Satellite image of Category 4 Hurricane Joaquin over the central Bahamas on the morning of Oct. 1. (NASA)

The 2015 Atlantic hurricane season lived up to its forecast — the pre-June seasonal forecasts all predicted below-average activity, and that is indeed what occurred. But one storm in particular, Hurricane Joaquin, stood out in this strong El Niño season, which tends to create an unfavorable environment for hurricanes in the Atlantic.

The season ended with a total of 11 named storms, four of which became hurricanes, and two of those became major hurricanes (Category 3 or stronger on the Saffir-Simpson scale). The 1981-2010 average numbers are 11.5 named storms, 6.1 hurricanes, and 2.6 major hurricanes.

Storm summaries

Ana was a short-lived “pre-season” storm that formed during the second week of May near the Southeast coast and moved inland around the South Carolina-North Carolina border. Bill and Claudette were also very short-lived tropical storms, though Bill did hit Texas as a tropical storm.

In late August, we finally saw the season’s first hurricane, Danny, which briefly achieved major hurricane status prior to dissipating over the Leeward Islands due to strong vertical wind shear.

Erika formed in the central deep tropics, entered the Caribbean, and had the potential to strike Florida, but strong wind shear and the mountains of Hispaniola and Cuba caused it to dissipate.


Fred formed from a vigorous tropical wave that exited the west coast of Africa and quickly became a hurricane. It is the easternmost known hurricane formation in the tropical Atlantic, and it prompted the first-ever hurricane warnings for Cabo Verde — also known as the Cape Verde islands. Hurricane Fred also gave us the first opportunity to view a hurricane over Cabo Verde in satellite imagery.

Grace, Henri, Tropical Depression Nine, and Ida were all somewhat forgettable.

Then came Joaquin at the end of September. Joaquin formed northeast of the Bahamas, moved southwest toward the Bahamas, then stalled over the central Bahamas as it rapidly intensified to a Category 4 hurricane. It was the most intense Atlantic hurricane since Igor in 2010. The most recent Atlantic storm that was stronger that Joaquin was Felix in 2007.

For three days, Hurricane Joaquin was considered a threat to the East Coast — the Mid-Atlantic and Northeast in particular — due to the consensus in forecast models that the storm would eventually track northwest into the U.S. But the European model, which is often hailed as the superior global weather model that “nailed” the Hurricane Sandy forecast, was suggesting the storm would track harmlessly northeast into the open Atlantic.

[What the Euro model’s Joaquin ‘win’ means for weather forecasting]

Unfortunately, the National Hurricane Center’s cone of uncertainty — the cone that surrounds the track and gives a range of possibilities for the storm’s future path — is not determined by the uncertainty in the storm at hand, nor is it determined by the vast range of plausible model solutions. It actually represents the average error in Hurricane Center track forecasts over the past five years. Each year the forecasts get better, so the cone shrinks. But in the case of Joaquin, the cone should have been much, much larger to encompass the true range of uncertainty, and the forecasts that were calling for the storm to turn away from the coast. Instead, the Eastern Seaboard anxiously began to prepare for a potential hurricane landfall.

By the third day, it became apparent that Joaquin was not going to be such an East Coast threat, and the rest of the models began to turn the storm away from the U.S. The Mid-Atlantic and Northeast sighed a collective breath of relief.

[Hurricane Joaquin among the strongest storms on record in Bahamas]

But even without U.S. landfall, Hurricane Joaquin still managed to leave its mark. It devastated parts of the central Bahamas, where it lingered as a powerful storm. And then as the storm tracked north, a plume of deep tropical moisture streamed off of the storm for several days, resulting in catastrophic flooding in the Carolinas.

Joaquin was the perfect example of how it only takes one hurricane to make a season memorable. Even during an otherwise generally quiescent season, this storm found a window of perfect environmental conditions to explode into a beast, and it did so near land. Not to diminish the severe impact it had on the Bahamas, it was fortunate that the same storm didn’t hit a more populated area of the basin.

[The meteorology behind South Carolina’s 1,000-year rainfall event]

Finally, Kate developed in early November and eventually reached hurricane intensity, becoming the 40th known hurricane to form in the Atlantic during the month of November since 1851. It was upgraded to a hurricane on Nov. 11, making it the latest hurricane formation since Epsilon on Dec. 2, 2005, and it was also the strongest November tropical cyclone since Tomas in 2010.

For the second year in a row, there were no hurricanes anywhere in the Gulf of Mexico or the Caribbean Sea. Prior to these two years, you have to go back to 1994 for the last time such a strange lack of activity was observed in those normally-active areas. Interestingly, these past two years both had a below-average number of named storms, and the last two consecutive years like that were 1993 and 1994.

[Before and after: Hurricane Joaquin leaves behind astonishing cold wake]

Although the number of storms was near normal, the El Niño-induced wind shear that dominated much of the basin kept their lifetimes suppressed, so in terms of accumulated cyclone energy, the season’s activity was just about 59 percent of an average season. But Hurricane Joaquin alone contributed nearly half of that.


