It was a quintessential English village, with simple houses along streets and fenced cottage gardens of precisely 100 m2 each. This is what the heart of England, the Cotswolds, looked like. Perhaps 85% of the population was local; the others were distant immigrants. The people were farmers and workers in the local flour mill. But although typical, this village would have an exceptional legacy. It was the only English village ever to be destroyed by a volcanic eruption.
The location gives a clue. For this was the England of the colonies, and the immigrants were people who had left their home in England – not always voluntary – and traveled halfway across the world to start a new life. This was about as far from England as it was possible to get. The fact that their new homeland was already occupied was not seen as a problem, in the colonial era. The local population could provide the labour for the new arrivals.
Some saw a wider responsibility, and took it on themselves to develop the local population, the Maori, by teaching them civilized ways. Te Wairoa, on the North Island of New Zealand, was designed and build as a model village which would help educate the Maori – and would help them become more English. The intention may have been good (although this is disputed), but it was misguided. A culture grows over centuries; transposing and imposing a culture from one country to another never works. English houses and factories, and English agriculture (wheat) were a poor match to the Maori way of life. Over time, conflicts grew about land use, with different Maori tribes claiming ownership. During these years Te Wairoa was abandoned, but eventually an arrangement was reached between the warring tribes and people returned. But the factory and church became abandoned and the original English colonizers, with all their good intentions, were slowly leaving. A new way of life developed, but it was very different from what either side of the divide had envisaged. Nature finally stepped in and brought the culture war to a close.
The location of Te Wairoa had been well chosen, in a fertile valley in between steep mountain sides and in between two lakes. One of those lakes, Lake Rotomohana, contained a marvellous and world-famous attraction: the Pink and White Terraces. Once the land dispute had been settled and the area became safe to visit, people came from far and wide. And the tourists kept coming. They would travel to New Zealand just to see this. Quickly, tourism became a better source of income than English agriculture. The viewing experience was well organised. A sign at the entrance to Te Wairoa listed the exact charges for guiding. Every day during the season, some 20 people would visit the Terraces on a strict itinerary: evening entertainment in the village, travel by canoe to visit the White Terraces at 11 am, lunch with boiled crayfish at one of the hot springs, afternoon bathing at the Pink Terraces, and return to Te Wairoa where there would be concerts and dancing. The guiding was largely done by women. The local Maori became well off. This also brought a new group of English with commercial interests, and they owned the local businesses and hotels. By the end, some 120 Maori and 15 Europeans lived in the village. Alcoholism had become a problem.
The end came suddenly. In the previous days, one of the Maori guides had noted an increase in hydrothermal activity at the lake. Apart from this, there was no warning. Soon after midnight on June 10, 1886, an earthquake swarm started, which increased in strength. The eruption itself began around 2am with a “roar like a tornado”. Te Wairoa found itself next to the second largest eruption to hit New Zealand since the arrival of the Maori. When morning came, the village no longer existed.
The mountain itself was hidden from view from Te Wairoa, and the main eye witness accounts are from further away. They describe the onset: “There was no sign of a storm, the wind was steady and the stars shining; but right over the eastern end of Ruawahia was a high, thin column of black smoke with a spreading top looking not unlike an immense mushroom. All this smoke cloud was blazing with lightning, which scintillated through every part of it and, shooting out from its dark edges fringed them with vivid light. The total height of this column when first seen would have been about 5000 feet; and it stood quite straight as though not subject to the action of any wind….”
(The quotation is taken from Keam 2015. Sources for the post are listed at the end.) The observer saw flames rising 500 meter high coming from the top of Tarawera mountain. And shortly after, perhaps around 3am, a large explosion was seen some 10 kilometer southwest of the mountain; the cloud, first white, then black, rolled across the landscape. It rendered the remainder of the eruption invisible.
This large explosion had happened under Lake Rotomahana, the lake southeast of Te Wairoa. The cloud that it generated was not just ash and smoke. The bottom of the lake had blown up, and the cloud brought a rain of mud, stones and cinders which lasted for 3 hours. Te Wairoa was buried under a meter of mud; other places were hit even harder. There are desperate descriptions of people trying in vain to support the roofs of their homes. Some died in the collapses, others escaped into the streets enduring the rain of mud and stones.
