Klotzbach / Gray 2011 forecasts (August is up)

[cycloneye]

17/9/5 are their numbers of this first forecast for the 2011 Atlantic Hurricane Season.As TSR did,they point out the little skill this December forecast has.

http://hurricane.atmos.colostate.edu/Fo ... ec2010.pdf
Here is the abstract of this report.

ABSTRACT
Information obtained through November 2010 indicates that the 2011 Atlantic hurricane season has the potential to be quite active. We estimate that activity will remain well above average, approximately at levels that were experienced in the average of the years between 1995-2010 that did not have El Niño conditions. We expect to see approximately 17 named storms, 9 hurricanes and 5 major hurricanes occur during the 2011 hurricane season. These numbers are based on the average of our statistical model, our analog model and qualitative adjustments and insights. Although there is significant uncertainty at this long lead time, we believe that El Niño conditions are unlikely, given the current upper ocean heat content anomalies in the tropical Pacific. Because we are predicting an above-average hurricane season in 2011, the probability of U.S. and Caribbean major hurricane landfall is estimated to be above the long-period average. This forecast is based on an extended-range early December statistical prediction scheme that utilizes 58 years of past data. The influences of El Niño conditions are implicit in these predictor fields, and therefore we do not utilize a specific ENSO forecast as a predictor. We do not expect to see El Niño conditions reemerge in 2011. At this point, we are uncertain whether La Niña conditions or neutral conditions are more likely for the 2011 hurricane season. Sea surface temperatures in the far North Atlantic remain at record warm levels, indicating that the active phase of the thermohaline circulation and positive phase of the Atlantic Multi-Decadal Oscillation is expected to continue.

[wxman57]

I read the forecast. To summarize, Klotzbach/Gray (mostly K) found that the best results (for a Dec forecast) occur when 3 predictors are used:

1. Oct-Nov Sea Surface Temps in the northern North Atlantic. Warmer temps there signify a strong AMO circulation and warm SSTs throughout the tropics. For 2010, the Oct-Nov SSTs there were 3 STANDARD DEVIATIONS above normal. That’s quite a lot! It’s a strong predictor of an active season.
2. November 500mb heights over northern Greenland. Heights greater than normal there mean higher pressure way up north, and lower pressure way down south in the Subtropical Atlantic. This would mean a Negative North Atlantic Oscillation, something similar to the 2010 season. Weaker Bermuda High, weaker easterly trade winds, warmer SSTs in the Main Development Region. Pressures in November of 2010 there were +1.1 Standard Deviations above normal, an indication of a –NAO for the hurricane season.
3. November SLP in the Tropical Pacific. Pressures there in November were +1.6 Standard Deviations above normal. Higher pressure there means stronger easterly trades and a continuing La Nina (at least no El Nino) for the summer. Also means sinking air there and more rising air in the Caribbean/Gulf.

So all 3 main predictors are indicating that 2011 will be significantly more active than normal.

Oh, and best analog years were 1956, 1961, 1989, 1999 and 2008. Of those, only 1956 was pretty tame.

1961 – Carla on TX coast, Esther into eastern Long Island
1989 – Hugo into SC, Jerry into Galveston Bay (and 2 weak TS here)
1999 - Bret into south TX, Dennis and Floyd East U.S. Coast, Irene into south FL
2008 – I think we remember 2008.

[Ptarmigan]

I read the forecast. To summarize, Klotzbach/Gray (mostly K) found that the best results (for a Dec forecast) occur when 3 predictors are used:

1. Oct-Nov Sea Surface Temps in the northern North Atlantic. Warmer temps there signify a strong AMO circulation and warm SSTs throughout the tropics. For 2010, the Oct-Nov SSTs there were 3 STANDARD DEVIATIONS above normal. That’s quite a lot! It’s a strong predictor of an active season.
2. November 500mb heights over northern Greenland. Heights greater than normal there mean higher pressure way up north, and lower pressure way down south in the Subtropical Atlantic. This would mean a Negative North Atlantic Oscillation, something similar to the 2010 season. Weaker Bermuda High, weaker easterly trade winds, warmer SSTs in the Main Development Region. Pressures in November of 2010 there were +1.1 Standard Deviations above normal, an indication of a –NAO for the hurricane season.
3. November SLP in the Tropical Pacific. Pressures there in November were +1.6 Standard Deviations above normal. Higher pressure there means stronger easterly trades and a continuing La Nina (at least no El Nino) for the summer. Also means sinking air there and more rising air in the Caribbean/Gulf.