The U.S. also made it through another year without a major hurricane landfall, Category 3 or stronger. The last major hurricane to make landfall was Wilma in October 2005, so it has now been ten years and counting.

[10-year major hurricane drought may be ‘arbitrary,’ but it’s far from meaningless]

This unprecedented “major hurricane drought” only applies to the U.S. though — other countries have experienced very intense landfalling hurricanes since 2005, as the map below illustrates (not shown on here is a white dot over the central Bahamas for Joaquin). Another related record continues to get shattered: Florida has not been hit by a hurricane of any category since Wilma in 2005.


Some more interesting facts on the 2015 season, compiled by Phil Klotzbach at Colorado State University:

  • Four hurricanes formed in 2015, which brings the combined 2013-2015 total to 12 hurricanes. This is the lowest three-year total since 1992-1994 (11 hurricanes)
  • Two major hurricanes formed in 2015, which brings the combined 2013-2015 total to four. No three-year average has been lower since 1992-1994 (two major hurricanes)
  • September ACE was only 11. Three-year (2013-2015) summed September ACE for the Atlantic was only 44, the lowest since 1912-1914 when only 29 ACE was recorded during September
  • June-October-averaged 200-850-mb vertical wind shear in the Caribbean (10-20°N, 90-60°W) was 28.5 knots which was the strongest on record (since 1979)
  • Joaquin was the first Category 4-5 hurricane to impact the Bahamas during October since 1866

Environmental conditions

Throughout the season, and in common formation areas across the Atlantic, vertical wind shear was greater than average, which is a well-known side effect of a strong El Niño. The Caribbean was affected the most, but the Gulf of Mexico and much of the deep tropical Atlantic was affected by stronger-than-average wind shear as well.

Anomalous wind shear magnitude (color shading) and direction (vectors) average between August 21 and October 21. (NOAA/CPC)
Anomalous wind shear magnitude (color shading) and direction (vectors) averaged between Aug. 21, and Oct. 21. Blue shading indicates higher than average wind shear, and the values are in meters per second. (NOAA/CPC)

A unique and confounding feature of the season was the anomalous warming of the entire tropical Atlantic — the hurricane season started out with the central and eastern Atlantic much cooler than average, but by the end, it was quite a bit warmer than average.


Monthly average sea surface temperature anomalies. (NOAA/ESRL/PSD)

Zooming out a little, Phil Klotzbach noted that there has also been a region of significant cooling in the north-central Atlantic between June and October, and agrees that this pattern is a “conundrum”.


Difference in sea surface temperature anomaly between June and October 2015. (NOAA/ESRL/PSD)

Forecast skill

One thing we always try to emphasize in hurricane forecasts is uncertainty, whether it’s something specific like what the “cone of uncertainty” means, or situation-dependent atmospheric predictability.

But when the dust settles, intensity and track forecasts made by the National Hurricane Center as well as computer models can be verified by averaging the errors of all of the 24-hour forecasts, 48-hour forecasts, and so on out to five days. This season presented some challenges.

First, let’s look at the average track errors:


Average track errors for the 2015 hurricane season. CLP5 is the baseline “climatology & persistence”, OFCL is the National Hurricane Center, AVNI and EGRI are the GFS and UKMET global models, and HWFI and GHMI are the HWRF and GFDL regional hurricane models. (B.McNoldy)

The good news is that the National Hurricane Center and model guidance outperformed climatology — which means that they did better than average, basically. The bad news is that at all forecast hours, the Hurricane Center’s 2015 track forecasts were worse than their average errors over the past five years, shown above by the black diamonds.

And, incidentally, this year’s track errors will get factored into the size of next year’s cone of uncertainty (it will use errors from the 2011-2015 seasons), which means the cone will get larger.

For intensity, the news is somewhat better. The National Hurricane Center did end up with lower forecast errors compared to the 2010-2014 average out to three days, but was slightly worse beyond three days. The dashed lines on this chart just show the bias — the solid lines are the absolute error.


Average intensity errors for the 2015 hurricane season. SHF5 is the baseline “climatology & persistence”, OFCL is the National Hurricane Center, DSHP and LGEM are relatively simple statistical-dynamical models, and HWFI and GHMI are the HWRF and GFDL regional hurricane models. (B.McNoldy)

Of course, these verification statistics are preliminary, based on the real-time data. The Hurricane Center will perform a post-season check, which could adjust the number of verifying cases and the average errors, though there likely won’t be any dramatic shifts from what is presented above.

Looking ahead

Although we are in the midst of one of the most potent El Niños in recorded history, computer models generally agree that it will come to an end by early next summer, and as of now, there is an equal probability of a neutral year, or even a La Nina, going into the heart of the 2016 hurricane season.

ENSO forecast through next summer, with probabilities of warm, neutral, and cool phases. (IRI)
ENSO forecast through next summer, with probabilities of warm, neutral, and cool phases. (IRI)

The 2016 Atlantic hurricane season begins on June 1, and the first names on the list are Alex, Bonnie, and Colin, and the list also includes two new names: Ian and Tobias. Ian replaces Igor, and Tobias replaces Tomas, both of which were retired after the 2010 season.

(Angela Fritz contributed to this post.)