NOAA hurricane forecast : 14-23 named storms

[TreasureIslandFLGal]

I think that folks are getting mixed up witih what NOAA has actually presented. Their product was not a forecast so much as a statistical solution based on a 70% likelihood of accuracy.
Based on the conditions and statistical representation of storm numbers, a bell curve of storm numbers was represented for the season. They were then given a criteria in which to report the resulting bell curve of probability. The numbers represent the high and low threshhold of numbers representing the volume of the statistical curve that results from the desired condition. One could also reach the conclusion, based on their statement, that there is a 15% probability of there being less storms than their lowest number, and a 15% likelihood that there will be more than 23 storms. In other words, we have a 30% chance that the actual number of storms this year will fall outside their statistically constrained number.
70% is pretty acceptable for accuracy. Would the public accept a 60% accuracy rate though? If they said, "we have a 60% chance that the numbers will be between, say, 16 and 21?" Or how do you think the public would react to a "forecast" of, "we have a 50% likelihood of having between 17-20" storms?" -that seems like a complete guess! -to those that don't understand the statistical bell curve.
What they have done is homed in on a statistically acceptable likelihood while still aiming for a publically acceptable constraint.

However, as a product to release to the public, they should have avoided the bell curve altogether and simply made a statement, such as, "there is a very high probability of an above normal to hyperactive storm season, where we could expect to see 4-9 more named storms than we would in a normal year."

(I just read this back to myself...boy am I a geek!!! But, I guess I did earn that A is Stats last year! haha. -and I've just seen how hard it is to try to explain it without using the formulas and drawings that would make it much easier to comprehend!)

[Nimbus]

So they are predicting an active season based on large scale long term patterns.
It would be great if they could show example years with similar initial early season conditions.
Does this look like a 2005 season setup?

[TreasureIslandFLGal]

It does look a lot liek the 2005 set up in many areas. Not exactly the same, but many key inputs are remarkably similar, or even more positive!
But, no year is exactly like another. We'll have to see just how the subtle differences play out over the season.
And we have to remember too that the season lasts months, and not just completely based on a single point in time when everything was measured. All the measurements will be different in another 2 months, and may prove to be more or less advantageous to a hyperactive storm season.

[KWT]

There are certainly factors that make this season seem like 2005, I still think something between 1995-1998 is a more likely outcome that being said, esp because both were La Nina seasons, unlike 2005 which was actually borderline El Nino during the summer (ENSO 3.4 was at 0.4C for most of it)

Still the general idea is that conditions aloft will be very condusive in general and SST's are pretty insane as well and thats why NOAA are so high on the upper end of thier forecast.

[tolakram]

I have no scientific data to back me up ... but I still say 2005 was 2005 because for whatever reason there were numerous opportunities (waves, fronts, other areas of low pressure) that were able to develop into tropical storms.

If we have perfect conditions but no spark then these forecasts will bust high. We just don't know what the exact conditions will be like. On the other hand if we have conditions aloft more favorable than 2005 and just as many opportunities it may be more active, which is very hard to imagine, but the odds of us experiencing the most active season ever in 2005 are still rather slim. We don't have a lot of data, on a planetary time scale, so we just don't know what the most active can be. Least active is easy

[KWT]

The reason why 2005 got above the 19-21 NS we've seen in other seasons pretty much was down to a very favourable NE Atlantic, where a good 6-7 or so storms managed to develop, I'd imagine the average could be as low as 1 for that part of the basin, maybe 2 at a push, so clearly 2005 was well above normal up in that part of the basin...

Those extra 6-7 storms are the difference between a 1933/1887/1995/1969 type numbers and 2005 type of numbers.

Simple as that IMO...

[Trader Ron]

William Gray was quoted yesterday, "The numbers are

going to go UP quite high" Are we looking at 20+ named

storms in next weeks update?

[Stormcenter]

Good post and I agree.

I think that folks are getting mixed up witih what NOAA has actually presented. Their product was not a forecast so much as a statistical solution based on a 70% likelihood of accuracy.
Based on the conditions and statistical representation of storm numbers, a bell curve of storm numbers was represented for the season. They were then given a criteria in which to report the resulting bell curve of probability. The numbers represent the high and low threshhold of numbers representing the volume of the statistical curve that results from the desired condition. One could also reach the conclusion, based on their statement, that there is a 15% probability of there being less storms than their lowest number, and a 15% likelihood that there will be more than 23 storms. In other words, we have a 30% chance that the actual number of storms this year will fall outside their statistically constrained number.
70% is pretty acceptable for accuracy. Would the public accept a 60% accuracy rate though? If they said, "we have a 60% chance that the numbers will be between, say, 16 and 21?" Or how do you think the public would react to a "forecast" of, "we have a 50% likelihood of having between 17-20" storms?" -that seems like a complete guess! -to those that don't understand the statistical bell curve.
What they have done is homed in on a statistically acceptable likelihood while still aiming for a publically acceptable constraint.

However, as a product to release to the public, they should have avoided the bell curve altogether and simply made a statement, such as, "there is a very high probability of an above normal to hyperactive storm season, where we could expect to see 4-9 more named storms than we would in a normal year."