So all 3 main predictors are indicating that 2011 will be significantly more active than normal.

Oh, and best analog years were 1956, 1961, 1989, 1999 and 2008. Of those, only 1956 was pretty tame.

1961 – Carla on TX coast, Esther into eastern Long Island
1989 – Hugo into SC, Jerry into Galveston Bay (and 2 weak TS here)
1999 - Bret into south TX, Dennis and Floyd East U.S. Coast, Irene into south FL
2008 – I think we remember 2008.

1961, 1989, 1999 and 2008 were bad years. 1961, 1989, 1999, and 2008 were bad for Texas. 2011 has me concerned as a result. Chantal was a hurricane when it hit Texas. I remember it more as a rainmaker.

[frigidice77]

What about 2005 and 2010 for texas?

[wxman57]

What about 2005 and 2010 for texas?

Not sure what you mean.

[frigidice77]

Hurricane rita?hermine and alex?I think emily too.

[WeatherWiseGuy]

Well, I'll probably get skewered for stating this opinion, but after this season I've decided I don't really like these forecasts. There was a lot of breathless reporting of how busy this season would be, which it was, but to most average weather information consumers this was a totally un-remarkable season. To this day, I still hear people make statements akin to the following: "Well, there was a 60% chance of rain today, but we didn't get anything. Those weather people hardly ever get it right." Unfortunately, I see the seasonal hurricane forecasts following the same path. Until there is enough skill that a forecast can be made with reasonable accuracy that there is an increased threat of direct impacts to this area or that area, I think these forecasts are better left in the halls of academia where their relative merits can be studied.

[wxman57]

Hurricane rita?hermine and alex?I think emily too.

Neither 2005 nor 2010 is an analog for the upcoming 2011 season.

[wxman57]

Well, I'll probably get skewered for stating this opinion, but after this season I've decided I don't really like these forecasts. There was a lot of breathless reporting of how busy this season would be, which it was, but to most average weather information consumers this was a totally un-remarkable season. To this day, I still hear people make statements akin to the following: "Well, there was a 60% chance of rain today, but we didn't get anything. Those weather people hardly ever get it right." Unfortunately, I see the seasonal hurricane forecasts following the same path. Until there is enough skill that a forecast can be made with reasonable accuracy that there is an increased threat of direct impacts to this area or that area, I think these forecasts are better left in the halls of academia where their relative merits can be studied.

I actually agree with you that there's very little value to these forecasts. Should the general public take any extra precautions if a busy season is forecast? No, the same precautions should be taken each season regardless of how many storms are forecast. A very "busy season" doesn't mean that there is necessarily any greater chance of being hit than with a "quiet" season. We've had some very major hurricane landfalls during relatively inactive seasons (Andrew, Betsy, Alicia, Audrey).

Dr. Gray originally began his seasonal forecasts back in 1983, I believe. One goal was to produce a long-range seasonal forecast of activity that was better than climatology. The forecasts could be used by insurance companies to estimate potential claims in the upcoming season. I don't believe that he ever intended the general public to use these forecasts go gauge how they should prepare for a season.

Dr. Klotzbach has recently begun working on predicting not only how many storms might form, but what areas might be at higher risk for an impact during the upcoming season. There is a lot of work that still needs to be done in that area, though. Best we can do now is to look at where the storms hit during the best analog seasons and predict that those same areas might be hit in the current season. That didn't work out too well this year, as 1998 was our best analog year.

