Entropic Man, as a Climate Scientist, you may consider it statistically significant that ALL Climate Scientists agree with your confidence in Climate Models that don't work, and can't be trusted to forecast or hindcast reliably against recorded evidence.

You have reduced Climate Science to levels of public confidence not even foreseen by tea leaf reading astrologers.

Climate Science might have some chance of evolution and survival, if only it could voluntarily reject all dependence on the Hockey Stick, which you are not allowed to do.

As someone who did trust Mann's Hockey Stick, because it was endorsed by International Science Experts of the prestigious United Nations funded IPCC, I can now appreciate how Climate Science is just damaging the credibility of all Science, and the United Nations plus all the people the United Nations was supposed to protect.

Climate Science needs grilling, roasting and toasting.

Steve Richards, supertroll

"EM, would you fly in a plane that was the result of a faulty model/simulator?"

I have.

"EM, would you drive a car whose breaking system had been modeled and manufactured via a faulty model/simulation?"

I have.

"So I will say, 'do you really really believe that anyone can use a faulty model/simulation for anything?'"

We all do.

As a software tester you already know that no complex software is perfect. It all has bugs and you never find them all.

You mention flight software. You will remember the Airbus that crashed full of journalists when the software locked into landing mode and refused to allow the pilot to overshoot.

The LA air traffic control system at LA crashed last year when a U2 entered it's airspace at a combination of height and speed which the software could not handle.

Last month the international airline check in system collapsed.

There is also fraud. How about the Volkswagen engineers who rigged the engine control software to clean up emissions under test conditions.

"It is interesting to note that the last time I looked, the market leader for testing systems LDRA did not work with fortran!!!"

Why not? Why was it published without the ability to test in a language which is still widely used?

Look at your own post. Your spell checker entered "except" where you intended to say "accept" and "breaking" where you intended to say "braking". "Modeled" should be spelled "modelled", unless you are an American.

Supertroll, your pension is already managed by an algorithm, as bug-ridden as all the rest. The same financial software manages cash flow between accounts and crashes regularly.

Come to think of it, our entire civilization is now controlled by algorithms.

Steve Richards, you cannot test for every possible permutation of inputs that these algorithms might encounter, nor can he guarantee that the systems analysis on which they are based is flawless. You therefore cannot guarantee that they do not have major bugs.

The whole concept of perfect software is complete bollocks. I know it and you know it.

Golf Charlie

"Climate Models that don't work, and can't be trusted to forecast or hindcast reliably against recorded evidence."

You are mistaken. The models hindcast reliably. Over the timescale from run dates to the present they have forecast reliability.

That is how the validity of the models has been tested and confirmed.

So why in the bowels of Michael Mann EM do you trust climate models to be able to remotely predict the climate of the future when they have consistently been unable to match the past without resort to fudges? Or changing the past to fit.

"The whole concept of perfect software is complete bollocks. I know it and you know it.".
The software that allows me to watch pay-for TV works pretty well in my experience, and that used to send me utility bills is near perfect, dammit!

EM, is this as close as we get to you admitting GCMs are faulty?

Re LDRA not producing a tool to verify fortran software: scientific types have no clue when it comes validating their software in the normal accepted sense, so do not buy said s/w.

I have encountered many times genius level software developers who tell me they never make mistakes! Its quite common the higher up the tree you go.

You admit to knowledge of s/w induced catastrophic faults in modern passenger aircraft. Even with all of the checking and verification that goes into the development of flight read s/w - DO-178B and all of the chores surrounding compliance and the extra workload generated, faults do get through and problems arise.

Serious s/w developers use a variety of techniques to minimise bugs.

The first being every function/procedure/class is tested in isolation, just on its own with every concievable input value, and its outputs are tested - this is called unit testing.

The main program can be built up from a collection of these functions - we could call the collections module, we do the testing at module level, a bit more difficult due to all of these pesky functions inside.

The main program could be built up of these modules, which include all of these functions inside.

The testing of the main program is now much more difficult since we will have few inputs to vary and not many outputs to check.

