Nobel Prize-winning Intergovernmental Panel on Climate Change says wild weather will worsen

Lots of hypotheses,few validations.Corbyns theory is obviously valid and confirmed thousands of times.
He's basically banned from the Britsh press or rather censored.They can't threaten their political Cap&Trade stance with reality.
Currys website?
Id bet its because Corbyn isn't a climatologist but is from astrophysics.
That's some serious upstaging to acknowledge.

 

Can you show us what specifically his prediction was? How detailed was it?
Instead of telling us after an event that Corbyn got it right, why don't we list out his predictions ahead of time and keep tally?
Or are his predictions still only available to those that pay enough money?

 

Corbyn predictions are for pay. He did predict the russian heat wave. This was in the time period of a weather not a climate prediction, so I doubt this would help predict results of climate change. During the 2003 heat wave in europe there was also a blocking event. Some blamed warm sst, but the 2010 had more moderate sst temperatures. These were looked into in the NOAA analysis. Some like corbyn think these jet stream kinks are predictable by solar cycles. This may be a good avenue to investigate from a weather prediction point of view.

 

I guess my problem is that I actually read the detail and try to understand it. That's why my comments always seem to be so less certain and self-assured. I'll make this short.

A) Of course they left off 2010 when calculating trend. It says so right in the Realclimate posting.

B ) Moscow versus region temperature was directly addressed, third graph in the Realclimate posting. But to understand why there is disagreement ...

C) You'd have understand why Rahmstorf highlighted the issue of linear versus nonlinear trend (or my statement, that when you fit a straight line from 1888 to 2009, the linear trend is blind to the post-1980s temperature rise)

D) For the rest of it -- why doesn't Rahmstorf's purely statistical analysis incorporate "blocking" as a potential cause, the answer is, because it's a purely statistical analysis. There is no climate model in that analysis.

A deeper answer is, what is blocking? It's a description of the state of the atmosphere, which makes it kind of a stretch to call it a cause. It's one step removed from saying that Moscow had a long heat wave because a lot of hot air stayed around there for a long time.

E) Finally, I'll point out that the Dole analysis uses a general circulation model as the tool for the investigation. To give a wholehearted endorsement of the results here would seem, at least indirectly, to be an endorsement for the usefulness of GCMs.

 

That guy Corbyn's a piker. You want to see this guy:

Whitepapers - Weather Trends International

"Overall the U.S. accuracy was 87% for temperatures and 74% for precipitation but the U.K. was overall our most accurate country with 81% accurate temperatures and 87% accurate precipitation forecasts."

I mean, wow. Whatever that means.

Cosmic rays and solar magnetic fields? For-get-it. This guy's approach is based on

" ... a trade secret formula that has been constantly improved over the past 20 years. It’s fundamentally built in 100 plus years of history, math, statistics and about 100 other variables. The unique piece is that once we make a forecast we don’t change it, so it’s much more reliable, 80% or better, even though it’s a year-ahead forecast for thousands of locations in every country across the globe. "

Makes me wonder how many more long-term weather predicting organizations there are, and whether any of them uses anything other than a secret method. I mean, pretty cool that two organizations, using two apparently total unrelated methods, both produce super-accurate long-term forecasts. I'm not exactly sure what that tells me, but it seems to suggest that the actual details of the method ... apparently don't matter? Or something.

 

Well you are definitely self assured. Did you look at the graph you gave me, and the trend line? The trendline clearly includes 2010.
OK, I have now skimmed the paper to gather these facts.
http://www.pik-potsdam.de/~stefan/Publications/Nature/rahmstorf_coumou_2011.pdf
The graph never appears in the paper. In it they say they use this analysis but then redid the statistics because using 2010 would include "confirmation bias". The graph is therefore misleading since it obviously uses 2010 in the trend line. I don't think we need to go any further. I'll look at the graph again when it is corrected for this biases and others the authors state in their own paper. I'll leave my points 2 and 3 to things the authors looked at but are also to my understanding not included properly in this graph. Any conclusion you draw from biased dats is well, biased. In this case it makes the authors point seem much more forcefull and convincing, but it may still be convincing in a non-biased graphic.

Blocking is a description of the jet stream over a period of time. The thing that makes the russian heat wave important is the heat for a duration not simply that a new high temperature was reached. Climatologist gave causation to blocking for both this event and the 2003 european heat wave in which more people died. These two extreme weahter events are linked by climatologists. In 2003 theories that the blocking may have been caused by ghg came about and were put into models. These models did not indicate the russian warming blocking event. It may be that we need better models or that warming has little or no responsibility of the 2010 event.

The pnas paper does not attempt to model any physical climate variability other than a gausian distribution of temperature about a trendline of recorded temperature. If you are skeptical about blocking you can assign a probability to blocking being the cause and multiply this by the probability that it would not have been severe if temperatures had been lower then add the probability of other causes and high temperatures needed to replicate these. This is a physical model, and the authors seem to not attempt any physical model but gausian variation over temperatures. They also seem to only be hunting for high temperature records and not the duration of these heat waves that makes them weather events.

