Geochemistry, Mineralogy, Petrology & Volcanology

Imaggeo on Mondays: A sunrise over Kelimutu’s three-colour lakes

Imaggeo on Mondays: A sunrise over Kelimutu’s three-colour lakes

Volcanoes are undeniably home to some of the most beautiful landscapes on Earth. It doesn’t take much imagination to picture slopes of exceedingly fertile mineral rich soils, covered in lush vegetation; high peaks punching through cloud cover offering stunning vistas and bubbling pools of geothermally warmed waters were one can soak ones worries away.

What about strikingly coloured crater lakes? You’ll have to travel to Kelimutu volcano, on the Indonesia island of Flores, to catch a glimpse of those.  But the journey is guaranteed to be worth it. Picture three deep pools of water, at times turquoise blue; at others emerald green and even blood red!

The andesitic to basaltic (this simply means that the rocks which form the volcano are depleted in silica, sodium and potassium bearing minerals – compared to other types of igneous rocks that is – and you’ll predominantly find pyroxene, plagioclase and hornblende in them) volcano is capped by the three colourful lakes, formed as a result of a powerful ancient volcanic eruption.

In stratovolcaones (those which are cone shaped) the intensity of an eruption(s) can be so great that once all the magma, ash and rock in a caldera is erupted the edifice can no longer hold itself up and collapses in on itself, in a process known as a caldera collapse. When this happens, it is not uncommon for the crater left behind to gradually fill with water, both from within the volcano and from precipitation and other external sources.

What is unusual about the Kelimutu lakes is that they are very striking in colour, and even more remarkably, their colour changes over time! It is of great interest to geologists since it is rare that these lakes can have different colours even though they are from the same volcano and are located side by side at the same crest.

According to Indonesian folklore, these lakes are the resting places of the ancestors of the Indonesian people.

  • Tiwu Nuwa Muri Koo Fai (Lake of Young Men and Women) – This lake is turquoise.
  • Tiwu Ata Polo  (Bewitched Lake) – Home to those who have been evil in life. This lake is usually red or brown
  • Tiwu Ata Mbupu ( Lake of Old people) – This lake is usually blue/green

The reason for the changing colour of the waters is hotly debated. Some argue that it is fumaroles beneath the lakes which emit volcanic gases like sulphur dioxide, which are to blame. The fumaroles create upwelling within the lakes, forcing denser mineral rich water from the bottom of the lakes upwards and this interaction causes the visible colour changes in the lake. Others argue that it is the changing levels in the oxygenation, as a result of the injection of volcanic gases, of the waters which drives the colour fluctuations.

While the mystery is resolved, all that is left is to visit the enigmatic lakes, as Danielle Su (author of today’s imaggeo on Mondays image and researcher at the University of Western Australia) did. Danielle’s research typically deals with upwelling around oceanic islands in the Indian Ocean so it was exciting to see the parallels of the upwelling mechanism replicated within these volcanic lakes.

‘Upwelling generates high primary productivity in the ocean by bringing deep nutrient rich water to the surface and can be identified in remotely sensed data by the colour of the phytoplankton chlorophyll-a signatures. Although the source and output is different, the physics is similar and I really enjoyed finding this similarity in such different environments,’ describes Danielle.

The morning hike requires some commitment but the view from the peak makes it all worthwhile as the first rays of sunlight casts a glow over the volcano’s summit lakes.

‘When you see something so beautiful in nature, the questions take a backseat for a while because deconstructing it seems to diminish it temporarily. But when you do go back to the science to understand the process, admiration then changes to appreciation, an appreciation of how the complexity of the natural world constantly challenges our curiosity.’

Imaggeo is the EGU’s online open access geosciences image repository. All geoscientists (and others) can submit their photographs and videos to this repository and, since it is open access, these images can be used for free by scientists for their presentations or publications, by educators and the general public, and some images can even be used freely for commercial purposes. Photographers also retain full rights of use, as Imaggeo images are licensed and distributed by the EGU under a Creative Commons licence. Submit your photos at

Volcanic darkness marked the dawn of the Dark Ages

Volcanic darkness marked the dawn of the Dark Ages

The dawn of the Dark Ages coincided with a volcanic double event – two large eruptions in quick succession. Combined, they had a stronger impact on the Earth’s climate than any other volcanic event – or sequence of events – in the last 1200 years. Historical reports reveal that a mysterious dust cloud dimmed the sun’s rays between in 536 and 537 CE, a time followed by global societal decline. Now, we know the cause.