Now, do you think that climateers go to all of this trouble to test/verify that their shiny climate model functions as intended?

Have you any evidence of testing/verifying to share, with the extensive test suites of inputs and outputs.

I would really like to see it.

Now EM, you have stated that from time to time in the very recent past there have been aircraft incidents due to s/w issues.

How confident that all of the CMIP models are as robust as aircraft s/w?

When you evaluate the worth of the two fields, aircraft failures could affect say 500 people, climate model failures affect whole economies leading to ruin of millions.

So should climate modelling be done properly, or if you begin to see that with our current level of understanding of the physics processes involved, we should turn the climate models off for a few decades until science catches up?

That is how the validity of the models has been tested and confirmed.

Oct 28, 2017 at 11:44 AM | Unregistered Commenter Entropic man

Tested and confirmed as what?

Mann's Hockey Stick was 1998, so before the Y2K bug kicked in, and all of Climate Science since has been dedicated to prove Mann right. He wasn't was he?

It has been a disaster for the Planet and all life forms, animal and vegetable, apart from China etc who couldn't care more about their rapid economic development, at everyone else's expense - literally.

I have nothing against China etc, if anything, my respect for the Chinese has risen, as my mistrust of the EU has fallen.

With so many EU jobs exported to SE Asia, BREXIT can be blamed on corrupted Climate Science, and it is possible that Corbyn worked that out a few years ago.

Trump was elected by the former workforce of the USA swinging away from Clinton.

The Green Blob has been very keen to attribute everything bad to CO2. Can you see how much "bad" has been caused by Climate Science trying to save Mann?

we should turn the climate models off for a few decades until science catches up?

Oct 28, 2017 at 2:45 PM | Steve Richards

It would be the best way of testing the models, and Climate Science. We have had 20 years with, now let us try 20 years without. The first two decades of the 21st Century may be known as the age of stupid, when Green Dinosaurs ruled the Earth.

Radical Rodent,
There is more evidence (unadjusted) to support the theory that Climate Scientists and their models are to blame for bad stuff, than Climate Science has ever found to blame manmade CO2 for anything (apart from a hangover after a hard night of conferencing at a Climate Science jaunt, watching sea level not rise, on a tropical island beach.)

Meanwhile, back at the thread.....

WUWT have got Lapse Rates etc
https://wattsupwiththat.com/2017/10/28/the-atmospheric-lapse-rate-and-molar-density/

WUWT also have "philosophy" and "consensus"
https://wattsupwiththat.com/2017/10/28/philosophy-uncertainty-probability-consensus-the-ipcc-and-all-that/

"Consensus is King. A great deal has already been said on this website (WUWT) regarding the CAGW campaigners’ obsession with the supposed 97% consensus amongst scientists. It has been quite rightly pointed out that science doesn’t work that way; it isn’t a democracy."

"Such a misunderstanding is perhaps to be expected amongst journalists, politicians and activists who are just looking for an authoritative endorsement of their views, but it is all the more shocking to see that the IPCC also holds to the belief that consensus can manufacture truth. "

"Whilst you and I may struggle to come to terms with the subtleties of probability and uncertainty, the IPCC appears to have had no trouble in arriving at a brutally simplistic conclusion. In guidelines produced for its lead authors to ensure consistent treatment of uncertainties, one can find advice that is tantamount to saying, “A thousand flies can’t all be wrong when they are attracted to the same dung heap”.

So, I went to the WUWT lapse rate thread. Blimey. Nobody can agree on anything. So many definitive statements of fact at odds with each other. I'm not going to pick one as true. I do note that they can't agree on where the tropopause lies. Never mind 'top of atmosphere'. I'll wait til it settles down...

Rhoda, over at WUWT, is there a consensus that nothing can be agreed on now about Lapse Rates, and could not have been agreed on when Mann decided to erase the MWP and LIA, because the Science was already settled?

Surely there must be some reliable Climate Science that is pre Mann?

