Climate Change

[Aramike]

And my ultra quicky response is that a year (or in this case what looks like a few months) does not a trend make.

Indeed.

However, on the geological timescales we're talking about, neither do a few decades.

[NeonSamurai]

Originally Posted by Skybird

In vain, Neon. Don't know how often I have pointed out the very same over the years. It simply gets ignored as if it never was explained by anyone.

I know, but I seem to have a penchant for tilting at windmills

Originally Posted by Aramike

Indeed.

However, on the geological timescales we're talking about, neither do a few decades.

Depends on which scales we are looking at, But yes at the geological level that is true. Though I would point out that Global Warming theories come from many different scientific fields including areas where looking at time periods of a decade or 2 is more appropriate.

[baggygreen]

One thing I haven't had a satisfactory answer to is this: Why is 1979 deemed so important? Why is the the norm only taken from 1979? Who says that what we're experiencing over the past 500 years is 'normal'?

the vostok ice core showed that the temperatures we're experiencing are certainly not normal in the context of the earth's past 400,000 years.

The average temperature for the past 400,000 years, which is shown by that ice core sample, shows we're currently at an temperature well above the mean. so why 1979??

We all know that if the earth hits an ice age we're buggered anyway.

what i think i'm trying to get at is ultimately, who is honestly arrogant enough to say that the planet now must stay withini a narrow band of temperatures, when it has changed so much in the past. Surely noone really thinks that we're going to be able to stop any apparent warming now, and then stop cooling later on??

[Skybird]

Sigh... I often said that over the years. It is not only about the extremes at the upper and lower limit that define the band inside which climate phenomeneons take place. It is about the speed of chnage in them. And what we currently see is an acceleration in warming of global temperatures that is so quick and fast that you cannot explain it with natural causes, only. This acceleration is often been described as a three digit factor.

The acceleration must have been caused by something outside the range of natural causes. And it seems that the beginning of this trend of drastically accelerated warming correlates with the the emission of industrial gas. This correlation is extremely high, meaning: the two points of time marking both, beginning of that acceleration, and beginning of industrialisation, are very, very close to each other. That high correlation makes the probability that both are just coincidence (random chance) extremely unlikely.


Attention everybody: Acceleration. Correlation. Try to engrave these two words in your longterm memories now, dear children. Two or three years of explaining it over and over and over and over again should be enough for even the most tiresome of minds. Certainly I become tired of explaining it over and over and over and over again myself.

[Torvald Von Mansee]

[tater]

^ does that mean that climate scientists consult statisticians and programmers, or do they just write their own code since they did take that FORTRAN class once back in the mid 80s?



tater

[NeonSamurai]

Many scientists are trained statisticians for statistical processes relating to their field. I sure as hell am, though psychology is rather stats heavy. And most labs will employ programmers too if they make use of computer programs that are not pre-built. Of course it is not unheard of for people to have more then one area of job training, I for example was a computer tech before entering my current profession.

[tater]

BTW, regarding the notion of "average temperature" that one might get from ice cores—it's not an observation, but a calculation. It is a value derived from proxies by assumptions, and is not a direct measurement. Ten people with ten assumptions could get differing temps from the same data. Calibration to the present might, or might not mean good calibration with periods farther back depending on the model. I'm quite wary of over generalized conclusions.

Even modern "average temp" is a calculated value. Weighted, too. It's not like someone takes a giant thermometer and shoves it in Bangladesh (or wherever the ******* of the Earth actually is) and gets a temperature measurement. Even satellite data has to be manipulated to provide such a figure—how does one weight over land vs ocean, does the satellite cover the whole earth, or does it miss certain regions due to orbital inclination?

So the notion of some long term global average temp is somewhat silly, IMO, particularly when the proxy is ice cores. What IS the proxy there, are they using layer thickness (what if the ice melted one year?)? Are they using dissolved CO2—doesn't that require assumptions about CO2 absorption leading or following warming (there are reasonable arguments that dissolved CO2 is released as a response to ocean warming, for example)? Tree ring proxies have certainly been troublesome statistically (they basically throw away any data that doesn't support their conclusion instead of throwing all the data in, and seeing if it works or not).

