David Pyle

David Pyle is a volcanologist, and Professor of Earth Sciences at the University of Oxford. His first encounter with volcanoes was at the age of 7, when he visited Villarrica, Chile, shortly after an eruption. David studied geological sciences at the University of Cambridge, and later completed a PhD on the 'older' eruptions of Santorini, Greece. After a short post-doc at the California Institute of Technology, David returned to a lectureship in Cambridge. In 2006, he moved to his current post in Oxford. David tweets at @davidmpyle

The Botanic Gardens of St Vincent and the Grenadines

The Botanic Gardens of St Vincent and the Grenadines

The oldest Botanical Gardens in the western hemisphere lie on the outskirts of Kingstown, St Vincent, in the Windward isles of the West Indies – and what a gem they are. As the ironwork above the entrance declares, the gardens were founded in 1765.

The entry to the St Vincent Botanic Garden; the oldest in the western hemisphere

The entry to the St Vincent Botanic Gardens; the oldest in the western hemisphere

The original ambition of Robert Melville, the then Governor in Chief of the Windward Isles, was to establish a horticultural research station for ‘the cultivation and improvement of many plants now growing wild and the importation of others from similar climates” which would ” be of great utility to the public and vastly improve the resources of the island”. The early years of the Botanical Garden saw the rapid establishment of cinnamon, nutmeg and mango trees, among others,  leading to awards from the newly formed Royal Society for the encouragement of Arts, Manufactures and Commerce, which was offering a number of medals and monetary prizes for the promotion of agriculture in the then colonies of the British Empire.

Central walk 1824

A view of the Central Walk of the Botanic Gardens from an 1824 sketch by Lansdown Guilding; image from the Howard’s history of the Botanic Garden (Geographical Review, 1954)

The Botanic Gardens flourished under the direction of its first two superintendents,  George Young (also Medical Officer for St Vincent) and later Alexander Anderson, a surgeon and botanist. Anderson was, among other things, responsible for the establishment of breadfruit on St Vincent (delivered by Captain William Bligh, formerly of HMS Bounty, and later HMS Providence – both ships designed for botanical missions). He also wrote an early account of a visit to the volcanic crater of Morne Garou (now called the Soufrière of St Vincent).

central walk

A view of a part of the Central Walk of the Botanic Gardens at the present day.

After 1819, the gardens fell into decline and were not reinstated until 1890.  In the latter years of the nineteenth century, the gardens were re-established under the curatorship of Henry Powell, to include a botanical station for experimenting on the development of new crops and new agricultural techniques, with experimental plots distributed around the island of St Vincent. By this stage, the production of two economically-significant crops of sugar and arrowroot were already in decline, and considerable efforts were made to develop the potential for Sea Island cotton growth on the islands. Scientific results from these experiments were reported in great detail annually, so there is a wonderful archive record of the subsequent impact of the devastating eruptions of the Soufrière of St Vincent in 1902-1903 on agriculture across St Vincent. The experiments on cotton growth also meant that the agriculturalists of the day were well prepared to diversify into this new crop following the eruption.

The Botanic Gardens remain under the jurisdiction of the Ministry of Agriculture of St Vincent and the Grenadines, and the legacy of its (nearly) 250 year history can be appreciated in the diversity of the mature trees and plants that can be enjoyed every day by visitors to the gardens.

Flower of the Cannonball tree, Couroupita guianensis

Flower of the Cannonball tree, Couroupita guianensis

Red ginger flower

Flower of the Red Ginger Plant, Alpinia purpurata


Fruit of the nutmeg tree

Lizard on the lignum vitae tree, Guaiacum officinale

Lizard on the lignum vitae tree, Guaiacum officinale

Cycad, St Vincent

Cycad plant, Botanic garden, St Vincent

Further reading

Anderson, J. (1785). An account of Morne Garou, a mountain in the island of St Vincent with a description of the volcano on its summit in a letter from James Anderson (surgeon) to Mr. Forsyth His Majesty’s Gardener at Kensington, communicated by the Right Honourable Sir George Yonge, Bart. F. R. S. Philosophical Transactions of the Royal Society 1785, 75, doi: 10.1098/rstl.1785.0003.  As noted in the Dictionary of National Biography, the author’s name was clearly incorrect in the published letter. 