I can't see any consensus over lapse rates, but then these things are not decided at WUWT. What I can see is that of the two explanations for dummies which we have been fed neither the back-radiation one nor the TOA/lapse rate one actually work. Do the climate models use one or the other? If they indeed model actual physics rather than take climate sensitivity as an input, do they them reveal a prediction in their output of either back-radiation or TOA/lapse as a heating mechanism? Or neither? Or both?

Alternatively, is the forcing figure in the models derived from a table lookup against CO2 concentration? What does CMIP5 require?

Rhoda

Think of the tropopause, stratopause and mesopause as altitude bands, not spot altitudes. Since all three are levels in the atmosphere where the lapse rate changes sign I like the definition of these bands as transitions where lapse rate=0.

Out in the physics departments things settled down long ago, with radiation and lapse rate reinforcing or in opposition.

The surface is the main place where radiation from the sun is converted to heat, so it is the warmest part of the atmosphere.

The tropopause is the altitude at which the atmosphere becomes transparant enough to radiate to space at IR wavelengths. This allows CO2 and water vapour to cool the atmosphere. Below the tropopause thel negative temperature gradiant and negative lapse rate reinforce each other.

The stratopause is the altitude at which the atmosphere becomes transparant to UV. This is the altitude at which oxygen absorbs UV to make ozone and warm the atmosphere. Below the stratopause conduction overcomes convection The atmosphere stratifies with a positive lapse rate.

Above the stratopause oxygen, ozone and CO2 all radiate to space. The normal negative lapse rate with altitude resumes.

At the mesopause the air is so thin that radiation absorbed from the sun exceeds the energy radiated from the atmosphere.From here on up the effect gets stronger with altitude so the lapse rate goes positive again.

Did you notice that the article did not mention radiation at all, nor did most of those commenting. This produced an enormous variety of hypotheses, most of them with no basis in physics. This is an old problem among those who do not accept the role of absorbtion and emission of radiation in the atmosphere.

Rhoda

Both "explainations for dummies" are simplifications. In fact both processes take place and interact, but that makes it too complex to pass the barmaid test.

Are you familiar with the concept of vectors in physics? You might get a better handle on the climate of you stop trying to explain the whole thing in terms of one variable and think of all the different processes as vectors pulling or pushing temperature in different directions.

Oct 29, 2017 at 5:41 PM | rhoda
Entropic Man

Experts in theoretical physics seem to have failed the Practical and changed their mind

http://notrickszone.com/2017/10/29/new-working-paper-advent-of-computer-modelling-has-corrupted-climate-science/#sthash.YlDHugna.dpbs

"Unfortunately the advent of computer modelling has corrupted climate science into believing models are now the main source of knowledge, even though it’s not uncommon for models to have systemic deficiencies [Santer, et. al. 2017]. Theories always had a bad press [Rabinovich, et. Al. 2012], but many scientists seem to be confused about the difference between a model and a theory [Hug, H., 2000]. "

https://wattsupwiththat.com/2017/05/25/no-santer-et-al-have-not-refuted-scott-pruitt/
(SVP = Santer versus Pruitt)
"SVP would be rejected by any reputable publication. Nevertheless, it passed the supposed peer review by Nature Scientific Reports in a mere 29 days: Received on March 6, Accepted on April 4, Published on May 24; praised and brandished by WaPo and ThinkProgress on the very same day. The publication date might have been selected to influence the G7 meeting. The Nature Group announces median times for its review process (https://archive.is/G3tT4). The median time from submission to acceptance is 124 days. The median time is not very useful metric, so we investigated actual times to acceptance of the last 500+ articles, published in the Nature SRep from May 16. Less than 1% of these articles were accepted within 30 days from submission. All articles in this exclusive club, except for SVP, report experimental research (supporting material is available). This exceptionally quick acceptance, combined with the invalid scientific methodology of the article, suggest that Nature SRep accepted SVP without bona fide peer review, or despite negative recommendations of the reviewers."

