While the Atlantic hurricane season officially spans June 1through November 30, 87% of the activity historically occurs in the core months of August, September, and October, with a peak around September 10.
Several months ago, the forecast from several groups was for above average activity, and that has partially panned out (see “Four groups predict very active 2013 Atlantic hurricane season”).
The season thus far
Thus far, we have had four tropical storms, but no hurricanes. Looking at the past 50 years (1963-2012), the average formation dates of the fourth named storm and first hurricane are August 23 and August 10, so in terms of just quantity, the season is ahead of par.
But, in terms of Accumulated Cyclone Energy (ACE), 2013 is essentially average for this date. We’re at 6.6 compared to the 1981-2010 average of 7.1 for this date. The end-of-season norm is 104.
I’ve described ACE in several previous posts, but as a refresher, ACE is a standard metric used for objectively measuring tropical cyclone activity. It doesn’t take into account exactly how many hurricanes form, or exactly how long a storm was a major hurricane, or how large a storm was, but rather it simply squares the peak wind speed of any storm exceeding 40 mph at each 6-hour interval and adds them all up. Therefore, it’s possible that a very long-lived tropical storm could contribute more ACE than a short-lived low-end hurricane. This season, although we’re above average in terms of number of named storms, those storms haven’t lasted very long nor have they been very strong.
A look ahead
Looking ahead, we can use current conditions as well as short- and long-term predictors to guide our expectations. In terms of current conditions, I created a series of four maps (see below), each showing a difference from the average.of some atmospheric variable that influences tropical storm formation. ( The “current condition” is the past two weeks, and the “average” is those same two weeks but over the 1981-2010 period.)
Sea surface temperatures: somewhat favorable for development
In the figure above, the upper-left panel is sea surface temperature. SSTs are running generally above normal across much of the deep tropics, but only slightly; while the area around the Bahamas and Florida is quite a bit cooler than average. Warmer than normal sea surface temperatures act as fuel for storms.
Surface pressures: somewhat favorable for development
In the lower-left panel is surface pressure, and that is generally normal to low across the deep tropics, which is also a positive signal for hurricane formation. Lower than normal pressures promote rising air and the organization of storms.
Mid-level humidity: somewhat unfavorable for development
The upper-right panel is the mid-level humidity, which has been somewhat moist immediately off the coast of Africa (although recent dusty conditions may alter that), but then drier from the central Atlantic and especially into the western Caribbean. Dry air is unhealthy for storms, as it stabilizes the atmosphere.
Low-level moisture: somewhat unfavorable for development
Finally, the lower-right panel is the precipitable water, which is mostly a measure of low-level moisture. That too has been slightly high near the African coast (although recent dusty conditions may alter that), but then noticeably low across the rest of the tropical belt. Low-level moisture helps sustain developing storms.
No La Nina or El Nino (neutral conditions) – neither favorable nor unfavorable for storm development
Another player is the El Nino Southern Oscillation (ENSO) which varies on longer time scales. It has positive (El Nino), negative (La Nina) and neutral phases, all of which can have important effect on Atlantic tropical weather.
The temperature of the water in the equatorial eastern Pacific Ocean helps determine the ENSO phase. When the eastern tropical Pacific is cooler than average, ENSO is in a La Nina phase, and tropical Atlantic activity is enhanced; conversely, when that area is warmer than average (El Nino), tropical Atlantic activity is generally suppressed. When ENSO is neutral, tropical Atlantic activity is more of wild card.
Modeling groups around the world are working on improving predictions of ENSO, and the latest series of forecasts is shown below. It’s basically neutral now, and the models don’t show a big shift away from that anytime soon. Visually, there’s a consensus that through the end of this hurricane season (“SON” is September/October/November of 2013) the neutral phase should persist.
Global models: Minimal tropical activity projected in next 5-10 days, perhaps picking up mid-August
Global dynamical models such as GFS and ECMWF are run out for a couple of weeks, and as of today, there are minimal signs of any noteworthy activity anywhere in the Atlantic. Certain models or certain ensemble members from a model spin something up in the 5-10 day range, but there are no robust, multi-model, multi-run signals. Also, a huge plume of dry dusty air is streaming off of the African continent right now. This “Saharan Air Layer” will likely suppress any easterly wave development for a while (see “Huge African dust storm squashes Atlantic hurricane potential”).
The Madden-Julian Oscillation, or MJO, is an upper-atmospheric phenomenon that can enhance or suppress disturbances over a large area for weeks at a time. It can be tracked around the world.
In the figure below, the purple and red lines trace the observed phase and amplitude (location and strength) of the MJO since June 20, and the green and yellow lines are forecasts starting on July 30 and ending on August 13.
When the MJO is in phases 1-2, activity in the Atlantic typically picks up, and when it’s in phases 6-7, Atlantic activity is suppressed. In other phases, the effect is fairly neutral. This forecast is from the GFS ensemble model, and shows that the MJO is currently almost non-existent, but is forecast to head into phases 8 and/or 1 – this suggests that perhaps MJO-related forcing will act to enhance any activity in the Atlantic by mid-August.
Qualitatively combining all of these observations and forecasts, I would still expect the upcoming core of the season to be above average, but probably not significantly above average. Tomorrow, my good friends Bill Gray and Phil Klotzbach at Colorado State University will be issuing their much more statistically-based quantitative outlook for the remainder of the season.
* Brian McNoldy is a senior researcher at the University of Miami’s Rosenstiel School of Marine and Atmospheric Science.