cheezyWXguy wrote:gatorcane wrote:Looks like the long-range Euro AI model and GFS busted big-time in the long-range with Erin moving west impacting Florida and then the Gulf or being in the vicinity of. They did, however, accurately predict a significant hurricane forming in the long-range. Modelling has a long way to go predicting not only genesis but track especially once you get beyond 168 hours (a week) and especially over 200. There is a reason why the NHC forecast and outlooks only go out 7 days. That being said, as we see just about every year, we can expect more major hurricane landfall hits for Florida in the long-range forecasts especially by the GFS (who some here call the "Giving Florida Storms" model). It's when we get within 180 hours and especially within 168 hours with the modeling that can be more concerning. Even then sometimes the track can shift within a 3-day forecast even up to the day of landfall.
I hope no one on here is getting too emotionally invested in landfall points for this system as of now, for specifically the reasons you mention. Main themes at this point are that the chances of storm formation look to be going up, and that formation looks to occur in an area where the likelihood of land impacts for someone is notably higher. Anything more specific or farther out in time is hardly more than entertainment value at this time. Those who live on hurricane prone coastlines should be getting prepared, if for no other reason than that we are in the climatologically favored period where storms like this can occur. Otherwise, plenty of time to watch the model wars.
I think AI/machine learning, if set up properly, will learn from its mistakes and get better.