The same Santer who wrote in 2007:
From: Ben Santer <santer1@llnl.gov> To: P.Jones@uea.ac.uk Subject: Re: FYI Date: Wed, 25 Apr 2007 16:58:29 -0700 Reply-to: santer1@llnl.gov <x-flowed>
"Dear Phil, I looked at some of the stuff on the Climate Audit web site. I'd really like to talk to a few of these "Auditors" in a dark alley. They seem to have no understanding of how science is actually done - no appreciation of the fact that uncertainty is an integral part of what we do. Once again, just let me know how I can help...."

Wikipedia notes his "Honours"

Santer received a B.SC. in Environmental Sciences and a 1987 Ph.D. in Climatology from the Climatic Research Unit of the University of East Anglia.[1]

In 1998 Santer was awarded a MacArthur Fellowship for research supporting the finding that human activity contributes to global warming. He has also received the Ernest Orlando Lawrence Award and a Distinguished Scientist Fellowship from the U.S. Department of Energy and the Norbert Gerbier/MUMM award from the World Meteorological Organization.[1] He ranked twelfth amongst climate scientists in a 2002 assessment of most cited scientists in the field of global warming.[3]

In 2011, Santer was elected as a fellow of the American Geophysical Union[4] and as a member of the National Academy of Sciences.[5]

Can it be assumed that Santer misplaced his trust in Theoretical Physics, Computer Generated Climate Models, and perhaps now regrets not appreciating Climate Audit?

Rhoda

You build basic physics into the model.

Forcings such as solar insolation, albedo, volcanic and industrial pollution are presett for hindcasts, or randomised for forecasts.

CO2 is preset according to the RCP you are testing.

Back radiation, TOA, lapse rate, climate sensitivity and temperature are not set in advance. They are all outputs which emerge from your calculations.

You build basic physics into the model.

Back radiation, TOA, lapse rate, climate sensitivity and temperature are not set in advance. They are all outputs which emerge from your calculations.

Oct 30, 2017 at 12:26 AM | Entropic man

If that is the case, is it reasonable to assume that consistent errors in calculating back radiation, TOA, lapse rate, climate sensitivity and temperature, are why the models are all overheating, fairly consistently? Find the error(s) and some of Climate Science may survive budget cuts.

Steve Richards

"How is it scientific in any sense of the word to publish results from models when the model makers know that a) the model is incomplete ."If the models were produced by under grads , grads or post docs as a means of research, fine, "

The model is incomplete because it is a tool for research. If we already knew all the answers, or had lots of spare Earth's to do controlled trials, there would be no need for models.

Models are produced by under grads , grads or post docs because they understand the science and because they are cheap.It is also why they tend to stick to Fortran instead of something more sophisticated. A typical research grant is about £250,000, of which half goes on facilities and housekeeping. I don't know your fees ,but I doubt that a software house could write much code for £125,000.

"b) the results are used to change the world"

The existing models are not intended to change the world, just as research tools. The problems come when the politicians demand to know what future climate will by like as a guide to making policy. The scientists can only answer using the tools they have.

I agree with you that if models are to guide policy, a lot more resources should be put into them. Perhaps you could estimate the cost of contracting a software house to writie a computer model which meets you professional standards. Call it 10 million lines of code, plus testing and verification?

I did a quick search and found costs around $4 per debugged line. That would bring a 10 million line climate forecasting tool in at $40 million. No university department could afford that, though a government might.

The existing models are not intended to change the world, just as research tools. The problems come when the politicians demand to know what future climate will by like as a guide to making policy. The scientists can only answer using the tools they have.
Oct 30, 2017 at 2:00 AM | Entropic man

So now we know:
1. Climate models have never been fit for purpose
2. Climate Scientists have deliberately programmed themselves AND their models to over exaggerate
3. It has always been about money, but more money could never have bought more honest science from Climate Scientists, Cook's 97% Consensus has proved that.

What is the time slice for these models and are they all the same in terms of that and smallest area considered? I am aware you can't model below certain limits because of computer power, but what are those limits?

Rhoda

As you say, there is a tradeoff between the spatial resolution and the temporal resolution of a weather or climate model.

