
The imagery put the source near three known features, one subsurface and two barely above the surface of the Pacific. These were Havre Seamount (subsurface), Havre Rock (70 m above the water), and L’Esperance Rock (also slightly above sea level).
The seismic / hydroacoustic events were more difficult to locate, though they most certainly indicated something happening on the ocean floor. There were 157 events recorded 17 – 18 July. Waveforms of the events were very similar. These included 68 events M 3.0 – 4.6. The mean location of these events was at Havre Seamount.
Eric Klemetti and Robert Simmon discovered the first signs of the eruption by examining satellite imagery and correlating dates and times with the hydroacoustic events. They concluded that the main pulse of eruption only lasted a couple days but was strong enough to put a thermal pulse on the surface of the ocean visible from space from a depth over 700 m.
By July 21, the eruption was mostly over, and with it pumice production to the surface. By Aug. 13, the raft covered an area 450 x 250 km. October coverage was closer to 22,000 km2.
Floating bits of material produced by subsurface volcanic activity are not uncommon.
The 2011 – 2012 eruption of El Hierro produced lighter than water blobs of pumice called pumice clasts / lava balloons that reached the surface and floated briefly before sinking. The balloons were ellipsoidal in shape, up to a meter long by 40 cm in diameter. They were a shell a few centimeters thick with several layers of bubble density inside. Depth of formation was 300 – 1,000 m. Formation took place on an active conduit or lava lake on the bottom. These were also seen from the 1998 – 2001 Screta Ridge eruption off Terceira Island in the Azores.
Pumice rafts are yet another volcanic hazard caused by gas-impregnated pumice. They require some portion of the eruption to be underwater and the eruption to be appreciably gas and water-rich. Like most things volcanic, they also have a beneficial impact on wildlife and plant life, providing a way for that life to spread far and wide as the winds and currents move them thousands of miles. They also provide ways for species to spread between isolated island ecosystems.
Pumice rafts are quite long lasting. A 2006 Home Reef eruption in Tonga produced pumice rafts that had traveled 5,000 km mostly west in eight months. The rafts are also used by ocean-based, deep water, shallow water, and some land based wildlife and plant life to distribute itself throughout the widely scattered islands. The Home Reef rafts distributed 80 species of marine life. Some of the pumice from this eruption floated for over 2 years.
The majority of rafting events over the last 200 years took place in the Pacific Ocean, with the Tonga – Keermadec region being the most prolific over the last century. Krakatau and Tambora put significant pumice rafts into the Indian Ocean.
The first closely studied pumice rafting vent was the 2006 eruption of the Home Reef Volcano in Tonga. Pumice from it made it to the eastern shore of Australia up to 900 days after the eruption. The study documented the biological cargo carried by the raft during its time at sea. This eruption was sub-Plinian, in the VEI 2-3 range, with a plume 7 – 15 km tall. The eruption built a pumice cone 75 m tall which was quickly removed by wave action, not unlike what we see with Bogoslof.
The Home Reef pumice raft was produced on Aug. 7, 2006 and initially move NE towards Vava’u Islands of Northern Tonga. The raft dispersed into extensive linear features called stringers or windrows, tens to hundreds of kilometers long, covering 1,600 km2y mid-Sept. Pumice strandings from this raft were observed for 20 months following the eruption. Total volume of magma producing this pumice is estimated at 0.16 km3 DRE.
Average velocity of movement is a combination of wind and current action. Strong trade winds for the Home Reef rafting moved it an average of 20 km / day.
The Havre Seamount eruption produced a 400 km2 raft in a single day. Initial thickness of the raft was over a meter thick. By Aug. 10, there were 0.6 m of pumice above the surface. A second rafting event near Tonga on Aug. 12 created additional rafting pumice.
The Kermadec Islands are the emergent tops of a line of stratovolcanoes making up the Kermadec volcanic arc. They lie 800 – 1,000 km north of New Zealand’s North Island. They are a similar distance SW of Tonga. There are four main islands or island groups with a total area above the water of 33 km2.
Water around the islands is quite deep, with depths of 1,000 m common. Perhaps a third of the area surrounding the islands is 5,000 m deep. Volcanic activity is not confined to the Kermadecs and extends throughout the Tongas to its north.
Climate of the region is subtropical. Dense subtropical forests cover the largest island Raoul. There are no native land mammals on the islands, though cats, rats and goat were introduced over the last few centuries which devastated populations of birds and the forests. New Zealand designated the Kermadec Ocean Sanctuary in 2015, over 620,000 km2 of islands and surrounding oceans.
