Ex NOAA/NWS Met forecasts Hurricane Seasons 4 years out

This is the general tropical discussion area. Anyone can take their shot at predicting a storms path.

Moderator: S2k Moderators

Forum rules

The posts in this forum are NOT official forecast and should not be used as such. They are just the opinion of the poster and may or may not be backed by sound meteorological data. They are NOT endorsed by any professional institution or STORM2K. For official information, please refer to products from the National Hurricane Center and National Weather Service.

Help Support Storm2K
Message
Author
User avatar
AJC3
Moderator-Pro Met
Moderator-Pro Met
Posts: 2648
Age: 53
Joined: Tue Aug 31, 2004 7:04 pm
Location: West Melbourne, Florida
Contact:

Ex NOAA/NWS Met forecasts Hurricane Seasons 4 years out

#1 Postby AJC3 » Wed Dec 18, 2013 4:45 pm

http://www.gainesville.com/article/2013 ... 9704?tc=cr

Ocala man forecasts hurricanes

By Joe Callahan
Staff writer
Published: Monday, December 16, 2013 at 1:23 p.m.

David Dilley has spent decades building a computerized weather forecast model that he says can predict the volatility of a hurricane season up to four years in advance. Dilley, 68, an Ocala resident who owns and operates Global Weather Oscillations Inc., recently unveiled his computer model concept, which he touts as a one-of-a-kind long-range forecasting tool that relies on weather cycles. The National Oceanic Atmospheric Administration uses several short-term weather cycle-type oscillation models — as well as La Nina or El Nino influences — to forecast six months to a year into the future. NOAA does not use weather cycle data to predict hurricanes four years out.

Dilley, a former NOAA meteorologist who worked in Boston for two decades, says his models can predict hurricane activity years ahead. He sells his expertise to clients such as insurance agencies. Those companies use the hurricane forecasts before deciding when, or whether, to expand into coastal markets down the road. Dilley said he has gathered decades of weather data that help identify specific weather cycles, which in turn help him predict the frequency of hurricanes in the Atlantic Ocean and the Gulf of Mexico.

Dilley says his models have accurately predicted hurricane activity in each of the past five seasons. Using the model, Dilley projects the activity in 11 different Atlantic and Gulf Coast zones. He has found that each of the zones has varying weather cycles — up to about 50 years each. And each zone's cycle has its own smaller weather cycle. Once all of the cycles within cycles are discovered, a pattern for each zone emerges. After analyzing the data, Dilley then projects hurricane and tropical storm probabilities for each of the 11 zones. Dilley said his agency, unlike the major prognosticators, predicted a slow season in 2013 and an active season in 2012. Dilley believes his prediction model, called “Climate Pulse Technology,” proves that weather cycles are the most accurate long-range hurricane forecast tool in the market.

Like NOAA, Colorado State University, which is known nationally as a leader in hurricane forecasting, predicted a very active hurricane season in 2013. It turned out to be one of the least active in decades. Because of CSU's inaccurate forecast, as well as a few others in recent years, lead researchers Phil Klotzbach and William Gray are having trouble raising research funding. CSU may bow out of the forecast game in February for the first time in three decades. “It's hard to ask for money when you have had your worst forecast,” Klotzbach told the Insurance Journal in a story published Nov. 21. “It is like batting .180 and then going into free agency.”

Dilley will not share every aspect of his computerized prediction system, or predictions in specific zones throughout the United States. That is information he sells to clients. Until he publishes his research, or can sit down with other meteorologists to share how he arrives at his forecasts, Dilley's claims will be in doubt by some in the science community. “I don't know if it is luck or real science,” said Scott Cordero, meteorologist in charge of the National Weather Service in Jacksonville. “But we, as scientists, have to be skeptical” of being able to predict hurricanes four years into the future.

Dilley is so confident in his system that he recently gave his 2014 hurricane season forecast. Dilley predicts that the 2014 hurricane season, which runs June 1 through Nov. 30, will be an above-average season. He said there will be 17 named storms and eight hurricanes, three of them Category 3 or greater.

Dilley said that while man is playing some role in global warming, he believes most climate changes are primarily attributable to weather cycles. He disputed the notion that hurricanes occur randomly and are impossible to predict. “They are no random hurricanes,” he noted. “Everything comes in cycles.”

