MOUNT ETNA has traditionally been thought of as a safe volcano. Because of the fertile volcanic soils, more than 20 per cent of Sicilians live on its slopes. Major eruptions, happening at least once a decade, destroy homes and farm buildings; vines, olive groves and lemon trees vanish beneath lava and ash. The swellings and sinkings of the surface of this volcano hold the key to understanding how it erupts. Precise measuring of Etna’s shape could allow us to predict better both eruptions and the damaging landslides that they may trigger.
But the methods of prediction refined on Etna do not help the people of Sicily alone. The great majority of the world’s active volcanoes are situated in developing countries. Competition for living space and valuable farmland mean that more and more people are living within the danger zones of these volcanoes. The methods that volcanologists have used on Etna could be applied across the world, making warnings of eruptions more reliable and saving lives.
Advertisement
Rising from the east coast of Sicily, between the cities of Messina to the north and Catania to the south, Etna is an impressive volcano by any standard. Its summit stands 3375 metres above sea level, while its circumference of around 100 kilometres encloses over 1100 square kilometres of lava, ash and cinders. Etna is the largest active volcano in Europe, and the largest built on continental crust anywhere in the world. The four summit craters continually release gases, and there are significant eruptions of lava every two or three years. The last sign of life was in October last year, when lava flows threatened towns and villages high on the slopes of the volcano, and heavy falls of ash blocked major roads in northeast Sicily.
It is the knowledge that something is always happening, or about to happen, that attracts researchers to Etna in large numbers. Compared with many other volcanoes, it is easy to reach, having good roads and an airport nearby, so that a monitoring team can travel to the volcano quickly should it erupt unexpectedly. In spite of a concerted research effort over the past two decades by British and European scientists, much remains to be discovered about its internal plumbing, the system of cracks and pipes through which molten rock, known as magma, makes its way to the surface. Unlike the Hawaiian volcanoes, Mauna Loa and Kilauea, which stand on thinner ocean crust, and have fairly simple plumbing systems, Etna seems to be more complex. It has changed several times since volcanic activity started there about 300 000 years ago. The magma which erupts from Etna in the form of lava, ash and scoria (a form of lava containing many bubbles of air, looking like cinders), forms by melting of rock in the Earth’s mantle, the layer immediately beneath the crust, at depths more than 30 kilometres below the surface.
In 1980, Alastair Sharpe, with colleagues from the University of Cambridge and the University of California at Los Angeles, used deep seismic sounding techniques to show that Etna’s magma does not rise directly from the mantle to the surface. Instead, it accumulates in a network of small fractures, forming a reservoir of magma at a depth of around 20 kilometres, near the base of the crust. From this reservoir, magma rises in a near-continuous column towards the summit region. Gases that were dissolved in the molten rock when it was deeper in the Earth escape through open vents around the summit. On occasions, these craters fill or partly fill with magma to form lava lakes, or they may erupt explosively, throwing out molten rock, cinder and ash.
The most destructive eruptions on Etna, however, take place on the slopes of the volcano. They tend to happen a few times per decade, sometimes accompanied by earthquakes, or eruptions from a crater near the summit. These flank eruptions arise when magma beneath the summit exerts such pressure on the walls of the pipe or fissure in which it flows that the rock cracks. The fracture then fills with magma and grows horizontally and outwards towards the surface. If the pressure of magma rising from below is sufficient, the fracture will break the surface of the ground some distance from the summit, and molten rock will erupt from the fissure. In some cases, however, the magma does not reach the surface, but solidifies below ground to form a body of rock called an intrusion. Intrusions that come from these fissures have a particular shape, forming upright, usually vertical, blade-like sheets of igneous rock. They are known as dykes, and have played a major role throughout much of the history of Etna by feeding eruptions on its flanks. There are hundreds of ancient, solidified dykes, superbly exposed in the cliffs surrounding the Valle del Bove. This is a caldera, a large volcanic crater formed by collapse or explosion on the southeast slope of Etna.
Most of the major lava eruptions in historic times have come from dykes feeding eruptive fissures. The greater the depth at which a dyke forms, the lower the altitude of the resulting fissure at the surface and, generally speaking, the greater the damage to property, because more people live on the lower slopes. One of the most extensive eruptions of lava in recent history happened in 1669, when a fissure stretched from the summit area to the village of Nicolosi some 15 kilometres away and 2400 metres below. In places, the crack was 2 metres wide, containing glowing lava, according to contemporary reports. Lava from this eruption reached the coast and destroyed part of the city of Catania, 12 kilometres from the vent, after flowing over the city walls.
Researchers now hope to be able to predict the areas at highest risk from this sort of eruption by monitoring gravity and magnetic changes, seismic activity, thermal anomalies, the chemistry of the gases coming from the vents and the distortion of the volcano’s surface that results from the flow of molten rock below ground. Ideally, they will acquire data that can be added to maps which show the degree of hazard for different areas around a volcano. This takes into account features such as valleys, which might channel a lava flow towards a particular village, and so increase the local risk. A technique for measuring the movement of the surface above rising magma was developed on the volcanoes of Hawaii and Japan (see Box overleaf). It has since proved valuable in predicting the time and place of eruptions on Etna.