I'm not saying that a temperature history isn't interesting, but I think that the complexity of the system is such that a "global average" is not terribly meaningful. The average temperature of a star, for example, is similarly meaningless. The core temp is a function of the type of fusion cycle it is running, and the surface temp a function of convection, reradiation, etc. At least that temp is actually observable directly—though the corona is yet another region to look at. Complex systems and the idea of a single number to describe them are an odd combination.

[tater]

Originally Posted by NeonSamurai

Many scientists are trained statisticians for statistical processes relating to their field. I sure as hell am, though psychology is rather stats heavy. And most labs will employ programmers too if they make use of computer programs that are not pre-built. Of course it is not unheard of for people to have more then one area of job training, I for example was a computer tech before entering my current profession.

Note the CRU guys, and Mann, et al. You can look at their (horrible) code. They finally hired a guy who was a programmer to go over the code and try to clean it up. His "readme" notes are available to read. I feel sorry for the guy. Note that IPCC is based on work with the (still unfixed) code. It's amateurish at best.

They were not trained in either except whatever was required for their degrees. My astrophysics degree virtually "came with" a math minor, for example, though programming was entirely optional.

I'm not saying they are wrong, but they simply do not respond to reasonable criticism of their methodology. Heck, they don't even publish their methodology. They claimed that they were bombarded with FOIA requests and it was slowing their real work. They got maybe 3 or 4 such requests over a few years, and all were because they stated conclusions derived entirely from computer models and never once published the models, OR the exact data sets used. Yes, 90-something % of the data is online, but which they choose to use matters.

Since important, public policy is involved, is it not reasonable to ask for them to have to "show their work?"

I think they knew how awful their code looked, and didn't want anyone to see it, frankly. Hard to make sweeping statements with certainty when the remarks in your own code say to ignore vast stretches of time in the results, and that it will be added back by hand (using some completely invisible methodology).

I read many of the CRU emails when they came out, and it paints a terrible picture of their discipline, frankly, and does their goals no good. I frankly don't trust them at this point—least many of the major names that appear again and again in the literature that seems to drive the political end. I'm open to AGW, but I remain (appropriately, IMHO) skeptical.

Less politics, more, and more open science, IMO.

tater

PS—money is a powerful force to drive certain conclusions, and cannot be ignored. The amount spent on "climate science" is orders of magnitude higher than it used to be. A friend of mine who is an entomologist told me that a way increase chances to get funds is to tie research to "climate change." He does comparison of Asian and New World termites... but the money pours in if he simply adds some talk about differential climate change to the grant. It's certainly a motivator for confirmation bias. Back in the '80s, I knew tons of guys working on SDI. Everyone was writing SBIRs, etc to milk that cow, too. I know my wife's friends (she's a PhD toxicologist of physician) in medical research were all over retroviruses back when HIV was the cash cow, too. Money can be good, or corrupting.

[CaptainHaplo]

Tater - the readme you are talking about is a roadmap for how the data used was so faulty it boggles the mind. Plugging in temperatures from "sources" that had no identifier - either code or lat/long coords.... Who knows if they were even real - especially since a number of them came from data "subsets" that had no original sourcing. Heck, the problems the guy had just getting the data to be accepted by various routines to spit out subset results show how questionable the data itself was. I am still reading through it, but I am beyond appalled. Anyone with a modicum of knowledge on the subject could read that and be horrified that the data that came out was used.

[SteamWake]

Since the Eruption of the Volcano in Iceland I figured it was time to bring up this old topic again

Originally Posted by Financial Times

A key piece of evidence in climate change science was slammed as “exaggerated” on Wednesday by the UK’s leading statistician, in a vindication of claims that global warming sceptics have been making for years.

http://www.ft.com/cms/s/0/162b0c58-4...44feab49a.html

[August]

Originally Posted by SteamWake

http://www.ft.com/cms/s/0/162b0c58-4...44feab49a.html

Got to be registered to read the article.

[SteamWake]

Originally Posted by August

Got to be registered to read the article.

Huh your right.. you can get to it through drudge.

I would copy paste it but it has specific copyright language.

Originally Posted by Professor David Hand, president of the Royal Statistical Society

“It used a particular statistical technique that exaggerated the effect [of recent warming],” he said.

Thats pretty much it in a nutshell though.

[Gerald]

By,PAUL KRUGMAN

Why didn’t climate-change legislation get through the Senate? The triumph of greed and cowardice.


http://www.nytimes.com/2010/07/26/op...ml?ref=opinion

[Lord_magerius]

My bad...

*Turns oven down*