Volcano Risk: Have we Learned any Lessons?

Don’t miss the exhibition Life and Death: Pompeii and Herculaneum still showing at London’s British Museum for a few days. For a place so closely associated in everyone’s mind with sudden death, the curators managed to put the emphasis on life, taking you through the mores and daily life of Pompeians. 

The only obvious reminder of death is the mosaic of a skeleton holding wine jugs, welcoming guests into a Roman dining room – a memento mori, a reminder to seize the day while we can. Until you are faced with the plaster casts of Pompeians, frozen in death by the pyroclastic flow.

Vesuvius had a long history of violent eruptions, but had been dormant since 680BC, luring its neighbours into a sense of false security. So it took Pompeii and Herculaneum by surprise, despite what may have been warning signs: a series of earthquakes in the previous decades, and bradyseism (uplift or susbidence linked to movements of the magma chamber). In his book on Natural phenomena, Seneca didn’t see any link between earthquakes and volcanic activity. Our knowledge has obviously improved since, so are we better prepared?


People still chose to live around volcanoes because the perceived risk pales by comparison to the immediate benefits: fertile soils, access to minerals and geothermal energy.  Around 10% of the world population lives within 100km of a volcano that was active in the last 10,000 years.

A previous WRN study highlighted that the ten most dangerous volcanoes in Europe could affect 2.1 million people with a combined exposed residential property value of $85 billion. Vesuvius comes first, posing the greatest risk to life (1.7 million people) and property ($66.1 billion residential property).

Understanding / Forecasting

Exposure concentration around volcanoes is there to stay.

There isn’t much hope in defence mechanisms against the various volcanic hazards: lava flows, lahars, pyroclastic flows, and ashcloud (in increasing order of reach). Volcano event response is more about flight that fight.

So naturally, most of the investment has been on understanding to obtain better forecasting methods.

With close to 1500 known ‘active’ volcanoes (active in the last 10,000 years), instrumentation and monitoring of every single one would be expensive – the rise of affordable GPS sensors has helped.

But land-based monitoring stations, which offer valuable real-time information, are expensive and subject to budget cuts, as in Alaska recently.

Satellite technology, used so far mostly for event response (e.g. tracking ash cloud movements), has the potential to allow global constant monitoring (detecting temperature or land surface changes). However, coverage is still limited, and the information is not real-time.


Volcanoes have the potential to have an impact on a global scale.

Tambora in Indonesia was the source of history’s largest explosive eruption in 1815, causing an estimated 71,000 fatalities, and global climatic effects. 1816 was known as the ‘year without a summer in many parts of the western world, causing crop failure, emigration, and epidemics.

More recently, Eyjafjallajökull in Iceland reminded us of this potential in 2010, without the immediate death toll and such dramatic consequences.


Providing adequate insurance for volcanic risks is tricky for the insurance industry. For example, the standard 72 hours clause is at odds with the average 7 weeks duration of an eruption. And there are currently no commercial volcano models.

The Global Volcano Model initiative (which amongst others, Munich Re and Willis are supporting) is trying to fill this vacuum, as an “international network that aims to create a sustainable, accessible information platform on volcanic hazard and risk. GVM will provide systematic evidence, data and analysis of volcanic hazards and risk on global, regional and local scales, and will develop the capability to anticipate future volcanism and its consequences”.

A key activity (which Willis is involved in) is understanding the impact of a major explosive eruption: considering local effects, impact on supply chains, but also the likelihood of pronounced climate changes associated with such an eruption in the following years, affecting the likelihood of other hazards (storms, floods etc).

About Hélène Galy

Hélène is Managing Director of Willis Global Analytics, where she is Head of Proprietary Modelling. Since 1998, s…
Categories: Analytics, Natural Catastrophe | Tags: , , ,

Leave a Reply

Your email address will not be published. Required fields are marked *