But one must understand how a models work, they are based on numerical grid data being generated by a computer using sonde data, surface obs, other models, some use Sat/Rad data and commercial aircraft data. All this data even though is great is still very sparse, especially over the ocean. On top of that the grid domain especially in the global models is very large. Tens of kilometers, even the meso models have a big domain in contrast to the tight small nuances of a tropical cyclone. So the models so a great job at time of forecasting big things, like Ridge and trough in the long wave pattern. They do a bad job of forecasting thunderstorms, lake effect snow, and hurricanes. Meso scale features that are outside the domain of most models. That¡¦s why models are ¡§guidance¡¨ and disecting them is as much of a art as a science. Plus you can never escape what is ¡§actually¡¨ happening on Satellite, radar and current surface and plane observation. One thing the NWS and NHC seem to forget! ƒº
In tropical cyclones you have to be aware of the current situation and the physics of what is actually taking place, boundary layer effects, AIR/LAND/SEA interaction and small scale features within the CDO and eye wall. All things no model will ever pick up, that were the skill of the forecaster comes in. I love all the modern technology I use everyday to forecast, but many Mets are blinded by the flood of information and forget to sometimes look out the window! So to speak, experience and trial and many errors makes you a better forecaster.
Boy if I had a dollar for every time the ¡§models¡¨ miss the CAD here in the Carolinas I¡¦d be rich. But I know what to look for in the model that hints at the CAD. I like to call it the GFS CAD, which looks nothing like a CAD, but there is a still a distinct signature to see.
I¡¦ve learned the same when it come to Tropical Meteorology
I need this storm to go away!
