[OT] How science is not done

"Martin Brown" wrote in message
...
Peter Webb wrote:

"Martin Brown" wrote in message
...
Peter Webb wrote:

Have you got one which predicts cooling from 1820 to 1850 when fed
1820s data?

They will if they take vulcanism into account. I take it that you
deliberately chose a period immediately following the Tamborra super
eruption.

No.

A model would have to know about the volcanic aerosols.

OK.

You don't know if there were massive volcanic eruptions in the 19th
Century, and that is why climate models don't work for the 19th Century?

No. A model has to be given knowledge of the volcanic aerosol
contributions if it is to get the right answers.

This is your claim. To prove it, you must prove that it does provide the
right ansewrs when it does know the volcanic aerosol contributions. Say
from 2000 to 2209, assuming the model was developed before 200.


There are uncertainties about the extent of the Tamborra inputs. I
personally would love to know why it didn't trigger the same sort of polar
stratospheric cloud displays in the UK as Krakatao did in the 1880's. It
was a much bigger bang.

So you are saying that the climate models are untestable before a certain
date, because the aerosol contributions cannot be determined.

What was that date roughly? What is claimed period of applicability of the
model?




By the 19th Century, Europeans had discovered the entire world, and in
most cases there were European observers. Independently of this, surely
the local population would have noticed massive volcanic eruptions in
their neighbourhood. I would also have expected geological evidence for
massive volcanic eruptions 150 years ago.

Are you really telling me that the reason that climate models don't work
when applied to 19th Century is because their were massive volcanos in
the 19th Century of which we have no records? Where?

Not at all. The models would work reasonably well if they were given the
right inputs. That is in part how they have been calibrated.



Yeah, I calibrated my model in the same way, by changing the formula until
it matched experiment.

The problems arise in determining accurately what the climate was like in
that historical period when the available global observational constraints
are nothing like as good as in the modern era.


The "modern era". So your claimed period of applicability of your model is
what date range?



I just want to see if any of them predict cooling, ever, or whether
they are all hardwired to produce only warming.

Of course they do if either insolation decreases or GHG forcing
decreases. Equally if you keep adding more GHG then the climate warms.
This should not be a surprise to anyone. Put an extra blanket on the bed
and it helps to keep the heat in.

Also, some papers which compare past predictions of climate with
measured results would be great; my main problem with climate science
is this whole agreement between theory and experiment thing.

OK. If you are serious. And I very much doubt that you are then take a
look at the printed paper copy of the IPCC Science Case Climate Change
2001 p447 has specific predictions for Arctic Sea ice from GFDL and
Hadley models of the day as compared with actual observations to 1998.

To 1998?

The reason I chose this specific paper is that the printed version is
locked to exactly what they predicted out to 2040 with the models and data
available up to 1998. Their predictions are a slight under estimate of the
extent of polar ice loss seen to date.

I would of course be very interested in seeing the specific predictions made
in 1998 and the actual measured data since.

Your statement that the experimental data does not in fact match the
theoretical predictions does not leave me with much confidence.

I can also see how this would provide additional evidence the earth is
warming, but I don't dispute that it. It has been warming for over 150
years. The theory of anthropogenic warming demands an anthropogenic
component. The mere fact that the climate is changing doesn't mean that man
has anything to do with it, the earth has spent long periods either warming
or cooling long before homo sapiens appeared.


This is actually a bit worrying since it means that the CO2 feedback is
more aggressive than the 1998 models have accounted for.

So the 1998 models have been proved wrong.

Again, I stresss that I am seeking succesful predictions of climate science,
not incorrect ones.



These models were the justification for a world-wide treaty which called
for huge changes to the structure of the world's industry.

Your evidence is a paper published in 2001 which correctly "predicts" sea
ice levels for 1998, which were in fact already known to the author in
2001 when he correctly predicted them.

NO. CAN'T YOU READ? I THINK YOU ARE BEING DELIBERATELY OBTUSE.

The prediction made in 1998 goes out to 2040 and is so far a pretty good
match to the observations - the GFDL one is pretty close to reality.


Got a link?


Prior to 1960 the sea ice was fairly steady with annual fluctations. It is
only since GHG forcing became non-neglible in the past few decades that
sea ice has taken a serious hit with a systemtic downward trend.


I would love a table of predicted sea ice versus measured.

What else did the model predict, other than sea ice levels - ocean
temperature, cyclonic activity, upper atmosphere temperature, other stuff
like that? How well did it predict these other variables? At what confidence
level did it eliminate the null hypotheses of simply random variation?

Of course, if you are only the claiming that climate science is valid only
at predicting sea ice levels, then forget the rest of trhat stuff, and a
table of predicted ice levels versus actual measure ice levels will do just
fine.




It is now 2009. What did the model predict for 2001 - 2009, and how well
does it match observation?

From the graph I have in front of me it predicted a 5% per decade decrease
in Arctic sea ice and the modern observations show:

http://nsidc.org/data/seaice_index/

That is a strange graph in some ways. Partly because the bottom axis goes
from 1970, but data is omitted prior to 1980.

But as far as I can tell, it is a graph of sea ice concentration over the
last 29 years, with a dotted line emphasising the fact that this has been
generally decreasing over this period.

But you claim that some paper written 10 years ago also predicted that sea
ice levels would drop, and they did.

I might point out that I am not exactly gobsmacked by this, the earth is
indeed warming, that sea ice levels were predicted to drop and did is hardly
surprising.

I would also of course be interested in the predictions and experimental
results for a considerably longer period than 29 years. I don't know how
much more I can ask for, as you have not yet told me what the period of
validity of the climate models is (ie when volcanic aerosols could be
measured well enough). I thought there was a lot of anecdotal **** about the
North West passage being open last century, when the world was much cooler,
suggesting different and presumably more complex variables driving sea ice
than global temperature.



Writing a paper in 2001 which predicts the past accurately is not
independent experimental verification. It should be easy to gain such
verification; just look at te climate changes aftewr the paper was
written and see if they were correctly predicted.

Were they?

YES. Why don't you go and read the literature?


Well, its just that its such an important and contentious issue that I would
expect experimental verification would be easy to find on the net, if it
exists that is.

If I google for experimental proof of evolution or special relativity (to
use somebody else's examples) then in addition to a bazilion crank sites
there are some that provide examples of very specific predictions of
evolution and SR that have subsequently been confirmed experimentally.

Try that for climate science, and ... well, lets just say I haven't seen
anything vaguely resembling lists of specific predictions made by specific
models compared to subsequent measurements ... I woud have thought that even
you believers in AGW would be interested in such a table; how else do you
decide which model was correct, if not by seeing which ones best matched
experimental results?




We are now a decade further on and you can see that they have the trend
and rate of decrease of sea ice about right when compared to current day
observations. GFDL is a bit more aggressive than Hadley and nature it
turns out went beyond the pessimistic prediction of GFDL.

See for example the NSIDC comparison of the later models with actual sea
ice data. The conclusion is basically that we are warming the poles
somewhat faster than the preferred IPCC models predict.

http://nsidc.org/news/press/20070430_StroeveGRL.html


So your evidence that the predictions of climate science are correct is
in fact an article which points out every single one of the 18 models
analysed got completely the wrong answer?

This is a strange approach to determining scientific fact. Normally, when
an experiment fails to match a theory, this is evidence the theory is
wrong. You seem to think that the fact that every one of the 18 models
made demonstrably wrong predictions proves they are all correct.

You are back to sophistry again. The models make predictions that are
specific and have been validated.


Fantastic! That is exactly what I am after!


They are not perfect - nothing in the real world ever is but they describe
the overall system sufficiently well that we can have confidence in their
predictions.

Well, hold it, you got me all excited, now you are already underselling it.

Let me guess. These models say the earth is warming. As it has been warming
now for over 150 years, that seems a reasonable bet. At worst, a prediction
made 10 years ago that temperatures would continue to rise for the next 10
years (ie up until now) would have a fifty-fifty chance of being true or
false.

What kind of experiment establishes something to be true when there is a 0.5
chance it was just a random correlation?

For that matter, where does the "A" in AGW enter into this equation,
exactly?



Were you the dude who taught "University level science" for 30 years? Do
you prove to your students that other scientific theories are correct by
pointing out the theory consistently fails to explain experimental data?

You are deliberately twisting my words for your own malign purposes here.
It won't wash. The literature is all out there - if you want to find how
the models were verified and validated then go away and do it!



But you yourself can't provide any good links. OK.



I don't believe in climate science because the models don't match
experimental data.

You will have to be more specific. The climate science models do have
significant predictive power.

Well, gee, OK. Ten years ago it was 1999. What were the predictions of
the main climate models of 1999 for the world tempertaure over the last
10 years? To what confidence level do they exceed chance?

If you want an answer to that you can work it out for yourself.

So nobody has bothered already to see if predictions match experiment,

That isn't what I said at all. There is a lot of work going on to test the
models against actuality, and to refine them by running other planets and
simplified test cases.


Do these other planets have people living on them burining fossil fuels?

It would seem an extremely simple task to "test the models against
actuality"; you simply list the specific predictions made by the theories 10
years ago against measured data for the last 10 years, and see how well they
worked.

Surely this must exist somewhere on the net, aren't people who believe in
AGW interested to see which models have subsequently been proved most
accurate? How else do you decide which one is correct?


excepting of course the study you provided a link to which shows that
they do not.

If I wanted "proofs" of Special Relativity, I can find them easily. If
somebody said that nobody had ever bothered to verify the predictions of
SR against experiment, and I had to do it myself, I would be very curious
as to why this basic checking had not occurred.

The basic checking has occurred. You insist that you are not satisfied
with it and cannot be arsed to go and look at any of the references you
have been given. It is your choice to be wilfully ignorant.

I haven't been given any links.

Sorry, there was a link to a graph (a very tiny graph) of sea ice levels
over the last 29 years showing they were generally decreasing, and you told
me that one of the climate models you knew about also predicted 10 years ago
that sea ice levels would decrease.




Or you could read the IPCC Science Report which deals in detail with
most of the factors and remaining uncertainties in climate modelling.

http://ipcc-wg1.ucar.edu/wg1/wg1-report.html

Or, you could just simply provide the predictions of the models of what
would happen in the future, and the actual experimental results.

A report which lists all the reasons a model may be wrong (ie the
uncertainties) is not going to make it more likely I will believe it is
true.

The scientific position is to build a descriptive model of the planet and
its climate. The scientists are honest. The deniers for hire are not.

Ohh, wandering away from science to your true love and area of expertise,
conspiracy kookdom.



It is frustrating that a handful of maverick scientists and a denier for
hire PR machine can hold sway on a credulous population of scientifically
illiterates, but that is how the world is at present.

And that explains why it is so hard to find sites which list the specific
predictions of different climate models and compared them to subsequent
data? Because these evil corporations are censoring the internet? Somebody
already said that some of them used to work for tobacco companies, were any
of them foreigners as well?



I would ask the same of the models used for the Kyoto protocol, which
was 2 years earlier.

I hope that's not too specific - really any experimental evidence at
all would be a good start.

Try the polar ice distributions then. The specific predictions are made
and they have come true in spades.

I thought the article you posted said exactly the opposite, that all 18
models tested were wrong?

But, OK, where are the predictions of the model (whatever model you
believe correct) and the subsequent experimental data compared?


The best models still under estimate the extent, but the sense is very
strongly for global warming and more specifically strong polar warming
as the polar albedo changes.

So the best models are wrong, and this proves that climate science is
correct?