I'll give you an example of this type of reasoning. Say you believe stress causes ulcers and red headed wives cause stress. If for example there is another explanation that bacteria cause ulcers and stress may be contributing, it would be wrong to run a monte carlo model with only red headed wives and find there is a 80% chance that someones ulcer was caused by being married to someone with that hair color. We can test for the bacteria and it is always there for a certain type of ulcer and is not dependant on wives hair color. Now stress is a contributing factor but to ignore the bacteria and say the cause of Bob's ulcer is his red headed wife is reaching far beyond the pale. I am not saying that warming can be ruled out, but we do know the climatic pattern for these extreme events and it seems a bogus journey not to consider these in the hypothesis. In treatment of the problem, antibiotics seem much more effective than divorcing the wife. If this technique of simulation does actually predict events and help us deal with them I will find it usefull. If it simply is there to assign blame but can not accurately predict anything except there will be more record highs than lows, well duh, that isn't very interesting or important.

The claim is because simulation based on a limited model have a higher probability of coming to a single event that happened using only one degree of freedom, this degree of freedom is causation. This is something that is normally rejected in science. The alternative hypothesis that actually modeling all the variables including those that are believed to cause climate variation should be considered. We may be in a brave new world where the nobel prize winning pope decides the null hypothesis and stnadards of data. At a minimum though I would think any scientific investigation of the matter

 

For those that don't think atmospheric blocking has anything to do with the russian heat wave and does not need to be modeled, here is a layman's explanation in the IPCC friendly NYT.

Climate 'CSI' Team Takes on Russian Heat - NYTimes.com

It should be noted some in the climate community have been trying to find cause and effect for over 20 years.

The above is the most moldy analysis I can find. Here is the current NOAA description
http://www.esrl.noaa.gov/psd/csi/events/2010/russianheatwave/prelim.html

Note 4 standard deviations is much higher than any local warming from heat island or ghg,so if local warming or global warming are to be blamed for this extreme weather some mechanism is needed.

 

Even the "Old farmers Almanac" is right occasionally, and some people swear by it's prognostication!

Icarus

 

I'm on my cellphone and it hard to post.
Go back to my 1st post #12,and look at the link at the bottom.
This is a pdf which on the upper right is the Oct. forecast page released 30 days prior.For detail its astounding.Nothing vague or wiggle room.

 

mojo: "Currys website?
Id bet its because Corbyn isn't a climatologist but is from astrophysics.
That's some serious upstaging to acknowledge."

OK, but before betting you should take a look around there. Lots of things get discussed. Not just peer-reviewed things. One frequent poster there has is own theory about the structure o fthe sun. It does not intersect much with climatology (most of the time). For me, this undoes the astrophysicist upstaging argument for me.

Now, I can't force you to sign up there and add your assessment of Corbyn to their discussions. But I can't think of a reason why you'd not.

Seems to me you've given your best shot to Corbynizing PriusChat, and it hasn't panned out. People here like to see the details, or understand the mechanisms, ideally without paying the author's subscription fee. Yes, it sounds rough, but there it is.

But a credulous audience awaits you there, so why not? Bonus is that if Corbyn is right (or even almost) you can be among those honored for saving the world.

With this, and free-to-reanalyze temperature data, and several other things just laying there. I continue to wonder why AGW skeptics don't do more. We've heard the talk, but what about the walk?

 

Thanks, it was interesting to see.
So he gave no expected snowfall totals. His area for the storm missed the area where the snow fell, and hit a lot of areas where there was no snow. No reports of large hail in the south that I remember or could find.

It looks to me like he is using the fortune teller/psychic strategy. Tell the mark a bunch of things. If you are right on any of them, the mark will remember that one part and think you are really accurate.

 

Well, in a nutshell, there's the difference between you and me. To you, the trendline clearly includes 2010. To me, I think you can't possibly know that without having access to the data. Which I don't have.

This exact question was asked and answered in the posting.

"[Response: That's exactly what they did. They removed 2010 from the trend calculation and computed their numbers based on 1910 to 2009 and then also from 1880 to 2009. The 2010 value is not included--Jim]"

 

and how it is different from people who have a null hypothesis that global warming doesn't exist?

 

IMHO all graph shows that the chances of maximums are more likely during upturn then downturn. No more no less. Doesn't really prove that upturn is result of GW, not a part of natural cyclic pattern. BTW Russian climate variations had been linked to NAO cycles.. for example the harsh winters of 1940s were complemented by mild winters in east Canada.

Also not sure use of Moscow [urban] data is a good idea as there is too much man-made viability (not to bring the reduced in last decade methane and sulfur dioxide pollution); the study would have been more credible if it used data from wider geographical area.

 

The first point is well taken, and gets to the key question of what does "trend" mean here. (The second point is also well taken, but in practice the temperature trends were the same -- that's the third graph in the Realclimate posting).

On that first point, I don't want to fall into the either-or trap - either it's the AMO or it's global warming. As most of these studies are shaking out, what you see is that laying natural variation on top of trend leads to extreme events.

