It's Official! 2013 Slowest Hurricane Season In 3 Decades.

I am certainly thankful for the title of the tread. Concerning the round file it might be good to know what we are talking about.

Seasonal hurricane forecasts are empirical multiple regressions based on anticipated values for sea surface T, atmospheric moisture levels, and atmospheric wind shear (not a complete list). They are made months in advance. With the possible exception of the Florida State U model they are not directly based on mechanisms. As such they have little in common with physics and energy balance models used for climate projections.

Seasonal hurricane forecasts (all of them, made by everyone, apparently) are the ones that did so poorly for 2013 North Atlantic. There was a pretty good previous string, but not 2013. I honestly don't know if this is done for other hurricane basins.

Those, I presume, are Treb's subject here.

The other type are track and intensity forecasts for individual storms. They operate at small scales of time and space, and predict where particular storms will go, and how to develop. Their physics makes them similar to climate models (though they are not global, and focus on the massive short-term energy exchange between ocean and air, not 'external' atmospheric heating). These models have steadily improved. They are not all equal, and I guess nobody really understands why the European model is better than the rest.

Calling these things all the same, and discarding them all because one simply ignores the fundamental differences, is not necessarily praiseworthy.

It is also glaringly obvious that something (at least one thing) is missing from seasonal hurricane forecasts. Perhaps by looking at more hurricane basins (differently affected by ocean oscillations ENSO etc) they can be improved. Discarding them and starting over may not be the most productive way forward though.

 

A paper "Seasonal tropical cyclone forecasts" by Camargo et al. 2007 can be downloaded from the Colorado State hurricane site. It describes the situation as of 2007. Statistical models are produced for several basins, and links are included (though I have not tested they are still live). Dynamical models (more akin to climate models as I suggested above) were, at that time, not making seasonal forecasts of storm numbers and strength.

There are several other related papers linked at the same site, that may provide a lot more info about the statistical forecasting process. But it seems to me that they assess several factors (including SST and surface pressures in different regions), find a previous year that had similar conditions, and base predictions on that.

I hope it is clear to most of the readers here that this approach is quite different from a climate model.

Since the thread title did not restrict us to the N Atl basin, I looked again at the Weatherbell (Maue) website. The W N Pacific was only at 91% of the climatological ACE average value, even with Haiyan now included. Global ACE is 72% of average, so it's still a fortunate year. Except for those affected by the years' storms, of course.

The first forecasts for N Atl 2014 will come in early December, if they follow previous schedules. With the 2013 'bust' and ENSO still stuck near neutral, those are going to interesting to read...

 

My understanding in 2013 there was substantially more African dust over the Atlantic leading to lower surface temperatures. We've already seen the effect of reduced aerosols from the EU cleaning up their smokestack emissions. So one wonders why there was more African dust?

There were a couple of days after 9/11 when all USA airflights were suspended. I remember reading a report that the lack of contrails led to cooler than expected surface temperatures. Clouds apparently keep the heat in by suppressing radiation into space . . . much like CO{2}.

Bob Wilson

 

I have read also that there was a lot of dust this year. Don't quite see the direct mechanism, but more important: These stat models work by finding an earlier year that is similar to use as an example. We don't have seasonal total Sahara dust loads for very many years in the past. If that pops up as a significant factor, the stat models won't have much to work with.

You'd do better getting the opinion of Klotsbach or one of the other 'practioners' in this field though.

 

I don't know the mechanism but there have been many reports of other aerosol sources including volcanos contributing to cooling cycles. My thinking is CO{2} as a single molecules can absorb IR and re-radiate it. In contrast aerosols as significantly larger particles would require bulk warming to radiate anything like CO{2}. Convective cooling alone would prevent them from radiating IR.

Although Einstein said "God does not play dice . . ." the weather is not so predictable. Aerosol events may require periodic adjustments to the models but the basic physics remains the same. I've seen inflection points in the record versus the models that were attributed to volcanic aerosols.

Bob Wilson

 

Bob, no offense, but we may not have the horsepower here to figure out how to add dust to these seasonal forecast stat models.

Let Treb's thread be the place to discuss whether, after this year's notable failure of N Atl hurricane forecast models, an entirely different group of climate models ought to be round filed. That was his original thesis, and oddly enough, he hasn't been back.

 

I've got to say, dust sounds more like an excuse than an explanation.

In 2000 the idea was huricanes were difficult to predict, and had multidecadal cycles, that were not well understood. In 2005 the theory put forth was that ghg killed the conventional wisdom, and all that mattered was ghg. That would break the cycles and major huricanes would be worse in the future. Since that theory, we have... ta da. in the north atlantic an extremely quiet period. Its actually the quietest period in recorded history, but history is quite short. Now this new monkey wrench pretends to explain why the theory ghg causes huricanes is still right, but it must be the dust is getting in the way.

Its an excuse. We do have a great deal of activitity trying to model the saharan dust and ocean ocilations, but if all that mattered was the ghg, then shouldn't it be responsable for the ocillations that are then causing or not causing the extreme weather? I say its time to reject the ghg more major huricanes hypothesis. It was just there to scare people into legislation. Political.

Let's get down to honestly modeling these ocean oscilations, which appear to have little to do with ghg. That should both help get better climate models and better huricane prediction.

