Why Should I Care?

As I undertake final preparation on a presentation related to seasonal forecasting for next week’s American Meteorological Society Annual Meeting, it seemed like a good time to look back at last year’s seasonal forecast for the North Atlantic Hurricane Season.  In previous posts (1st, second), we looked at what various forecasting groups were saying the 2009 hurricane season would look like.  We have now put a month between us and the official end of the Atlantic Hurricane Season, so it is time to determine how well these forecasts panned out.

The 2009 season’s nine tropical storms and three hurricanes ended below both the ten year averages of 15.4 and 7.9, as well as the 50 year averages of 10.9 and 6.1.  The season was a bit strange in its lopsided nature.  The normal seasonal activity tends to follow the general pattern seen in the image to the right, with a peak sometime around September 10th.  While activity leading up to the peak seemed in-line with normal levels of tropical cyclone activity, the essentially dormant 2nd half of the season was certainly unusual, particularly given the first half performance.  For instance, in the previous 10 years, only the years of 2008 and 2004 had an imbalance even close in comparison.  Both of those years had 9 named storms between 7/1 and 9/10 and then 6 named storms through the rest of the official season, much closer to some definition of a ‘balanced season’.

Now it is likely that if you followed the hurricane season that you heard mention of El Niño taking root in the pacific ocean that this would likely influence the level of tropical cyclones.  This also appeared to influence the change we saw between spring and early summer forecasts that we discussed back in the June 15th post.  In cases where groups issued multiple forecasts, the levels of expected activity were reduced.  As you can see in the plot to the right (originally from the June 23rd post), with the exception of the UK Met Office, the forecasts were generally centered around 11 tropical cyclones.  Interestingly, all the forecasts contained 9 named storms within their predicted ranges, but none of the forecasts were centered near that result.

So the short answer is no one missed it, but then again no one nailed it either.  All the forecast means were within 17 to 24% of the 9 named storms.  Now when each of these groups verifies their forecast for the year, they will be able to stay that their forecasts were accurate to their own standards.  However, are their standards acceptable to you?  For that matter is the number of storms really the most important aspect? 

When this year’s hurricane season gets a little closer, we will likely take a look at another measure related to tropical cyclones called ACE (Accumulated Cyclone Energy).  This measure is really geared at trying to help us understand the total power within the hurricane season captured in the various storms.  It makes it a bit easier to compare hurricane seasons when for instance most of the storms are short lived and weak (like 2009′s 3 hurricanes and average storm lifespan of less than 5 days) and a year like 1975 that also had 9 named storms, but 6 of which were hurricanes and the average lifespan was in excess of 8 days.  We will also discuss why ACE itself (even though often used in the scientific community) may not be the best way to compare the power between storms and/or seasons.

Next time you will get a little inside coverage of the American Meteorological Society Annual Meeting, hopefully there will be some presentations more interesting than mine for me to update you on.  As always – read, learn, question and come to your own educated decisions!