Wednesday, October 27, 2021

Hurricanes and snowfall, revisited

Many, many years ago, I pondered the relationship between hurricanes and snowfall in Houston, as there seemed to be a prevalent perception among Houstonians that a hurricane striking the Houston region during the summer made it more likely that a snowfall would occur that following winter. I based my post on a now-deleted blog entry from a now-retired local TV meteorologist, who at the time wrote: 

Is there a connection between southeast Texas hurricanes and snowstorms? I don't know. But I do know there are larger patterns within the atmosphere that we don't completely understand. If someone had a lot of time and a lot of money, this would be an interesting weather connection to investigate.

As it turns out, Space City Weather recently did some investigating about this phenomenon and determined that there is no such correlation: 

One of the most common winter forecast refrains I’ve heard since moving to Houston in 2012 is that if we have a hurricane in summer, we almost always have snow in the subsequent winter. 2021 saw us get hit with Hurricane Nicholas, so obviously that means we should get ready for an 1895-style snowball fight, right? In words of the great philosopher Lee Corso, “Not so fast my friend.”

After defining what "hit by a hurricane" means (and explaining, for example, why Harvey does not qualify), Eric Berger and Matt Lanza do some statistical analysis:

Since 1895, the data (plus 2021) suggests Houston has seen snow in 23 percent of all winters, hurricane or no hurricane. Based on the hurricane data, seven out of 30 winters following a hurricane hit since 1895 have seen snow, placing our odds at—wait for it—23 percent. The takeaway? It’s fun to say that Houston sees snow in winters following a hurricane. The statistics say that is false, and the odds of snow in a post-hurricane winter are perfectly identical to the odds of snow in any other winter.

The belief that a hurricane hitting Houston in the summer causes a snowfall the following winter is an example of illusory correlation; i.e. the perception of a relationship between two variables that in fact have none. While this particular one may have brought comfort to Houstonians - "the hurricane was horrible, but at least this means we'll get snow this winter" - it is nevertheless an illusion.

Bummer.

No comments: