As you may have heard, some of the global weather models are insisting on tropical storm and hurricane development in the Atlantic Basin at some point during the next week. Some bloggers are posting forecasts 240 hours (or more) into the future and anticipating the possibility of a hurricane along at the East Coast at the end of next week.


Long-range computer models simulate the development of a tropical storm or hurricane approaching the U.S. in the 5-10 day time-frame. Such forecasts - while not implausible - must be viewed with considerable skepticism. (Left graphic from NOAA’s NCEP; right graphic from StormVistaWxModels.com)

Potentially threatening forecasts like these aren’t too surprising, given that we are approaching the most active phase of the hurricane season. It’s the time of year when some of these models in their long-range predictions are notoriously unafraid of spawning phantom tropical cyclones out of thin air.


New tropical disturbance “AL?” may develop in the eastern Atlantic. Another disturbance in the southern Caribbean, AL93, has a 30% chance of developing into a tropical storm in the next 48 hours. (National Hurricane Center)

But before we jump on the bullish forecasts for AL?, either in support of their guidance or in complete disregard of it, it is important to remember a couple of things.

Reliably predicting tropical cyclone genesis at any lead time, let alone properly capturing the large-scale environment surrounding it even a few days prior, is hardly skillful. So a high false alarm rate, though troubling, is not particularly unexpected. Yet, on the other hand, there have been some recent successes.

Just last week, when these same models advertised similarly suspicious tropical activity several days ahead of time, we saw three (albeit relatively innocuous) tropical storms: Emily, Franklin, and Gert.

And the first demise of Emily over Hispaniola (before it was born again over the Bahamas) was forecasted quite accurately by the Euro and NOGAPS prediction systems well ahead of time. But how about this go-around? Should we keep a particularly wary eye on the tropics right now?

To the extent that there are global-scale mechanisms in place – namely the Madden-Julian Oscillation (MJO) - that will likely support an active tropical cyclone period in the Atlantic Basin through the end of August, the answer is yes. In the upcoming days and weeks, the likelihood of development may indeed be higher than average for this time of year based on this information alone. Beyond that, however, there is way too much uncertainty to expect any particular thunderstorm cluster to mature into a tropical cyclone unless it’s already nearly there.

This is partly because, as mentioned in our last blog, the Atlantic Basin is highly capable of presenting insurmountable challenges to seemingly qualified candidate tropical systems. It is often too dry and too windy for their own good. Right now, a huge area of extremely dry Saharan air is comfortably situated in a large swath just north of the region AL? is expected to migrate through.

Any disturbance that dares to develop in the tropical Atlantic will have to deal with this harsh environment nearby. Consider, for the first time in recorded history, we’ve had seven tropical storms develop this year but not one has intensfied to hurricane strength. The tropical Atlantic has simply not been kind to these systems.

As for AL?, until those global weather forecasts form a broader multi-model consensus, both in their control runs and in their individual ensemble members, we should watch development in model projections more than a day or 2 away with a highly critical eye.