The reason I don't think AMO or some other periodic phenomenon is an adequate explanation is that the current AMO is unexceptional, while the Moscow temperatures broke a 70-year-old record by 3C (about 5.4F). That's a fairly large outlier, if the underlying long-term temperature trend really is stable.

So if you plot AMO either unsmoothed or smoothed:



Or smoothed, here:



The current AMO is not materially different from the values seen in the 1930s, yet Russian temperatures substantially exceeded the prior record.

One of the problems with AMO and other periodic phenomena is that you'll see people plot the raw sea surface temperatures and call that AMO. That combines long-term warming with the periodic cycle, and basically says there is no warming, there is only AMO -- which is absolutely not how scientists measure it.

So I don't think AMO, by itself, is an adequate explanation.

And I've dismissed the folks who, in effect, by using raw sea surface temperatures and calling that AMO, combine global warming with the cyclical phenomenon.

That said, while Rahmstorf's writeup was pretty neutral about this being an statistical analysis of trend plus variation, it is certainly being interpreted as a study of the impact of global warming. (It's not an attribution study, it's a statistical study, but ... that point is probably getting lost). And, interpreted as such, it would be an outlier. Most of the other true attribution studies have found that (e.g.) GHG increased have raised the odds of extreme events on the order of 20% or so, depending on the particular study. Unsurprising, as warming has barely started. While the measure here may or may not be directly comparable, it certainly is an outlier if it gets interpreted in the same light as true attribution studies. Maybe, if this is going to be interpreted as a global warming study, the more conservative approach would not be to evaluate relative to no trend, but relative to the 1930s temperature "hump" moved forward in time, but not increased. In effect, relative to just re-playing the prior AMO peak with no underlying trend. (You'll note that the trend line ends up about 2C above the prior 1930s trend). I'm going to post on Realclimate and ask him this question. If I get an answer, I'll post here.

 

I have extensive experience in modeling, so it is easy for me to tell from the curve shape that the trend line includes 2010. If you get some modeling experience you may be able to spot the shapes. I don't mean to be insulting here, and apologize if my tone is a little rude.

The paper also explains the source of the publicly available data, so no problem in getting the data and applying the algorithms. You do need to use information from the pnas paper though and not just the article if you want to try this for yourself. I did not do this work though, and you don't have to recreate the curve to confirm it included this bogus data. Its right there in the article that you said you read.

This is curious language in explanation. The paper explained that this would bias the data, making it invalid, but they chose to include it here. I assume they expect readers of realclimate to not be critical of bad data. Then they characterize the data containing the extreme whether event they are trying to model as just random noise. The paper if you read it had the authors analyse the urban heat effect and they estimated that it accounted for 1/3 of the warming, my point B was they should have included this in the graph and trend line.

 

You were right, I was wrong, the actual analysis properly excluded 2010, the graph in the article did not, and I misread that. As they state it, presumably the actual warming trend was a bit smaller near the end of the series than is shown in the graph in the posting. On the heat island effect, I believe you are still missing the point. While there is an annual effect, there doesn't seem to be one in July, for the time period they studied. It seems limited to the winter months. That's the bar chart that has been posted a couple of times already in this thread. So that subtracting the annual average, from the July value, was not sensible.

 

No, problem, I think the article was written with the intent for you to come away with that idea. I just was persistant in the point because sites like real climate, WUWT, etc seem to consistantly post misleading graphs when they could post the correct data, but it would not be as convinciyng. Again I don't think the author has any excuse for doing this, and it does make you look at the data differently. As I said before, the pnas paper clearly states this is wrong and does not use this biased statistical method.

The authors make a case in the paper that the urban heat island affect is overstated by GISS adjusted data. I agree with this point. You do need to read the pnas paper, and I posted a link earlier. They estimate that the urban heat island effect in moscow in july is about 1/3 of the giss estimates. I have not verified this number, and it is difficult to compute, but this is the number the authors estimate. The bar chart really does not tell you what the real effect is. I only am commenting that they author'a should use their estimate instead of 0 which they do in this chart. Now GISS provides adjusted and unadjusted data and it is not easy to make a july adjustment. In this case there is a fairly easy mathematical technique. You can take 2/3 of the unadjusted temperatures and add 1/3 of the adjusted temperatures to compute your July adjustments. This will not be nearly as good as making july only adjustments but it is much better than assuming 0. In the NOAA preliminary assessment that you referenced they suggest using the 17 stations in the Moscow area to get rid of some of the heat island affect. In the pnas paper the authors do give you enough information to do these adjustments and recalculate, for this objection.

 

by any means I did not try to frame discussion into either/or corner; they are not mutually exclusive, and in reality trends could and do combine.. we had somewhat similar discussion in spring on tornado topic. But let's be honest scientists must look at this sort of things to find the truth, and lawyers will look into them find excuses. Unfortunately many discussions with deniers fall into category "..somebody touched the wall thermostat is on, there is an e-mail, so the data aren't valid, you see?" Basically they come with preconceived motion and only looking for facts to support them.

Appreciate digging out info; I'd have to go over it after we finish wine

 

There was no response to my posting.