Skip to content. | Skip to navigation

Updates by Email
Join Our List
Privacy Policy
Personal tools
You are here: Home CBI Blog Mt St. Helens, Black Swans, and Other Surprises

Mt St. Helens, Black Swans, and Other Surprises

Posted by Robert M. Scheller at May 22, 2009 12:00 AM |

This week marks the 29th anniversary of the eruption of Mt. St. Helens (MSH) in Washington state. MSH is a reminder of the capacity of the future to alarm and surprise us. The challenge for scientists at CBI is how to effectively communicate surprises to the public and land managers.

This week marks the 29th anniversary of the eruption of Mt. St. Helens

(MSH) in Washington state. As a resident of Portland, Oregon, the sight of MSH is a frequent reminder both of the power of the eruption and the capacity of the future to alarm and surprise us.

The eruption of MSH is categorized by forest ecologists as a Large Infrequent Disturbance (LID) .  The eruption was soon followed by the Yellowstone fires of 1988 , another LID. More recently, Nassim Nicholas Taleb coined these events 'Black Swans' in his book of the same name.  Despite the name, LIDs are common in the U.S., one or two per decade, although they do not recur in the same general vicinity.

At CBI, we are often tasked with developing  scenarios of the future.  LIDs are easy to incorporate in scenarios - simply ensure that your models have the capacity to stochastically, albeit rarely, simulate an enormous fire, hurricane, whatever.

However, LIDs are difficult to communicate.  The public and land managers

St. Helens
are trained to expect gradualism.  When our results project a LID at some point in the future, the first reaction is often, "Your model must be wrong!  A disturbance that large has never happened and could never happen here."

Model projections are always wrong, partly because the future will always be unpredictable.  But if a model predicts a LID, it is not evidence that the model is erroneous - rather it may be evidence that our assumptions about a predictable, stable future are ill founded.

One of the values of modeling the future is the ability to surprise people.  It is preferable to surprise the public and land managers with a computer simulation and enable them to become better prepared, rather than to allow Black Swans to dictate our natural resource management policies as has so often happened in the past.

Document Actions
powered by Plone | site by Groundwire and served with clean energy