Sunday, November 17, 2013

The 2013 Hurricane Season: something of a dud

No, this isn't about the University of Miami Hurricanes, though it easily could be, as the football team started the season strong but now has dropped three games in a row, including yesterday's loss to Duke, the first time the Blue Devils have beaten the Hurricanes since the latter joined the ACC. But I don't want to talk about that, about Hurricane football, because it's just too painful right now.

What this piece is about is the 2013 Atlantic hurricane season, the one that began on June 1st and will officially conclude in just about two weeks, on November 30th. Forecasters predicted an above average season this year. Dr. Gray at the Colorado State Department of Atmospheric Science predicted 18 named storms, with 9 of those reaching hurricane status, and 4 of those 9 reaching major hurricane status: 18 S, 9 H, 4 M. Forecasters at North Carolina State predicted 13-17 S, 7-10 H, 3-6 M. At Florida State University: 12-17 S, 5-10 H (they do not break out major hurricanes in their models). The Tropical Storm Risk consortium--a group of experts on modeling and risk management--predicted (after rounding) 16 S, 8 H, 4 M. Finally, the forecasters at NOAA predicted 13-20 S, 7-11 H, 3-6 M.

Overall, the predictions suggest at least 13 named storms, of which at least 7 would become hurricanes. And of those 7 hurricanes, the expectations were for at least 3 major (category 3 or higher) hurricanes. So how are they doing? Not too well. So far, there have been 12 named storms in the Atlantic basin (a number of which barely attained the required strength for naming and degenerated shortly afterwards). But only 2 developed into actual hurricanes: Humberto on September 8th in the far eastern Atlantic, and Ingrid on September 14th near the gulf coast of Mexico. Humberto was a category 1 hurricane for all of four days and never threatened land. Ingrid was a category 1 hurricane for only two days, yet it did do substantial damage when it made landfall in Le Pesca, Mexico (make no mistake, hurricanes are serious business, no matter their category).

Given that there seems little chance of realizing the predictions for hurricanes and major hurricanes, we should all--especially those of us on the Atlantic seaboard--breathe a sigh of relief. After all no one, or least no sane person, hopes for dangerous storms, wants to see more hurricane activity.

That said, we should be concerned about the apparent across-the-board failures of the predictive models in use here. No one did a good job predicting the 2103 hurricane season. Dr. Ryan Maue at WeatherBell Analytics maintains a running total of Accumulated Cyclone Energy (ACE), worldwide and by region. ACE is based not only only the strength of storms, but also on their longevity, how long they last. A look at the current totals indicates just how much various experts over-estimated the 2013 season. To date, the total ACE for the Atlantic basin is only 29, just 29% of the average year-to-date ACE of 99 for the years 1981-2012. The second chart on this page shows the average ACE on each day of the hurricane seasons. As should be obvious, late November is not a period that tends to add much to the total ACE. Thus, it is likely that the end total will remain close to where it is now, at around 30% of the historical average for ACE.

Yet forecasters and their models predicted an above average year for 2013. Given that the year has not only been below average--and far below predictions--in terms of numbers of storms and hurricanes, but has been deeply below average in terms of ACE, it seems reasonable to question the bases of these various models, to ask if the models are actually working in the least. I'm no expert on this stuff, just an interested observer, but it seems to me that the knowledge base of the experts--all of them--is lacking in some way. I know their models take into account a large number of variables, but maybe they're missing some. Maybe--just maybe--they've been missing some all along and have never really had a firm grasp on how to make these predictions, except in a superficial way. Maybe worldwide weather events are a little more difficult to predict, months before they are supposed to occur.

Cheers, all.