• Question: can you use science to predict when a volcano will erupt?!!

    Asked by clogerina01 to Philippa on 14 Jun 2011.
    • Photo: Philippa Demonte

      Philippa Demonte answered on 14 Jun 2011:


      Hi @clogerina01 – the short answer is that it is impossible to /predict/ when a volcano will erupt (or likewise when an earthquake will happen). Scientists can only attempt to /forecast/ the likelihood of an eruption occurring.

      Some of the signs which indicate that a volcano might erupt include the following:

      * increased seismicity (little earthquakes) due to the magma rising and gases trying to escape the system

      * inflation of the volcano’s flanks (sides) as the system fills with magma, or deflation as the magma either drains out or erupts somewhere within the system

      * an increase in first carbon dioxide, then sulphur dioxide, then water vapour / steam, and finally gases such as chlorine and fluorine into the atmosphere. The type and quantity of volcanic gases released by a volcano gives us a rough indication of how deep or shallow the magma is within the volcano’s system.

      These and other signals were used very effectively by the U.S. Geological Survey (USGS) to forecast the big eruption of Pinatubo in the Philippines in 1991 (go check out the footage on YouTube – it’s pretty awesome!) However, the problem is that another volcano might show all of these signs, but then not erupt. Alternatively the volcano might appear to be dormant (asleep), and then very suddenly re-awake and erupt with very little warning, usually because of a lot of water getting into the volcanic system (This causes the magma to cool very rapidly, the gases within the magma therefore can’t escape easily, and then the pressure builds up until the volcano erupts explosively).

      Every volcano behaves differently, which is why it is so important to monitor them, and then to do what’s called modelling. This involves using computer programmes to simulate the possible reactions of a volcano to different scenarios so that we can be prepared for any situation.

      There is a scientist called Warner Marzocchi who works at INGV, the Italian institute responsible for monitoring Vesuvius volcano. He is currently using computer programmes based on mathematical probability and actual data (measurements) taken from the volcano over time in order to test and improve volcano eruption forecasting. This is not the only method though. Making a map of which volcanic rocks are found where on a volcano, and likewise mapping ash fall deposits from previous eruptions can tell us a lot about /how/ a volcano might erupt in the future. Another branch of volcanology known as geochemistry looks at very small, thin samples of volcanic rocks through a microscope to analyse what the rocks are made of – this tells us what kind of magma is in a volcanic system, and therefore the way in which a particular volcano might erupt. Likewise the size of the grains in the rock tells us whether the magma at that volcano was able to cool slowly in the magma chamber, or if it was erupted quickly. Clues about the past can tell us much about the future.

      Hope this helps answer your question.

Comments