There are some key ways in which a forecast for tropical storms and hurricanes is unlike any other forecast. First, the ability to predict when and where a tropical weather system will form before it actually does is close to nil. This is different from other low pressure systems like winter storms or spring frontal systems which are well predicted to form days or even over a week in advance. This is something we can't do with tropical systems very often.

The prediction of how soon an already existing tropical system will become a tropical storm or hurricane-- IF at all-- is I would estimate 50-50 at best and even less beyond 72 hours. A prediction for how strong a storm will be in 5 days is likely to be way off the mark, either forecast way too strong or much too weak. On occasion this will even be the case forecasting intensity just 12 hours in advance.

Almost all tropical cyclones eventually re-curve to the North and then Northeast, the hard part is determining when and where that re-curvature will take place. Very few head on a more or less strait West path for their entire existence. See the graphic.

From a statistical climate viewpoint 97% of Gulf hurricanes DO NOT hit New Orleans, and 95% of hurricanes in the Atlantic stay just off the East Coast.

Intensity forecasts over the 2-5 day period is, on average, off the mark by one WHOLE CATEGORY of storm classification. We do a much better more accurate job in forecasting track paths but there is uncertainty all along the forecast track. The hurricane center used to go out only 3 days, now they go out 5 days, and are behind the scenes experimenting with a 7 day track projection. However, keep in mind that the track error for the CENTER of the storm grows by about 45 miles per day in any direction!

Yet slow steady progress has been made. The accuracy of the 5 day forecast track has IMPROVED by 50% over the past 2 decades.

When you look on the web or on TV to those hurricane track forecasts and see the center line of the forecast, it is surrounded by a shaded error zone. This is known as the "cone of uncertainty".  The mid point is the forecast but could be off in distance and direction ANYWHERE in that shaded cone of error. At day 5 of the forecast the track error is on average OFF THE MARK by 225 miles! Please note: that "cone of uncertainty or average error zone" is not storm specific, to any storm being reported on or forecast but is based on the average of all track forecast errors from the previous year!

That is just one reason you (and the news media) should NOT focus so much on where the center or eye of a storm will strike land. The other good reason not to focus on just the center of the storm is that they are so large, the impact of tropical storm systems will be felt long before the center hits land and far away from that center. Typically the worst of a storm will be to the right of its path. So a storm CENTER passing well to your left could have a BIG IMPACT on your location, while a storm center passing very close to you BUT TO YOUR right may have MINIMAL EFFECTS on where you are located! But every storm is different.

As for Isaac at the time I type this some interesting notes: The current forecast track is CONTRARY to weather history. As 80% of the closest 10 matches to Isaac's location did not enter the Gulf of Mexico. However, top analogs to top similar overall weather patterns do support a track just West of the main Florida Peninsula. Much will depend on how Isaac interacts with the Caribbean Island Mountains which in the past have greatly disrupted storms and altered direction and intensity. The heat content of the region is very high however so there is great POTENTIAL for strengthening IF it does not get sheared apart by land first.

Numerical forecast guidance equations, including the GFS, European and UKMET models forecast a general west-northwest track right into Saturday.

After Saturday, the forecast guidance is in pretty good agreement in forecasting a gap in the ridge of high pressure to develop across Florida and the eastern Gulf of Mexico. This gap will cause Isaac to turn northwesterly by Sunday and Monday. The most recent run of the GFS model forecasts a landfall in the Florida Panhandle near Apalachicola on Tuesday night, which is a shift to the west of its previous forecasts. The latest European model forecast now forecasts a landfall near Pensacola on Wednesday, which is a shift to the east of its previous forecasts and is in good agreement with the GFS model. The UKMET model also seems to suggest a landfall somewhere between Mobile and Apalachicola on Wednesday. The latest track consensus guidance forecasts a track that takes Isaac across southwestern Haiti on Friday afternoon and evening and then along the northeastern coast of Cuba during Saturday. After that, the consensus guidance forecasts a track that takes Isaac right across the lower Florida Keys on Sunday night before tracking into the eastern Gulf of Mexico on Monday.

As of now looks like Isaac will make its final landfall on the Florida Panhandle between Fort Walton Beach and Panama City or around Apalachee by Wednesday morning. As I said since the ocean heat content in the eastern Gulf of Mexico is quite high significant intensification is quite possible. Should this occur, then a significant and formidable hurricane could be making landfall on the Florida Panhandle by Wednesday morning with the remnant low center heading toward Birmingham, AL Thursday morning with scattered heavy showers along and east of its path.

But we still can not rule out a path over or east of Florida yet. So all areas from southeastern Louisiana to the Florida Peninsula need to closely monitor the progress of Isaac as this is going to be a large storm in its overall size and its effects will be felt over a large geographical area; meaning, tropical storm conditions or at least Gale force winds may occur across much of the Florida Peninsula even though the storm is tracking just offshore. There is certainly the possibility of significant flooding in parts of FL, MS, AL, GA, NC, SC, TN DEPENDING on the future path and speed of Isaac or its remnant low after landfall.

BEWARE THE WEB... the internet lets anyone post anything including raw weather data or to make "mock up" weather data that's not real. One thing that happened a few days ago was because of the RNC being held in Tampa next week, a single computer model run showing a direct hurricane hit on Tampa Bay got a lot of play because somebody highlighted it. But that same day and the next day too, they could have just as easily posted the single computer model run that showed a CAT 5 Katrina-Camille strike on New Orleans almost to the day of the Anniversary of Katrina. But both of these models were extreme "outliers" and WERE NEVER close to the official forecast track by the hurricane center OR ANYBODY ELSE. But amateurs including people in the news business ran with it anyway.

Keep in mind there are over 50 numerical equation models we look at to forecast tropical weather systems, many of them not on the internet. Anybody would be a fool to pick out JUST ONE of those and post it to the web or believe it if they see it. Stick to meteorologists and you won't waste your time chasing a good headline but maybe not the best forecast.

Local officials had a mock hurricane drill for the RNC in Tampa back in May. In their exercise a category 3 storm with 111 mph winds was bearing down on the city on day 2 of the convention. Their recommendation was to cancel the rest of the convention.

It has been about 7 years since a major hurricane has hit Florida, that was Wilma in 2005. The Tampa/St. Pete metro area has been spared a major hit for 90 years. Back in 2004, Hurricane Charley was heading toward Tampa and then abruptly turned west and slammed into Port Charlotte.

I may not blog on this daily as I will be already giving you my thoughts on Isaac in the newscasts and local forecast on the radio, and things can change every 4-6 hours with each model run and new data from hurricane hunter aircraft and other types of probes sent into the storm like unmanned weather drones and robotic weather "boats".