1 September 2016
Look at The Cone- NOT the line Inside It
Posted by Dan Satterfield

Hurricane Warnings are now posted for Florida, and concern is growing that a tropical storm will be moving up the East Coast beaches on Labor Day Weekend. I’ve been looking at different model runs tonight and there is actually a growing spread instead of agreement. NOAA runs a 12km model called the NAM (WRF core for geeks reading this) but they also run the model at a lower resolution a bunch more times and see what the track looks like. I am posting it below. This run is from 21 GMT Wednesday. Notice that the position varies A LOT!
Well, what’s your forecast?
I posted the NHC forecast above and I think it’s pretty good. There is one other model that is stalling the storm off of the Eastern Shore of Maryland and intensifying it (Where I just happen to be typing this, next to a snoozing Yorkie named Riley- probably the cutest dog on Earth, but still gathering data to support that hypothesis). Scary, but that’s just one run, of one model.
The take away here is to focus on the cone not the line. That cone is not there to look good, it is a mathematical expression of uncertainty. It’s width is based on average position errors in past NHC forecasts of tropical cyclones. You are taught in science that any result or measurement that does not have a calculation of the uncertainty with it is, WORTHLESS.
The NHC is doing this (although very few Americans realize it). Synoptic forecasters like me who work in broadcast, and local NWS forecaster give high deference to the NHC forecasts because that’s ALL they work on. If I disagree, I’d say so, but I’d still give them the NHC forecast, and tell viewers what I wrote in this paragraph. My job is to concentrate on the local effects, and believe me that will be tough enough.

Tide forecasts are now reflecting the winds…These will likely change of course, depending on storm track, intensity.