The best models reproduce the right behaviour, but the amount of change
being observed is actually slightly larger than GFDL predicts. Taking the
polar ice as a clear observable with very specific predictions from the
models:

AGW Deniers predict: no change

No. I predict a linear decrease, identical in rate to that observed on
average for the last 30 years.

I don't deny the climate is changing. Always has, always will.

I do dispute whether there is any strong evidence that humans are a major
contributer to this, simply because I haven't seen any, and I do dispute
whether the existing computer models are correct, because they don't seem to
work.


GFDL Model predicts: -5% per decade
Observed data shows: -6 +-2 % per decade


So how many decades ago were these predictions made! Or do they include
"predictions" of what has already happened?

Lets take a simple, direct case. As you picked sea ice, what levels of sea
ice cover did the models predict 10 years ago, compared to the sea ice
levels that have subsequently been measured? Not the average including data
already known when the model was constructed, just the specific prediction
portion. If you also provide actual sea ice levels for a longer period - say
50 years - I can use this to model the natural variability, and hence
determine a confidence level for any correlation not just being random. All
standard stuff, of course, in a very wide range of sciences, but as I can't
find anything like this on the net maybe I have to do it myself (though I
still wonder why).


Who are you going to believe? The observations are intrinsically noisy.


If the effect is too small to measure against background noise, then maybe
it doesn't exist.



It isn't going to be disproven because their key predictions are already
coming true. There is a bit of a worry that the best models at present
are too conservative and that nature has a few surprises in store.


Well, lets see the predictions of the dominant models of the late 1990s,
and the subsequent experimental verification

Somebody did bother to check if the models were correct, didn't they?

Yes. But if you want to look at it in detail you are going to have to
visit a library. There is some material on verification and validation of
the models on the Hadley site. The other global repository of climate
model testing verification and validation is at LLNL.

http://www-pcmdi.llnl.gov/about/index.php

You are on your own probing in there. I do have some sympathy for your
position that a simple explanation of the reason why almost all scientists
are agreed on the reality of AGW is needed.


As you imply, that site is useless for my purposes.


There are just too many dittohead "junkScience" sites out there peddling
obvious lies but which sound plausible to the general public.

List of predictions, list of previous and subsequent measurements, what I
want is pretty simple. Much like I can find on evolution, relativity, QM,
plate tectonics, just about every important theory which I believe to be
true.



In my experience of "science", if somebody spends a great deal of effort
developing a testable mathematical model, the results of that model are
compared to subsequent experimental data to see if they are correct. This
is the defining characteristic of science and the scientific method; it
is what separates science from religion.

No disagreement there. The problem here is that whenever some evidence of
the models correctness is presented you pretend that it isn't enough.

Tell me about other predictions that have been made about how the world
will end in 100 years unless something is done immediately that have
ever been shown to be true?

How do you think this is even remotely relevant?

Because the climate scientists are making predictions about what will
happen in 50 or a hundred years time. AFAIK, of the thousands of
confident predictions of world disaster within 50 or 100 years, none has
ever been proved correct.

Popular though they have been through history.

I suppose Australia deserves it extreme drought conditions then.


Not actually as bad as the 1890s drought, as it now turns out.

Though here in Syney we did have our warmest August in almost 10 years.



What I would do is use Fourier analysis to extract any low frequency
signal, then polynomial fit the difference. I can easily make the poles
uninhabitable for you.

You use such long words. How impressive. NOT!
The models only permit you to alter the composition of the atmosphere
and total solar irradiance.

Again the fact that the models ignore data which would seem relevant
would indicate that they are wrong, not that they are correct.

You seem to consistently argue *against* the AGW models being correct -
you posted an article which said that the 18 main models all made wrong
predictions as to sea ice, you posted a link to an article which listed
the "uncertainties" which may make the models incorrect, you point out
that the model which you are using ignores key inputs.

I am prepared to discuss the science. What is known, what is uncertain and
what is not known at present. Leading edge science always comes with some
level of uncertainty - there is no way of avoiding that.


I guess you are trying to gently break the news that the experimental
evidence is not actually that good.



However, the basic principles are now clear. Adding CO2 at an ever
increasing rate will get us into trouble in the relatively near future.

And if we do not add CO2 at an ever increasing rate, we won't get into
trouble?

I know there is a lag effect in cycling CO2 through the atmosphere, but in
the long term, today's global temperatures look pretty close to perfect. So
surely the level of CO2 emissions that can be tolerated in the distant
future while maintaining today's temperature is an important target. What
level of anthropogenic CO2 production would be required for global
temperatures to eventually stabilise at today's levels?




Again, I am looking for places where experiment matches predictions, and
NOT stories which point out that theory does not match experiment, and
possible reasons why climate models are demonstrably wrong.

Then go and read the IPCC Science Report. It deals with comparisons of the
past records with the models and how they compare. as I have pointed out
before if you get a paper copy of the 2001 report you can compare the
predictions and graphs made at that time with the actual observations
today for yourself without having to rely on outside help.


So nobody has ever thought to post evidence that climate science makes
correct predictions somewhere on the internet? Or has it been removed by
maverick right wing corporations?

But it does exist, right?


It is quite hard to drive Earth over the edge. To that extent the
environmentalists are crying wolf, but they do have a point - the time
for profligate waste of energy is over now. We should be doing a lot
more to conserve energy.

Maybe, but for me that is an economic and geo-political argument, and not
proof that AGW is correct.

I am beginning to think that you don't understand anything about how CO2
affects the atmospheric energy balance by blocking outgoing long wave
thermal radiation. It doesn't make sense to deny the possibility of AGW.


I don't deny the possibility of AGW. I just say I have not seen any
experimental evidence that would make me believe its true.




Add enough CO2 to the Earth's atmosphere and you would eventually end up
with something like Venus but with a surface temperature of about 690K
instead of 750K (representing our greater distance from the sun). Models
won't allow you to push it that far since they limit it to burning fossil
fuels and not decomposing all the carbonate rocks.

As I mentioned, I'm really just more interested in the match between
prediction and experiment.



I don't really feel I need to know much more about climate science
theory as such; I know the connections between CO2, ocean acidity,
albedo, vulcanisation and cloud cover are deep and mystical. My lack of
knowledge is where the theory has been experimentally verified.

You choose to remain wilfully ignorant. THAT IS YOUR PROBLEM.

And I am desperately trying to remedy it.

Where can I find retrospective studies which compare the predictions of
various climate models with subsequent experimental results, so I can at
least decide which climate models (if any) could at least possibly be
correct?

As I have said earlier. GFDL looks reasonable, but the models are being
refined and will inevitably get better in the future. The key point here
is that even a relatively simple model can show why adding CO2 will make
the Earth get warmer. And none of the "scientific" sceptics worthy of the
name scientist deny this any more.


But the earth is getting warmer anyway.

Between the CO2 input and the temperature output, there are about a million
lines of computer code , and a huge collection of arbitrary and post-hoc
assumptions about the transfer of heat, CO2, water vapour, cloud cover,
methane, albedo, vegetation, ice, fresh water, ocean currents, volcanic
aresols (apparently), and agriculture, the chances of all of which being
substantially correct are approximately zero, which is born out by the fact
that the models are **** at making predictions which actually turn out to be
true.





I am posting now to make sure you cannot mislead any other weak minded
individuals than might be reading this thread.

Ha ha.

Perhaps providing some independent experimental verification would be a
better tactic?

If we were arguing about evolution or SR, and you maintained they were
incorrect, I would galdly supply you with hundreds of links to dozens of
sources of experimental verification. It is the existence of this
independent experimental verification which convinced me - and those
people who beieve in the scientific method - that SR and evolution are
correct theories.

And yet you can still find plenty of particularly electronics engineers
that still do not believe in relativity a century later. The first GPS
satellites had a disable relativistic corrections feature because the
engineers did not believe the physicists.

So, what have you got for climate science?

Whatever it is it will never satisfy you.


No, a simple list of the specific predictions made by the models used for
Kyoto, and the subsequent experimental data will do.



Here is what you said: "Whether or not it fails before 1850 [during a
coling period] really isn't relevant."

You latch on to random cherry picked points and labour them. Vulcanism
plays a certain part in climatic behaviour - Tamborra in 1816 year
without a summer and more of interest to me Krakatoa in 1883-4-5.

Actually, prior to this post, I don't think I had mentioned volcanos,
let alone cherry picked data about them.

You deliberately chose a period immediately following one of the most
violent periods of vulcanism in recorded history and demanded that
climate models should match its cooling effect.


No, I didn't even know it was one of the most violent periods of
vulcanism in recorded history (what does that mean, by the way, is
recorded history 2000 years? Instead of terms like "recorded history",

You have to go back about 75000 years for a bigger bang with the Toba
super volcano that was close to a human extinction event and is widely
believed to be responsible for our relatively low genetic diversity.

http://hvo.wr.usgs.gov/volcanowatch/2005/05_04_28.html

couldn't you mention specific time periods.). I picked it because it was
a cooling pewriod.

Does climate science predict that large scale vulcanism cools the earth?

Of course. Any high altitude aerosol or dust reflects incoming light.

Don't the models include this?

Some do some don't. It was one of the notable discreprancies in the early
models - vulcanism was not well handled for the super volcanoes that put
fine material high up into the stratosphere and alter climate by changing
incoming and outgoing transmission of light and IR.

So the models that didn't include it were very wrong.

Were any of these models used for Kyoto?



What other cooling forces are omitted from the models?

I think at present none of them include the Keeling tides (which is AIUI
still not widely accepted). It can be either a warming or cooling effect
so it averages out longer term. Basically it says that larger tidal range
from changes in the lunar orbital elements makes more turbulence and
shunts heat deeper into the oceans. Their paper seemed to be fairly
convincing from an astronomers point of view.

How do you know that their are no other cooling forces? Do these two
things - vulcanism and Keeling tides - explain all previous cooling periods?




http://www.geolsoc.org.uk/gsl/site/G.../page2972.html

Toba is generally reckonned to be the volcano that very nearly wiped
out humanity - cutting our ancestors gene pool down to a small set.

As does the variation in the Earths orbital elements and continental
drift. These longer term geological timescales are important mainly
over millenia. Although there is some evidence that shorter timescale
Keeling tides may be able to cause short scale climate variability.


Yeah, yeah, I know, the theory is magnificently complicated, wheels
within wheels, almost but not quite perfect and complete, probably way
to difficult for me to even begin to comprehend ...

That seems highly likely since you have set out to ignore any
observations or data that would conflict with your preconceived ideas.

No, I am looking for "observations or data" that matches the preconceived
ideas in climate models.

There are no preconceived ideas in the climate models. They are ab initio
simulations of the system behaviour. If you vary things then they respond
according to a set of linked differential equations.

These are also known as experimental verification.

Have you got any?

Personally no. But if you really want to look the LLNL hold the work.


So experimental verification exists, its just that you don't personally have
any, and nobody has ever gotten around to putting some up on a website.



**** it. If neccesary, I will polynomial fit every single data point
exactly, which must guarantee me at least equal first place.

It will also be the most amazingly discontinuous looking mess ever. You
are trying to be clever and failing dismally. The key to model fitting
is to use the fewest free parameters to explain the observations.

My model has 200 data points which were used to formulate the model
(technically, they are determined by previous experiment and are hence
"fixed", not "free").

But to match those points exactly you need a degree of freedom in the
model for every statistically independent point measured.

I agree with your central premise that there should be the fewest number
of arbitrary inputs; this is just a version of Occams razor.

The model which you believe is most correct used how many data points as
input parameters? Significantly less than 200? How many were temperature
data, how many were CO2 concentrations, etc?

We seem to agree this is important. You criticise my model for using 200
data points in its formulation, implying this is too high. So how many
does your model use?