Despite the fact some may doubt his forecasting, one insurance company has relied on Dilley's service since 2005. One CEO, who is now retired, said Dilley and Dilley's company, Global Weather Oscillations, possibly saved his company millions in 2011 and 2012.
It was just after 2004's hurricane onslaught when Florida insurance executive Don Cronin began looking for more accurate long-range hurricane forecasts. Forecasts from once-reliable sources, such as universities, were becoming unreliable, and Cronin needed a more accurate look into the future. After all, accurate hurricane forecasts along the Atlantic Ocean and Gulf Coast can help save insurance companies millions of dollars annually.

At the time, Cronin was the chief executive officer of United Property and Casualty, based in St. Petersburg. Cronin, now retired, heard a presentation from Dilley. Dilley spent decades researching weather cycles from the last century and gave a presentation to Cronin's staff in 2005. Cronin said the weather-cycle presentation made sense. He remembered thinking that, if it were true, his company could know years in advance of the potential expansion areas to avoid. Cronin hired Dilley's company and never regretted the move, he said. Cronin believes Dilley's forecasts likely saved United Property and Casualty millions of dollars in 2011 and 2012. Cronin began debating in 2010 whether to expand into Massachusetts the following year. The state had approved his insurance company's plans for expansion. Once he got approval, Cronin consulted Dilley and Dilley's company about the Northeast zone outlook.

Dilley's model predicted the Northeast would be a hot zone for hurricane activity in 2011. “I told him to wait (to expand) in 2011,” Dilley recalled. He also shared his concern about another active year in the Northeast in 2012. Based on Dilley's advice, Cronin postponed the expansion into Massachusetts. That year Hurricane Irene hit North Carolina, skirted the coast and crossed Long Island into Massachusetts, causing nearly $17 billion in damage. “It was because of David's forecast that I decided not to expand that year,” Cronin said. The company decided not to expand in 2012 either, in part because Dilley predicted another hot zone in the Northeast and the likelihood of a major storm. That storm turned out to be Sandy, which caused $70 billion in damage to the region, including New Jersey — where United Property and Casualty is now expanding. “It is important to the insurance industry to have that kind of insight,” Cronin said.

Though some people may not agree with Dilley's assertion that weather cycles can be used to predict hurricanes, Cronin relied on Dilley's data because it was more consistently accurate than Colorado State University's, “which was getting it wrong more often than not,” Cronin said. Dilley said he believes the time is now to get more people interested in his company. He believes emergency management officials, big-box stores and law enforcement could benefit. Dilley said it was in 2009 that he really began to see the accuracy of his models paying off.

And his work is gaining attention. Dilley is now working with Wall Street Network, which has built online risk management solutions for companies on a website called XtremeGis — http://www.xtremegis.com. Wall Street Network will use Dilley's company and Accuweather to compile outlook packages to sell to clients. On the XtremeGis site, there is a PowerPoint slide portraying one of Dilley's forecasts. Accuweather will supply the short-term forecasts and he'll supply the long-term forecasts, Dilley said of the partnership. “There is no doubt in my mind that weather comes in cycles,” Dilley noted. “And I think what I have going here proves that.”

For Cordera, the NOAA meteorologist in charge, he would love to sit down with Dilley and see his product. And if turns out to be as accurate as Dilley portrays, Cordera says he would be happy “to take a class” from Dilley. “I do have severe reservations about a product that claims it can predict (hurricane season) four years out,” Cordera said. “I wish his research was published. I would love to see it.”
0 likes   

Alyono
Professional-Met
Professional-Met
Posts: 6343
Joined: Fri Apr 26, 2013 3:52 pm
Location: Texas Coast

#2 Postby Alyono » Wed Dec 18, 2013 5:45 pm

as much as it would be of scientific benefit to publish... the guy is making $$. No need to change a viable business model
0 likes   

terstorm1012
S2K Supporter
S2K Supporter
Posts: 1273
Age: 36
Joined: Fri Sep 10, 2004 5:36 pm
Location: Millersburg, PA

Re: Ex NOAA/NWS Met forecasts Hurricane Seasons 4 years out

#3 Postby terstorm1012 » Mon Dec 23, 2013 11:48 am

Has he offered actual results compared with reality?

I have to say I'm pretty skeptical too, without more data.
0 likes   

Frank2
Category 5
Category 5
Posts: 3806
Joined: Mon Jul 25, 2005 12:47 pm
Location: Florida

#4 Postby Frank2 » Mon Dec 23, 2013 4:09 pm

His name is appropriate - that'd be a dilly of a fictional forecast if it's made 35,040 hours out (365 x 4 x 24)...