There had been many more explosions, but they had been unobserved behind the cloud of mud. A 16-km long, straight rift had opened, running from the mountain through Lake Rotomahana and beyond. The explosions that occurred as the rising magma met the ground water thundered across the landscape. They were heard 130 km away on the North Island, and even people in Christ Church, on the South Island, noted the distant explosions. In Auckland, the explosions were first heard around 2:30am, and lasted until after 4am. At the coast the black cloud brought darkness; even ships at sea found themselves under a rain of ash.
When the eruption was over, around 6:30am and after only 4 hours, some 120 people had died. It had been the end of the tourist season and only three tourists had been staying at the local hotels. But there had been a Maori wedding that evening and guests had come for that. A few of the smaller Maori dwellings survived better than the large houses, but most of the 70 houses, the village hall, the hotels and the factory were destroyed. Survivors had been propping up the roofs against the weight of the mud. Other had fled the collapsing buildings. In the morning people started digging out the survivors. 11 inhabitants of Te Wairoa had died. But there had been several small settlements around the shore of Lake Rotomahana, and those had been completely destroyed. Here, only a single person survived.
Te Wairoa was never rebuild, although for a while it still attracted tourists who came to view the destruction. Te Wairoa is now mainly known as New Zealand’s best known archeological site.
A mountain split
The eruption had occurred along a straight and fairly narrow rift. It started at the northeast side of the summit of Tarawera, where afterwards there were two craters. From here the rift passed southwest through the extended summit where the three domes had split completely. The summit was covered in pyroclastic ejecta which in places reached 75 meters in thickness. After a short gap, the rift continued along the southwest slope down the mountain. Beyond that there was a further rift sections and craters, which terminated at the newly formed Rotomahana crater. After another gap, there more craters further southwest for another 3 kilometers, in the Waimangu section. The penultimate one was Waimanga (Black crater).
The summit rift showed 13 separate craters, and as many as 50 vents. The Rotomahana section had at least 4 craters, and the Waimangu section had a line of craters which still showed minor hydrothermal explosions years after the eruption had ended.
The Waimangu section remained hydrothermally active for even longer. Black crater surprised everyone when fourteen years after the eruption it suddenly developed a large geyser. The geyser is still listed as the largest known: at times it erupted 400 meters high. The geyser activity ended after 4 years. Nowadays it is known as Echo crater and although the geyser is gone, there is still a hot spring which erupted mud as recently as 2016.
The summit eruption was plinian and it was basaltic. That is an unusual combination. Basalt tends to erupt effusively, without large explosions, because it does not have much volatile content which can drive explosions. And whereas basaltic plinian eruptions are already rare, Tarawera is unique in doing it simultaneously along such a long rift with so many individual eruption locations. And even stranger, Tarawera is not a basaltic volcano. Previous eruptions here produced rhyolite, with only very minor basaltic components. It appears that often, the rhyolite eruptions were triggered by a small basaltic intrusion, but the eruptions came from the activated rhyolite magma chambers. But in this case, the basalt came up and bypassed any rhyolite magma that may have been present. A dike was emplaced underneath the summit (possibly two dikes of which only one reached the surface) and very rapidly progressed along the rift direction. At the lowest point along the rift, where Lake Rotomahana was located, it produced massive phreato-magmatic explosions (i.e. water-magma interaction). The Rotomahana crater left by these explosions was 2.5 kilometers across. Further down-rift the explosions were equally watery but smaller.
And the explosions at the summit should not be underestimated either. Here, no water was involved. The scoria that was ejected from the summit covered an area of 10,000 km2. The wind blew the ejecta north and northeast, towards Whakatane. The plinian eruption appears to have come from four of the 13 summit craters, with eruptions speeds of 300 m/s. The other craters erupted with less intensity, and build smaller scoria deposits within about 500 meters of the rift.
The structure and composition of the basaltic ejecta indicates that the magma had been relatively cool, at 1100C (this is the temperature in the dike underground – not the lava ejection temperature which is rarely known) and had been fairly shallow (1-2 kbar, or 2-3 km depth). The reason why there was only 1-2 hours of earthquake activity prior to the eruption was that the dike was already quite close to the surface. But there was no existing conduit, and instead the dike ripped open the entire summit. The northeast-southwest direction which the dike and rift followed is a common alignment at Tarawera. The sides of the rift show some sideways shift, perhaps caused by a combination of some extension and rotation. The stress field made it easy for the dike to push its way out and up along this path.