Volcanologists studying ground deformation on Etna want to identify bodies of magma rising beneath the summit region, and dykes growing away from the summit towards its flanks. In 1975, John Murray, now at the Open University in Britain, set up a line of markers across the summit of the volcano. He then surveyed their positions twice each year, and discovered that vertical displacements of bench marks to the south of the summit reflected the movement of molten rock within the volcano. All the surface around the summit rose, as if inflated like a balloon, when bodies of magma moved upwards immediately beneath it; the slopes deflated after magma erupted.
Murray recorded a pattern of inflation in the early 1980s and began to suspect that an eruption on the southern flank might be imminent. In order to provide more detail on the deformation of the ground associated with this expected eruption, Andy Pullen of Imperial College, London, established a second network of markers in 1981, to measure horizontal deformation on the upper part of the southern flank. This network, which is now operated by the Volcanic Studies Unit of the West London Institute, has been considerably enlarged in the past few years to provide increased monitoring of activity in the area.
Murray’s suspicion proved well founded in March 1983, when an eruptive fissure fed by dykes cut through Pullen’s survey network, generating extensive flows of lava which threatened villages on Etna’s southern slopes. This eruption galvanised the authorities into an attempt at diverting the flow using explosives. Although the lavas stopped flowing soon afterwards, researchers are still debating whether or not this was simply coincidence.
After the opening of the 1983 fissure, Murray and Pullen remeasured both their surveying networks. They found that they could trace the path of the dyke responsible for feeding the eruption, and, by calculating the distortion of the ground that they would expect above different intrusions, find out its size, shape and depth below ground.
In 1975, James Dieterich of the US Geological Survey, and Robert Decker of Dartmouth College in New Hampshire, used computer models to show that the precise form of patterns of uplift and sinking above a dyke underground could help researchers to define the height, depth, and tilt of the intrusion. The surface above a dyke distorts in a distinctive pattern as the molten rock below squeezes away from the core of the volcano; a long narrow zone immediately above the intrusion sinks, and the ground to either side rises.
Predictions fall short
In the case of the 1983 dyke, the central zone sank by as much as 125 centimetres, and the sides rose by up to 40 centimetres. Dieterich and Decker calculated a displacement curve for a vertical dyke 950 metres high, starting 475 metres below ground, which almost matched the observations made on Etna. The actual dyke was probably about this size and shape. But the surface above the centre of the 1983 dyke had sunk more than the model predicted. This is because the rocks have not stretched in quite the way that the model assumed. Dieterich and Decker’s model applies only to elastic deformation, and does not take into account other types of distortion, such as faulting and fracturing, which probably cause the extra depression above the dyke on Mount Etna.
Murray’s observations of vertical displacement, supported by data on horizontal deformation derived by Pullen, showed that the 1983 dyke grew from beneath the summit towards the southeast. As it approached the rim of the Valle del Bove caldera, it deviated to run in a southerly direction, parallel to the western cliff wall. Lava erupted when the dyke met the surface at a height of about 2500 metres.
In the network to measure horizontal displacement, bench marks on either side of the path of the dyke had become farther apart; the intrusion of molten rock below ground had fractured and displaced the surface above by about 2 metres horizontally, in addition to the vertical movements. By modelling the movements of all the stations in his network relative to a base station which he considered stable, Pullen demonstrated that this displacement was not uniformly distributed on either side of the dyke. Instead, most of the horizontal distortion was to the east of the dyke. In particular, a huge block of the western cliff wall of the Valle del Bove had moved outward.
The reasons for this became clear after Pullen and I carried out a series of experiments at the West London Institute. We made solid gelatine models of volcanoes in different shapes, and injected them with coloured water to represent magma rising to the surface. The water rises until it builds up sufficient pressure to crack the gelatine. It then flows outward in the cracks, forming vertical sheets which represent dykes. One gelatine cone represented Etna, with a steep side resembling the cliff wall of the Valle del Bove caldera. Dykes changed their paths as they approached the ‘cliff wall’, and ran parallel to it. This was exactly the behaviour of the 1983 feeder dyke adjacent to the Valle del Bove western cliff wall. Fissures from the 1978-9 and the 1985 eruptions also followed paths which turned to run parallel to the cliff wall of the caldera. K K The explanation for this turns out to be straightforward. Both the steep side of the gelatine model, and the Valle del Bove cliff wall, represent vertical or near-vertical surfaces which are ‘falling outward’ under the force of gravity. This creates tension in the gelatine and the rocks near the cliffs. Fractures form most easily parallel to the vertical surface because of the extra gravitational pull. The shape of Mount Etna is an important control on where and how lava escapes in this sort of flank eruption.