Anderson, T., Flett, J., McDonald, (1903) Report on the Eruptions of the Soufriere, in St. Vincent, in 1902, and on a Visit to Montagne Pelee, in Martinique. Part I. Philosophical Transactions of the Royal Society of London. A200, 353-553

Howard, R.A. (1954) A history of the Botanic Garden of St Vincent, British West Indies. Geographical Review 44, 381-393.


I visited the Botanic Gardens during a STREVA project workshop on St Vincent; more about this in a later post. Many thanks to Errol for his masterful guidance and tour.

Remote sensing of volcanoes and volcanic processes


The spectacular front cover of the Geological Society of London Special Publication 380 – with many thanks to Elspeth Robertson and ESA for this SPOT5 image of Longonot volcano, Kenya.

A major goal of volcanological science is understand the processes that underlie volcanic activity, and to use this understanding to help to reduce volcanic risk. Advances in instruments and techniques mean that scientists can now measure many different aspects of the behaviour of  restless or active volcanoes, including seismicity (to detect magma movement at depth, for example); deformation (often reflecting pressure changes at depth); and emissions of heat and gas.  With the exception of seismicity, which requires sensitive instruments placed close to the volcano, many of these measurements can now be made remotely using instruments on board satellites or aircraft.

Remote-sensing techniques have transformed our capacity to detect, monitor and measure volcanic activity worldwide. In the past 35 years, applications have moved from the first satellite remote-sensing observations of the rise and spread of a volcanic plume and volcanic gases from an explosive volcanic eruption (the April 1979 eruption of the Soufriere of St Vincent in the Caribbean); to the current situation where constellations of satellites are used to provide routine monitoring of volcanic gas emissions, volcanic hotspots and volcano deformation. As well as dramatically improving our ability to monitor the progress of volcanic eruptions, these techniques also help us to understand better how volcanoes work by providing long-term data on what happens at volcanoes when they are not erupting; and by making it possible to compare the behaviour of different volcanoes in ways that are simply not possible from the ground, or with ground-based observations.

In a new Geological Society of London Special Publication, we have brought together a selection of papers that give a broad perspective of the current state of the art in the remote sensing of volcanoes and volcanic processes. The 14 papers in the volume focus on the observation, modelling and interpretation of satellite-remote sensing of volcanoes: from surface deformation, to thermal anomalies, gas fluxes and eruptive plumes. Many of the papers take a broad perspective, reviewing current techniques and applications, or demonstrating the potential to investigate volcano behaviour and volcanic activity at regional to global scales.  Papers in the volume also show the ways in which people are now trying to go from these observations to a deeper understanding of underlying processes, by integrating observations with theoretical models and computer simulations of volcano behaviour; and then to use these insights to advance the potential for eruption forecasting. We hope that this Special Publication will find a wide and appreciative audience out there!

An illustration of some of the applications of remote-sensing techniques to a volcano during a hypothetical eruption cycle.

An illustration of some of the applications of remote-sensing techniques to a volcano through a hypothetical eruption cycle, across wavelengths ranging from the Infrared (I.R.), through the Visible (Vis.) and Ultraviolet (U.V.), to radar (ca. 2.5 – 30 cm in volcanic applications). Earth Observation (EO) techniques now allow the detection and analysis of a spectrum of different aspects of volcano behaviour at both non-erupting and  erupting volcanoes. From the introduction to the Geological Society Special Publication 380 (Pyle et al., 2013).  The seismic event rate trace is schematic, but based on observations at Mt St Helens in March – May 1980. 


Editing a volume of this scale requires a lot of support from a lot of people. On behalf  of the editors (Tamsin Mather, Juliet Biggs and myself), we would like to thank the authors of all of the contributions for their hard work and for entrusting their manuscripts with us; we would like to thank the very many reviewers who selflessly gave up their time to provide the feedback and constructive criticism of the papers that is the key part of the peer-review process; and we would also like to thank the staff of the Geological Society’s Publishing House, and in particular Angharad Hills, Tamzin Anderson and Hannah Sime, who shepherded this v0lume from start to finish, and who have turned our initial idea into such a wonderful physical volume. Finally, and on behalf of all of the authors, we would like to acknowledge the many individuals, institutions and agencies who have provided the facilities, funding, imagery and datasets which have underpinned all of this work.


