2013 Hurricane Season Forecasts
June 3, 2013
In addition to the review of storm names for the season, it is also my custom to post a summary of forecasts of activity for the forthcoming hurricane season.
If one reads over my posts from past seasons, there will find some of my misgivings and caveats regarding these forecasts. The purpose of the annual posts isn’t necessarily to highlight the forecasts as an uber-important piece of data. It is merely a convenient one-stop summary of forecasts for the season and those of the recent years past.
The forecasts I track here come from the universities Colorado State, Florida State,North Carolina State as well as NOAA, the United Kingdom’s Meteorological office, and the consortium Tropical Storm Risk. Note that is not the authoritative collection of forecasts; as the years have gone by more and more organizations release forecasts. The forecasts I’ve opted to follow are for reasons of long track record and easily accessible forecast verification (in the cases of CSU, NOAA, and TSR), notable performance (NC State in 2006), and parochialism (FSU, though as we shall see, its performance has justified its presence here). While my compilation of forecasts from past season may give the impression that the UK forecast is quite new, that is not the case; for a number of seasons they issued a July-November forecast that was incompatible for comparison with the other forecasts.
While some forecasts provide more predictions than others (with variance in how the numbers are presented), I’ve reduced the forecasts to core numbers. That is, single numbers for predictions of named storms, hurricanes, major hurricanes, and Accumulated Cyclone Energy. Note that for organizations that issue multiple forecasts over the course of a year, I’ve used the “final” pre-season forecast. As a sort of performance baseline, I’ve also included the trailing average of the previous 5 seasons for the forecast parameters.
Without further rambling, here are the forecasts for 2008-2012, with links to NOAA’s season summary for each year.
|5 season trailing
|5 season trailing avg||160||154||119||102||111|
One misconception that some people have of these forecasts is that “every season is forecast to be above-average”. As one can see from 2009, when the inhibiting effects of the El Niño were correctly anticipated, that is not the case. It is the case, however, that we have been in an ongoing multi-decadal period of increased hurricane activity. As such, forecasts of activity, as well as actual activity, will be above-average. Relatedly, some think that when these forecasts fail, it’s because they consistently over-predict the level of activity and serve to “hype” the storm threat. The forecasts from 2012 show that to not be the case; everyone’s forecast under-predicted the level activity.
Something that one may note when reviewing these numbers is that there isn’t that much spread in the forecast numbers, especially in the case of storm numbers. There is a bit more appreciable differences in the forecasts for Accunulated Cyclone Energy. In its four seasons of existence, the forecast from FSU has had the most accurate ACE forecast three times (though it’s one failure was rather bad; it was the highest forecast in 2011).
Here are the forecasts for 2013:
|5 season trailing
As we can see, the forecasts are in general unanimous in the expectation of an active season; the consenus number of storms forecast matches the average of the past five seasons exactly. Again, one sees a bit of spread in the ACE forecast, Colorado State’s forecast is a fair bit higher than everyone else’s (though if one examines the range of numbers in NOAA’s forecast, it will be found that their “high-end” number is higher, yet). There’s a suggestion in the ACE forecasts that storms on aggregate will be a little more intense and/or longer lived than those of past seasons.
In general, forecasters do not see any obvious inhibiting factors for the season to come. Of the forecasts that have explicit input parameters (vice being based off a wide-scoped computer model), nearly all parameters therein are tilting to the plus side for storm activity.
As I mentioned earlier, there are many other forecasts besides these; the Washington Post’s Capital Weather Blog had a post that linked to many of the ones that I did not include here.