You’ve probably heard of Katla, Eyjafjoell’s more-pronounceable Big Sister. If you’ve seen any of the popnews in response to a press release about a report I haven’t gotten my paws on yet, you’ve probably heard of their history of Eyjafjoell erupting, then Katla erupting even more violently shortly later.
Weirdly, although the press release barely mentions Katla, and even points out that current observations do not point to an imminent eruption, you will get an entirely different story from most media.
Yes, I totally agree with the report that governments should think about the consequences of ongoing eruptions, and figure out a plan in case of Katla erupting well in advance of it being an actual problem. And yes, although vulcanologists can’t pinpoint exact to-the-minute eruption times (and frequently die trying to get the data to figure that out), they’re good at determining that a volcano will soon erupt at least a week, usually more several weeks or months, in advance of the eruption. And no, despite the popnews reports, Katla is not at risk of erupting today, tomorrow, or the next day.
So, how can you tell when Katla, or any other volcano, is going to erupt? Look for changes such as:
– earthquakes (increased frequency or intensity) or volcanic tremors
– ground swelling or cracking (from a full magma chamber)
– more or different gases venting
– changes to local springs (new ones, changes in acidity or chemicals…)
– melting of snow or ice (…which is really easy to see for a glacier-covered volcano!)
And as far as statistics goes: A 3-for-3 eruption pairing history is a good start to a pattern, but it’s a very short history to make solid claims from.