Per the below from the article:

"Like NOAA, Colorado State University, which is known nationally as a leader in hurricane forecasting, predicted a very active hurricane season in 2013. It turned out to be one of the least active in decades. Because of CSU's inaccurate forecast, as well as a few others in recent years, lead researchers Phil Klotzbach and William Gray are having trouble raising research funding. CSU may bow out of the forecast game in February for the first time in three decades. “It's hard to ask for money when you have had your worst forecast,” Klotzbach told the Insurance Journal in a story published Nov. 21. “It is like batting .180 and then going into free agency.”

On one hand I'm sorry to read that but on the other hand our guesses during the Fall proved correct - like their Asian counterparts, they have found it better to not continue with seasonal forecasts than to be consistenly wrong. I'm sure the media will not be happy to hear this non-news news (lol), because it means less sensationalism for them to captalize on come the May sweeps ("Forecasters predict above active season - what you need to know!"), and again, again - that's why the NOAA Directors of my time were very much against seasonal or long-range forecasting, because there are just too many variables that can direct affect the outcome. There might be every positive meteorlogical indicator that calls for an active season - but it just takes one volcano to change everything, so again, the Earth is just too complicated to make forecasts months or years in advance...
0 likes   

tolakram
Admin
Admin
Posts: 13322
Age: 54
Joined: Sun Aug 27, 2006 8:23 pm
Location: Florence, KY (name is Mark)

Re:

#5 Postby tolakram » Tue Dec 24, 2013 11:20 am

Frank2 wrote:His name is appropriate - that'd be a dilly of a fictional forecast if it's made 35,040 hours out (365 x 4 x 24)...

Per the below from the article:

"Like NOAA, Colorado State University, which is known nationally as a leader in hurricane forecasting, predicted a very active hurricane season in 2013. It turned out to be one of the least active in decades. Because of CSU's inaccurate forecast, as well as a few others in recent years, lead researchers Phil Klotzbach and William Gray are having trouble raising research funding. CSU may bow out of the forecast game in February for the first time in three decades. “It's hard to ask for money when you have had your worst forecast,” Klotzbach told the Insurance Journal in a story published Nov. 21. “It is like batting .180 and then going into free agency.”

On one hand I'm sorry to read that but on the other hand our guesses during the Fall proved correct - like their Asian counterparts, they have found it better to not continue with seasonal forecasts than to be consistenly wrong. I'm sure the media will not be happy to hear this non-news news (lol), because it means less sensationalism for them to captalize on come the May sweeps ("Forecasters predict above active season - what you need to know!"), and again, again - that's why the NOAA Directors of my time were very much against seasonal or long-range forecasting, because there are just too many variables that can direct affect the outcome. There might be every positive meteorlogical indicator that calls for an active season - but it just takes one volcano to change everything, so again, the Earth is just too complicated to make forecasts months or years in advance...


They did not find it "better", they simply lack the funds to do it as stated above. Sorry if I misread what you wrote.
0 likes   
M a r k
- - - - -

The posts in this forum are NOT official forecasts and should not be used as such. Posts are NOT endorsed by any professional institution or STORM2K.org. For official information and forecasts, please refer to NHC and NWS products.

User avatar
hurricanetrack
HurricaneTrack.com
HurricaneTrack.com
Posts: 1689
Joined: Tue Dec 02, 2003 10:46 pm
Location: Leland, North Carolina
Contact:

#6 Postby hurricanetrack » Tue Dec 24, 2013 2:18 pm

Nope, you read it correct. He wrote precisely this:

"hey have found it better to not continue with seasonal forecasts than to be consistently wrong"

I assure you Phil and Bill did not say, "you know, it is probably best that we stop issuing seasonal forecasts because we have been consistently wrong".

In fact, Phil indicated in the insurance journal article that due to the lack of success in the 2013 forecast that it has been difficult to raise funding going forward. To that I say too bad, and by that I mean, why give up now? You cannot get better at something if you do not keep trying. It may take another 30 years to have something that can really benefit the public but if no one is working at it, no matter the reason, then it will never happen. It IS a solvable problem, just like breaking the Sound Barrier was.
0 likes   

User avatar
Cyclenall
Category 5
Category 5
Posts: 6452
Joined: Thu Jun 08, 2006 10:01 pm
Location: Ontario Canada

Re: Ex NOAA/NWS Met forecasts Hurricane Seasons 4 years out

#7 Postby Cyclenall » Wed Dec 25, 2013 7:44 pm

If he was clever, he would have released the seasonal numbers at the same time the rest of the weather community did showing a slow season against every single other forecast. Then, by the end when everyone was wrong but him and his model he would rise to incredible fame in the TC sphere. His business would go up considerably. He would have to then have a track record and show some mets the process to further verify its not dumb luck.

