Over the past week or two I have followed a few conversations on weather bulletin boards considering the accuracy of hurricane forecasts. Usually, if you hear a hurricane forecast, it is couched with mentions of how hurricane forecasts are very difficult to make.
Let’s use layman’s terms: inaccurate.
This is a given. It is a part of my field I wish was better – though I understand why it’s not.
After Hurricane Charley, there has been a lot of thought given to how people in the Fort Myers area weren’t prepared for the storm even though they were in the Hurricane Warning Area. Maybe… just maybe… it has something to do with how we on TV, and the Hurricane Center, display these forecasts.
The projected path is a line and is bordered by an area of possible error, which expands over time. The thought is, get rid of the line and just show the larger area that displays our margin of error.
It’s a good idea and I’m going to implement it in my system later today.
The problem with the line is, it makes people think we’re more accurate than we really are. Here’s an example of a similar situation. Last winter, the news director for one of my competitors told a magazine reporter that her meteorologists could predict snowfall to the fraction of an inch. What probably happened was, she saw computer outputs that went to 2 places to the right of the decimal point and thought, because it was printed that way, it would be that way!
In order to produce forecasts we often make assumptions and use numbers that imply more accuracy than we have. It’s my job to keep those numbers hidden and just show the usable, trustworthy results. She had seen something not meant as a final product and without any background in meteorology latched onto it.
That’s the problem with the track line. Yes, it’s in the center, but to dwell on it implies more accuracy than we really have.
Blogger’s note: I continue to display ‘live’ links to the latest hurricane info on the right side of this page. It’s a neat feed from the Hurricane Center which I update about 100 times a day.
One thought on “Showing Hurricane Forecasts Differently”
Speaking of inaccuracies, a certain channel whose name I won’t mention (not 8) seems to use their computer model equivalent as the be all and end all. They will say, almost definitively, that rain will occur when and where the model shows. It pains me, because I imagine the people who expect that to happen as stated. It rarely does. People don’t understand that it should be used as a general idea, nothing more.
I also think that the viewers blow certain scenarios out of proportion, therefore making the weathercasters look bad. For instance, I was in New Hampshire when Charlie and Bonnie hit the US. As their remnants moved up the East Coast and got to New Hampshire, I was informed that we were being hit by a hurricane. When we got about an hour’s worth of rain, I was informed that the weather people had no clue what they were talking about, since we had nothing like hurricane conditions.
Obviously, no one would have claimed a hurricane had been headed, full fledged to New Hampshire, but all it took was someone to miss the word “remnants” or disregard it, and voila! it suddenly has a very different meaning.