I doubt over the time range where we have good climate data that anything
beyond a quadratic fit is justified.


So none of the equations used in standard climate models are more complex
than quadratic equations?

Far out. I assumed they would be full of PDEs, exponential functions, stuff
far more complex than polynomials of degree 2.


Arctic Sea ice is as unambiguous as any. Very specific prediction of the
models and borne out by observations.

But didn't you post an article which said the 18 dominant models all got
this wrong?

They mostly underestimate the effect. In other words the AGW observed now
is strong than our preferred models predicted in the past.

Well, warming may be stronger, but still you sneak in that additional letter
"A" in front of "GW", when nothing you have shown demonstrates it ...



Why don't you post an article which shows the predictions of arctic ice,
and the actual results?

The diagram in the IPCC report 2001 comes from Vinnikov et Al 1999
http://www.sciencemag.org/cgi/conten...cetype =HWCIT


But these are predictions of sea ice from 1953 to 1998.

Didn't they already know these when they developed the computer model?



If you have a Science subscription you can read it for free otherwise I
strongly suggest you look the paper up in a public library. Online
scientific journals are ludicrously overpriced for the casual user.


So, really, you are telling me that climate science has excellent
experimental verification, but nobody has bothered to post it on the
internet (or the evidence is being supressed by tobacco company employees?).



You want a model which better matches a certain data set?

You give me the data set, and I will give you a model which precisely
predicts it. You can't do any "better" than a perfect fit.

ANOVA says that you can.

ANOVA says that you can do better than a perfect match between experiment
and theory?

****, I would be delighted with a perfect match alone, let alone better
than perfect.

Where can I find a link to these better than perfect predictions of
climate science?


At least if you want to get something that sensibly represents the
system behaviour inbetween the data points that you have observed. Over
fitting data is a classic naive mistake of badly trained beginners in
scientific data analysis.

Yeah, in practice I wouldn't use a polynomial of degree 200. I would
least-squares fit a polynomial of degree 100, that should give me pretty
accurate answers.

Try it and see what happens in between the constrained points. BTW very
few numerical algorithms are stable for a polynomial of degree 100. Excel
these days cannot get much more than a cubic fit right.

Gee, I know about curve fitting, thank you.

I'm pretty sure I can get a polynomial fit to within 0.1 degrees accuracy on
every data point - does your model do that well?



It can be done with Chebyshev polynomials though.

As to whether it "sensibly represents the system behaviour", that is
really just asking if its predictions match experimental results not
already known when the model was constructed.

Which is exactly the question I am asking about your theory.

It isn't my theory. I am trying to provide you with some answers on the
basis that I think you may have a point that the scientific community is
not communicating adequately with the general public. And that the gap is
being filled with dittohead BLOGS and deniers for hire.

I wish you would provide pointers to where the theoretical predictions of
climate science are successfully compared to experimental evidence.

LLNL are the central clearing house for model comparisons.


Just the models used for Kyoto, their predictions for 1998 to 2009, what
actually happened.




The science is clear enough now. It is one thing to argue about the
finer points of some feedback mechanism, but quite another to deny
reality on the grounds that you don't like the answer.

Ohh, so there is strong independent experimental verification?

Cool! Can I see it?

It is almost enough to point at Venus and note that CO2 is triatomic. All
polyatomic molecules have potential as IR absorbers aka GHG.


So if you feed a clmate model with the parameters relevant to venus, it
works correctly, even though this is a different planet and its data was not
used in constructing the model?



The details of the models do not have to be perfect for the rising CO2
(and other polyatomic GHGs like CH4 and N2O) to be attributed to our
warming planet.


CO2 is rising because the planet is warming. I have heard people say that.
There seems to be some experimental data supporting this.


Logical inference allows you to conclude that if the heat input from the
sun has not got stronger (and we have accurate satellite flux records over
the past forty years) and the planet is warming then heat is escaping more
slowly. This conclusion is inescapable and even sceptics like Baliunas and
Soon concede this in their scientific papers.

Regards,
Martin Brown

Pity that nobody has bothered posting experimental verification of the
predictions of climate science anywhere on the web, particularly if there is
so much of it. And damn those right wing corporations and their tobacco
company employee minions, for stopping it being published.

On Sep 9, 10:35*am, "Peter Webb"
wrote:
"Martin Brown" wrote in message

...





Peter Webb wrote:

"Martin Brown" wrote in message
...
Peter Webb wrote:

Have you got one which predicts cooling from 1820 to 1850 when fed
1820s data?

They will if they take vulcanism into account. I take it that you
deliberately chose a period immediately following the Tamborra super
eruption.

No.

A model would have to know about the volcanic aerosols.

OK.

You don't know if there were massive volcanic eruptions in the 19th
Century, and that is why climate models don't work for the 19th Century?

No. A model has to be given knowledge of the volcanic aerosol
contributions if it is to get the right answers.

This is your claim. To prove it, you must prove that it does provide the
right ansewrs when it does know the volcanic aerosol contributions. *Say
from 2000 to 2209, assuming the model was developed before 200.

There are uncertainties about the extent of the Tamborra inputs. I
personally would love to know why it didn't trigger the same sort of polar
stratospheric cloud displays in the UK as Krakatao did in the 1880's. It
was a much bigger bang.

So you are saying that the climate models are untestable before a certain
date, because the aerosol contributions cannot be determined.

What was that date roughly? What is claimed period of applicability of the
model?



By the 19th Century, Europeans had discovered the entire world, and in
most cases there were European observers. Independently of this, surely
the local population would have noticed massive volcanic eruptions in
their neighbourhood. I would also have expected geological evidence for
massive volcanic eruptions 150 years ago.

Are you really telling me that the reason that climate models don't work
when applied to 19th Century is because their were massive volcanos in
the 19th Century of which we have no records? Where?

Not at all. The models would work reasonably well if they were given the
right inputs. That is in part how they have been calibrated.

Yeah, I calibrated my model in the same way, by changing the formula until
it matched experiment.

The problems arise in determining accurately what the climate was like in
that historical period when the available global observational constraints
are nothing like as good as in the modern era.

The "modern era". So your claimed period of applicability of your model is
what date range?





I just want to see if any of them predict cooling, ever, or whether
they are all hardwired to produce only warming.

Of course they do if either insolation decreases or GHG forcing
decreases. Equally if you keep adding more GHG then the climate warms..
This should not be a surprise to anyone. Put an extra blanket on the bed
and it helps to keep the heat in.

Also, some papers which compare past predictions of climate with
measured results would be great; my main problem with climate science
is this whole agreement between theory and experiment thing.

OK. If you are serious. And I very much doubt that you are then take a
look at the printed paper copy of the IPCC Science Case Climate Change
2001 p447 has specific predictions for Arctic Sea ice from GFDL and
Hadley models of the day as compared with actual observations to 1998..

To 1998?

The reason I chose this specific paper is that the printed version is
locked to exactly what they predicted out to 2040 with the models and data
available up to 1998. Their predictions are a slight under estimate of the
extent of polar ice loss seen to date.

I would of course be very interested in seeing the specific predictions made
in 1998 and the actual measured data since.

Your statement that the experimental data does not in fact match the
theoretical predictions does not leave me with much confidence.

I can also see how this would provide additional evidence the earth is
warming, but I don't dispute that it. It has been warming for over 150
years. The theory of anthropogenic warming demands an anthropogenic
component. The mere fact that the climate is changing doesn't mean that man
has anything to do with it, the earth has spent long periods either warming
or cooling long before homo sapiens appeared.



This is actually a bit worrying since it means that the CO2 feedback is
more aggressive than the 1998 models have accounted for.

So the 1998 models have been proved wrong.

Again, I stresss that I am seeking succesful predictions of climate science,
not incorrect ones.



These models were the justification for a world-wide treaty which called
for huge changes to the structure of the world's industry.

Your evidence is a paper published in 2001 which correctly "predicts" sea
ice levels for 1998, which were in fact already known to the author in
2001 when he correctly predicted them.

NO. CAN'T YOU READ? I THINK YOU ARE BEING DELIBERATELY OBTUSE.

The prediction made in 1998 goes out to 2040 and is so far a pretty good
match to the observations - the GFDL one is pretty close to reality.

Got a link?

Prior to 1960 the sea ice was fairly steady with annual fluctations. It is
only since GHG forcing became non-neglible in the past few decades that
sea ice has taken a serious hit with a systemtic downward trend.

I would love a table of predicted sea ice versus measured.

What else did the model predict, other than sea ice levels - ocean
temperature, cyclonic activity, upper atmosphere temperature, other stuff
like that? How well did it predict these other variables? At what confidence
level did it eliminate the null hypotheses of simply random variation?

Of course, if you are only the claiming that climate science is valid only
at predicting sea ice levels, then forget the rest of trhat stuff, and a
table of predicted ice levels versus actual measure ice levels will do just
fine.



It is now 2009. What did the model predict for 2001 - 2009, and how well
does it match observation?

From the graph I have in front of me it predicted a 5% per decade decrease
in Arctic sea ice and the modern observations show:

http://nsidc.org/data/seaice_index/

That is a strange graph in some ways. Partly because the bottom axis goes
from 1970, but data is omitted prior to 1980.

But as far as I can tell, it is a graph of sea ice concentration over the
last 29 years, with a dotted line emphasising the fact that this has been
generally decreasing over this period.

But you claim that some paper written 10 years ago also predicted that sea
ice levels would drop, and they did.

I might point out that I am not exactly gobsmacked by this, the earth is
indeed warming, that sea ice levels were predicted to drop and did is hardly
surprising.

I would also of course be interested in the predictions and experimental
results for a considerably longer period than 29 years. I don't know how
much more I can ask for, as you have not yet told me what the period of
validity of the climate models is (ie when volcanic aerosols could be
measured well enough). I thought there was a lot of anecdotal **** about the
North West passage being open last century, when the world was much cooler,
suggesting different and presumably more complex variables driving sea ice
than global temperature.



Writing a paper in 2001 which predicts the past accurately is not
independent experimental verification. It should be easy to gain such
verification; just look at te climate changes aftewr the paper was
written and see if they were correctly predicted.

Were they?

YES. Why don't you go and read the literature?

Well, its just that its such an important and contentious issue that I would
expect experimental verification would be easy to find on the net, if it
exists that is.

If I google for experimental proof of evolution or special relativity (to
use somebody else's examples) then in addition to a bazilion crank sites
there are some that provide examples of very specific predictions of
evolution and SR that have subsequently been confirmed experimentally.

Try that for climate science, and ... well, lets just say I haven't seen
anything vaguely resembling lists of specific predictions made by specific
models compared to subsequent measurements ... I woud have thought that even
you believers in AGW would be interested in such a table; how else do you
decide which model was correct, if not by seeing which ones best matched
experimental results?





We are now a decade further on and you can see that they have the trend
and rate of decrease of sea ice about right when compared to current day
observations. GFDL is a bit more aggressive than Hadley and nature it
turns out went beyond the pessimistic prediction of GFDL.

See for example the NSIDC comparison of the later models with actual sea
ice data. The conclusion is basically that we are warming the poles
somewhat faster than the preferred IPCC models predict.

http://nsidc.org/news/press/20070430_StroeveGRL.html

So your evidence that the predictions of climate science are correct is
in fact an article which points out every single one of the 18 models
analysed got completely the wrong answer?

This is a strange approach to determining scientific fact. Normally, when
an experiment fails to match a theory, this is evidence the theory is
wrong. You seem to think that the fact that every one of the 18 models
made demonstrably wrong predictions proves they are all correct.

You are back to sophistry again. The models make predictions that are
specific and have been validated.