For a given amount of computing power you can have a finer grid and more levels in the atmosphere or a faster update .

The Met Office weather model runs on a 10km grid with 50 levels and updates on a 3-5 minute cycle. A typical run is from 1 day in the past to 10 days ahead, about 3100 time cycles. To do multiple runs faster than real time they use a supercomputer.

Climate models have a lower spatial resolution, perhaps a 25km grid and fewer levels. They update on a time cycle of 30 minutes to six hours. Since a climate model might run from 1980 to 2050 that is somewhere between 85,000 and 1 million cycles. Since they tend to run on PCs they are slow. A single run can take weeks.

You see the difference. Weather models are optimised for maximum detail over a small number of cycles. Climate models are optimised to run with less detail over a large number of cycles.

Climate science has particular challenges. It is not an experiment based science. There are many cyclical changes, some longer than the lifetime of an observer. It is very complex. Modelling is about the only way that our climate can be understood. Unfortunately, climate science is just a few decades old and its evolution has been flawed in a number of ways. The result is that today it has become a pseudoscience.

Normal practice involves proposing a hypothesis that can be falsified. Climate science only has models. These models are incorrectly regarded as experiments and frequently a piece of work with a conclusion turns out to be entirely based on models including the data that was used in the first place. This is pseudoscience.

I was first alerted to the state of climate science when I discovered that the early scientists attributed all warming to carbon dioxide. Natural variability was deemed unimportant. Nothing could offset CO2 induced warming. Previous warming events such as the MWP were not just ignored, there were attempts to erase inconvenient results. The 300 year warming from the LIA was ignored even though it clearly exists.

None of this is science in the real world, but it is at the heart of pseudoscience climate change. The models all run hot because they have a linear relationship between the carbon dioxide increase and temperature. When the divergence from reality became a problem, they increased the aerosol level to offset the warming. The model simulated our climate.

Except it didn’t. The aerosol level was completely unrealistic. Tweaking a model until it hindcasts a particular set of data correctly is not any sort of achievement either. Climate models are a long way from being predictive and may never be successful.

So where does this leave us? Climate science is a pseudoscience because incompetent people hijacked it and made it a model with a single dominant driver then convinced politicians that policy should be based on the flawed model output.

We have not progressed very far from that position today because it now has a trillion dollar industry to support and who is going to question the emperor’s clothes?

SC. I must disagree. Firstly, we presumably do agree that an understanding of climate and of climate change is important, secondly that we need to do something more than wring our hands and wail that the subject is too hard. The only way we can advance our understanding is by simplifying the subject and by using complex models. Such models should be considered as gross approximations and (at present and perhaps always) totally unsuitable for long term predictions. This view avoids the question of whether climate models have been devised to confirm a postulate - that CO2 is the major control. Climate change modelling is not invalid science but it seems to be misapplied. However without an ability to accurately forecast, the chances of major funding would evaporate and political support would vanish.

SC Further reflection suggests I that I disagree with another part of your last post, namely:
"Normal practice involves proposing a hypothesis that can be falsified. Climate science only has models. These models are incorrectly regarded as experiments and frequently a piece of work with a conclusion turns out to be entirely based on models including the data that was used in the first place. This is pseudoscience."

Let's take Galileo's experiments of rolling a ball down a slope. It would be relatively easy to produce from a number of critical experimental results a mathematical model that would incorporate controlling factors such as the gravitational constant, air resistance, friction angle of slope, length of slope etc. This model then could be used to accurately predict results (speed of ball, length of travel etc.) for all possible combinations of variables. The validity of the model could be checked by further experiments. Now let's make predictions for experiments made on the surface of Mercury or Mars. Change values for gravity, air resistance and produce results. But oh now I hear you cry those results are based on models so they can't be scientific. I can't falsify the results by experiment so it must be pseudoscientific. Yet similar, but much more complex predictions were made that resulted in successful landings or orbits of those planets. Popper's reliance on falsification for determining good science is overblown.