17/8/3 is a relativity safe call for 2014. The only part I find highish is the 17 named storms but because every year now including 2013 has way more named storms, 17 isn't a stretch. I'll make note of this and when November 30th 2014 rolls around, I'll see if this guy has it in the bag or not.

Alyono wrote:as much as it would be of scientific benefit to publish... the guy is making $$. No need to change a viable business model

He is 68 and cannot stand not making a huge amount of money.

terstorm1012 wrote:Has he offered actual results compared with reality?

I have to say I'm pretty skeptical too, without more data.

He should do that and keep the more precise data for his business. Think of it as advertising.
0 likes   

Frank2
Category 5
Category 5
Posts: 3806
Joined: Mon Jul 25, 2005 12:47 pm
Location: Florida

Re: Ex NOAA/NWS Met forecasts Hurricane Seasons 4 years out

#8 Postby Frank2 » Thu Dec 26, 2013 2:49 pm

Some sharp comments here - I based my "better" term on the FACT that CSU stated that asking for funding would be "...like batting .180 and going into free agency", so based on that comment it's apparent CSU decided it better (or best) to not go forward, because in asking for funding right now puts themselves and the grantors in an awkward situation and is probably best resolved by adopting the "live to fight another day" solution by stepping back for now...

As we all know they had a busted forecast in 2006 and again this year, and really not very much better in all other years since 2007, and again no doubt it's because there are too many variables - as they used to say at NOAA, the "what happens if a butterfly flaps it's wings in the forest" way of thinking, that in nature there are too many things that can affect cyclone formation...

The sound barrier was another issue - the only fear was what would happen if it were crossed, similar to ships falling off the edge of the world in the old days, but those are constant laws of nature, unlike the many environmental variables that can affect a hurricane...

Frank
0 likes   

User avatar
Andrew92
S2K Supporter
S2K Supporter
Posts: 3233
Age: 33
Joined: Mon Jun 16, 2003 12:35 am
Location: Phoenix, Arizona

#9 Postby Andrew92 » Fri Dec 27, 2013 4:06 am

The following post is NOT official forecast and should not be used as such. It is just the opinion of the poster and may or may not be backed by sound meteorological data. It is NOT endorsed by any professional institution or storm2k.org. For official information, please refer to the NHC and NWS products.

Another one of my (in)famous long posts coming....be warned!

I have read this article and mulled over it for a little while and have come away with a mixed opinion. On one hand, I want to say that, if indeed true, it is quite impressive that he foresaw the Northeast taking significant hurricane hits in 2011 (Irene) and 2012 (Sandy). Also if true, it is very intriguing that he nearly nailed the 2013 season as being a lot quieter than everyone else predicted.

That said, I have a few reservations about his methodology. For one thing, I would like to see an archive of data going back to about when he first started taking this project on. I know he has one for 2013 that shows he predicted nine named storms with no hurricanes. On the other hand, I know that this is something that needs to be purchased in order to see the best data around, but without clear data going back then and comparing that data to the other agencies and what really happened each year, I scratch my head just a little. I did take a look at what the packages include and I also did not see that information anywhere in there. That said, a Google search for his name shows that he did accurately predict a strong El Nino coming into 2009.

One other thing that I would like to see is actual testimonials from his customers, but not on his website. What do his clients think of the predictions he has made? Can they attest to his accuracy that he totes about? Where has he come up short?

I especially ask questions like this because with the PDO suddenly going from -87 recently to reaching positive territory, and a lot of models picking up on this happening, that an El Nino appears to be on the way. Yet here we have someone predicting 17 named storms, with 8 hurricanes and 3 majors in the Atlantic for 2014. If this prediction is to pan out, the one way I could see this really be reasonable is if 2014 happens to be a Modoki year. It is very difficult to ignore that 1969 and 2004 fit that bill and yet were still similarly extremely active. So is this the type of situation that David Dilley is going with? Or is he going with a traditional El Nino that causes less favorable conditions, but produces lots of very short-lived storms in the Atlantic? (ala 2002, with maybe two months of the September-ish activity seen that year)

Then again, I know one member did mention on the El Nino thread that there are murmurs that indeed 2014 could be a Modoki year. I don't think we'll really know entirely what it's going to be like until it gets closer, but some kind of an El Nino seems way too likely. The last one was in 2009, five years ago, and the longest period in the satellite era between El Nino's of any type is five years, between 1997 and 2002.