Fantastic! That is exactly what I am after!

They are not perfect - nothing in the real world ever is but they describe
the overall system sufficiently well that we can have confidence in their
predictions.

Well, hold it, you got me all excited, now you are already underselling it.

Pity that nobody has bothered posting experimental verification of the
predictions of climate science anywhere on the web, particularly if there
is
so much of it. And damn those right wing corporations and their tobacco
company employee minions, for stopping it being published.


Why don't you just make up some more bull****? You have been
repeatedly provided with everything you have asked for: the papers,
model documentation, the mathematics, the model output, and
verification data.

_______________________________
Umm, no. You have several times claimed this, but is simply not true. If you
have it, please post it, maybe the original post didn't come through to my
ISP.



Your response? just more bull****. You were provide
documentation that proves a common link between techniques and people
who led the charge to delay action on tobacco and those leading the
charge to delay action on climate change.

____________________________
Well, its just that I have heard similar conspiracy proofs of so many things
that I believe to be untrue, that I am starting to doubt whether these sorts
of conspiracy theory proofs are actually true. For example, exactly the same
sorts of arguments "prove" the moon landings were a hoax, because a video
production company used by NASA specialised in outer space special effects.
I have also seen a few "proofs" that Special Relativity is wrong which rely
on a Jewish conspiracy. Its almost like the existence of one of these
conspiracy theories is evidence the underlying premise is wrong, because if
there was better scientific evidence they wouldn't rely on guilt by
association.


Both cases the reason for
delay is to continue make a profit for a very few people at the
expense of everyone else. More proof that not only do you lack even a
minimal amount of intelligence, but are greedy as well.

________________________________
Is this the scientific evidence that climate science models have predictive
power that you think has already been posted? Because its not really
scientific evidence, its more some sort of deranged conspiracy theory.
Please don't bother sharing with me your similar proofs that fluoridation of
water supplies is evil, that 9/11 was a controlled demolition, or that SR is
incorrect. I see these sorts of "proofs" all the time on the internet, and
don't need to see any more. Not for AGW or anything else.

Peter Webb wrote:

"Martin Brown" wrote in message
...
Peter Webb wrote:

"Martin Brown" wrote in message
...
Peter Webb wrote:

Have you got one which predicts cooling from 1820 to 1850 when fed
1820s data?

They will if they take vulcanism into account. I take it that you
deliberately chose a period immediately following the Tamborra super
eruption.

No.

A model would have to know about the volcanic aerosols.

OK.

You don't know if there were massive volcanic eruptions in the 19th
Century, and that is why climate models don't work for the 19th Century?

No. A model has to be given knowledge of the volcanic aerosol
contributions if it is to get the right answers.

This is your claim. To prove it, you must prove that it does provide the
right ansewrs when it does know the volcanic aerosol contributions. Say
from 2000 to 2209, assuming the model was developed before 200.

I am beginning to lose patience. Most of the answers that you seek are
covered in the IPCC FAQ at a level that should be understandable by
anyone with high school physics and a bit of perseverence.

http://ipcc-wg1.ucar.edu/wg1/Report/...Print_FAQs.pdf

In particular you should note that the inference that AGW is real does
not rely on the details of the forward projection global climate models.
It is based on observational constraints over the past century and the
requirement that the Earth's energy budget has to balance.

Conservation of energy is an extremely powerful tool and the energy
balance models do not care about the internal details of the system. If
you imagine a sphere around the Earth then by accounting for everything
that comes in or goes out through that sphere you can compute the
average temperature of the Earth.

A model that fits the observations adequately has to include:

GHG forcing (positive)
Aerosols (negative)
Solar Variation (positive)

Over the past century the bulk of the GHG forcing has been in the past
3-4 decades where GHG concentrations really rise quickly. The solar
contribution has been rising erratically. Both factors are roughly
equally important, but the GHG forcing can be ignored before about 1960.

Aerosols take into account vulcanism and sulphur emissions from power
stations during the acid rain era. You can even choose sceptic
scientific papers on this. There is no significant scientific
disagreement on this amongst the scientists.

So you are saying that the climate models are untestable before a
certain date, because the aerosol contributions cannot be determined.

No. I am saying you have to be careful. The climate model researches are
testing their models against all available paleoclimatology data. Ice
core isotope ratios provide a proxy for global temperature and the
volume of water in the oceans for instance but you have to be careful
when using inference from indirect measurements.

What was that date roughly? What is claimed period of applicability of
the model?

The model is generally applicable, but there may not be enough
observational data to constraint it.

OK. If you are serious. And I very much doubt that you are then take
a look at the printed paper copy of the IPCC Science Case Climate
Change 2001 p447 has specific predictions for Arctic Sea ice from
GFDL and Hadley models of the day as compared with actual
observations to 1998.

To 1998?

The reason I chose this specific paper is that the printed version is
locked to exactly what they predicted out to 2040 with the models and
data available up to 1998. Their predictions are a slight under
estimate of the extent of polar ice loss seen to date.

I would of course be very interested in seeing the specific predictions
made in 1998 and the actual measured data since.

Your statement that the experimental data does not in fact match the
theoretical predictions does not leave me with much confidence.

Look. You are never satisfied. They get the right sense of change and in
about the right amount to within the experimental error. I suppose even
when there is no Arctic ice cap you will still not believe in AGW. And
an ice free north polar route is on the cards in a decade or so if
current trends of warming continue or accelerate (as seems likely).

I can also see how this would provide additional evidence the earth is
warming, but I don't dispute that it. It has been warming for over 150
years. The theory of anthropogenic warming demands an anthropogenic
component. The mere fact that the climate is changing doesn't mean that
man has anything to do with it, the earth has spent long periods either
warming or cooling long before homo sapiens appeared.

I suggest you read the FAQ particularly sections 7.1, 8.1 & 9.2. They
deal with your objections in some detail.

This is actually a bit worrying since it means that the CO2 feedback
is more aggressive than the 1998 models have accounted for.

So the 1998 models have been proved wrong.

Again, I stresss that I am seeking succesful predictions of climate
science, not incorrect ones.

The prediction was successful. They err on the side of caution to avoid
being branded as scaremongering so it isn't too surprising that the
models tend to underestimate future change. I suspect there will always
be an underestimate systematic bias in IPCC model predictions because of
the nature of committees.

These models were the justification for a world-wide treaty which
called for huge changes to the structure of the world's industry.

Your evidence is a paper published in 2001 which correctly "predicts"
sea ice levels for 1998, which were in fact already known to the
author in 2001 when he correctly predicted them.

NO. CAN'T YOU READ? I THINK YOU ARE BEING DELIBERATELY OBTUSE.

The prediction made in 1998 goes out to 2040 and is so far a pretty
good match to the observations - the GFDL one is pretty close to reality.


Got a link?

I gave you one. The original paper. Primary scientific literature is not
all on the web (or if it is not free access).

It is now 2009. What did the model predict for 2001 - 2009, and how
well does it match observation?

From the graph I have in front of me it predicted a 5% per decade
decrease in Arctic sea ice and the modern observations show:

http://nsidc.org/data/seaice_index/

That is a strange graph in some ways. Partly because the bottom axis
goes from 1970, but data is omitted prior to 1980.

But as far as I can tell, it is a graph of sea ice concentration over
the last 29 years, with a dotted line emphasising the fact that this has
been generally decreasing over this period.

But you claim that some paper written 10 years ago also predicted that
sea ice levels would drop, and they did.

I might point out that I am not exactly gobsmacked by this, the earth is
indeed warming, that sea ice levels were predicted to drop and did is
hardly surprising.

But you have to ask why is the Earth warming. We can *measure* the solar
flux. It hasn't changed by anything like enough to explain the last few
decades. And the changes to the atmospheric concentrations of GHG match
the energy balance required to fit the observations extremely well. Any
reasonable practitioner would conclude that solar variation and GHG are
now significant since about 1970 and that with the rapidly increasing
levels of CO2 the latter forcing will become ever more important.

I would also of course be interested in the predictions and experimental
results for a considerably longer period than 29 years. I don't know how
much more I can ask for, as you have not yet told me what the period of
validity of the climate models is (ie when volcanic aerosols could be
measured well enough). I thought there was a lot of anecdotal **** about
the North West passage being open last century, when the world was much
cooler, suggesting different and presumably more complex variables
driving sea ice than global temperature.

See FAQ 3.1

Writing a paper in 2001 which predicts the past accurately is not
independent experimental verification. It should be easy to gain such
verification; just look at te climate changes aftewr the paper was
written and see if they were correctly predicted.

Were they?

YES. Why don't you go and read the literature?

Well, its just that its such an important and contentious issue that I
would expect experimental verification would be easy to find on the net,
if it exists that is.

Why. The models are incredibly complicated and the experts haggle over
the best way to do things and interpret the results. I have reasonable
confidence that they know what they are doing. The full climate models
are useful for predicting likely future scenarios but they are not
needed to prove that AGW is occurring.

Try that for climate science, and ... well, lets just say I haven't seen
anything vaguely resembling lists of specific predictions made by
specific models compared to subsequent measurements ... I woud have
thought that even you believers in AGW would be interested in such a
table; how else do you decide which model was correct, if not by seeing
which ones best matched experimental results?

Certain of the general principles particularly of the much simpler
energy balance models provide the evidence needed.

The full global circulation models may or may not be entirely accurate
in their simulations. But it doesn't matter. The main conclusions still
hold. You cannot attack AGW by attacking the simulations. There is
observational data that demonstrates there is siginificant additional
warming and that the GHG forcing component matches the required energy
budget. And noone can find another plausible explanation.

We are now a decade further on and you can see that they have the
trend and rate of decrease of sea ice about right when compared to
current day observations. GFDL is a bit more aggressive than Hadley
and nature it turns out went beyond the pessimistic prediction of GFDL.

See for example the NSIDC comparison of the later models with actual
sea ice data. The conclusion is basically that we are warming the
poles somewhat faster than the preferred IPCC models predict.

http://nsidc.org/news/press/20070430_StroeveGRL.html


So your evidence that the predictions of climate science are correct
is in fact an article which points out every single one of the 18
models analysed got completely the wrong answer?

This is a strange approach to determining scientific fact. Normally,
when an experiment fails to match a theory, this is evidence the
theory is wrong. You seem to think that the fact that every one of
the 18 models made demonstrably wrong predictions proves they are all
correct.

You are back to sophistry again. The models make predictions that are
specific and have been validated.

Fantastic! That is exactly what I am after!

They are not perfect - nothing in the real world ever is but they
describe the overall system sufficiently well that we can have
confidence in their predictions.

Well, hold it, you got me all excited, now you are already underselling it.

Let me guess. These models say the earth is warming. As it has been
warming now for over 150 years, that seems a reasonable bet. At worst, a
prediction made 10 years ago that temperatures would continue to rise
for the next 10 years (ie up until now) would have a fifty-fifty chance
of being true or false.

Mores sophistry. You are determined not to look at the evidence that
might alter your preconceived ideas.

For that matter, where does the "A" in AGW enter into this equation,
exactly?

The GHG forcing component is from us changing the atmospheric
composition mostly CO2 with some CH4, CFCs and N2O.

Were you the dude who taught "University level science" for 30 years?
Do you prove to your students that other scientific theories are
correct by pointing out the theory consistently fails to explain
experimental data?

You are deliberately twisting my words for your own malign purposes
here. It won't wash. The literature is all out there - if you want to
find how the models were verified and validated then go away and do it!

But you yourself can't provide any good links. OK.

I provide the links and you refuse to look at them. Coming back again
and again with the same questions. It is becoming tiresome.

