*UPDATE: FSU has made a change, the press agent writes to me with this update:
The lead researcher on the hurricane forecast has requested a change to the headline to make it clearer, and I have made that change to the release posted on EurekAlert.
http://www.eurekalert.org/pub_releases/2012-05/fsu-hsi053112.php
The preferred headline is “Hurricane season is here, and FSU scientists predict a near-normal one.”
Kudos to them for making this change. – Anthony
==============================================================
The was released today from Florida State University sans Dr. Ryan Maue. Their forecast is at odds with NOAA’s near normal forecast and Klotzbach’s and Gray’s slightly below-average activity forecast(also released today). Given the history, I’m thinking we’ll see the media jump on the FSU “active” forecast headline and minimize the others, as “average” or “below normal” headlines don’t generate clicks and views. Numerically, they aren’t much different. (updated for clarity) – Anthony
Hurricane season is here, and FSU scientists predict an active one
Scientists at the Florida State University Center for Ocean-Atmospheric Prediction Studies (COAPS) have released their fourth annual Atlantic hurricane season forecast, using a unique computer model with a knack for predicting hurricanes with unprecedented accuracy.
Hurricane season begins June 1 and runs through Nov. 30.
This year’s forecast calls for a 70 percent probability of 10 to 16 named storms and five to nine hurricanes. The mean forecast is for 13 named storms, seven hurricanes, and an average accumulated cyclone energy — a measure of the strength and duration of storms — of 122. These numbers are based on 51 individual seasonal forecasts conducted since May 25, 2012, using sea surface temperatures predicted by National Oceanic and Atmospheric Administration.
The forecast mean numbers are slightly below the 1995-2010 average of 14 named storms and eight hurricanes, and reflect the possible emergence of El Niño conditions in the tropical Pacific and cooling surface water temperatures in the tropical North Atlantic.
“There is still uncertainty in the magnitude and timing of the emergence of the warming waters in the tropical Pacific along with the cooling of the tropical North Atlantic waters,” said lead scientist Timothy LaRow. “These factors combined will to a large extent dictate the level of tropical activity.”
LaRow and his colleagues at COAPS use a numerical climate model developed at Florida State to understand seasonal predictability of hurricane activity. The model is one of only a handful of numerical models in the world being used to study seasonal hurricane activity and is different from the statistical methods used by other seasonal hurricane forecasters. FSU is the only university in the United States that ussues [sic] a seasonal hurricane forecast using a global numerical atmospheric model.
The COAPS forecast is already gaining recognition for its accuracy only three years after its launch. The 2009 forecast predicted eight named storms and four hurricanes, and there ended up being nine named storms and three hurricanes that year. The 2010 forecast predicted 17 named storms and 10 hurricanes, and there were actually 19 named storms and 12 hurricanes. The 2011 forecast predicted an average of 17 named storms and nine hurricanes, and there were actually 19 named storms and seven hurricanes. Re-forecasts conducted using data since 1982 show that the model has a mean absolute error of 1.9 hurricanes and 2.3 named storms.
Related:
FEMA loses the global warming plot on hurricanes
NOAA predicts a near-normal 2012 Atlantic hurricane season
Klotzbach and Gray’s tropical forecast page
David, their record of predicting hurricanes is good so your cynicism is unfounded.
“average” or “below normal” headlines don’t generate clicks and views”
Well, in the long-term, when people get inured to man bites dog, dog bites man will become a story again.
I think using numbers of storms is no longer a good method given the way storms are named in the middle of nowhere. Instead the ace index is the way to go and at 70-90. Weatherbell is the lowest of the bunch. My forecast calls for over 2/3rds of the ace index to be attained north of 25 north this year and over 50% of it within 300 miles of the US coast. We have been touting in close development since April, had maps on line on May 4 of the development of Able of May of 1951 off the Carolinas saying we should look for that in 10-15 days , followed by a system out of the Caribbean. My point is that fast starts like this have happened before, like 1968 with Abby Brenda and Candy all by june 25 in 1968 and 4 by mid July in 1997 and the season shut down.
I presented a new technique idea at ICCC7 on using 400 mb temps in March over the tropics as a precursor matching up the lowest ace years with this year, so this will be an interesting test. At the very least, FSU, myself, and most of the other vendors put their forecast out in real time for all to see, rather than have research that is not seen until someone puts out a paper.
also showed how using the AMSU 400 mb temps, one can predict the tornado trend from the year before… even after the rip roaring start, our forecast called for activity overall to be near normal and shut down of this season. Folks on our paid site saw that.
Impossible to hide a decline when you are in the public eye and I applaud FSU and others for their straightforward ideas we can all see in real time.
I was on Cavuto tonight outlining the season and firing warning shots at anyone trying to attribute any long overdue major eastern hurricane to global warming. The nonsense about Irene last year was as bad as ignorance or deception ( they either do not know the history, or they do, and simply lie) on that matter as I have seen in the climate fight.
In the meantime, this should be very interesting. Dr Maue just joined us so the forecast is still my fault if its wrong… Ryan should not be blamed for my shortcomings, and he knows how much respect I have for FSU!
Is there a 2-handicap on the numbers? I mean “season” is designated so….