The ejecta at Rotomahana were different from those from the summit craters. This was the so-called Rotomahana mud. It too contained basalt, but this was only around 20% of the ejecta. The rest was pulverized ground. The soil and rock around Tarawera are from older, rhyolitic eruptions. This was taken up in the ejecta, which became a mix of fresh basalt and ancient, cold rhyolite. In fact the same mixture is seen at the summit but with a much higher percentage of basalt, of up to 85%.
The Rotomahana mud formed a dune field, extending up to 6 kilometers from the new crater. This indicates a base surge. The fact that the distance observer saw the explosion causing first a white and then a black cloud also points at this. Base surges are common in phreato-magmatic (shallow underground) explosions. The gas vents vertically out of the crater hole, flows over the crater edge and moves out horizontally over the ground, as an expanding ring around the explosion column. The surge cloud hugs the ground while moving out at speeds of typically 50 m/s (100 miles/hour, if you prefer these units). The surge carries the ash, mud and even stones, well beyond distances to which the rocks could be thrown by the eruption. Te Wairoa was hit by such a base surge.
The deposits change from the deep mud near the lake to a volcanic ‘flour’ further away. The two have the same composition, but the water-soaked material dropped out of the surge closer to the lake, with the dry ash traveling further. This is also the cause of the change from a white (vapour) to a black (dust) cloud. Layering of the ejecta shows that the surge dried out later in the eruption, with a thin layer of flour dropped on top of the thick mud. Whether this was because the water in the lake had all been used up, or that the mud ceased to be carried as far as the eruption declined, is an open question. The dry flour consists mainly of old rhyolitic deposits, remobilized by the explosion, with some 20% fresh basalt. The temperature of the dry flour may have been as high as 150C: trees in the region were scorched, showing the base surge was hot (unlike ash fall which is cold). But it was not as hot as a pyroclastic flow would have been, and perhaps the water in the mud kept the temperature down around the lake since none of the reports of the eruption mention heat.
The eruption was very destructive. In those four hours, some 1.3 to 2 km3 of pulverized rock, scoria and lava was ejected, making this a VEI 5 eruption. The numbers are based on several different and independent measurements of the thickness and extent of the deposits. About half the volume came from the summit eruption, and half from the Rotomahana explosion. But most of the lava came from the summit: Rotomahana contained less than 1/5th of the ejected basalt. This is an indication that the power of the Rotomahana explosion came from the interaction with the lake water. The basaltic lava itself had relatively little volatiles (less than 2% water content), and the plinian summit eruption instead was driven by the straight pressure on the dike. In contrast, the lake provided volatiles and turned what would otherwise have been a smaller explosion far along the rift into a borderline VEI-5. Phreato-magmatic eruptions are far more dangerous than the amount of lava involved may suggest. At Tarawera, all of the casualties came from the lake explosion.
The region has changed in the 135 year since the eruption. The volcanic grey, impassable wilderness has greened. The craters are filled with water, and formed new lakes with new outflows. Not all of the 1886 craters are still recognizable in the landscape. Rotomahana crater too has became a lake, different in shape to lake that went before. The level of the lake is much higher than it was before the eruption: the original outflow was blocked after the eruption. New thermal fields and hot springs have developed. There are now geysers along the western side of Lake Rotomahana. As the lake level rises and falls, different geysers become active; geyser activity is seen if the exit is no more than 3 meters above the level of the lake. When the lake drops, geysers higher up may still discharge warm water but do so without steam explosions, while new geysers appear lower down.
But this hydrothermal activity cannot undo the damage of the eruption. Te Wairou lost its village – and its biggest attraction.