All of this said, I haven't looked at his predictions for the years after, but I can say that I am very concerned about the 2015 season at a minimum. No hurricanes hit the US in 2013, and now 2014 just seems like it will have some kind of El Nino. I know in one post I mentioned that when this is the case, the Gulf of Mexico is the usual place watch in those years. Well, what about the future? It has been pretty much beaten to a pulp in some of my previous posts that the first year after a Modoki and the second year after a traditional El Nino have always seen at least one hurricane with a pressure of 960 mb or lower hit the US (my personal definition of a major hurricane, regardless of wind speed and Saffir/Simpson category, as just about every storm with that low a pressure has proven to be quite destructive).

But in this case, even in the case of a traditional El Nino, if the El Nino is over in 2015, you could look to examples of 1984, 1995, and 2003 as years that followed El Nino events in which the US was spared a hurricane in the last year before that event (if that makes sense). I'll break that down. 1984 was the first year after the El Nino years of 1982 and 1983, and no hurricanes reached the US in 1981. No hurricanes hit the US in 1990 either, and then came a long, four-year El Nino event that finally ended in early 1995. Finally, the US was spared any hurricanes in 2001, but 2002 saw an El Nino form that ended in early 2003.

What happened in 1984, 1995, and 2003? In 1984, it should have hit as a major hurricane, and I know I have seen the preliminary report as they were called then (now Tropical Cyclone Report) for Diana and am pretty sure I read that a coastal station did report winds of about 115 mph, though it may have also been on top of a tower of some sort, where winds are a bit higher. In any event, if not a major hit, it was a close to one as you can get without being one. Opal smashed in the Pensacola area in October 1995 as a category 3 storm with a pressure of 942 mb. Finally it was Isabel making headlines in 2003 in North Carolina and Virginia, striking as a category 2 in winds alright, but with a pressure of 957 mb, just meeting the 960 threshold. This is an admittedly very small sample size, but you could also look at this with 1963 being a Modoki year after 1962, when the US was spared, and see that 1964 that Hilda was a major for Louisiana and Dora was a borderline major for the Jacksonville area.

On the other hand, it is also quite possible that any event that takes shape in 2014 lasts another year or longer. However, reading into this bit of history tells me that 2015 will probably not be such a friendly year for the US if 2014 is any El Nino, traditional or Modoki. Then again, the years 1983 and 1992 would also have to be looked at, which featured Alicia and Andrew respectively, two storms I think we all know quite a bit about at this point - only further cementing my concerns for the US in 2015. Maybe going Modoki and seeing David Dilley's prediction of an active hurricane season would work out better, so long as the worst storms in 2014 stay far away from land. Perish the thought, but if 2014 is a traditional El Nino and quiet in the Atlantic, it could easily mean 2015 and 2016 could both be rough years.

I know I went off quite a bit from the topic of David Dilley to my own research, but the point of it is it just makes me wonder what he must say for that year, given what my own historical research is telling me for what 2014 and 2015 might be like. David does come across as a very smart man on the surface with quite the background in meteorology. I actually want to believe that he is onto something with this prediction. But I also need to think critically and wonder what his clients are saying about his past predictions, and if there is access to archived predictions and comparing those predictions with everyone else's for each year and reality. Saying the Northeast would take hard hits in 2011 and 2012 a year or two in advance has to mean he knows something that a lot of other people do not. I think I agree so far the most with Cyclenall that it's best to keep his predictions in mind and come back to it in a year. Then, when it comes, regardless of if he is right or wrong, we can look at his 2015 prediction and try it all over again. Finally, we repeat this cycle until we determine if indeed we should look at him more closely or not.

-Andrew92
0 likes   

tolakram
Admin
Admin
Posts: 13322
Age: 54
Joined: Sun Aug 27, 2006 8:23 pm
Location: Florence, KY (name is Mark)

Re: Ex NOAA/NWS Met forecasts Hurricane Seasons 4 years out

#10 Postby tolakram » Fri Dec 27, 2013 7:44 am

Frank2 wrote:Some sharp comments here - I based my "better" term on the FACT that CSU stated that asking for funding would be "...like batting .180 and going into free agency", so based on that comment it's apparent CSU decided it better (or best) to not go forward, because in asking for funding right now puts themselves and the grantors in an awkward situation and is probably best resolved by adopting the "live to fight another day" solution by stepping back for now...



They ARE asking for funding. :)
0 likes   
M a r k
- - - - -

The posts in this forum are NOT official forecasts and should not be used as such. Posts are NOT endorsed by any professional institution or STORM2K.org. For official information and forecasts, please refer to NHC and NWS products.


Return to “Talkin' Tropics”

Who is online

Users browsing this forum: Hurricaneman, joey, NONAME, Sciencerocks, SootyTern, toad strangler, Torgo, weathaguyry and 29 guests