The Volcano of Santorini

Santorini is one of the top destinations in the world. Famous for its unique natural beauty and its romantic character, Santorini will steal your heart. At first, Santorini is a top honeymoon destination. Additionally, many groups of friends choose Santorini for their holidays. They usually rent hotel rooms or luxury villas, at reasonable prices. But what is the history behind this amazing natural beauty of Santorini?

The island of Santorini is an outcome of volcanic eruptions over about 2 million years, during which the island has changed its form and size repeatedly. Before any volcanic activity took place (caused by the subduction of the African plate under the Eurasian one), there was a small island in the area, similar to the other Islands of the Cyclades. This non-volcanic island forms the Profitis Ilias and Mesa Vouno mountains, as well as the shists at Athinios harbor and in the Plaka bay connected to Profitis Ilias mountain through the ridge of Pirgos, the Gavrilos ridge (the "windmill hill" south of Emborio) and the isolated rick of Monolithos.

Volcanism in the area of Santorini started about 2 million years ago when the first eruptions occurred from the seafloor in the area of the Akrotiri Peninsula and probably also at the location of the Christiania Islands 20 km SW of Santorini.

The activity built dacitic lava domes that eventually formed a series of islands, still noticeable in the hills of the Akrotiri peninsula. In a second stage, a stratovolcano (Peristeria volcano) was formed in the northern part of Santorini, parts of which are still visible in the cliffs and slopes of Mikro Profitis Ilias and Megalo Vouno.

Starting about 400,000 years ago, the focus of volcanic activity shifted to where it has been since then, the center of the present-day caldera. The most characteristic type of activity over the last 400.000 years has been the cyclic construction of shield volcanoes intruded by large explosive and catastrophic events like the Minoan eruption about 3600 years ago.

In detail, the volcanic evolution of Santorini can be split into six main stages (Druitt and others, 1989):

• Akrotiri Volcanoes (approx. 2 mio - 600,000 years ago)
• Cinder cones of the Akrotiri peninsula (around 600 - 300 ka)
• Peristeria Volcano (530 - 300 ka)
• Products of the first eruptive cycle (360-180 ka)
• Products of the second eruptive cycle (180 ka - 1613 BC)
• Kameni shield (1613 BC - present)

there are some (still minor) signs that the volcano of Santorini could wake up in a medium future (months to years). Predicting IF and if yes, WHEN exactly there will be a new eruption is impossible - volcanoes are and will always be unpredictable.

It is quite certain, though, that the volcano will erupt in the future again, because it is an active volcano and far from extinct.

Based on the frequency of the past eruptions, a statistical forecast can be attempted: during historic time, Santorini has erupted in 197 BC, 46/47 AD, 726, 1570-1573, 1707-1711, 1866-1870, 1925-1928, 1939-1941, and last in 1950.

Looking at these dates, there have been 9 eruptions over 2000 years, and perhaps a few that are not preserved in geologic or historic records. One arrives at an ordinary eruption interval of about 200 years for this period. On the other hand, the intervals between eruptions are all but regularly spaced: for instance, about 700 years seem to have passed between the eruption in Roman times (46 AD) and the probably very large eruption in 726, of which little is known except that it was said to have covered the Aegean Sea with pumice.

Another long interval of about 700 years followed until Mikri Kameni was born in 1570, which was a rather small dome-building eruption.

After 1570, eruptions followed with smaller intervals, notable the series of activity in 1866-70, 1925-28, 1939-41 and 1950. Since then, a significantly larger time interval has passed until now (62 years).
A major limiting factor is the number of data points to make a reliable statistic out of this: only 4 eruptions and 4-time intervals are available if we only look at the last 150 years, and any purely statistical approach on such a few numbers of samples is nearly useless. If we only had this statistical approach, the long-term prediction would be limited to say that it will explode in the future, which could be just a day away or thousands of years.

As one of our commentator's volcanologist, J. Stone from the University of Bristol (who are involved in monitoring Santorini at the moment) kindly pointed out, predicting volcanic eruption is not limited to the WHEN, but also the important questions of HOW and WHERE, which hazards may occur, etc. This has been a major task of modern volcanology and in the recent decade, much study has been devoted to this subject, and many more sophisticated approaches are available (e.g. Baxter et al, 2008) which use Bayesian statistics, expert judgment and event trees to model possible eruptions. In this way, volcanologists can communicate the uncertainty, but also provide something useful for planners and decision-makers.

Volcanologists may not necessarily be able to predict the "when" with much temporal resolution, but they can forecast probability of eruption scenarios with some success, in other words, how likely eruptions are within a given timeframe. Crucially, science is now able to predict relatively accurately where eruptions and hazards may occur with high spatial resolution.

To make an example for Santorini: over the past several hundreds of thousands of years, there have been at least 12 Plinian eruptions, spaced at intervals of an average 20-30,000 years, some of which were followed by caldera collapse, and interrupted phases of effusive and smaller eruptions that built new shields (such as Nea Kameni), and contributed to filling the caldera. It is reasonable to believe that Santorini is now in a shield-building phase in the long term, following the last Plinian eruption, the famous Minoan eruption 3600 years ago. Statistically, it is very unlikely that the next eruption will be of such magnitude if it occurs in a timeframe of years to a few centuries from now. It is supported by the argument that the volcano is believed to have long rest periods (thousands of years) preceding such eruptions.

The worst-type scenario for an intra-caldera eruption that could be imagined with any substantial likelihood could then be the type of the little-known explosive eruption in 726AD, which was possibly sub-plinian. Such an explosion would be severely damaging the island and its infrastructure, threaten people, and probably require a large-scale evacuation. Fortunately, this scenario is not the most likely type of next eruption at all.