Since my last update on Monday at noon, Erika has formed into a tropical storm and is tracking toward the Bahamas. Erika is currently 285 miles east of the island of Antigua in the western Caribbean. Tropical storm watches and warnings are out for many of the northern Caribbean islands. However, the headliner is that for the first time in three years, Florida is under the National Hurricane Center’s (NHC) forecast track cone of uncertainty. With a Florida landfall probability of 30%, there is a very real possibility that the 10-year Florida landfalling hurricane drought could end.

However, I must remind all readers about the overall uncertainties with forecasting hurricanes more than five days in advance. The uncertainty exists for a few reasons. First, the NHC track forecast errors over the past five years are 180 miles at day 4 and 240 miles at day 5, which are represented in the NHC cone of uncertainty. Secondly, intensity errors can also be large. Finally, this season is an El Niño year, which seem to be notorious for major hurricanes that start late in their careers. Betsy in 1965, Alicia in 1972, and Andrew in 1992 all took time to develop, but each El Niño year resulted in a major hurricane.

To demonstrate how a forecast can change look at what happen 10 years ago today as Katrina was moving off the Coast of Florida. In the loop below you will notice how the NHC forecast cone of uncertainty on Auguest 26th 11 am 2005 change in just a 12 hour period. This pivotal shift in Katrina’s forecast track occurred on Aug 26, 2005 moving from a FL panhandle landfall to LA/MS.

It should also be noted drastic forecast improvements of 40% have been made to hurricane track since 2005 which is show in the following plot of Katrina than and what it would be now.



The forecast models for Erika are currently all over the place. Many meteorologist and various media outlets may continue to play model roulette, but the fact remains that Erika’s forecasted track and intensity is uncertain. Further, as I highlighted on Monday, we have to consider Erika’s overall weather pattern in light of the fact that the peak of the hurricane season is just 15 days away. Forecasters need to consider El Nino characteristics which include tropical waves and storms that may struggle in the main development region of the Atlantic Ocean. But as these storm move closer to the U.S. coastline, they will be more likely to develop and strengthen. These factors can’t be seen in any one forecast model run.

With Erika still 1,500 miles away from Miami, tropical storm Erika is still disorganized. However, Erika has plenty of time get organized. At this time, several factors support a gradual strengthening of Erika:

  1. The vertical wind shear, which can tear storms apart, is not very strong. However, until Friday, wind shear will continue to limit development. So if Erika can survive wind shear this week, as discussed in Monday’s write up, the change in the weather systems will cause shear to decrease as Erika tracks closer to the U.S.
  2. Less environmental dry air is getting wrapped into the circulation, which can disrupt thunderstorm development.
  3. The ocean temperatures below the storm are becoming increasingly warm.
  4. The ocean heat content is increasing along the storm’s forecast track.

Lots of hurricane SST heat potential in the Bahamas, SST’s are 29-29.5 degrees Celsius. Could mean rapid intensification

Here’s how I’d break down Erika’s forecast options at the moment:

  • Death in the Caribbean (5% chance)  – Failure to reach the mainland U.S.; storm stays further south and tracks over Puerto Rico, Hispaniola. But it’s looking less likely at this point.
  • Weak tropical storm landfall in the Florida panhandle (15% chance) – Maintain a west-northwest heading, avoiding the Greater Antilles, and gradually intensifying and heading for south Florida as a tropical storm (then tracking into the Gulf of Mexico and becoming a hurricane along the west coast of Florida).
  • Hurricane landfall in south Florida (30% chance) – A Category 1 or 2 hurricane near south Florida on Monday (as suggested by the historically accurate European model (ECMWF), along with the high-resolution U.S. Hurricane Weather Research and Forecasting (HWRF) model). The HWRF model has been especially insistent on this.
  • Stalled hurricane (50%) chance. – Maybe at this time, the most likely scenario would be a hurricane tracking over the Bahamas near Florida where it will get trapped by weak atmospheric steering currents for a few days.



850mb wind and surface pressures from four dynamical models, all valid on Sunday evening. (


It is important to note “impacts” of hurricane can occur many hundreds miles away from the cone of uncertainty: rain, storm surge, and strong winds could certainly occur outside of this cone.    In the next post I can start to focus in on the insured impacts.

Fun Fact:
In addition to frequent flights into Erika by hurricane hunters, a NASA Global Hawk drone was scheduled to embark on a 24-hour mission in and around the storm on Wednesday.