We did see quite a few similarities to 1998 in terms of the late start and the busy part of the season lasting through October, but there were no Gulf Coast or East Coast landfalls as there were in 1998.

Current seasonal forecasts are really just issued because people are naturally curious as to what the upcoming season will bring. Right or wrong, they equate forecasts of a busy season to an increased risk of being impacted. I look forward to the forecasts, myself, but I pay closest attention to the analog years rather than the predicted number of storms.

[Ptarmigan]

Well, I'll probably get skewered for stating this opinion, but after this season I've decided I don't really like these forecasts. There was a lot of breathless reporting of how busy this season would be, which it was, but to most average weather information consumers this was a totally un-remarkable season. To this day, I still hear people make statements akin to the following: "Well, there was a 60% chance of rain today, but we didn't get anything. Those weather people hardly ever get it right." Unfortunately, I see the seasonal hurricane forecasts following the same path. Until there is enough skill that a forecast can be made with reasonable accuracy that there is an increased threat of direct impacts to this area or that area, I think these forecasts are better left in the halls of academia where their relative merits can be studied.

I actually agree with you that there's very little value to these forecasts. Should the general public take any extra precautions if a busy season is forecast? No, the same precautions should be taken each season regardless of how many storms are forecast. A very "busy season" doesn't mean that there is necessarily any greater chance of being hit than with a "quiet" season. We've had some very major hurricane landfalls during relatively inactive seasons (Andrew, Betsy, Alicia, Audrey).

Dr. Gray originally began his seasonal forecasts back in 1983, I believe. One goal was to produce a long-range seasonal forecast of activity that was better than climatology. The forecasts could be used by insurance companies to estimate potential claims in the upcoming season. I don't believe that he ever intended the general public to use these forecasts go gauge how they should prepare for a season.

Dr. Klotzbach has recently begun working on predicting not only how many storms might form, but what areas might be at higher risk for an impact during the upcoming season. There is a lot of work that still needs to be done in that area, though. Best we can do now is to look at where the storms hit during the best analog seasons and predict that those same areas might be hit in the current season. That didn't work out too well this year, as 1998 was our best analog year.

We did see quite a few similarities to 1998 in terms of the late start and the busy part of the season lasting through October, but there were no Gulf Coast or East Coast landfalls as there were in 1998.

Current seasonal forecasts are really just issued because people are naturally curious as to what the upcoming season will bring. Right or wrong, they equate forecasts of a busy season to an increased risk of being impacted. I look forward to the forecasts, myself, but I pay closest attention to the analog years rather than the predicted number of storms.

I am always prepared active or not. It just takes one. Devastating storms have happened in less active seasons. Let see.
1900-Great Galveston Hurricane 7/3/2 ACE:84 No satellite, so it could be possible 1900 was more active.
1972-Agnes 7/3/0 ACE:28
1983-Alicia 4/3/1 ACE:17
1992-Andrew 7/4/1 ACE:75
1994-Alberto and Gordon. Should of been retired. 7/3/0 ACE:32

All of them were El Nino seasons.

I see 2010 more of 1999 because one storm formed in June, Arlene. Also, I see 2010 like 2001 season as the same reason I see 1999. Then it goes quiet until August with Bret and from there it is active. 1999 was a devastating season. 1998 started late and that was really a bad year because of Georges and Mitch. Some late starters were very active.
1950-Second most active after 2005 in terms of ACE. 13/11/8 ACE:243
1961-Carla 11/8/7 ACE:205
1969-Camille 18/12/5 ACE:158
1998-Georges and Mitch 14/10/3 ACE:182
2000 15/8/3 ACE:116
2004-Charley, Frances, Ivan, and Jeanne 15/9/6 ACE:224

I like using analogs too. As for 2004, I think its analog years would be 1886 and 1969. 1886 because multiple storms made landfall on Texas and Florida. Four hurricanes hit Texas, the most until Florida got hit by four in 2004. The reason I think 2004 is like 1969 is because they were El Nino and very active.