I would say climate modeling of the ocean definitely was a highlight of research to be done.

What should be round filed is trenberth for using grey liturature to pretend scientific evidence was ghg would cause more major huricanes in the future. You fix the models with better data, but please throw out the politics in the IPCC that chose grey liturature over scientific papers in this area. That means the chairman has to go. He decided to go with the grey when protests were made, and picked an industry study. Choosing Insurance reasearch to decide on bad weather risks, is like listening to the tobacco institute on the medical safety of ciggarettes. Anyone that does it should feel dirty.

 

Aerosols are often not predictable:

• volcanoes
• grounding all flights over the USA
• EU decides to clean up their smokestacks
• Perhaps another Asian country might make a similar decision?

I'm OK with a model getting 'adjusted' to reflect reality . . . those times when pure chance throws the dice in unusual distributions. God may not play dice but Mother Nature loves the game.

Well if he won't discuss the wealth of economists . . . <grins>

Bob Wilson

 

IPCC grey literature is OT here, so I apologize in advance.

Terrestrial processes is the one large section of the new IPCC report that I feel competent to assess, because I know the field well. There, grey lit was cited sparingly and appropriately. I talked about that earlier. It does not mean that such material was used appropriately throughout the report. Somebody could certainly examine the hurricane section (more broadly, extreme events) to see if they were done well.

Another area that seems appropriate for 'auditing' would be paleotemperatures over a one to few thousands year time scale. IPCC has stayed with 'conventional wisdom' here, they cite a lot, but I am not sure that they cite (and weight) each and every study appropriately. NIPCC and C3 headlines, in my opinion, cite selectively and interpret some studies in ways apparently contrary to the original authors. The point is, neither neither approach need be taken at face value.

The original data exist, various temperature proxies, various time intervals, various geographic coverages. The technique of meta analysis is pretty good, and AFAIK hasn't been applied to this area. It would yield a range of temperatures (probability distributions) for each slice of time and area. The distributions would be narrow where (when) the studies agree, and broader where they don't.

Not claiming it would be a simple task, and not volunteering. But the first step, collecting the primary publications, has already been done. Just combine the literature cited by IPCC, NIPCC, and C3headlines.

AustinGreen and I certainly agree that the oceans are inadequately represented in climate models. How to fix it is another matter. I think the hundreds of years of ENSO records (based on barometers in different locations) should be helpful. As I understand it, PDO and AMO are defined in similar ways so ought to present similar data histories.

The relatively early invention (and deployment) of mercury barometers. I can't quite say 'to the rescue' because I don't know enough about the modeling and for sure, it won't be me doing it. But there is a recurring theme in scientific research and I want to sermonize (briefly) about it. (1) some gadget gets invented (2) people make careful measurements and record them in some durable manner. With no particular thought about how the data may be used later, they are essentially stamp collectors (3) somebody else comes along later and develops a powerful analysis that could not have been done without those 'stamp collections'.

 

The IPCC's history has not been good with respect to grey literature. There are two cases that come to mind, and have been the source of the harshest critisicm. In both cases

1) The grey literature was not simply grey because peer review was not completed, it was written by groups that stood to gain from the port of view.
2) The chapter head lacked expertise in the area, but ignored scientific criticism by those that had expertise.
3) The IPCC chairman was brought into the discussion, and rejected the scientific critism, and stood by the chapter authors. He did not follow up with experts to determine if the grey lituratue even passed the smell test.

The two chief peices are the insurance lobbies notation on higher costs of huricanes (not higher frequencies), and an environmental groups pamphlet on melting of himilayan glaciers. The first piece is in the process of being discredited by peer reviewed science, and the new data of lower frequency of major huricanes. The second piece was disputed by the scientist reported to have made the claim. When asked he said he was misquoted and the melt rate was too fast. While the same chairman holds his post, and intends to include grey literature for political purposes, I don't trust any of it comming out of the IPCC. Trenberth, the chapter author on huricanes, was recently on the PBS newshour, and admitted that major huricanes may indeed have lower frequencies (but slightly more energy) in the future.

The cheif purpose of the hockey stick reconstruction was to put out the point of view that natural variation is very small compared to ghg. Our recent period 1997-2013, shows that we have had strong natural variation. We do not need paleo proxies here, but better modeling of this natural variation, primarily ocean oscilations, ocean mixing, and its effect on global tempeartures and weather.

Indeed North Atlantic Huricanes have a multidecadal pattern that is dependant on these oscilations especially ENSO and AMO, and these affect bob's drought conditions in west africa that cause more dust, which reducees huricanes. If you want to link ghg, you probably need to find a link between ghg and ENSO and/or AMO.

Data is sparse, approximately 120 years for enso, amo, temperature, ghg, and major huricanes. Smaller huricanes have even a shorter amount of history.
Hurricane Archive | Weather Underground
For atlantic huricanes, 1893-1900 was a particularly deadly time period, then 1926-1935, then a calm until 1963-1972, then calm until 1996-2005. Data is sparse.

+1

Yes and as a american, my extreme weather is highly correlated with ENSO and AMO. Whether these are affected by ghg or not is an interesting question, but modeling them better would help weather prediction, aiding things as far appart as water management and which crops to plant.

 

There is a new ENSO vs century-scale warming paper in Nature this week