Pyle DM, Mather TA, Biggs J (eds) 2013. Remote sensing of Volcanoes and Volcanic Processes: Integrating Observation and Modelling. Geological Society, London, Special Publications, 380. ISBN 978-1-86239-362-2.

The volume is available to subscribers through the Geological Society’s Lyell Collection, and can be purchased via the Geological Society’s online Bookshop.

Abbreviated contents list (the full list is available via the Lyell collection):

Pyle, DM et al. – Remote sensing of volcanoes and volcanic processes: integrating observation and modelling – introduction (Free content)

Ebmeier, SK et al. – Applicability of InSAR to tropical volcanoes

Wauthier, C et al. – Nyamulagira’s magma plumbing system inferred from 15 years of InSAR

Aoki, Y et al. – Magma pathway and its structural controls at Asama volcano, Japan

Segall, P – Volcano deformation and eruption forecasting

Blackett, M – Review of the utility of infrared remote sensing for detecting and monitoring volcanic activity

Zaksek, K et al. – Constraining the uncertainties of volcano thermal anomaly monitoring using a Kalman filter technique

Jay, JA et al. – Volcanic hotspots of the central and southern Andes as seen from space by ASTER and MODVOLC, 2000 – 2010

van Manen, S et al. – Forecasting large explosions at Bezymianny volcano using thermal satellite data.

Hutchison, W et al. – Airborne thermal remote sensing of the Volcan de Colima lava dome from 2007-2010

Carn, SA et al. – Measuring global volcanic degassing with the Ozone Monitoring Instrument

McCormick, BT et al. – Volcano monitoring applications of the Ozone Monitoring Instrument

Grainger, RG et al. – Measuring volcanic plume and ash properties from space

Pieri, D et al. – In situ observations and sampling of volcanic emissions with NASA and UCR unmanned aircraft

Friday Field Photo – Soufrière Hills Volcano, Montserrat in 1998

The lava dome of the Soufrière Hills Volcano, Montserrat, February 1998

The lava dome of the Soufrière Hills Volcano, Montserrat, February 1998, viewed from Perches Mountain.

View of the steaming dome of the Soufrière Hills Volcano (SHV), Montserrat, in February 1998, just at the beginning of the first pause in the eruption which began in 1995. Since that time, the volcano has gone through another 4 cycles of slow lava extrusion,along with a number of major episodes of dome collapse. The volcano remains active, and closely monitored by the Montserrat Volcano Observatory. Montserrat is a major focus of an ongoing NERC and ESRC-funded research project aiming to Strengthen Resilience in Volcanic Areas (STREVA), and SHV is one of the highlights of Volcanoes Top Trumps!

Volcanoes Under the Ice

The ice-filled summit crater of volcan Sollipulli, Chile.

The ice-filled summit crater of volcan Sollipulli, Chile.

A fascinating story has emerged this week from a paper in Nature Geoscience by Amanda Lough and co-workers (Lough et al., 2013), on the discovery of a new volcano deep beneath the ice of the Western Antarctic Ice Sheet (WAIS).  The discovery is partly a story of scientists looking in a place where no-one had looked before; this case, using a network of seismometers, as a part of POLENET/ANET – a polar earth observing network. The instruments were laid out across a section of Western Antarctica, which is known to contain some large and relatively young volcanoes. The surprise from the data was the discovery of a deep source of seismic activity (earthquakes), in a location that didn’t match any of the existing volcanoes. The authors were able to show that the earthquake signals look like those from volcanoes, and not like those made from breaking rock or creaking ice; and they were also clustered into small swarms, both in time and space. All of this evidence points to the idea that these earthquakes relate to the movement of magma deep in the crust beneath a small volcano that was not previously known, and which hasn’t yet emerged above the ice. This is fascinating ‘discovery’ science, and poses all sorts of interesting questions about how and why the volcanoes in this part of Antarctica exist in the first place – a topic where little is known, but where there is a very timely recent paper by Wesley LeMasurier.  Of course, much of the press coverage has focussed on speculation about whether, or not, there might be calamitious effects for the West Antarctic Ice sheet in the case of an eruption (The short answer is no!).