Volcanic activity in the Kermadecs include hydrothermal vents, seamounts, and volcanic landforms such as cones and calderas. There have been six deep-sea expeditions to survey the region in detail over the last 15 years. More than 50 new submarine volcanoes were discovered in the last decade.
Although the islands are mostly uninhabited, they do sit underneath air traffic routes linking New Zealand with Tonga, Western Samoa, Niue and the western US. There are regular shipping routes that traverse the arc. Based on the sizes of volcanic eruptions, tsunamis are also an ongoing threat.
Many of the Kermadec submarine volcanoes are comparable in size and form to terrestrial volcanoes in New Zealand. At least eight of the seamounts are over 2,000 high. One of them rises within 65 m of the surface. The largest volcano is 25 km in diameter. At least 16 more are 10 km in diameter. Macauley Island is part of a large volcanic complex at least 35 km in diameter. Its rim is capped with more than 30 recently active cones that have sent lava flows down the outer rim and into the caldera. Seamount numbers appear to increase as we approach White Island off northern NZ.
Subsurface investigations of seamounts along the Kermadec – Tonga Arc identified numerous calderas, sector collapse events, and volcanoes with eruptive volumes in the 35 – 60 km3 range. Unfilled caldera volumes around 18 km3 exist. The Healy caldera near northern New Zealand has been tentatively correlated with a 600–year old beach pumice deposits. Monowai is one of at least 26 major volcanic centers on the Kermadec arc. It has numerous eruptions since 1877. At least one of these was a sector collapse with associated direct blast. The amphitheater was refilled with subsequent volcanic activity.
Primary magmas along the arc are felsic basaltic and basaltic andesites.
Raoul Island is the largest island along the arc. It was originally constructed out of basaltic andesites. Magmas shifted to dacites some 4,200 years ago. The island was originally a seamount and it has multiple violent caldera forming eruptions over the last 10,000 years. One eruption 1,700 years ago was primarily basaltic. Raoul eruptions over the last 4,100 years have mostly been VEI3s, with a VEI 5 -6 some 2,300 years ago. The larger eruptions also emplace the majority of its ejecta as pyroclastic flows and tephras.
Macauley caldera is the second largest island in the Kermadecs. Its highest point is 238 m above sea level. It has no known historic eruptions, though there were a pair of submarine eruptions within 50 km of it. The island is primarily basaltic lava flows with thick deposits of dacitic ignimbrites. The last of these erupted some 1,800 years ago. This eruption was in the VEI 6 range and is the largest eruption in the Western Pacific between NZ and New Guinea over the period. There have been larger eruptions on the NZ mainland during that period.
Curtis Island is 35 km SSW from Macauley. It is a small island less than a kilometer across built mostly from welded pyroclastic tuffs. There are active fumaroles on its crater floor. There are neighboring active seamounts along the same ridge. There are numerous additional rocks, banks, and seamounts as you work your way south.
The Rumbles are a group of active seamounts north of White Island named because they were the source of subsurface noise recorded north of NZ. Most are unnamed, though a few are numbered. There are also numerous seamounts as you approach White Island that show evidence of flank collapses and lateral blasts. Dredged samples include dacites and andesites.
Havre Seamount was initially mapped in 2002. It was a 5 km diameter caldera rising a kilometer from the surrounding sea floor. Its crater was 800 m deep. The caldera had a wall with a bulge toward the interior of the caldera. Speculation was that this bulge was a possible location for a future eruption or undersea avalanche.
The seamount lies 30 km NW of Havre Rock in the Kermadec volcanic arc. Minimum depth is 650 m. Maximum depth is 1,750 m. Total relief is 1,100 m. Havre measures 22 km x 25 km.
The seamount is located in the Havre Trough, a 2,700 m deep depression between the Kermadec Ridge to the east and the Colville Ridge to the west. The seamount was mapped after the 2012 eruption, with a new volcanic cone 240 m tall on the rim of the caldera. Top of this cone gets within 1,100 m of the surface.
The 2012 eruption of the Havre volcano is the largest known deep-ocean eruption. It is one of the few recorded submarine eruptions with rhyolite magma. And it is largest and deepest silicic submarine eruption of its type ever recorded. The size and depth of the eruption are significant. Before the eruption, the extreme hydrostatic pressure at depth was thought to preclude explosive volcanic eruptions. We saw a similar surprise with explosive eruptions along the Gakkel ridge at depths of 4,000 m.
The pumice started hitting the shores in March 2014. Estimates are that the eruption produced a cubic kilometer of pumice for rafting.