The Pink and White Terraces
It was know as the 8th wonder of the world. This title may be queried. The original, ancient list had contained nine world wonders, leaving no vacancy at number 8. And there is competition for the title: many things have been called the 8th wonder. The strangest one, perhaps was the suggestion by Albert Einstein of compound interest: he reportedly said “He who understands it, earns it; he who doesn’t, pays it.” Even King Kong was nominated at one time. The Taj Mahal and the Grand Canyon are on the list. The Taj Mahal would qualify, as the original list was about marvels of engineering, not those of nature. But all visitors agreed that the Pink and White Terraces were a world-class marvel. Therefore, after the Rotohamana explosion, people quickly went to see the damage to them.
They could not even decide where the Terraces had been. Not enough of the landscape had survived for them to get their bearings. The Auckland Evening Star newspaper described the scene, shortly after the eruption: “.instead of a splendid sheet of water, there was opened out immediately beneath our feet, its edge not 250 yards distant, a huge crater, belching out showers of mud and stones from innumerable yawning mouths, amid dense volumes of steam and smoke, with a din and roar and rattle baffling description. Stones were being ejected high into the air from eleven separate orifices or small craters, on the side nearest to us, and the volumes of steam and smoke prevented further vision into the centre of the old lake site. A partial clearing away of the vaporous envelope, however, occasionally gave a brief glimpse into the gloomy recesses of the great crater revealing only a bed of steaming seething mud in flats and hillocks, bubbling and spouting in ceaseless ebullition. A small patch of discoloured water was dimly distinguishable in one part, but the lake was gone—not only the water, but the bottom driven out, scooping the bed to a depth of at least 250 feet below the old level…. The great crater was over a mile long and half a mile wide.”
And together with the lake, the Pink and White Terraces too were gone. People have been looking for the remnants ever since. The largest explosion happened very close to the White Terraces and it is unlikely any part of it survived. But the Pink Terraces were further away: are they perhaps just buried underneath the 10 or 15 meters of mud? The ejecta covering Te Wairoa and surrounding areas included many silica fragments which resembled the material of the Terraces, but we do not know whether they come from one or both Terraces. Every now and then a report comes out that remnants of the Terraces have been located deep under the lake. But these reports have not been confirmed and the claims do not agree with each other. The lake has risen by 30 meters or more and any remnant would now be under water. But this rise came slowly after the eruption, and the early searches had a dry view of any surviving areas. Even if they had been buried by deep Rotomahana mud, part of them could easily have been uncovered as the mud was washed away. The steps between Terraces could be several meters. It therefore seems likely that the Pink Terraces too were mostly or fully destroyed. Only one of the original ancient world wonders survives, so this puts the Terraces in good company. But there is always hope and perhaps one day the remnants will be uncovered.
Volcanoes continuously redevelop the land they occupy. What they destroy may well over time reform. But so far, the famous Terraces show no sign of being recreated.
The name Lake Rotomahana means ‘warm lake’ and this is already an indication of the hydrothermal activity in the region. The White Terraces, on the northeastern shore of the lake, were fed from the Te Tarata geyser 30 meters above the lake. (The same name was also used for the Terraces themselves.) The silica-rich water descended from here to the lake, depositing the Terraces in the process with some 50 large pools and many more small ones. The Pink Terraces were smaller and on the opposite shore, and came from a spring also located 30 meters high, along a steep valley. The Pink Terraces were not pink everywhere: they were marble white at the bottom, but turned pink and rose higher up; the pool at the top was cobalt blue. The upper pools of the Pink Terraces had the best temperature for bathing.
Similar cinter terraces exist elsewhere in the world, for instance at Mammoth Springs in Yellowstone and Pamukkale in Turkey. Those at Rotomahana were larger and (reportedly) more impressive. It was also unique to have two such wonders on the same lake.
Before the eruption, the lake was bordered on its eastern side by a 50-meter high ridge. This ridge may have been the wall of an older caldera. The Te Tarata geyser was located on the slope of this ridge. The sketch below is the only map in existence of the old lake. The Terraces are the larger yellow areas near the bottom left (the Pink Terraces) and the centre (the White Terraces). The Pink Terraces were fed by a spring with a temperature of 80C, while the White Terraces came from a boiling intermittent geyser. The other yellow areas were fed by cooler springs which had deposited silica on the valley floors but which had not developed the silica basins of the Terraces. Temperature matters.