That isn't what I said at all. There is a lot of work going on to test
the models against actuality, and to refine them by running other
planets and simplified test cases.

Do these other planets have people living on them burining fossil fuels?

It would seem an extremely simple task to "test the models against
actuality"; you simply list the specific predictions made by the
theories 10 years ago against measured data for the last 10 years, and
see how well they worked.

Surely this must exist somewhere on the net, aren't people who believe
in AGW interested to see which models have subsequently been proved most
accurate? How else do you decide which one is correct?

The ensemble of the models agree on certain fundamental characteristics.
I don't know which one of them is correct. If I had to bet then I think
GFDL looks promising.

excepting of course the study you provided a link to which shows that
they do not.

If I wanted "proofs" of Special Relativity, I can find them easily.
If somebody said that nobody had ever bothered to verify the
predictions of SR against experiment, and I had to do it myself, I
would be very curious as to why this basic checking had not occurred.

The basic checking has occurred. You insist that you are not satisfied
with it and cannot be arsed to go and look at any of the references
you have been given. It is your choice to be wilfully ignorant.

I haven't been given any links.

You have but you don't follow them.

Sorry, there was a link to a graph (a very tiny graph) of sea ice
levels over the last 29 years showing they were generally decreasing,
and you told me that one of the climate models you knew about also
predicted 10 years ago that sea ice levels would decrease.

Or you could read the IPCC Science Report which deals in detail with
most of the factors and remaining uncertainties in climate modelling.

http://ipcc-wg1.ucar.edu/wg1/wg1-report.html

Or, you could just simply provide the predictions of the models of
what would happen in the future, and the actual experimental results.

A report which lists all the reasons a model may be wrong (ie the
uncertainties) is not going to make it more likely I will believe it
is true.

The scientific position is to build a descriptive model of the planet
and its climate. The scientists are honest. The deniers for hire are not.


The best models still under estimate the extent, but the sense is
very strongly for global warming and more specifically strong polar
warming as the polar albedo changes.

So the best models are wrong, and this proves that climate science is
correct?

The best models reproduce the right behaviour, but the amount of
change being observed is actually slightly larger than GFDL predicts.
Taking the polar ice as a clear observable with very specific
predictions from the models:

AGW Deniers predict: no change

No. I predict a linear decrease, identical in rate to that observed on
average for the last 30 years.

So now you have to explain what is causing this change in temperature.
Excessive motion of fairies and goblins will not do!

There is a law of conservation of energy and radiative equilibrium.

In essence the Earth's global average temperature over the long term has
to balance so that energy received is equal to energy radiated. If we
alter the emissivity of the atmosphere (which is exactly what GHG
forcing does) then the global temperature has to rise to compensate.
This is what is being observed. The physics is very sound.

I don't deny the climate is changing. Always has, always will.

But you have to explain why it is changing. We have a strong law of
conservation of energy here that you are seeking to violate.

On geological timescales we have powerful driving forces that come from
the changes in the Earths orbital elements and continental drift (as
well as vulcanism).

I do dispute whether there is any strong evidence that humans are a
major contributer to this, simply because I haven't seen any, and I do
dispute whether the existing computer models are correct, because they
don't seem to work.

ITYM You haven't looked.

However, the basic principles are now clear. Adding CO2 at an ever
increasing rate will get us into trouble in the relatively near future.

And if we do not add CO2 at an ever increasing rate, we won't get into
trouble?

The amount we have already added means that temperatures will continue
to rise for a while even if we stopped completely tomorrow.

I know there is a lag effect in cycling CO2 through the atmosphere, but
in the long term, today's global temperatures look pretty close to
perfect. So surely the level of CO2 emissions that can be tolerated in
the distant future while maintaining today's temperature is an important
target. What level of anthropogenic CO2 production would be required for
global temperatures to eventually stabilise at today's levels?

We can't realistically meet that target. I suspect we will do well to
avoid doubling the CO2 concentration from where we are now. And I rather
doubt that anyone has the political will to do it.

Again, I am looking for places where experiment matches predictions,
and NOT stories which point out that theory does not match
experiment, and possible reasons why climate models are demonstrably
wrong.

Then go and read the IPCC Science Report. It deals with comparisons of
the past records with the models and how they compare. as I have
pointed out before if you get a paper copy of the 2001 report you can
compare the predictions and graphs made at that time with the actual
observations today for yourself without having to rely on outside help.

So nobody has ever thought to post evidence that climate science makes
correct predictions somewhere on the internet? Or has it been removed by
maverick right wing corporations?

But it does exist, right?

As I said LLNL hold the main repository for model intercomparison but it
is a field for experts.

I am beginning to think that you don't understand anything about how
CO2 affects the atmospheric energy balance by blocking outgoing long
wave thermal radiation. It doesn't make sense to deny the possibility
of AGW.

I don't deny the possibility of AGW. I just say I have not seen any
experimental evidence that would make me believe its true.


Add enough CO2 to the Earth's atmosphere and you would eventually end
up with something like Venus but with a surface temperature of about
690K instead of 750K (representing our greater distance from the sun).
Models won't allow you to push it that far since they limit it to
burning fossil fuels and not decomposing all the carbonate rocks.

As I mentioned, I'm really just more interested in the match between
prediction and experiment.

OK. Now maybe we can get somewhere. Ignore the complex full global
climate models (which try to completely model the Earth with regional
and 3D modelling of all the various systems) and concentrate on the
simpler energy balance ones. The latter are very robust in terms of
limited assumptions and clear evidence of an AGW forcing component. The
Baliunas & Soon paper I referred you to originally was written by
climate sceptics and even with their best efforts they could not explain
away the observations without including GHG forcing after 1970.

As I have said earlier. GFDL looks reasonable, but the models are
being refined and will inevitably get better in the future. The key
point here is that even a relatively simple model can show why adding
CO2 will make the Earth get warmer. And none of the "scientific"
sceptics worthy of the name scientist deny this any more.


But the earth is getting warmer anyway.

Between the CO2 input and the temperature output, there are about a
million lines of computer code , and a huge collection of arbitrary and
post-hoc assumptions about the transfer of heat, CO2, water vapour,
cloud cover, methane, albedo, vegetation, ice, fresh water, ocean
currents, volcanic aresols (apparently), and agriculture, the chances of
all of which being substantially correct are approximately zero, which
is born out by the fact that the models are **** at making predictions
which actually turn out to be true.

THe global models are trying now to predict regional changes to climate.
But you don't need to believe in those to obtain clear evidence for AGW.

I am posting now to make sure you cannot mislead any other weak
minded individuals than might be reading this thread.

Ha ha.

Perhaps providing some independent experimental verification would be
a better tactic?

If we were arguing about evolution or SR, and you maintained they
were incorrect, I would galdly supply you with hundreds of links to
dozens of sources of experimental verification. It is the existence
of this independent experimental verification which convinced me -
and those people who beieve in the scientific method - that SR and
evolution are correct theories.

And yet you can still find plenty of particularly electronics
engineers that still do not believe in relativity a century later. The
first GPS satellites had a disable relativistic corrections feature
because the engineers did not believe the physicists.

So, what have you got for climate science?

Whatever it is it will never satisfy you.

No, a simple list of the specific predictions made by the models used
for Kyoto, and the subsequent experimental data will do.

The FAQ has most of this.

We seem to agree this is important. You criticise my model for using
200 data points in its formulation, implying this is too high. So how
many does your model use?

I doubt over the time range where we have good climate data that
anything beyond a quadratic fit is justified.

So none of the equations used in standard climate models are more
complex than quadratic equations?

You are deliberately misinterpretting what I said. The outputs and
trends they predict do not justify anything more than a quadratic fit.
The internals are extremely complex fluid in cell models with code to
handle details that occur on finer scales than the main model grid.

Far out. I assumed they would be full of PDEs, exponential functions,
stuff far more complex than polynomials of degree 2.

It is. You are deliberately misinterpretting my answers.

Arctic Sea ice is as unambiguous as any. Very specific prediction of
the models and borne out by observations.

But didn't you post an article which said the 18 dominant models all
got this wrong?

They mostly underestimate the effect. In other words the AGW observed
now is strong than our preferred models predicted in the past.

Well, warming may be stronger, but still you sneak in that additional
letter "A" in front of "GW", when nothing you have shown demonstrates it
...

You mean that whatever I post and whatever evidence is presented you
will parrot that phrase ad infinitum.

The details of the models do not have to be perfect for the rising CO2
(and other polyatomic GHGs like CH4 and N2O) to be attributed to our
warming planet.

CO2 is rising because the planet is warming. I have heard people say
that. There seems to be some experimental data supporting this.

The experimental evidence proves *exactly* the opposite. Anyone you have
heard saying that CO2 is rising now because the planet is warming is one
of the pathological liars I warned you about. The changing isotopic
composition of atmospheric CO2, the increase in CO2 concentration and
the corresponding decrease in O2 (which is now also measurable with
sufficient precision) show clearly that we are responsible for the
changes in the atmosphere. And also that at the moment the oceans are
still mopping up some of the CO2 we emit (ISTR about 45% of it). The
seas are getting more acidic as a result which is bad news for corals.

The isotope ratio change occurs because burning fossil fuels releases
carbon with a distinctive isotopic signature (less C13 than normal).
Life concentrates the lighter isotopes.

The graph of deltaC13 is online at Scripps
http://scrippsco2.ucsd.edu/images/gr...13_mlo_spo.pdf

You can even see how during the recessions of 84, the early 90s and now
the isotope ratio trend stabilised (harder to see the change of gradient
in the concentration record). Keelings work on atmospheric CO2 is
definitive in this area:

http://scrippsco2.ucsd.edu/program_h...e_lessons.html

Logical inference allows you to conclude that if the heat input from
the sun has not got stronger (and we have accurate satellite flux
records over the past forty years) and the planet is warming then heat
is escaping more slowly. This conclusion is inescapable and even
sceptics like Baliunas and Soon concede this in their scientific papers.

Pity that nobody has bothered posting experimental verification of the
predictions of climate science anywhere on the web, particularly if
there is so much of it. And damn those right wing corporations and their
tobacco company employee minions, for stopping it being published.

I am not sure what more can be done. You clearly refuse to look at the
evidence and I cannot simplify what is a complex subject any more.

Regards,
Martin Brown

"Martin Brown" wrote in message
...
Peter Webb wrote:

"Martin Brown" wrote in message
...
Peter Webb wrote:

"Martin Brown" wrote in message
...
Peter Webb wrote:

Have you got one which predicts cooling from 1820 to 1850 when fed
1820s data?

They will if they take vulcanism into account. I take it that you
deliberately chose a period immediately following the Tamborra super
eruption.

No.

A model would have to know about the volcanic aerosols.

OK.

You don't know if there were massive volcanic eruptions in the 19th
Century, and that is why climate models don't work for the 19th
Century?

No. A model has to be given knowledge of the volcanic aerosol
contributions if it is to get the right answers.

This is your claim. To prove it, you must prove that it does provide the
right ansewrs when it does know the volcanic aerosol contributions. Say
from 2000 to 2209, assuming the model was developed before 200.

I am beginning to lose patience. Most of the answers that you seek are
covered in the IPCC FAQ at a level that should be understandable by anyone
with high school physics and a bit of perseverence.

http://ipcc-wg1.ucar.edu/wg1/Report/...Print_FAQs.pdf


I looked through this voluminous report, which contains haundreds of graphs
of measuerements of various temperatures and similar data, usually with
trend lines usefully added, but not one of the graphs or tables (as far as I
could see) showed the prediction of any model at all, and nor could I find
any data whatsoever on the predictions of various models anywhere in the
report.