One broader reason for my interest in the paper is that it links to the long-standing question of what happens when large volumes of ice are removed from volcanoes? On Earth, the main example of this relates to the end of the last glaciation, which began between about 18,000 and 20,000 years ago. The current hypothesis, which has been around for some time, goes like this. During the peak of glaciation, vast areas of the land surface can be buried under great thicknesses of ice – perhaps as much as hundreds of metres to kilometres. Then, during deglaciation, the ice retreats rapidly reducing the pressure on the rocks beneath. So far so good, but how could taking away some ice have any effect on volcanoes? One clue to the answer may come from Iceland. This is a part of the world where the mantle is hot enough at shallow enough depths, that there is a column of partially-molten rock that probably extends from about 40 km depth down to about 100 km depth. Here, where the rock is above its melting point, Charlie Langmuir and Dan McKenzie have shown that that amount of melt increases by about 1% for a pressure drop of 100 MPa. So removal of a 1000 m thick column of ice (at an Icelandic volcano) will translate into a pressure drop of 10 MPa, and an increase of melt fraction of 0.1% along the length of the 60 km melt column. This is to a change in the amount of melt produced of 60 m. This might not seem like much – but when you consider the full size of the melting region in the mantle, this could correspond to a considerable volume of melt produced simply as a result of ice removal. In Iceland, there is geological evidence for a pulse of eruptive volcanism that occurred very shortly after deglaciation – around 11,000 years ago. 

Moving back to Antarctica, the current idea is that the volcanoes beneath West Antarctica might also be fed from a plume, similar to that beneath Iceland and the Canary islands. So perhaps it is plausible that a similar explanation might also apply in Antarctica, and there might be a feedback on geological timescales between large-scale mass changes in the ice sheet, and the magma supply from depth? What is not clear, yet, is whether this same mechanism applies to the volcanoes that form above subduction zones. Here the melting process is a little different, driven primarily by the arrival of water into the mantle. So it is not clear how pressure changes due to changing ice load might lead to changes in melt production deeper down. It is also not yet clear, from the geological evidence, whether or not there were changes in rates of eruptions at the subduction zone volcanoes that were once covered with ice. This remains a work in progress!


Lough, AC et al., 2013, Seismic detection of an active subglacial magmatic complex in Marie Byrd Land, Antarctica,  Nature Geoscience (Advanced Online Publication, 2013/11/17/) http://dx.doi.org/10.1038/ngeo1992

Watt, SFL et al., 2013, The volcanic response to deglaciation: Evidence from glaciated arcs and a reassessment of global eruption records, Earth-Science Reviews, 122, July 2013, Pages 77-102, http://dx.doi.org/10.1016/j.earscirev.2013.03.007

Friday Field Photos: Eruptions at Lokon-Empung volcano, Indonesia

Friday Field Photos: Eruptions at Lokon-Empung volcano, Indonesia

This week I am at a workshop near the twin-peaked volcano Lokon-Empung, in Sulawesi, Indonesia. True to form (it is the most active volcano in Sulawesi), Lokon has been rather active, with fairly frequent small explosions forming some small but dramatic ash plumes. The active vent is not at the summits of either Lokon, or Empung, but instead at the crater called Tompualan, which lies in the saddle between the two peaks. Lokon-Empung had been quiet since July 2013, but re-awoke with a large explosion on the morning of September 9th. Below, I show a series of photos of explosions on September 10th and 12th, taken from Tomohon, about 5 km away. These explosions are moderately energetic, but short-lived – lasting at the most just a few minutes. Volcanoes like Lokon may have hundreds of explosions of this scale every year during their active phases.

Lokon Empung volcano.

View of Lokon (left peak) and Empung (right peak). The active crater of Tompualan lies in between the two. 10 September , 6.01 am local time.