And the temperature was impressive. The geologist Hochstetter once camped on the small Puqi island in Rotomahana. He was not comfortable: “The whole ground is . . . so warm from below that I started from my couch unable to bear it any longer”. He put a thermometer into the ground. When pulling it out, hot steam came from the hole.
It is notable that the various cool springs, hot springs and geysers were some 30 meters above the lake level. That is very different from the current situation where the new geysers are found below 3 meters above the lake level (albeit with a much higher lake level.) It is clear that the water supply for those pre-1886 springs and geysers did not come from the lake, but benefitted from much higher pressure than provided by the water. The geyser water had been in contact with hot rocks, which were likely to have been at some depth. Indeed, ground water can circulate to kilometers depth. Rising to the surface, it avoided the lake. Instead the water came up through the ridge next to the lake, perhaps following an old eruption conduit.
Plausibly, that the water got its pressure from Tarawera, and came down underground from the summit, following a flow channel at some depth. This underground direction which also brought it in contact with the warm (even hot) heart of the mountain. The dike of the 1886 eruption followed the same direction. When the magma came, it found the water already along the rift, ready to power the explosions.
This water flow must have been stable over a long time. Estimates for how long it may have taken to build up the Terraces range from 1000 to 10,000 years. And this happened while the lake level fluctuated and the volcano cycled through its phases of activity. The Terraces formed in a most unlikely location. Geyser fields normally indicate a deeply dormant volcano.
Where did the silica that formed the Terraces come from? The region was covered in old rhyolitic deposits, and rhyolite has a very high percentage of silica. Silica is slow to dissolve in water, however this becomes much faster when the water is very hot: the solubility peaks at around 300C, which obviously requires high pressure. As the water reaches the surface, it rapidly cools and the silica precipitates again, covering the valley floor in silica. This happens in the Hidden Valley in nearby Taupo, as shown in this video. The colours in the video are caused by microbes. In contrast, the colour of the Pink Terraces was intrinsic to the silica. The precise origin is not known, but trace elements in the silica such as iron are likely to blame. (Even gold has been suggested, perhaps a bit optimistic!)
To turn a silica-rich water flow into the Terraces requires a largely flat area where the water can be caught behind small barriers. The ponded water itself now creates a level surface where the silica is deposited on the floor of the pool. As it flows over the barrier, it also deposited here. The barrier grows higher and eventually the standing water becomes a proper pool and a Terrace. The dripping water on the outside of the step barrier can cause stalactite-like formations, which did indeed exist on the Pink and White Terraces. Building such structures takes time. A shorter-lived flow will only cause a white surface, as in the Hidden Valley.
Heat measurements have shown that the hydrothermal system beneath the Pink Terraces still exists. The hot water now mainly exits below the lake, but there are also geysers along the western shore. Gravity anomalies suggest there may be a basaltic dike hiding here underneath the old rhyolite ridge, a remnant of the 1886 eruption. But there is no heat signature at this location of the White Terraces, and its hydrothermal system was likely fully destroyed in the explosions. If the Terraces ever do re-form, it will not be in the same location. And probably, they never will. We are left with memories of a misguided model village and a world-class attraction.
Post and future
Tarawera has a dangerous side. Over many years it build up a world wonder, only to destroy it in just a few hours. This was never a safe place to build a model English village. The land disputes were settled once and for all, once the volcano showed what it was capable of. But the Maori could have known. The 1886 eruption was the second largest eruption since the New Zealand settlement. The largest eruption had happened some 500 year earlier. This early eruption is a key moment not just in the settlement of New Zealand, but even in the peopling of the Pacific. But that is another story.
Albert, November 2021
Te Wairoa, The Buried Village: A Summary of Recent Research and Excavations. Alexy Simmons, 1991. Australian Journal of Historical Archaeology, Vol. 9 (1991), pp. 56-62
The Tarawera eruption, Lake Rotomahana, and the origin of the Pink and White Terraces. Ronald Keam, 2015, Journal of Volcanology and Geothermal Research, 314, Pages 10-38
Tarawera 1886: an integrated review of volcanological and geochemical characteristics of a complex basaltic eruption. Michael C. Rowe, Rebecca J. Carey, James D. L. White, et al., 2021. New Zealand Journal of Geology and Geophysics, 64:2-3, 296-319,