In particular you should note that the inference that AGW is real does not
rely on the details of the forward projection global climate models. It is
based on observational constraints over the past century and the
requirement that the Earth's energy budget has to balance.


I understand that the models used past data in their construction, and were
in fact designed partly to reproduce already known data.

Scientific theories are tested on their predictive ability, in particular
how well they are verified by subsequent experiments, temperatures in this
case.

12 years and a bazillion dollars after Kyoto, surely somebody has bothered
to check how well each model predicted the last 12 years. In the 400 pages
and 27 Mbytes of the IPCC FAQ that you posted, I didn't see anything like
that at all.


Conservation of energy is an extremely powerful tool and the energy
balance models do not care about the internal details of the system. If
you imagine a sphere around the Earth then by accounting for everything
that comes in or goes out through that sphere you can compute the average
temperature of the Earth.

Gee, so climate science adheres to the first law of thermodynamics. That is
a huge relief. Does it also adhere to all the others as well?

You consider this to be "evidence" let alone "proof" !



A model that fits the observations adequately has to include:

GHG forcing (positive)
Aerosols (negative)
Solar Variation (positive)


No, I can create a model that fits the observations perfectly using as
stated inputs:

World canola oil production (positive). This is a proxy for the amount of
land dedicated to cereal crop production, which affects evaporation rates
and albedo.

Average wind speeds in the tropics (negative). These act to redistribute
heat.

Average real cost of a daily newspaper. I have no idea of what the causative
mechanism is here, but let me assure you that if you give me enough data
points I can fit any curve I like to them.




Over the past century the bulk of the GHG forcing has been in the past 3-4
decades where GHG concentrations really rise quickly. The solar
contribution has been rising erratically. Both factors are roughly equally
important, but the GHG forcing can be ignored before about 1960.

Aerosols take into account vulcanism and sulphur emissions from power
stations during the acid rain era. You can even choose sceptic scientific
papers on this. There is no significant scientific disagreement on this
amongst the scientists.


You just love pedantically trying to explain your complex and (frankly)
tedious theories, when what I have asked for and not received is any
independent experimental verification they are true. The obvious benchmark
case are the specific predictions made by the models used to develop the
Kyoto protocol 12 years ago, and the subsequent data.

Or any equivalent data from any other mainstream model; prediction vs
actual.


So you are saying that the climate models are untestable before a certain
date, because the aerosol contributions cannot be determined.

No. I am saying you have to be careful. The climate model researches are
testing their models against all available paleoclimatology data. Ice core
isotope ratios provide a proxy for global temperature and the volume of
water in the oceans for instance but you have to be careful when using
inference from indirect measurements.


Wow. Last time I hearda scientific theory get so complicated so quickly was
after Ptolmey and others started plotting the movements of the planets
accurately, and the numver of crystal spheres inside crystal spheres started
growing like crazy.

So, with all of these input variables considered, and such meticulous detail
placed on the development of the algorithms, I guess they must work really
well. Have you any examples?



What was that date roughly? What is claimed period of applicability of
the model?

The model is generally applicable, but there may not be enough
observational data to constraint it.

Oh, OK, I will ask the same question your way.

At roughly what time in history did there become enough observational data
to constrain climate science predictions?

Remember, my original question was roughly what is the domain of
applicability of today's climate models, that is what I am really trying to
find out.



OK. If you are serious. And I very much doubt that you are then take a
look at the printed paper copy of the IPCC Science Case Climate Change
2001 p447 has specific predictions for Arctic Sea ice from GFDL and
Hadley models of the day as compared with actual observations to 1998.

To 1998?

The reason I chose this specific paper is that the printed version is
locked to exactly what they predicted out to 2040 with the models and
data available up to 1998. Their predictions are a slight under estimate
of the extent of polar ice loss seen to date.

I would of course be very interested in seeing the specific predictions
made in 1998 and the actual measured data since.

Your statement that the experimental data does not in fact match the
theoretical predictions does not leave me with much confidence.

Look. You are never satisfied.

Because you won't give me any evidence.


They get the right sense of change and in about the right amount to within
the experimental error.

Well, what is that supposed to mean? What is the "right sense of change"?
That they predicted it would get warmer, and it did? So therefore these
million lines of computer program are correct, even if they are wrong?



I suppose even when there is no Arctic ice cap you will still not believe
in AGW. And an ice free north polar route is on the cards in a decade or so
if current trends of warming continue or accelerate (as seems likely).

Or maybe it will grow bigger. I am no expert, but I assume that at various
times over the last few million years its size has changed at various times
without any help from us.




I can also see how this would provide additional evidence the earth is
warming, but I don't dispute that it. It has been warming for over 150
years. The theory of anthropogenic warming demands an anthropogenic
component. The mere fact that the climate is changing doesn't mean that
man has anything to do with it, the earth has spent long periods either
warming or cooling long before homo sapiens appeared.

I suggest you read the FAQ particularly sections 7.1, 8.1 & 9.2. They deal
with your objections in some detail.


Regretably, on a theoretical basis, and none of them appear to cite any
experimental verification at all.

Which is more the sort of thing I am after - not reasons it is plausible,
experimental verification. Lots of plausible things are untrue.



This is actually a bit worrying since it means that the CO2 feedback is
more aggressive than the 1998 models have accounted for.

So the 1998 models have been proved wrong.

Again, I stresss that I am seeking succesful predictions of climate
science, not incorrect ones.

The prediction was successful. They err on the side of caution to avoid
being branded as scaremongering so it isn't too surprising that the models
tend to underestimate future change. I suspect there will always be an
underestimate systematic bias in IPCC model predictions because of the
nature of committees.

If you could just post some experimental verification, instead of just
saying some exists but apparently not entirely accurate ...




These models were the justification for a world-wide treaty which
called for huge changes to the structure of the world's industry.

Your evidence is a paper published in 2001 which correctly "predicts"
sea ice levels for 1998, which were in fact already known to the author
in 2001 when he correctly predicted them.

NO. CAN'T YOU READ? I THINK YOU ARE BEING DELIBERATELY OBTUSE.

The prediction made in 1998 goes out to 2040 and is so far a pretty good
match to the observations - the GFDL one is pretty close to reality.


Got a link?

I gave you one. The original paper. Primary scientific literature is not
all on the web (or if it is not free access).


And on such an important topic. With so many people (apparently) taken in by
the hucksters of the right wing corporations spreading misinformation, why
hasn't anybody who believes in climate science just summarised the
predictions and actual measured data for the main (or any) climate models
and posted it on a web page, so people can see the evidence for themselves?

Not enough money in the budget to put up a web page summarising the
independent experimental verification of climate science?



It is now 2009. What did the model predict for 2001 - 2009, and how
well does it match observation?

From the graph I have in front of me it predicted a 5% per decade
decrease in Arctic sea ice and the modern observations show:

http://nsidc.org/data/seaice_index/

That is a strange graph in some ways. Partly because the bottom axis goes
from 1970, but data is omitted prior to 1980.

But as far as I can tell, it is a graph of sea ice concentration over the
last 29 years, with a dotted line emphasising the fact that this has been
generally decreasing over this period.

But you claim that some paper written 10 years ago also predicted that
sea ice levels would drop, and they did.

I might point out that I am not exactly gobsmacked by this, the earth is
indeed warming, that sea ice levels were predicted to drop and did is
hardly surprising.

But you have to ask why is the Earth warming.

Excellent question. It has been warming for over 150 years. Why?


We can *measure* the solar flux. It hasn't changed by anything like enough
to explain the last few decades. And the changes to the atmospheric
concentrations of GHG match the energy balance required to fit the
observations extremely well. Any reasonable practitioner would conclude
that solar variation and GHG are now significant since about 1970 and that
with the rapidly increasing levels of CO2 the latter forcing will become
ever more important.

Damn, I thought you were going to tell me why it has been warming for over
150 years. Whatever it is, it aint anthropogenic CO2, because 100 years ago
it was negligible.

Why do climate scientists think the earth has been warming for about the
last 170 years?



I would also of course be interested in the predictions and experimental
results for a considerably longer period than 29 years. I don't know how
much more I can ask for, as you have not yet told me what the period of
validity of the climate models is (ie when volcanic aerosols could be
measured well enough). I thought there was a lot of anecdotal **** about
the North West passage being open last century, when the world was much
cooler, suggesting different and presumably more complex variables
driving sea ice than global temperature.

See FAQ 3.1

Here is the entirety of what FAQ has to say about sea ice:

"sea ice thickness and extent have decreased in the Arctic in all seasons,
most dramatically
in spring and summer;"

It doesn't even say over which years, let alone provide any numbers, let
alone (and this what I asked for) the predictions of the theories 12 years
ago and the subsequent measured results.

That is absolutely basic information to degermining if the models had or
have predictive power, which is what makes them true.



Writing a paper in 2001 which predicts the past accurately is not
independent experimental verification. It should be easy to gain such
verification; just look at te climate changes aftewr the paper was
written and see if they were correctly predicted.

Were they?

YES. Why don't you go and read the literature?

Well, its just that its such an important and contentious issue that I
would expect experimental verification would be easy to find on the net,
if it exists that is.

Why. The models are incredibly complicated and the experts haggle over the
best way to do things and interpret the results. I have reasonable
confidence that they know what they are doing. The full climate models are
useful for predicting likely future scenarios but they are not needed to
prove that AGW is occurring.


Well, yes, to "prove" AGW is occuring you at least need to show the
predictions of the AGW models correlate with subsequent experimental data;
it is the basic requirement of being a science.



Try that for climate science, and ... well, lets just say I haven't seen
anything vaguely resembling lists of specific predictions made by
specific models compared to subsequent measurements ... I woud have
thought that even you believers in AGW would be interested in such a
table; how else do you decide which model was correct, if not by seeing
which ones best matched experimental results?

Certain of the general principles particularly of the much simpler energy
balance models provide the evidence needed.

The full global circulation models may or may not be entirely accurate in
their simulations. But it doesn't matter. The main conclusions still hold.
You cannot attack AGW by attacking the simulations. There is observational
data that demonstrates there is siginificant additional warming and that
the GHG forcing component matches the required energy budget. And noone
can find another plausible explanation.


No one can find another plausible explanation?

And you fed these thousands of data points into millions of lines of code,
and the computer model told you to start building windmills ... this will
all seem very amusing and quaint in 200 years time.



We are now a decade further on and you can see that they have the
trend and rate of decrease of sea ice about right when compared to
current day observations. GFDL is a bit more aggressive than Hadley
and nature it turns out went beyond the pessimistic prediction of
GFDL.

See for example the NSIDC comparison of the later models with actual
sea ice data. The conclusion is basically that we are warming the
poles somewhat faster than the preferred IPCC models predict.

http://nsidc.org/news/press/20070430_StroeveGRL.html


So your evidence that the predictions of climate science are correct is
in fact an article which points out every single one of the 18 models
analysed got completely the wrong answer?

This is a strange approach to determining scientific fact. Normally,
when an experiment fails to match a theory, this is evidence the theory
is wrong. You seem to think that the fact that every one of the 18
models made demonstrably wrong predictions proves they are all correct.

You are back to sophistry again. The models make predictions that are
specific and have been validated.

Fantastic! That is exactly what I am after!

They are not perfect - nothing in the real world ever is but they
describe the overall system sufficiently well that we can have
confidence in their predictions.

Well, hold it, you got me all excited, now you are already underselling
it.

Let me guess. These models say the earth is warming. As it has been
warming now for over 150 years, that seems a reasonable bet. At worst, a
prediction made 10 years ago that temperatures would continue to rise for
the next 10 years (ie up until now) would have a fifty-fifty chance of
being true or false.

