Thursday, December 28, 2006

"Events.." in the natural world

Just over a year ago a gigantic earthquake hit the fault between the Indian and Burma tectonic plates. At 9.3 on the Richter scale it was the second largest earthquake ever recorded on a seismograph. A few minutes after the quake had stilled, the first tsunami came ashore on Sumatra, before spreading out across the Indian Ocean. Along a huge section of coast the surge of water flooded ashore- often for many miles. The US Geological Survey initially recorded the casualty toll as 283,100 killed, 14,100 missing, and 1,126,900 people displaced- one of the most destructive natural disasters ever recorded.

The Boxing Day Earthquake reminds us that the unexpected can come from many different quarters. Humans seem remarkably good at ignoring long term danger- until, that is disaster strikes.

There are several events that may at the time be considered "black swan events"- that is that they are inherently unpredictable- arguably 9/11 was one such event. However this is not so with many natural disasters. The problem is not forecasting what will occur, this is usually very clear, but when. For example Mount Vesuvius near Naples has been a regularly erupting Vulcano for all of human history- averaging an eruption about six times a century. However since 1944, the Mountain has been quiet. That the vulcano will erupt again is a certainty- when is highly unpredictable.

Similar statistics show up in the Kanto fault under Tokyo- regular large Earthquakes about every sixty years: the last one being the 1923 Great Kanto earthquake - Yes, one seems overdue at the moment. Other major cities, like San Francisco and Istanbul are also prone to large scale, destructive earthquakes, but on a regular scale that is longer than a single human life- so people forget about the risks. Floods and Storms are also certainties on a macro scale- but so unpredictable on a micro scale that preparation is almost always suspect: as New Orleans after Hurricane Katrina proved so brutally.

However, it is easy to add another unserious prediction to our survey: large scale natural disasters will occur in 2007, and humans will probably be unprepared for them.


Etzel Pangloss said...

One must be getting old Cicero, the tsunami was two years ago.

RK said...

Too simple an analysis I'm afraid. A sensible cost benifit analysis on large scale (but very unlikely) natural disasters will sometimes conclude that building a defence now is too expensive and that it is better to wait and pick up the pieces afterwards. It may seem harsh but there you have it. We only have so much money. The inevitable consequnce of this approach is that lives are lost in the disaster that could have been saved had a defence been built. If you don't consider how else this money could have been spent then it is obvious that the defence should be built whatever the cost. If you do think of how else you could spend the money involved then it is clear that lives could be saved and improved by it's application elsewhere.

For example there is a tsunami risk in the Altantic from a landslide in the Canary Islands but it is so small that preparing defences, building and maintaining a warning system would cost a lot of money. Should we spend this money on the Canary Island tsunami risk or the NHS?