Early morning ash plume, from an explosion in Tompualan crater, Lokon-Empung volcano. September 10th, 6.19 am local time. The explosion began at 6.15 am. The volcano observatory estimated the peak plume height at 1 km above the vent.


10 September, 6.35 am local time. The short explosion is now over, and the ash plume is dissipating.


Larger explosion at Tompualan crater on 10 September at 6.59 am local time. The explosion started shortly before this point, and the volcano observatory estimated the eventual plume height to be 2.4 km.


Tompualan crater, Mt. Lokon-Empung, 10 September, 7.00 am local time. The plume continues to grow.


The plume from Lokon-Empuang is now reaching its peak height, as it reaches the cloud base. 10 September, 7.01 am local time.


Mt Lokon-Empung, 10 September, 7.10 am. The main explosion has now ceased, and the plume is beginning to decay as ash particles fall out, and the plume is dispersed by winds.


Mt Lokon-Empung, 12 September. Fume from the Tompualan crater is barely visible against the background meteorological cloud.


The early stages of a small explosion from Lokon-Empung, 12 September, 12.37 local time.


Continued expansion of the ash plume at Lokon-Empung, 12 September, 12.38 local time.


The ash plume continuing to grow, September 12, 12.38 local time.


The explosion has ceased, and the ash plume is beginning to disperse. 12 September, 12.40 local time.

Continued ash venting, 13 September, 7.17 am local time.

Further reading

The Volcanological Survey of Indonesia has a website that documents the current state of activity at its many volcanoes.  Many of the local observatories have webcams – such as the one at Kakaskasen Volcano Observatory (KKVO) near Lokon.

Two recent blogs have documented the context of this volcanism, and the ongoing activity at Lokon:

Culture Volcan – journal d’un volcanophile.  Culture Volcan has also posted a nice time-lapse video of the 12 September activity on YouTube, based on the webcam from KKVO.

Joe Bauwen’s Sciency Thoughts blog

One year of volcanicdegassing

One year of volcanicdegassing

One year has passed since I first wrote a post for this occasional blog. Now, 12 months, 22 posts and 7500 page views later, here’s a quick look back. For me, this has been a way of using some of my back catalogue of field photographs, of fleshing out a bit of context around papers I have been working on, and adding a little commentary on some more topical aspects. I am pleased with the results so far, and will aim to keep it going for a little longer. In the meantime, thank you for reading and sharing the posts!

Sunset behind Mt Lokon, Sulawesi, Indonesia

Sunset behind Mt Lokon, Sulawesi, Indonesia

Top three posts of the past year.
A tribute to Barry Dawson
Sea-floor spreading: magmatism in the Afar, Ethiopia
An episode of volcanic unrest beneath Santorini, Greece

Friday Field Photos: the Southern Volcanic Zone of Chile

Friday Field Photos: the Southern Volcanic Zone of Chile

If you are ever in Chile and have the chance to take a mid-morning flight south from Santiago towards Puerto Montt or Concepcion, make sure you try and book a window seat on the left hand side of the plane.  Once the early morning cloud has cleared, you could be in for a treat as you fly along the ‘volcanic front’, with spectacular views of Chile’s brooding volcanoes popping up from the landscape. Be sure to take a map, too, so that you can work out which one is which. The pictures below are roughly in order, flying from north to south – and several major volcanoes of the chain aren’t included.

There are several things to notice about these volcanoes – they are often in pairs, either as distinct but closely spaced mountains (Tolhauca and Lonquimay), or as ‘twin peaks’ forming the summit of an elongated massif (e.g. Llaima, Mocho Choshuenco). Many of the volcanoes are also clearly very young structures – forming wonderfully characteristic conical shapes (e.g. Antuco, Villarrica, Osorno). These cones must be younger than 15 – 20,000 years (and perhaps much younger than this), based on what we know about when the last major glaciation in the region ended. These cones sit on top of the lower-relief and older parts of the volcanoes, many of which have been reshaped by caldera-collapse, perhaps shortly after the ice retreated during deglaciation. The accessibility of the volcanoes of the Southern Volcanic Zone of the Andes makes this a wonderful place to study volcanic processes and volcano behaviour, both at the scale of individual eruptions, as well on the regional scale.