Mores sophistry. You are determined not to look at the evidence that might
alter your preconceived ideas.


Evidence! Evidence! You mean like predictions of the theory that later
turned out correct? P.L.E.A.S.E post it.

Or is it only available at members only, paid web sites?



For that matter, where does the "A" in AGW enter into this equation,
exactly?

The GHG forcing component is from us changing the atmospheric composition
mostly CO2 with some CH4, CFCs and N2O.


No, I mean in the evidence, in tthe experimental results.

I don't deny te earth is warming, has been for 170 or so years.



Were you the dude who taught "University level science" for 30 years?
Do you prove to your students that other scientific theories are
correct by pointing out the theory consistently fails to explain
experimental data?

You are deliberately twisting my words for your own malign purposes
here. It won't wash. The literature is all out there - if you want to
find how the models were verified and validated then go away and do it!

But you yourself can't provide any good links. OK.

I provide the links and you refuse to look at them. Coming back again and
again with the same questions. It is becoming tiresome.


I looked at them. Indeed, above, in the very few places where the links you
provided said anything at all relevant to my question I have cut and pasted
what it said, which was almost but not quite nothing.

All of the links you have provided do one or both of the following:

* They provide voluminous information on different geophysical parameters
over periods ranging from typically 30 to 100 years. Many of them present
much of this anecdotally, they talk about different parameters over
different and seemingly completely arbitrary time periods, talk about
"general increase" and "somewhat higher" and use vague waffle terms of that
nature.

* They present a "plausibility" argument about anthropogenic CO2 being
important.

What I want is the predictions of the various theories at the time of Kyoto,
and the match with subsequent experimental data.

Have you got that, or any succesful predictions at all?



That isn't what I said at all. There is a lot of work going on to test
the models against actuality, and to refine them by running other
planets and simplified test cases.

Do these other planets have people living on them burining fossil fuels?

It would seem an extremely simple task to "test the models against
actuality"; you simply list the specific predictions made by the theories
10 years ago against measured data for the last 10 years, and see how
well they worked.

Surely this must exist somewhere on the net, aren't people who believe in
AGW interested to see which models have subsequently been proved most
accurate? How else do you decide which one is correct?

The ensemble of the models agree on certain fundamental characteristics. I
don't know which one of them is correct. If I had to bet then I think GFDL
looks promising.


Which of the ones from Kyoto turned in the best result, and how good was it?



excepting of course the study you provided a link to which shows that
they do not.

If I wanted "proofs" of Special Relativity, I can find them easily. If
somebody said that nobody had ever bothered to verify the predictions
of SR against experiment, and I had to do it myself, I would be very
curious as to why this basic checking had not occurred.

The basic checking has occurred. You insist that you are not satisfied
with it and cannot be arsed to go and look at any of the references you
have been given. It is your choice to be wilfully ignorant.

I haven't been given any links.

You have but you don't follow them.

All blah blah about how the theories are supposed to work, all very
complicated, obviously so because I haven't seen so many references to the
term "non-linear" since I studied Calc III all those years ago.

What they consistently fail to provide is a table of prediction versus what
actually happened for the climate models of a decade ago.





Sorry, there was a link to a graph (a very tiny graph) of sea ice levels
over the last 29 years showing they were generally decreasing, and you
told me that one of the climate models you knew about also predicted 10
years ago that sea ice levels would decrease.

Or you could read the IPCC Science Report which deals in detail with
most of the factors and remaining uncertainties in climate modelling.

http://ipcc-wg1.ucar.edu/wg1/wg1-report.html

Or, you could just simply provide the predictions of the models of what
would happen in the future, and the actual experimental results.

A report which lists all the reasons a model may be wrong (ie the
uncertainties) is not going to make it more likely I will believe it is
true.

The scientific position is to build a descriptive model of the planet
and its climate. The scientists are honest. The deniers for hire are
not.


The best models still under estimate the extent, but the sense is very
strongly for global warming and more specifically strong polar warming
as the polar albedo changes.

So the best models are wrong, and this proves that climate science is
correct?

The best models reproduce the right behaviour, but the amount of change
being observed is actually slightly larger than GFDL predicts. Taking
the polar ice as a clear observable with very specific predictions from
the models:

AGW Deniers predict: no change

No. I predict a linear decrease, identical in rate to that observed on
average for the last 30 years.

So now you have to explain what is causing this change in temperature.
Excessive motion of fairies and goblins will not do!


Why has the earth's temperature increased for the last 170 years?

Well, we can kick out anthropogenic CO2, because that didn't exist 170 years
ago.

As to the real reason, you tell me. I have no ****ing idea. Why did the
temperature of the earth increase between about 1840 and (say) 1900 when
cars and electricity started kicking in?


There is a law of conservation of energy and radiative equilibrium.

In essence the Earth's global average temperature over the long term has
to balance so that energy received is equal to energy radiated. If we
alter the emissivity of the atmosphere (which is exactly what GHG forcing
does) then the global temperature has to rise to compensate. This is what
is being observed. The physics is very sound.

Yeah, climate science does not actually conflict with the law of
conservation of energy, well done, that's a big test of any new theory.



I don't deny the climate is changing. Always has, always will.

But you have to explain why it is changing. We have a strong law of
conservation of energy here that you are seeking to violate.


Ohh, sorry, I don't want to break the law. If you see me potentially trying
to break any of the laws of thermodynamics, you pull me up quick.




On geological timescales we have powerful driving forces that come from
the changes in the Earths orbital elements and continental drift (as well
as vulcanism).


Ahh, vulcanism, what a great word.

I must admit I got a thrill when you said the models were wrong, because
they ignored vulcanism. Vulcan hasn't has this much power since Rome in 300
BC, when he was last worshipped.




I do dispute whether there is any strong evidence that humans are a major
contributer to this, simply because I haven't seen any, and I do dispute
whether the existing computer models are correct, because they don't seem
to work.

ITYM You haven't looked.

However, the basic principles are now clear. Adding CO2 at an ever
increasing rate will get us into trouble in the relatively near future.

And if we do not add CO2 at an ever increasing rate, we won't get into
trouble?

The amount we have already added means that temperatures will continue to
rise for a while even if we stopped completely tomorrow.


But then it would cool right down again, right? To pre-industrial levels, if
we stopped completely?



I know there is a lag effect in cycling CO2 through the atmosphere, but
in the long term, today's global temperatures look pretty close to
perfect. So surely the level of CO2 emissions that can be tolerated in
the distant future while maintaining today's temperature is an important
target. What level of anthropogenic CO2 production would be required for
global temperatures to eventually stabilise at today's levels?

We can't realistically meet that target. I suspect we will do well to
avoid doubling the CO2 concentration from where we are now. And I rather
doubt that anyone has the political will to do it.


Obviously not an answer to my question.


Again, I am looking for places where experiment matches predictions,
and NOT stories which point out that theory does not match experiment,
and possible reasons why climate models are demonstrably wrong.

Then go and read the IPCC Science Report. It deals with comparisons of
the past records with the models and how they compare. as I have pointed
out before if you get a paper copy of the 2001 report you can compare
the predictions and graphs made at that time with the actual
observations today for yourself without having to rely on outside help.

So nobody has ever thought to post evidence that climate science makes
correct predictions somewhere on the internet? Or has it been removed by
maverick right wing corporations?

But it does exist, right?

As I said LLNL hold the main repository for model intercomparison but it
is a field for experts.


So for those poor misguided masses, brainwashed by right wing corporations
and tobacco companies, who seek evidence of independent experimental
verification of the predictions of climate science ... nothing ? ...




I am beginning to think that you don't understand anything about how CO2
affects the atmospheric energy balance by blocking outgoing long wave
thermal radiation. It doesn't make sense to deny the possibility of AGW.

I don't deny the possibility of AGW. I just say I have not seen any
experimental evidence that would make me believe its true.


Add enough CO2 to the Earth's atmosphere and you would eventually end up
with something like Venus but with a surface temperature of about 690K
instead of 750K (representing our greater distance from the sun). Models
won't allow you to push it that far since they limit it to burning
fossil fuels and not decomposing all the carbonate rocks.

As I mentioned, I'm really just more interested in the match between
prediction and experiment.

OK. Now maybe we can get somewhere. Ignore the complex full global climate
models (which try to completely model the Earth with regional and 3D
modelling of all the various systems) and concentrate on the simpler
energy balance ones. The latter are very robust in terms of limited
assumptions and clear evidence of an AGW forcing component. The Baliunas &
Soon paper I referred you to originally was written by climate sceptics
and even with their best efforts they could not explain away the
observations without including GHG forcing after 1970.


What year was the model published, what predictions did it make, how well
did these match subsequent experimental data?



As I have said earlier. GFDL looks reasonable, but the models are being
refined and will inevitably get better in the future. The key point here
is that even a relatively simple model can show why adding CO2 will make
the Earth get warmer. And none of the "scientific" sceptics worthy of
the name scientist deny this any more.


But the earth is getting warmer anyway.

Between the CO2 input and the temperature output, there are about a
million lines of computer code , and a huge collection of arbitrary and
post-hoc assumptions about the transfer of heat, CO2, water vapour, cloud
cover, methane, albedo, vegetation, ice, fresh water, ocean currents,
volcanic aresols (apparently), and agriculture, the chances of all of
which being substantially correct are approximately zero, which is born
out by the fact that the models are **** at making predictions which
actually turn out to be true.

THe global models are trying now to predict regional changes to climate.
But you don't need to believe in those to obtain clear evidence for AGW.


By "evidence" do you mean succesful predictions of the theory?



I am posting now to make sure you cannot mislead any other weak minded
individuals than might be reading this thread.

Ha ha.

Perhaps providing some independent experimental verification would be a
better tactic?

If we were arguing about evolution or SR, and you maintained they were
incorrect, I would galdly supply you with hundreds of links to dozens
of sources of experimental verification. It is the existence of this
independent experimental verification which convinced me - and those
people who beieve in the scientific method - that SR and evolution are
correct theories.

And yet you can still find plenty of particularly electronics engineers
that still do not believe in relativity a century later. The first GPS
satellites had a disable relativistic corrections feature because the
engineers did not believe the physicists.

So, what have you got for climate science?

Whatever it is it will never satisfy you.

No, a simple list of the specific predictions made by the models used for
Kyoto, and the subsequent experimental data will do.

The FAQ has most of this.


No, it doesn't, none of the specific sections you cited conatins this or
anything remotely like it.

Section 8.1 is atually entirely on the subject of how reliable the models
are, but does not mention a single experimental verification of a
prediction, it exclusively about how well the models predict the past.

By the way, "Kyoto" appears exactly once in tghe entire document, on page
115 if you are interested.




We seem to agree this is important. You criticise my model for using
200 data points in its formulation, implying this is too high. So how
many does your model use?

I doubt over the time range where we have good climate data that
anything beyond a quadratic fit is justified.

So none of the equations used in standard climate models are more complex
than quadratic equations?

You are deliberately misinterpretting what I said. The outputs and trends
they predict do not justify anything more than a quadratic fit. The
internals are extremely complex fluid in cell models with code to handle
details that occur on finer scales than the main model grid.

Far out. I assumed they would be full of PDEs, exponential functions,
stuff far more complex than polynomials of degree 2.

It is. You are deliberately misinterpretting my answers.


So you will only let my model use quadratics, but you can use any damn
functions you want? You know that's not fair; all my solutions will have to
be algebraic, yours can be transcendental ...

Since when do models get based on a beauty contest run by you, they either
have predictive power or they do not, which in science dermines whther they
are possibly correct or definitely wrong.