As an aside, I thought South Florida (specifically FTL) was rainy last year, but good god, it might have rained every single day of May; or atleast the past few weeks. Only difference is they are late afternoon…or Prevening for Big Bang Theory fans, like 3-5pm at start. If we get a slow moving and/or rain heavy tropical cyclone, it’s going to be bad.
The link to the FSU site now says “Near Normal”, not active.
Is there an answer to the question: Which scientific journals will publish peer reviewed research only after raw data and methods are submitted and archived — in any field?
Andrew says:
June 1, 2012 at 4:31 pm
Ian W says: “The counts are inflated by the arbitrary decision to ‘name’ a storm that is somewhere in mid-Atlantic and for a brief period is assessed as passing the wind speed parameter for Tropical storm – whereas in the recent past it would not even have been counted as a storm.“
They get named because they are Tropical Storms. We are not over counting now. We undercounted before.
“* The error bars on these projections/predictions/forecasts are very wide. So wide that anything within 4 of average can be claimed as success“
They openly state how uncertainty their predictions are. They want you to know how close they think they can get. What’s wrong with that?
“Tell us how many land-falling hurricanes there will be and the state of landfall – No?“
This is at the frontier of current forecasting ability. If you don’t like that science isn’t to the point that it can offer such predictions, too bad. They do their best.
The counting of tropical storms looks then to be similar to the counting of sunspots. Previously, a ship would have had to be in the storm and assessed the wind against the Beaufort scale as ‘force 8 fresh gale’ (yes that’s all it is) or had someone monitor the anemometer and think it was significant enough to report the storm. Today the wind around a tropical depression gets briefly to >34kts when assessed by satellite metrics and it is ‘counted’ as a tropical storm. For people using the ‘active years’ as proxies in support of hypotheses of rising storm numbers – such as insurance companies setting premiums – this can be totally misleading. Perhaps, as solar science has done with counting sunspots, meteorology needs to keep a ‘traditional count’ as well as the satellite era count so that valid comparisons are made? This for many may be an academic ‘game’ like predicting sea ice minima, however for Florida and Gulf Coast households with continually increasing hurricane insurance based on “the increasing number of storms” it can mean significant financial losses.
There is also a huge difference between the damage caused by a ‘tropical storm’ which in wind terms often would not even merit a mention in many areas of the world, and a hurricane. SWAGed numbers for the number of hurricanes or land-falling hurricanes are actually taken seriously by people and used. (see http://tomnelson.blogspot.com/2012/05/remember-when-one-hurricane-was.html — note WUWT and others in the sidebar ). If it is known that the skill in forecasting storm activity in the Atlantic during hurricane season “is at the frontier of current forecasting ability.” then say so when you put out the forecasts. It’s not a case of “If you don’t like that science isn’t to the point that it can offer such predictions, too bad. They do their best.“; it is a case of ensuring everyone, including insurance companies and politicians, understand that caveat. But I see no such caveats in what is being published one even states: “Three More Tropical System Landfalls Forecast in the U.S” there is no mention that this is little better than a guess. I realize the difficulties of attempting to forecast the behavior of a chaotic system when not all the inputs are known and the values of the known inputs are probabilistic forecasts. However, these uncertainties need to be made plain when the annual ‘hurricane season guessing game’ kicks off.
The only metric that should truly count and is a fairly steady known in the landfalling major category. We have storms like Oct 9th, last year, Sept 11 2009 in NJ, sept 28, 2008 SC or June 1982 that caused damage to our coasts, and should have been named. My point in all this is that the naming criteria in the atlantic where we did not see storms before should be a separate season so to speak as the uptick in numbers due to increased observations is being used as a propaganda tool in agw. Its like saying hurricanes are more damaging now being used as agw propaganda, when its because of more people and inflation. Somehow, we have to determine how to make sure the line in the sand is not changing. I have had several proposals out on this, but they are ignored. I am baffled as to why it seems like its easy to name a storm in the middle of nowhere, but a rapid feedback/transition system in our back yard is ignored until it cant be ignored anymore.
Ian W says: “If it is known that the skill in forecasting storm activity in the Atlantic during hurricane season “is at the frontier of current forecasting ability.” then say so when you put out the forecasts.”
They’ve been forecasting Atlantic activity with decent skill for years. What is at the frontier is when, where, and how many will make landfall. What portion of the storms that form will make landfall seems to vary quite a bit, probably because whether a storm makes landfall depends on instantaneous conditions, whereas aggregated storm activity depends fairly reliably on large scale seasonal conditions.
I agree that some forecast centers could stand to be more upfront about the uncertainties in landfalls.
“Perhaps, as solar science has done with counting sunspots, meteorology needs to keep a ‘traditional count’ as well as the satellite era count so that valid comparisons are made?”
Something very much like this has been done:
http://www.aoml.noaa.gov/hrd/Landsea/landsea-et-al-jclim2010.pdf
Turns out, without the short lived storms unlikely to be identified in the past-and accounting for increasing density of ship tracks (and decreases during the world wars), apparent trends in storm numbers disappear.
DHS Warns of … Hurricanes
“The Department of Homeland Security’s latest concern is hurricanes. With the start of the hurricane season, Homeland Security Secretary Janet Napolitano is encouraging a so-called “Whole Community Approach.”
http://www.weeklystandard.com/blogs/dhs-warns-hurricanes_646437.html