Personally, this is how I go about tropical forecasting.
First, the key to all forecasting is using the forecast funnel. Take a look at the big picture -- the really big picture. I typically look at a hemispheric chart. (I know, some of these steps are typical of mid-latitude forecasting, which is what I normally do, but they do hold a key for tropical systems) Upstream ridges and troughs can play a role on the position of fronts and subtropical ridges which can all steer a cyclone in different directions. You saw this a few days ago in an Irene advisory.
THE FUTURE TRACK OF IRENE DEPENDS VERY MUCH ON THE EVOLVING LONGWAVE
PATTERN IN THE MID LATITUDES LATER ON IN THE FORECAST.
SPECIFICALLY...A BLOCKING RIDGE OVER ALASKA AND AN AMPLIFICATION OF
A DOWNSTREAM TROF OVER THE NORTH CENTRAL/NORTHWESTERN U.S. WILL
DETERMINE HOW ROBUST THE WESTERN ATLANTIC RIDGE WILL BE AND HOW
LONG IT STAYS IN PLACE. AT THIS TIME...GIVEN THE GOOD AGREEMENT ON
THE LONGWAVE PATTERN IN THE NCEP AND ECWMF ENSEMBLE MEAN SOLUTIONS
WHICH DEPICT A CLOSED ANTICYCLONE TO FORM OVER THE WESTERN ATLANTIC
AT THAT RANGE...THE TRACK OF IRENE HAS BEEN SHIFTED FARTHER SOUTH
AND WEST.
Once the macroscale features are determined, then I look at the synoptic scale across the basin and the relationship between it and the larger features. This will include atmospheric moisture, shear, water temperatures, etc. I think incorporate all the model guidance out there. I begin to look at the outlier models and try to determine why they are going “against the grain” of the other models. Then I start ordering the reliability of each model versus what I have determined from the synoptic and macroscale features.
Once those features are established and an idea of a forecast “cone,” then I being to focus on that “cone.” I pretty much look at the same things, but at a much closer scale. At this point, I start incorporating smaller features such as some bands of dry air or warm water eddies and cool pools left from previous storms. Each will throw off the track and intensity.
Afterwards, I begin to focus on the storm itself. I figure out a storm position and Dvorak estimate and then compare that to what the other agencies are thinking. I bring all sorts of satellite imagery into the mix – including microwave imagery. Once I establish the local storm’s set-up, then I can apply that to my forecast and adjust as necessary.
When a storm is approaching landfall, additional small factors also play a big role. Microscale features across the land can produce large differences. Some locations can pick up huge amounts of rainfall compared to other nearby areas. The same goes for wind, surge, and severe weather. All of these have to be fine-tuned in the small, local scale.
Now, I personally believe two heads are better than one, so after I complete my forecast and its associated reasoning, I’ll then apply it to what other professionals are saying. Sometimes another met will point out something minute that I may have missed. If I agree with the reasoning, or rather, feel that it will be a large enough influence and then I’ll adjust my forecast as needed. However, if I feel the other parameters are more dominate, then I’ll stick with my guns. It’s all a delicate balancing act with science, common sense, and a little bit of art.