The river Cachapoal runs out of the Andes mountains, past the city of Rancagua

The river Cachapoal runs out of the Andes mountains, past the city of Rancagua

The saddle-shaped volcanic complex of Planchon-Peteroa (35.2 S), which last erupted in 2011.

The saddle-shaped volcanic complex of Planchon-Peteroa (35.2 S), which last erupted in 2011.

Cerro Azul volcano, Chile.

The spectacular ice-filled summit crater of Descabezado Grande volcano, Chile, at 35.6 S. The last eruption from this complex was in 1932, shortly after an eruption of the  nearby volcano Cerro Azul (or Quizapu).

View across the volcanoes of Tolhuaca (or Tolguaca, near ground) and Lonquimay (38.3 S). Both volcanoes are young, but it is not known when Tolhuaca last erupted. Lonquimay last erupted from 1988-1990.

View across the volcanoes of Tolhuaca (or Tolguaca, near ground) and Lonquimay (38.3 S). Both volcanoes are young, but it is not known when Tolhuaca last erupted. Lonquimay last erupted from 1988-1990.


The young cone of Volcan Antuco, 37.4 S. Its last known eruption was in 1869.


Twin-peaked Llaima (38.7 S) is one of the most active volcanoes of southern Chile, and last erupted in 2009.


Volcan Sollipulli (39 S) has a spectacular ice-filled summit caldera, but is not thought to have erupted since the 18th Century


Panorama across three young volcanoes, looking east: Villarrica (39.4 S) in front; the snow-covered sprawl of Quetrupillan in the middle ground; and the peak of volcan Lanin, on the Chile – Argentina border, in the distance.


Villarrica, with a characteristic thin gas and aerosol plume rising from the open crater at the summit.


The twin-peaked volcanoes Mocho Choshuenco (39.7 S). Choshuenco, thought to be the older vent, is the angular crag nearer the camera; Mocho is the small cone in the middle of the summit plateau. Mocho last erupted in 1937.


Looking across a bank of cloud towards volcan Osorno (front, 41.1 S), and volcan Tromen, in the background. Osorno last erupted in 1869; Tromen is thought to have last erupted in 1822.


Volcan Calbuco (41.3 S), which last erupted in 1972.

Data source: information on the recent eruptions of these volcanoes is all from the Smithsonian Institution Global Volcanism Project.

Further reading:

CR Stern, 2004, Active Andean volcanism: its geologic and tectonic setting. Revista geologica de Chile 31, 161-206 [Open Access].

SFL Watt et al., 2009, The influence of great earthquakes on volcanic eruption rate along the Chilean subduction zone. Earth and Planetary Science Letters, 277 (3-4), 399-407.

SFL Watt et al., 2013,The volcanic response to deglaciation: evidence from glaciated arcs and a reassessment of global eruption records, Earth-Science Reviews 122, 77-102.

Acknowledgements: my fieldwork in Chile over the past 10 years has been funded by NERC, IAVCEI and the British Council. Many thanks to my parents for introducing me to Chile and its volcanoes at the age of 7; and to Jose Antonio Naranjo and many others at SERNAGEOMIN for facilitating our continuing work in the region.

August Anniversaries: the eruption of Krakatoa

August 27th marks the anniversary of the culmination of the great eruption of Krakatoa (or Krakatau) in Indonesia in 1883. This devastating eruption has become the archetype of a volcanic catastrophe, even though it was a geologically modest example of a ‘caldera forming’ event. The eruption of Krakatoa quickly made the headlines around the world, in part because newly installed undersea cables allowed the news of the event to be wired rapidly across the globe.

Precursory eruption of Krakatoa in May 1883. From Symonds (1888).

Precursory eruption of Krakatoa in May 1883, several months before the climactic events of August 1883. From Symons (1888).