Analog Seasons Going Back to 2000. Seasons I think that would resemble other seasons.
2000: 1950, 1961, 1998
2001: 1999, 2010
2002: 1951, 1953, 1957, 1980
2003: 1952, 1958, 1964, 1970, 1984
2004: 1886, 1953, 1960, 1967, 1969
2005: 1887, 1933, 1995, 2010
2006: 1951, 1953, 1957, 2002
2007: 1936, 1971, 1990
2008: 1886, 1961, 1989
2009: 1997
2010: 1961, 1998, 1999, 2001, 2005

[Hurricane Jed]

This is crazy but something I've noticed. Since 1960 if Texas is hit by a large, powerful, multiple hurricanes or any combination of the three the following season is an inactive season, no truly significant hurricanes strike the U.S. or no hurricanes hit the U.S. at all.

1961-Carla, large and powerful storm. 1962-Inactive and no landfalling U.S. hurricanes.
1967-Beulah also a large and powerful storm. 1968-Fast start but only 8 storms that season. Only Gladys made landfall as a hurricane and it was weak.
1970-Celia, powerful Category 3 hurricane. 1971-Most active season of the decade, but Edith and Ginger the season's only landfalling hurricanes didn't have much of an impact on the U.S.
1980-Allen, large hurricane hit south Texas. 1981-Active but no landfalling hurricanes in the U.S.
1989- Hurricanes Chantal and Jerry hit the northern Texas coast. 1990-Active season but storms weak and no landfalling hurricanes or tropical storms.
1999- Bret, Category 3 strike on south Texas. 2000- Active but no landfalling U.S. hurricanes.
2005-Rita, large Category 3 storm strikes near Texas/Louisiana border. 2006-El Nino year with no U.S. hurricane landfalls.
2008- Hurricanes Dolly and Ike but especially Ike extremely destructive along the Texas coast. 2009-El Nino year with no U.S. hurricane landfalls.


Anyone think there is anyway to explain this or is it just probably coincidence?

[Category 5]

Got to speak to Dr. Gray in person earlier. Start out by saying, what a first class guy, very nice, took 15-20 minutes of his time to talk with me. Second, he's got quite the sense of humor. Third, we touched on 2011, he is thinking that we will be in a neutral ENSO and obviously he showed that with the 17/9/5 prediction

[CrazyC83]

I think good recent analogs for the upcoming season will be 1996, 1999 and 2004. My current thinking still holds at 15/9/6.

[srainhoutx]

Update due to release on April 6th...

[cycloneye]

IMO,they will downgrade by one in the three categories as TSR did.

[wxman57]

April update is out. He reduced the numbers to 16/9/5 (from 17/9/5):

http://hurricane.atmos.colostate.edu/Fo ... pr2011.pdf

[Ivanhater]

Florida folks will not like this tidbit from the forecast

As an example we find that the probability of Florida being hit by a major (Cat 3-4-5) hurricane this year is 34% which is substantially higher than the yearly climatological average of 21%.

[cycloneye]

Is the first time that I see them mentioning in a sentence where I am sitting.

For the island of Puerto Rico, the probability of a named storm, hurricane and major hurricane tracking within 50 miles of the island this year is 50%, 26%, and 8%,respectivly.

[OuterBanker]

What is also interesting is that the same areas in which Accuweather assigned as higer risk are also in their sights too.

Roughly 1.5 to 1.7 times an average year (Tx coast, South Florida and the Carolinas)

Yes, I do realize that these areas are always a higher risk. But this years risk is even higher than normal for them.

http://www.e-transit.org/hurricane/welcome.html

[Stormcenter]

This is just "My opinion" here but I have to believe the credibility of their annual predictions has gone down considerably in the last 10 years. Let's face it predicting where,when,how many and how strong hurricanes will be before a season even starts is like throwing darts blindfolded. It's just NOT an art anyone will ever master no matter how many super computer models you use. Like I said just my opinion.