WMO: ". . . we cannot at this time conclusively identify anthropogenic signals in past tropical cyclone data."

Tenuc,
Climate models and weather models are structured very similarly, so you are correct that there isn’t a huge amount of difference between how they are forecast.

Leif Svalgaard (13:44:47)
“as an average, the climate has a lot less variance than the weather which is what makes it more predictable.”
Not so, climate is just as variable as weather but on a larger scale over a longer time.

Ric Werme (21:51:41) :
Steve Goddard (16:11:40) :
> Hurricanes are caused by differences in energy between the tropics and the poles.
Hurricanes form in the Hadley cell, that’s separated from the polar cell by the mid-latiude cell. They’re really heat engines powered by the difference in temperature between low and high altiudes, and by the amount of evaporation due to the sea level temperature.
Extratropical storms (like nor’easters) are powered by baroclinicity – the temperature difference between north and south adjacent air masses.
> They are nature’s way of blowing off steam from the tropics and transferring it to the poles.
That they do, and at a Southern New England weather conference someone talked about warmer arctics the winter after some big hurricanes brought a lot of heat north.
> It is not coincidental that the 2007 record Arctic melt came after two very active hurricane seasons.
Well, except that wind and ocean currents we responsible for flushing a huge amount of ice out during the 2007 summer. There was more to the ice melt than just hurricanes.
My reply;
Hadley cells, polar cells, and mid-latitude cells were created as a model for textbooks to teach a systematic approach to understanding general global circulation, to beginning students.
There are not really any separate patterns of “bands of cells” that circulate the earth as you see on Saturn or Jupiter.
These basic constructs were created to attempt to explain thing to novices, in steps they could under stand, in separate lesson blocks. The fact that they are still an integral part of the standard models, shows the lack of flexibility of the standard academically trained mind.
The Lunar Declinational tides pull tropical air masses off of the tropics, as the moon crosses the equator, headed toward either of the poles, driving the Meridional flow surges that produce most of the severe weather, Tornadoes in spring / summer, then shift to Cyclones, hurricanes as the Solar apparent declination peaks Maximum North and shifts South, (in reference to N.H. where I live.)
The smooth flow of air perturbed off of the tropics, does not stop to shift gears, as it crosses latitude lines. If you sort tornado production by lunar declination rather than date, and further by the 18.6 year Mn declinational variation, you will find the same as I have that there is a repeating pattern to the timing of the occurrence of outbreaks of tornado production, in the N.H. spring summer months. That coincides with the period around the culmination of the Lunar declination plus about three days as the inertia of the atmospheric perturbation off of the Equator, continues as the moon reverses and heads back toward the equator, pulling the “polar air mass back side” of the tidal bulge with it. The shear between the two driving the tornado production.
On the down side past mid summer the heat built up in the tropical oceans causes a much more moisture laden tropical air mass to be perturbed, into a declining Solar atmospheric tide, that is shifting back toward the equator changing the latitude, of the confrontational ringing out of the moisture and energy.
The timing of the dynamics of both tornado and hurricane generation still follows the same Lunar declinational tidal driving periodicities, I have been watching this effect for more than 25 years, and cannot get anyone else to look at it seriously, because ( the text books don’t mention anything about Lunar declinational tides in the atmosphere! ).
Applying this feature of the global atmosphere to the models WILL add usable definition to the timing of occurrence, location of area to expect the production of tornadoes and hurricanes, as well as the estimations of their strength and duration.
If the timing of the synod conjunctions of the outer planets were to be included in the analysis, and forecasting method they can be seen to enhance the production number and intensity of tornadoes , and can be seen to regulate the periods of strengthening and intensification of hurricane / cyclones on a global scale.
And yes the swings in Lunar declination drives more / less tropical moisture and heat toward the poles, most noticeably around the time when the Lunar Declination at culmination is close to 23.5 degrees in phase with the solar apparent declination when the primary and secondary atmospheric tidal bulges from both, combine to produce maximum el nino effects that affect the glacial surges in the polar regions, accounting for the known ~nine year periods in glacial surge production.
To know this and not be listened to simply because of a lack of letter around my name is extremely frustrating. Sometimes the laughing due to ignorance or lack of critical thinking about what you <(Subjective, use if it applies) and people like Dr. Curry think they know just drives me nuts.
I am processing tornado data now, and finding some interesting results, I will post it as soon as I can get this da** software to accept 160,000 cells of data with out crashing, to plot graphs.

There are no prizes for being right in the real world. If right, you are more likely to get a kick in the arse, or be crucified.

For example, consider feedback. If the models are wrong one year about growth/decline of Arctic ice, that affects the next year’s ice, and it gets worse each year. A mistake in year one can mean the difference between an ice age and a tipping point.
The very fact that models are able to generate tipping points is a testament to their instability and inaccuracy.

Pamela,
Do the statistical models use the hockey stick as their historical reference?

Leif Svalgaard (14:26:33) :
That is not elementary that is a flawed assumption. The U.S. average temperature is reached through averaging local temperatures. If you are not calculating these accurately no accurate U.S. average in the future can be reached. If you attempting to determine these via some other method, that is a huge joke. There is no way around getting an accurate answer without first doing the local calculations. The fact that the local calculations are useless more than a few days out makes any climate calculation years in the future just as useless.
I am beginning to see more of the problem. It is a lack of computational understanding. Computers are only right and wrong. The margin or error used in empirical experimentation is not valid on computer models. There is no close enough. The reason is simple the real world laboratory is 100% accurate and thus our lack of understanding or measurement error is a human failing but the experiment still behaved perfectly to the natural laws of physics. In a computer simulation you are dealing with an imperfect laboratory and thus until it is perfect no valid results can be obtained let alone on the time scales computer climate models are running and all the fudge factors associated with them.
Weather models work on short time scales because these are essentially projections based on known data and physics. They become more accurate based on the more data they can track over a larger area. Climate modelers claim to do the same thing but they are not simply tracking known data and physics they are using subjective mechanisms and unproven theories in an imperfect virtual reality laboratory. They are nothing more than a theoretical exercise and meaningless to reality.

Leif Svalgaard (23:57:29) :
But they are not the sum knowledge of anything otherwise all climate models would be identical. Climate models are the subjective opinions of the scientists creating them. Some scientist determines this theory is right and this is wrong and thus gets the results they want. That is not science, that is manufacturing the results you want.

Leif Svalgaard (01:00:43) :
Reproducible experiments that can be independently verified are not manufactured. Regardless of a scientists motives this sort of verification prevents bias from unknowingly corrupting the scientific process. Running computer code is not an experiment.
But climate modelers are manufacturing the results they want through inclusion and omission of theories, intentional or not. I would say some scientists get satisfaction in the manufactured results of their models.
Again why are all climate models not identical?

Leif,
You have it backwards. Weather models are extremely accurate these days up to about three days into the future. Climate models have demonstrated zero skill in any time period. The Met Office uses climate models for their seasonal predictions, and so far have done considerably worse than a coin toss.