(I just read this back to myself...boy am I a geek!!! But, I guess I did earn that A is Stats last year! haha. -and I've just seen how hard it is to try to explain it without using the formulas and drawings that would make it much easier to comprehend!)

[Stormcenter]

The funny thing is 10 years ago he would have never made a statement like that in my opinion. I can't see 20 storms this season but then again the NHC has lately been naming a lot of hybrid type storms they normally wouldn't have in the past so I guess it's possible. I still don't think it will be any close to the 2005 record. IMO

William Gray was quoted yesterday, "The numbers are

going to go UP quite high" Are we looking at 20+ named

storms in next weeks update?

[somethingfunny]

I've always thought these seasonal forecasts were worthless anyway. NOAA seems to be proving it with their range. As was said earlier, 1983 had four named storms, but one of them was Alicia. Why do we have seasonal forecasts?

[cycloneye]

This is what is more important than the numbers,the pattern shaping up for a busy season.

[Trader Ron]

The funny thing is 10 years ago he would have never made a statement like that in my opinion. I can't see 20 storms this season but then again the NHC has lately been naming a lot of hybrid type storms they normally wouldn't have in the past so I guess it's possible. I still don't think it will be any close to the 2005 record. IMO

William Gray was quoted yesterday, "The numbers are

going to go UP quite high" Are we looking at 20+ named

storms in next weeks update?

I answered my own question. I think CSU's update will be 20 named storms.

[wxman57]

I have no scientific data to back me up ... but I still say 2005 was 2005 because for whatever reason there were numerous opportunities (waves, fronts, other areas of low pressure) that were able to develop into tropical storms.

If we have perfect conditions but no spark then these forecasts will bust high. We just don't know what the exact conditions will be like. On the other hand if we have conditions aloft more favorable than 2005 and just as many opportunities it may be more active, which is very hard to imagine, but the odds of us experiencing the most active season ever in 2005 are still rather slim. We don't have a lot of data, on a planetary time scale, so we just don't know what the most active can be. Least active is easy

I agree the chances of 28 storms are slim. But FYI, we began counting the number of tropical disturbances (waves and other features) in 2005 to try to get an idea how many features are out there each season. The results were interesting. In 2005, we counted a total of only 65 waves and other disturbances. Of those, 28 developed into named storms. That's a very high percentage. In subsequent years the total number of waves/disturbances averaged about 70-80 per season. So there was no great increase in the number of disturbances in 2005, it's just that conditions were so favorable for development that many of them DID develop.

[wxman57]

The funny thing is 10 years ago he would have never made a statement like that in my opinion. I can't see 20 storms this season but then again the NHC has lately been naming a lot of hybrid type storms they normally wouldn't have in the past so I guess it's possible. I still don't think it will be any close to the 2005 record. IMO

William Gray was quoted yesterday, "The numbers are

going to go UP quite high" Are we looking at 20+ named

storms in next weeks update?

I answered my own question. I think CSU's update will be 20 named storms.

I think that's unlikely. I conversed with Phil today and for quite a while last week. He didn't give anything away as far as the June update, but I'd be surprised if he went over 17 named storms on next week's update.

[KWT]

17 NS would still be a big season, it'd be a top 10 season in terms of numbers, which is a big season!

Imagine in a normal pre-season period and hearing them go for something like 16-17 NS at this stage, it'd certainly be seen as amazing!

[Trader Ron]

17 NS would still be a big season, it'd be a top 10 season in terms of numbers, which is a big season!

Imagine in a normal pre-season period and hearing them go for something like 16-17 NS at this stage, it'd certainly be seen as amazing!

CSU is at 15 named storms. I think 16-17 is a given. I know they are very conservative with their updates. And yes, 16-17 is a HUGE number for any season.

[lonelymike]

It's nice to know what can happen, but what does happen is usually a whole different thing. I think NOAA should get out of the sensationalism business (ala Jeff Masters and Joe Bastardi) and stay in the business of explaining the science and the reason why. That's what their best at.

[tolakram]

I agree the chances of 28 storms are slim. But FYI, we began counting the number of tropical disturbances (waves and other features) in 2005 to try to get an idea how many features are out there each season. The results were interesting. In 2005, we counted a total of only 65 waves and other disturbances. Of those, 28 developed into named storms. That's a very high percentage. In subsequent years the total number of waves/disturbances averaged about 70-80 per season. So there was no great increase in the number of disturbances in 2005, it's just that conditions were so favorable for development that many of them DID develop.

That's a scary observation. So if conditions are as good as 2005 but we have the typical amount of disturbances this could be one heck of a year. Hopefully we'll find that La Nina years are somehow less conducive than neutral years.

[KWT]

Depends on how high it goes in terms of La Nina, typically La Nina's tend to be slower to kick off, though of course thats not always the case, much depends on the strength of the MJO waves as they come through.

Still I just can't see anything other then the ACE ending up hyperactive...but we shall see!

[CrazyC83]

Upper high in the NE Atlantic would bring up more SAL though, which would slow down activity, right?