Magma on the move
As long as magma continues to rise to the southeast of the summit, dykes growing eastward or southeastward stand a good chance of being diverted to run parallel to the rim of the Valle del Bove. This poses a double threat to the people who live and work around the volcano; they are in danger from lava, and potentially also from landslides. Subterranean magma is being channelled southwards, towards the centre of tourism on Etna. Despite the fact that lava erupted during 1983 and destroyed parts of the cable car system and several hotels and restaurants on the upper slopes, construction in the area continued. The eruption in 1985 also failed to halt the building activity. In October last year, a network of fractures 6 kilometres long, which started near the summit, cut across the main road just to the east of the tourist village.
Of potentially greater importance is the fact that a huge block of the cliff wall of the Valle del Bove is moving east because of the dykes which repeatedly inject along its rim. The block moved 2 metres during the 1983 eruption, and another 80 centimetres in 1985. The fracture system which opened in October 1989 was formed by yet another dyke, similar to the earlier intrusions. Although in this case magma did not reach the surface, and there was no eruption of lava, the rim of the caldera moved a further 80 centimetres east – a total displacement of 3.6 metres over seven years. If more dykes form in this area, the rim of the caldera will become increasingly unstable. If the cliff collapses in a piecemeal fashion, generating landslides which are confined to the Valle del Bove, then there is little risk to the people of local towns and villages. This part of the caldera rim forms a very straight cliff, however, suggesting that past collapse may have involved the whole of the western side at one time. Such a fall today could generate landslides with the potential to reach the densely populated coast of Sicily east of the caldera. The monitoring programme would have to be continued or even stepped up, before scientists could assess the risks of such an event occurring without warning on Etna.
The 1980 blast from Mount St Helens, and the Nevada del Ruiz volcanic disaster in Columbia during 1985, have shown the destructive power of landslides and mudflows produced during volcanic eruptions. Measuring the way the ground has deformed on Etna has proved to be a valuable method of predicting its eruptions, and assessing the risk of landslides. For most of the world’s active volcanoes, in contrast, the prediction record is either poor or nonexistent. This may be due partly to the difficulty of understanding volcanoes which have complex plumbing systems; it is more likely to result from a poor monitoring programme, and a lack of funds.
* * *
Tracking the tilt that tells when a mountain could blow its top
THE LINK between the way the surface of the ground deforms, and the movement of magma underground, was first established in Japan during the early part of this century. In 1914, the Japanese geophysicist Fusakichi Omori decided that the vertical movements of a series of bench marks in a survey network around the volcano Sakura-jima, near Kagoshima in southern Japan, were caused by the volcano swelling before it erupted, then subsiding after lava flowed out.
Similar movements were recorded at Kilauea on Hawaii in the late 1950s, using a system of tiltmeters designed to monitor small variations in the height of the ground. This apparatus was simple but effective, consisting of tubes filled with water, arranged in a triangle and connected at each apex to hollow pots mounted on concrete pillars. Changes in the relative height of the three pillars affected the level of water in each pot, and allowed tilts as small as 1 millimetre in a kilometre to be measured.
Ground movements measured using this system, and later more sophisticated networks of tiltmeters, showed that before Kilauea erupts, its summit usually swells, causing the ground to tilt outward, away from the summit crater. Following an eruption, the volcano deflates; the summit region subsides, and the gradient of the upper slopes decreases.
Horizontal movements of the ground associated with the flow of magma underground were first recorded on Kilauea in the 1920s by RM Wilson. He also undertook a triangulation survey which involved repeatedly measuring the positions of bench marks around the summit, so that he could compute any horizontal offsets between them. Using this technique, Wilson successfully recorded the deflation following Kilauea’s eruption in 1924, which caused a 1-metre horizontal move ment of his bench marks towards the centre of subsidence.
In 1964, the measurement of horizontal ground movement on active volcanoes was revolutionised by Robert Decker of Dartmouth College, New Hampshire, and his colleagues, using electronic methods to establish precisely the distances between stations in a triangulation network on Kilauea. The advent of this equipment not only allowed more accurate measurement of horizontal displacement of the ground, but also made surveying much quicker. This system of monitoring horizontal ground movement is now in use on active volcanoes all over the world, and has been employed in recent years to monitor the movement of magma within Mount Etna.
I and my colleagues from the Volcanic Studies Unit at the West London Institute have used both precise levelling surveys and electronic distance measurement on Mount Etna. With levelling, we can detect changes in height between survey stations of less than a centimetre.
For the distance measuring, we use an instrument known as an electronic distance meter, which bounces a beam of laser light, infrared, or microwave radiation off a reflector. This enables us to determine the precise distance between marker stations to within a few centimetres over 40 kilometres or more. With this quick, convenient method, we can detect the small but significant movements of the ground.
Bill McGuire is head of the Volcanic Studies Unit at the West London Institute.