The Krakatoa eruption was one of the first major eruptions to be intensively studied by scientists. The journal Nature published an editorial explaining the ‘Scientific Basis of the Java Catastrophe’ shortly after news of the eruption broke, and over the next few weeks published reports with the first descriptions and explanations of some of the many widespread effects of the eruption – from the appearance of great floating rafts of pumice, to the many oceanic, atmospheric and other phenomena that accompanied the event. In February 1884, the Royal Society set up the Krakatoa committee, chaired by a meteorologist George Symons, to collect information on ‘the various accounts of the volcanic eruption.. and its attendant phenomena’, including ‘authenticated facts respecting the fall of pumice and dust .. unusual disturbances of barometric pressure and sea level.. and exceptional effects of light and colour in the atmosphere‘. These results were published in 1888 in a wonderfully illustrated monograph, which still stands as one of the most complete accounts of a major volcanic eruption and its widespread effects.

 Drawings of views of Krakatoa rock samples in thin section, under a microscope.  This set of images are of lavas from Krakatoa. These are quite rich in crystals (white feldspar; green pyroxene), set in a fine matrix of glass (colourless to brown).  Images on the left: about 1 cm across; on the right – about 1 mm across.

Drawings of microscope views of Krakatoa rock samples in thin section. This set of images are of lavas from Krakatoa, from Symons (1888). These are quite rich in crystals (white feldspar; green pyroxene), set in a fine matrix of glass (colourless to brown). Images on the left: about 1 cm across; on the right – about 1 mm across.

 Drawings of views of 1883 Krakatoa pumice and ash samples in thin section, under a microscope.  The pumice samples are mainly made of glass (very pale colour) with gas bubbles.  The top 4 images are each about 1 cm across, and show the texture of lumps of pumice.  The bottom two images are each about 1 mm across, and show the ‘ash’ that fell over a thousand miles away from Krakatoa, onto the ship Arabella (left), and the ‘ash’ formed by grinding up a sample of pumice.

Drawings of microscopic images of 1883 Krakatoa pumice and ash samples . The pumice samples are mainly made of glass (very pale colour) with gas bubbles. The top 4 images are each about 1 cm across, and show the texture of lumps of pumice. The bottom two images are each about 1 mm across, and show the ‘ash’ that fell over a thousand miles away from Krakatoa on the ship Arabella (left), and the ‘ash’ formed by grinding up a sample of pumice. From Symons (1888).

The most celebrated impact of the Krakatoa eruption in terms of science was the recognition that the many optical effects that followed the eruption, including both spectacular sunsets (an example from London, in November 1883, below) and the discovery of ‘Bishops’ Rings‘, must be  the consequences of the global spread of volcanic pollutants, high in the atmosphere. We now know that the major constituents of this haze were tiny droplets of sulphate, forming a thin aerosol layer in the stratosphere which scattered and absorbed incoming solar radiation.

Sunset at Chelsea, 4.40 pm,  November 26th, 1883. From Symonds (1888).

Sunset at Chelsea, 4.40 pm, November 26th, 1883. From Symons (1888).

Although the Krakatoa eruption was one of the largest eruptions of the past 200 years, it is not exceptional in comparison to other eruptions from the geological record. In terms of eruption ‘size’ it rates as a ‘6’ on the Volcanic Explosivity Index, having erupted an estimated 12 cubic kilometres of magma. Eruptions of this sort of size occur once every 100 – 200 years around the globe; while the largest known explosive volcanic eruptions erupt many thousands of cubic kilometres of magma over a very short period of time. Krakatoa remains the best documented example of a ‘caldera-forming’ eruption, during which an entire volcanic edifice collapses. In this case, the eruption and the formation of the caldera had catastrophic consequences for tens of thousands of people along the shores of Java and Sumatra, as a series of major tsunamis triggered by these events swept ashore.


G.J. Symons (Editor, 1888), The eruption of Krakatoa and subsequent phenomena.  Report of the Krakatoa Committee of the Royal Society.  London : Trübner & Co.

Links to some other posts on Krakatoa

David Bressan’s excellent post for the Scientific American Blog

Bill McGuire’s post on Krakatoa for the OUP Blog

Jeremy Plester’s Weatherwatch piece for The Guardian

Cynthia Wood’s post for Damn Interesting

A page on the Krakatoa eruption from the United States Geological Survey