A 2015 expedition from University of Tasmania worked with Woods Hole Oceanographic Institute for a two-year study of the eruption and its aftermath. The eruption itself was complex. Recent submersible surveys discovered lava from 14 vents between 900 – 1,220 m deep. There are fragmented pumice clasts up to 9 m in diameter. 80% of the erupted volume ended up in the pumice raft and was transported far away from the volcano.
The original record of the eruption did not come close to capturing the scope of what actually happened. The eruption also covered an active biological community with pumice, devastating it. While every eruption is in some way deadly, they also create new places for life to spread. In this case, the eruption also created new hydrothermal systems whose surrounding biological communities are just getting started again less than three years after the eruption.
The eruption itself was initially thought to be exclusively explosive in nature, mostly due to the creation of the 400 km2 pumice raft over the course of a single day of activity. However, at the same time, the eruption also created ash, lava domes, and seafloor lava flows. Total volume of the eruption was in the neighborhood of 1.5 times the size of Mount St. Helens, at 1.5 km3. The composition of the erupted magma appears to have changed over time. This eruption is classed as a VEI 5, something that is thought to take place once or twice a century in the subaerial world.
Post eruption surveys found several new features on the volcano. These include a previously uncharted bulge on the western caldera wall and several new volcanic cones. The caldera floor was in places 50 m shallower than the initial survey in 2002.
The new cone was formed between 900 – 640 m (250 m high). It is 1.2 km in diameter and has a summit crater 120 m in diameter. The cone has four tongues of ejecta, each 50 m thick, extend 1.5 km from the summit. They are thought to be primary or resediment gravity currents of tephras.
The eruption formed a small cone field with eight new cones up to 300 m in diameter. These cones have bases below 940 m, and range from 60- 115 m in height.
A bulged area between 1,080 – 1,400 m was created on the inner wall of the caldera. It adds 150 m to the height of the caldera wall. The cause of the new feature is unknown, with possible causes including volcanic ejecta, extrusion of a dome, or intrusion of a cryptodome.
The dive team deployed an autonomous underwater vehicle in 11 dives that mapped over 50 km2 of seafloor. They used a remote underseas vehicle on an additional 12 dives totaling 250 hours inderwater to collect samples, imagery, and video. Some of that video is available at the linked videos.
The region of Havre Seamount is a spreading center between the Colville ridge and the Kermadec Ridge. The Colville Ridge is thought to be the original ridge with the Kermadec more recent. The area between the two trenches to the south is called the Harve Trough.
The Kermadec – Tonga plates were created 4 – 6 Ma. The Pacific Plate was subducting under the Australian Plate creating the Lau – Colville Ridge. That subduction stopped and crustal extension started some 6 Ma with several spreading centers created. The spreading centers created the Tonga and Kermadec Platelets. As the Tonga plate is spreading much faster in the north, it separated from the Kermadec Plate along a transform fault. A third microplate, the Niuafo’ou is NW of the Tonga Plate created by the faster extension in the Tonga Plate.
The Pacific Plate started subduction under the newly created platelets with the Tonga Trench to the north and the Kermadec Trench to the south. The Tonga Trench is the second deepest trench in the world, with the Horizon Deep at 10,800 m. The Kermadec Trench is not far behind at 10,000 m at its deepest. The Tonga Trench is also one of the fastest subduction zones in the world at 24 cm/yr.
One of the pieces of the Ontong – Java Large Igneous Province, the Hikurangi Plateau has starting to subduct under northern New Zealand along the southern portion of the Kermadec volcanic arc, which may provide large volumes of magma for volcanic activity in Northern New Zealand and the southern portion of the Kermadec arc. Finally, the Louisville Seamount chain is subducting under the Havre Plate. The area of subduction marks the thickest portion of both the Colville Ridge and the Kermadec Ridges.
The entire Kermadec – Tonga volcanic arc is one of the most prolific producers of pumice rafts over the last couple of centuries and has one of the highest concentrations of active seamounts in the world. Havre Seamount sits perhaps a third of the way south from line of impact of the Louisville Seamount Chain.
The Kermadec volcanic arc is home to an incredible array of active, explosive, undersea volcanic eruptions. Together with the Tongas, the arc is responsible for most of the pumice rafts created over the last couple centuries.
This area is intensely active, with calderas, lateral blasts, and other forms of explosive volcanic structures. Most of this volcanism is driven by subduction of the Louisville seamount chain to the north and the Hikurangu Plateau under northern New Zealand to the south. There is more than enough eruptible magma to power a vigorous volcanic arc for the foreseeable future.