Arctic Sea ice is as unambiguous as any. Very specific prediction of
the models and borne out by observations.

But didn't you post an article which said the 18 dominant models all
got this wrong?

They mostly underestimate the effect. In other words the AGW observed
now is strong than our preferred models predicted in the past.

Well, warming may be stronger, but still you sneak in that additional
letter "A" in front of "GW", when nothing you have shown demonstrates it
...

You mean that whatever I post and whatever evidence is presented you will
parrot that phrase ad infinitum.


Well, unless of course it does provide experimental verification of
anthropogenic CO2, in which case I would stop asking for it.



The details of the models do not have to be perfect for the rising CO2
(and other polyatomic GHGs like CH4 and N2O) to be attributed to our
warming planet.

CO2 is rising because the planet is warming. I have heard people say
that. There seems to be some experimental data supporting this.

The experimental evidence proves *exactly* the opposite.

Well, read what you posted. "rising CO2 ... attributed to our warming
planet".



Anyone you have heard saying that CO2 is rising now because the planet is
warming is one of the pathological liars I warned you about.

But I thought that was exactly what you said. Re-read it, its still in
context above.


The changing isotopic composition of atmospheric CO2, the increase in CO2
concentration and the corresponding decrease in O2 (which is now also
measurable with sufficient precision) show clearly that we are responsible
for the changes in the atmosphere. And also that at the moment the oceans
are still mopping up some of the CO2 we emit (ISTR about 45% of it). The
seas are getting more acidic as a result which is bad news for corals.

The isotope ratio change occurs because burning fossil fuels releases
carbon with a distinctive isotopic signature (less C13 than normal). Life
concentrates the lighter isotopes.

The graph of deltaC13 is online at Scripps
http://scrippsco2.ucsd.edu/images/gr...13_mlo_spo.pdf

You can even see how during the recessions of 84, the early 90s and now
the isotope ratio trend stabilised (harder to see the change of gradient
in the concentration record). Keelings work on atmospheric CO2 is
definitive in this area:

http://scrippsco2.ucsd.edu/program_h...e_lessons.html


But the paper was written after 1984 and presumably the early 90s, so its
actually just predicting the past?

That is one of my own specialities, and something my own models do
extraordinarily well.

I was more interested in whether they could predict the future as well.


Logical inference allows you to conclude that if the heat input from the
sun has not got stronger (and we have accurate satellite flux records
over the past forty years) and the planet is warming then heat is
escaping more slowly. This conclusion is inescapable and even sceptics
like Baliunas and Soon concede this in their scientific papers.

Pity that nobody has bothered posting experimental verification of the
predictions of climate science anywhere on the web, particularly if there
is so much of it. And damn those right wing corporations and their
tobacco company employee minions, for stopping it being published.

I am not sure what more can be done. You clearly refuse to look at the
evidence and I cannot simplify what is a complex subject any more.

I have looked. At every link you have provided, and at every page, and
internally searched for "kyoto", "prediction", "experiment" and
"experimental verification", and what I have found I have reproduced above.

I could do that because there was very nearly or completely nothing at all
relevant even vaguely to experimenatl verification.



Regards,
Martin Brown

You know, I'm beginning to suspect the reason that the developers of the
Kyoto climate models are not now 12 years later and on the verge of
Copenhagen screaming from the rooftops how correct they were is because none
of them got it even vaguely right.

Why else wouldn't they and the pro-AGW group generally be advertising the
accuracy of their predictions, were they not in fact ****?

An amateur's response. (For the love of it) :-)

a) The models of 12 years ago were probably too simplistic to predict
anything really accurately. The problems were too complex, the data
too limited and number crunching too inadequate. All of which forced
further and damaging oversimplification. Even the TV news used to talk
about the problems of modelling climate.
b) The industrial revolution was well under way in Europe even if
Americans were still busy massacring Indians. Is there no CO2
inheritance from that period to show in the figures?
c) Whether model predictions were realistic, or not, no country has
done anything about their Kyoto promises. The models might as well
have predicted the moon was green cheese for all the interest shown by
a "business as usual" global political vacuum lead by Mad George
"Burning" Bush and his industrious religio-fascist-propagandist Mafia.
d) Emotional attachment to cynicism is no better than emotional
attachment to general scientific agreement. The onus is on the doubter
to show faulty methodology or results. To argue otherwise is to place
yourself squarely alongside the trolls.
e) I believe it was agreed at the Swiss climate conference-3 to
release climate data to all those who have an interest. Start
modelling! :-)

"Chris.B" wrote in message
...
An amateur's response. (For the love of it) :-)

a) The models of 12 years ago were probably too simplistic to predict
anything really accurately. The problems were too complex, the data
too limited and number crunching too inadequate. All of which forced
further and damaging oversimplification. Even the TV news used to talk
about the problems of modelling climate.

The models of 12 years ago were to simplistic to predict anything accurately
?

So they weren't in fact verified by subsequent experimental data?

Shouldn't somebody point out on the eve of the Copenhagen summit that the
models used at Kyoto turned out to be incorrect, and there is no independent
experimental verification of AGW ?


b) The industrial revolution was well under way in Europe even if
Americans were still busy massacring Indians. Is there no CO2
inheritance from that period to show in the figures?

Gee, I don't know, the models are incredibly complex, god knows what they
chucked in as input.


c) Whether model predictions were realistic, or not, no country has
done anything about their Kyoto promises. The models might as well
have predicted the moon was green cheese for all the interest shown by
a "business as usual" global political vacuum lead by Mad George
"Burning" Bush and his industrious religio-fascist-propagandist Mafia.


Yeah, waste of time, the developing world will not sign-up, and if the West
signs up to cuts then all that will happen is that electricity generation,
aluminium smelting, fertiliser production and heavy industry generally will
move from developed countries to the third world.


d) Emotional attachment to cynicism is no better than emotional
attachment to general scientific agreement. The onus is on the doubter
to show faulty methodology or results. To argue otherwise is to place
yourself squarely alongside the trolls.


No, you want to change the world's economic system, the onus of proof is on
you.


e) I believe it was agreed at the Swiss climate conference-3 to
release climate data to all those who have an interest. Start
modelling! :-)

So, where have the agreements between the climate models and subsequent
temperature been published, so we can see if the models actually work?

On Sep 10, 3:39*am, "Peter Webb"
wrote:
*

Pity that nobody has bothered posting experimental verification of the
predictions of climate science anywhere on the web, particularly if there
is
so much of it. And damn those right wing corporations and their tobacco
company employee minions, for stopping it being published.

Why don't you just make up some more bull****? You have been
repeatedly provided with everything you have asked for: the papers,
model documentation, the mathematics, the model output, *and
verification data.

_______________________________
Umm, no. You have several times claimed this, but is simply not true. If you
have it, please post it, maybe the original post didn't come through to my
ISP.

let's see on Sept 9th at 12:32am you posted

"Great. I couldn't find where the preictions of the models were
compared to subsequent experimental data, pretty basic I know, but I
couldn't find it. Have you got a more specific link to where I can
find this for some or all models? "

Further you said on Sept 8th at 6:01 am you posted

"So, let me get this straight. In order to obtain experimental
verification
of climate science predictions, I have to download, compile, and learn
a new computer system, devise an experiment for myself, undertake the
data entry, produce the output, and then compare it to experimental
data which I presume I can download from somewhere."

You claim in your posts that you couldn't find where the model
predictions were compared with observed data. Yet you claim that you
would have "download, compile ..." That says you went to the PCMDI
site and looked at what was available. Once you looked at the site you
found you couldn't just make up more lies and so you whined about
having actually do science. So you clearly did get my post about where
to get the mathematics, documentation, source code, model output could
be found Got caught with pants down on that one didn't you? I guess
what you me to do is convert a 15gb compressed binary file from PCMDI
into ASCII and post the resulting file on USENET. Once again you are
just a lazy stupid whiner just like Brad, Danny, Nancy, Ed, Gerald and
AJ. You keep repeating the same bull**** over and over again expecting
people to stop believing observed fact and believe you.

"yourmommycalled" wrote in message
...
On Sep 10, 3:39 am, "Peter Webb"
wrote:


Pity that nobody has bothered posting experimental verification of the
predictions of climate science anywhere on the web, particularly if
there
is
so much of it. And damn those right wing corporations and their tobacco
company employee minions, for stopping it being published.

Why don't you just make up some more bull****? You have been
repeatedly provided with everything you have asked for: the papers,
model documentation, the mathematics, the model output, and
verification data.

_______________________________
Umm, no. You have several times claimed this, but is simply not true. If
you
have it, please post it, maybe the original post didn't come through to my
ISP.

let's see on Sept 9th at 12:32am you posted

"Great. I couldn't find where the preictions of the models were
compared to subsequent experimental data, pretty basic I know, but I
couldn't find it. Have you got a more specific link to where I can
find this for some or all models? "

Further you said on Sept 8th at 6:01 am you posted

"So, let me get this straight. In order to obtain experimental
verification
of climate science predictions, I have to download, compile, and learn
a new computer system, devise an experiment for myself, undertake the
data entry, produce the output, and then compare it to experimental
data which I presume I can download from somewhere."

You claim in your posts that you couldn't find where the model
predictions were compared with observed data. Yet you claim that you
would have "download, compile ..." That says you went to the PCMDI
site and looked at what was available.

__________________________
All correct so far.

Once you looked at the site you
found you couldn't just make up more lies and so you whined about
having actually do science.

________________________
No, I asked if somebody had ever bothered to compare the predictions of
climate science models with subsequent experimantal data. The site you
posted does not do that, or anything similar. Hence my disappointment.


So you clearly did get my post about where
to get the mathematics, documentation, source code, model output could
be found

_______________________
But you did not provide a link to where the predictions of the models were
compared subsequent experimental data, which would seem a very easy way to
work out if the theories are correct. (Its called the scientific method, and
the perhaps gratuitous use of the word "science" in "climate science"
suggests that it should be evaluated using the scientific method)


Got caught with pants down on that one didn't you? I guess
what you me to do is convert a 15gb compressed binary file from PCMDI
into ASCII and post the resulting file on USENET.

______________________
Somebody created a 15 Gbyte computer program and never tested it to see if
its predictions were verified by subsequent experimental data? And you want
me to use this to create my own experimental evidence, because none is
available on the internet?



Once again you are
just a lazy stupid whiner just like Brad, Danny, Nancy, Ed, Gerald and
AJ. You keep repeating the same bull**** over and over again expecting
people to stop believing observed fact and believe you.


____________________
Do you or don't you have a link to a site which lists the predictions of
even one climate science model and the subsequent experimental data, where
we can see how accurate its predictions were? Even better, have you got a
link which shows the match between the Kyoto predictions and actual climate
over the last 12 years?

On Sep 10, 8:31*am, "Peter Webb"
wrote:
"Martin Brown" wrote in message
...

Conservation of energy is an extremely powerful tool and the energy
balance models do not care about the internal details of the system. If
you imagine a sphere around the Earth then by accounting for everything
that comes in or goes out through that sphere you can compute the average
temperature of the Earth.

Gee, so climate science adheres to the first law of thermodynamics. That is
a huge relief. Does it also adhere to all the others as well?

That wasn't his point. His point was, asking for more experimental
verification that global warming is real is like asking for more proof
of the first law of thermodynamics.

We know that the Sun shines on the Earth. We know that atmospheric
aerosols reflect some of that sunshine away. We know that the night
side of the Earth radiates heat into space - at a rate that is
affected by global carbon dioxide levels.

So more "proof" of global warming is needed only _if_ one thinks that
climate science maybe _doesn't_ adhere to the first law of
thermodynamics.

John Savard