By combining state-or-the-art ice core measurements with historical records and a climate model, researchers from GEOMAR Helmholtz Centre for Ocean Research, Germany, and a host of international organisations showed that the eruptions were responsible for a rapid climatic downturn. The findings, published in Climatic Change, were presented at the EGU General Assembly in April 2016.

Explosive volcanic eruptions typically emit large volumes of ash and gas high into the atmosphere. The way this ash spreads depends both on how high up it’s propelled and the prevailing weather conditions. When it reaches the stratosphere, it has the capacity to spread far and wide over the Earth, meaning the eruption will have much more than a local impact.

Individually, these events were strong, but not that strong. Their combined force was what made their affect of the earth’s climate so significant. They occurred closely in time and were both in the Northern hemisphere.

Volcanic emissions reflect light back into space. Consequently, less light and, importantly, less heat reaches the surface, causing the Earth to cool. Diminishing sunlight following the eruptions resulted in a 2 °C drop in temperature, poor crop yields and population starvation. The drop in temperature led to a 3-5 year decline in Scandinavian agricultural productivity – a serious problem.

This double event had a major impact on agriculture in the northern hemisphere – particularly over Scandinavia. It’s likely that societies could withstand one bad summer, but several would have been a problem.

An ash covered plant via Wikimedia Commons.

An ash covered plant via Wikimedia Commons.

There’s agricultural evidence to support the theory too. Pollen records read from sediment cores can be used to work out when agricultural crops covered the land and when the land was ruled by nature. Scandinavian cores suggest there was a shift from agricultural crops to forest around the time of the eruption. There is some scepticism regarding the cause of this shift, but the implication is that when food decreases, so does the population, This means there’s no need to farm as much land, nor enough people to do so. In the absence of agriculture, nature takes over and trees once again cover the land.

By Sara Mynott, EGU Press Assistant and PhD candidate at the University of Exeter.

Sara is a science writer and marine science PhD candidate from the University of Exeter. She’s investigating the impact of climate change on predator-prey relationships in the ocean, and was one of our Press Assistants this year’s General Assembly.

Great walls of fire – Vitrification and thermal engineering in the British Iron Age

It’s long been recognised the peoples of European prehistory occasionally, and quite deliberately, melted the rocks from which their hilltop enclosures were made. But why did they do it? In today’s blog post Fabian Wadsworth and Rebecca Hearne explore this question.

Burning questions

Throughout the European Bronze and Iron Ages (spanning 2600 years from 3200 BC to 600 BC), people constructed stone-built, hilltop enclosures. In some cases, these stone walls were burned at high temperatures sufficient to partially melt them. These once-molten forts are called vitrified forts because today they preserve large amounts of glassy rock. First described in full in 1777, the origins and functions of these enigmatic features have been the subjects of centuries of debate.

Today, researchers generally agree that the glassy wall rocks are the result of in situ exposure to high temperatures in prehistory, similar in magnitude to those temperatures found in volcanoes on Earth. This was sufficient to partially or wholly melt the stonework, and the resulting melts are preserved as glass upon cooling.

There are still many outstanding questions concerning vitrified enclosures and forts, but the most immediate and arresting are: how and why were they burned?

Vitrified fort walls are mostly found in Scotland and are built from a diverse range of rock types

Vitrified enclosures occur throughout Europe but the best known examples are found in Scotland. Using the compilation created by Sanderson and co-workers (link provided below) we can map the distribution of forts, categorized by the rock-type from which they were built, and compare this with a simplified geological map of Scotland (from the British Geological Survey; Image 1). This shows that the building stone used in fort walls is not always the same and was more likely to be found locally.

Map of Scotland with simplified basement geology and cover-sediments marked. Vitrified fort positions are numbered such that 1- Finavon, 2- Craig Marloch Wood, 3- Tap O’North, 4- Dun Deardail, 5- Dunagoil, 6- Craig Phaidrig, 7- Laws of Monifieth, 8- Knockfarrell, 9- Dunskeig, 10-Dumbarton Rock, 11- Carradale, 12-Dun MacUisnichan, 13- Art Dun, 14- Mullach, 15- Trudernish Point, 16-Cumbrae, 17- Dun Lagaidh, 18- Sheep Hill, 19-Urquhart Castle, 20- Eilan-nan-Gobhar, 21- Eilan nan Ghoil, 22- Duntroon, 23- Torr Duin, 24- Trusty’s Hill, 25- Doon of May, 26- Castle Finlay, 27- Mote of Mark. (From the British Geological Survey).

