Introducing the “Modelcane”


Last Wednesday, all the Media-rology lit up with a potential Category 4 hurricane in the Gulf (headed toward Texas, no less) just in time for Memorial Day weekend. The cover shot today shows what was going around. This was a forecast from the American model a good two weeks out. We call these hurricanes from the model a “modelcane” and I’ll explain what’s going on in a moment. Suffice it to say that, first of all, there has never been a major hurricane in the Gulf of Mexico in May, so that would be highly unusual, and second, never trust a model more than a week out, especially for tropical systems that develop in the Caribbean’s west end. Trust me, things will change.

In fact, here is the current American Model, which you can see now develops a weak 42mph tropical storm heading toward Florida. Quite the difference from last Wednesday:

This is for Saturday to the following Friday before Memorial Day weekend. Courtesy

UPDATE: The European model has now come into the game with a weak (40mph) tropical storm down in the Bay of Campeche NEXT Wednesday, the 25th and it doesn’t last, getting buried pretty quickly to the south. See the animation below and still shot.

40mph minimal tropical storm development

So what is going on?

This is fairly typical from May to November for modelcanes to show up. What is called the CAG, or Central American Gyre, is an area of upper level circulation that the American model can depict as developing at the surface. I’ve drawn out on a current upper level wind map how large this is:


That circulation is overplayed by the American model which suggests a possible spin at the surface where the water is plenty warm--right now 26-28°C.

courtesy NOAA

So, warm water warms the air above it, which rises, condenses and forms clouds and rain. This warm rising air is known as convection. The model continues to depict the rising, spinning, warm, moist air as a tropical system and with nothing to stop it, suddenly you have a monster hurricane in the Gulf. This is known as Convective Feedback bias. Any model is best when using its equations to solve a forecast of something that exists or is likely to, not something that might exist a week or ten days from now! The European model is not so quick to spin these systems up (today was the first EURO run that had anything) and it will be interesting to see if future runs continue with a tropical system or if it goes the way of all things.

One day we may be able to forecast a hurricane’s development and movement two weeks out, but that isn’t today. So remember when these “modelcanes” go viral on the internet, slow down and take a second look as to WHEN these are forecast to show up. The National Hurricane Center gets interested in tropical development five days out. When something shows up on their graphical forecast, pay attention!


Email me and follow me on Facebook!

About the Author:

KPRC 2's chief meteorologist with three decades of experience forecasting Houston's weather.