Map of Scotland with simplified basement geology and cover-sediments marked. Vitrified fort positions are numbered such that 1- Finavon, 2- Craig Marloch Wood, 3- Tap O’North, 4- Dun Deardail, 5- Dunagoil, 6- Craig Phaidrig, 7- Laws of Monifieth, 8- Knockfarrell, 9- Dunskeig, 10-Dumbarton Rock, 11- Carradale, 12-Dun MacUisnichan, 13- Art Dun, 14- Mullach, 15- Trudernish Point, 16-Cumbrae, 17- Dun Lagaidh, 18- Sheep Hill, 19-Urquhart Castle, 20- Eilan-nan-Gobhar, 21- Eilan nan Ghoil, 22- Duntroon, 23- Torr Duin, 24- Trusty’s Hill, 25- Doon of May, 26- Castle Finlay, 27- Mote of Mark. (From the British Geological Survey).

How hot? How long?

A key question surrounding the melting and glass-formation processes that occur to form vitrified fort walls is what temperature was required and how long must the fires have burned? As we know, the required minimum temperature for melting rocks is the solidus, which varies by hundreds of degrees from rock-type to rock-type. Above the solidus, the partial melt fraction will increase until the material liquidus, above which all components of the rock are molten. In mineralogically diverse rocks, the partial melt fraction between the solidus and liquidus not only increases, but changes composition. Pioneering investigations of vitrified fort wall materials (for example, by Youngblood and co-workers) and others have explored this by looking at the composition of the glass that is formed when the partially molten fort walls are cooled. The composition of these materials and an understanding of the thermodynamics of the melting process yields temperatures of the prehistoric fires in question. Youngblood and colleagues found that temperatures were likely to be, on average, 900-1150 ºC.

In a recently published study, we used a different technique in an attempt to answer the same question. Samples of fort walls from Wincobank vitrified hillfort, Sheffield, UK, were used in high-temperature tests to measure the melting process in situ. We measured the amount of each mineral phase as it decreased above the solidus. This technique allowed us to extract additional information that previous investigators have not been able to probe. As well as suggesting a temperature window within which melting occurs, we were able to find the timescale of burning required to achieve the degrees of partial melting seen within Wincobank’s vitrified enclosure wall. Wincobank was constructed from a local sandstone; we found that the quartz in this sandstone was steadily removed upon experimental heating, matching the final quartz content of the enclosure wall rocks at a temperature window of 1050-1250 ºC for burning events of more than 10 hours.

When taken together, such investigations of the conditions required to form glass in prehistoric enclosure walls can more reliably inform the debate about why the fires were set in the first place. If glass is consistently found around a fort’s circumference (as is often the case, particularly in Scotland), and its particular building stone type dictates a heating event requiring a duration of 10 hours or more at peak temperature, it seems unlikely to us that such walls were burned accidentally or during periods of conflict or events of warfare (see below). In that case, if the enclosure walls were burned deliberately by the occupants, the outstanding question remains: why?

Why set fire to a stone wall?

In any archaeological investigation, a key goal is to attempt to explore and extrapolate the beliefs, motives, and desires of people in antiquity from their material culture. For practitioners of any discipline this is no easy task, with subjective interpretation of evidence and difference of opinion often resulting in vibrant discussion!

Consequently, numerous possible prehistoric motives for burning a fort wall to the point of melting have been posited. First is the possibility that the fires were lit during enemy attack or some other act of violence (mentioned above). Second, there’s the possibility that the fires were a product of deconstruction of the fort walls at the end of occupancy. Third, the conflagration was part of a ritual or display of prestige. Finally, there is the possibility that the fires were set with the intention of strengthening the stonework during construction. Each of these explanations has received attention; however, the last option had, until recently, been dismissed by previous researchers as largely unlikely, once it was recognised that heating rocks in general weakens them by the proliferation of microcracks from thermal stresses.

We revisited the idea that fort walls could have been burned during the construction in order to strengthen them. In our latest work we explored this in a simple way by showing that while the blocks in a fort wall will get weaker during high temperature burning, the more fine-grained rubble interstitial to these blocks will get much stronger. This strengthening occurs simply because the fine grained materials between larger blocks can fuse together by sintering when they are partially molten. And indeed, it is so often reported that large blocks are surrounded by a glassy mass that is fused to them (we show this in Image 2 from Wincobank fort, Sheffield, U.K.). We pointed out that this is contrary to the conventional view that fort walls must be weakened by the fires.

A block from the Wincobank enclosure wall in Sheffield, UK. This piece shows the typical feature where fine grained glassy material is welded to the larger, less altered blocks. In detail, this demonstrates that thermal gradients resulting from heating blocks of different sizes play an important role in determining which blocks melt and weld, and which blocks do not. (Credit: Fabian Wadsworth)

A block from the Wincobank enclosure wall in Sheffield, UK. This piece shows the typical feature where fine grained glassy material is welded to the larger, less altered blocks. In detail, this demonstrates that thermal gradients resulting from heating blocks of different sizes play an important role in determining which blocks melt and weld, and which blocks do not. (Credit: Fabian Wadsworth)

The debate rages on

We acknowledge that the strengthening effect does not rule out other motives. Indeed, the strengthening may be incidental to the true motive for the wall burning. It is also important to take into account the fact that, in many of the known examples of vitrified enclosures, where dated, the burning event takes place, in some cases, many hundreds of years after the fort’s initial construction.

A little-discussed possibility which is gaining momentum is that some of these forts may not have been forts at all. The very term “fort” is loaded and implies inherent military purpose, which remains a hypothesis with little solid evidence to its claim. Rather, they may have been monuments that were built and burned as displays of power and prestige or in some ritual event. In periods of our history where only the stones upon which we can base our suppositions remain, it is difficult to differentiate between these possibilities.

These acknowledgements highlight not only that there are outstanding questions requiring future investigation, but that each fort is different and there need not be a common explanation for them all.

The debate continues and, although the evidence sheds new light on the possible truths, we still do not know why Iron Age peoples throughout Europe set fire to stone enclosures and stoked those fires to volcanic temperatures. Rocks melt and crystallize and re-melt in volcanoes frequently and as a matter of natural process. To combine our understanding of these rock-forming materials and Earth processes as they are melted in anthropogenic conflagrations is essential to understand these curiosities of our Iron Age.

By Fabian Wadsworth, PhD student Ludwig Maximilian University of Munich, and Rebecca Hearne, Department of Archaeology, University of Sheffield

The Wincobank hillfort is in Sheffield in South Yorkshire, U.K. If the stone walls of its  enclosure were deliberately burned   this feature potentially extends Sheffield’s heritage of high temperature expertise – exemplified by the once-prolific steel works of the city – much farther into the region’s past than hitherto imagined.

Further reading

Imaggeo on Mondays: Moving images – Photo Contest 2016

Since 2010, the European Geosciences Union (EGU) has been holding an annual photo competition and exhibit in association with its General Assembly and with Imaggeo – the EGU’s open access image repository.

In addition to the still photographs, imaggeo also accepts moving images – short videos – which are also a part of the annual photo contest. However, 20 or more images have to be submitted to the moving image competition for an award to be granted by the judges.

This year saw seven interesting, beautiful and informative moving images entered into the competition. Despite the entries not meeting the required number of submissions for the best moving image prize to be awarded, three were highly ranked by the photo contest judges. We showcase them in today’s imaggeo on Mondays post and hope they serves as inspiration to encourage you to take short clips for submission to the imaggeo database in the future!

Aerial footage of an explosion at Santiaguito volcano, Guatemala. Credit: Felix von Aulock (distributed via

During a flight over the Caliente dome of Santiaguito volcano to collect images for photogrammetry, this explosion happened. At this distance, you can clearly see the faults along which the explosion initiates, although the little unmanned aerial vehicle is shaken quite a bit by the blast.

Undulatus asperitus clouds over Disko Bay, West Greenland. Credit: Laurence Dyke(distributed via

Timelapse video of Undulatus asperitus clouds over Disko Bay, West Greenland. This rare formation appeared in mid-August at the tail end of a large storm system that brought strong winds and exceptional rainfall. The texture of the cloud base is caused by turbulence as the storm passed over the Greenland Ice Sheet. The status of Undulatus asperitus is currently being reviewed by the World Meteorological Organisation. If accepted, it will be the first new cloud type since 1951. Camera and settings: Sony PMW-EX1, interval recording mode, 1 fps, 1080p. Music: Tycho – A Walk.

Lahar front at Semeru volcano, Indonesia. Credit: Franck Lavigne (distributed via

Progression of the 19 January 2002 lahar front in the Curah Lengkong river, Semeru volcano, Indonesia. Channel is 25 m across. For further information, please contact me (



Get every new post on this blog delivered to your Inbox.

Join other followers: