Could the AGW problem be because the model was written wrong

Question

I was guided to this link today and he seems to be saying that because the people did not include good base real time data in their model the output cannot match real world climate or weather either forwards or backwards from the startup date. This is a PDF and the main researcher is Greek and his model is real world with in most cases 100 years of validated base data for the model to work from.

http://www.atypon-link.com/IAHS/doi/abs/10.1623/hysj.53.4.671

Answer 1

The General/Global Circulation Models use the laws of physics to predict the climate - not weather.  The global is divied into 1/2 degree by 1/2 degree lat-long cells with ~10 vertical layers (the old models 20 years ago - new may have higher resolution).The laws of physics (conservation of mas, momentum etc) are applied to a parcel of air that is representative of what is typically observed in a cell. The initial values are supplied to begin with, the estimated at the same time of every cell and layer on the globe at 1/2 hour intervals for a 100 yr time horizon.  It takes a supper computer to solve the integral and differential calculus equations.

The initial values defined at input cover a large region.  There is a large number of initial values that describe the weather are typical for every grid cell on earth on any given time on a day.   The model is pun many times using many combination of initial values. That is why they are representative of climate, not weather.  climate is basically the long term average of weather.  These model can't predict weather, especially on a particular day at a particular time for a particular city.  The information provided is relevant globally for the entire time horizon modeled..

It is not just 1 model or 1 developer and very unlikely that all the models are very wrong.  Multiple models exist written by different climatological research groups.  The NOAA's Geophysicsal Fluid Dynamics Laboratory GFDL, NASA Goddard Institute for Space Studies (GISS), OSU (Oregon State University), and National Center for Atmospheric Research (NCAR) in the US developed models 20 years ago and have constantly reviewed and refined them since.  The United Kingdom has some too.  I just don't remember the agency.  This is the way scientist check each other - each use basic knowledge to independently develop answers to the questions, then evaluate their answer and compare them to other answers.  They then find consistencies and inconsistencies and determine the reasons for each.  A consistent thread of each is that increasing CO2 from human activity is causing the Earth's atmosphere to warm and this will have real effects on climate that can threaten ecological and human welfare.

I agree that models do not always provide the 100% accurate  information.  Modelers are the very first to point this out and that is why they make multiple models and use multiple runs to generalize what the impacts could be.  You can't be specific about time and location with a model designed to investigate global impacts.

I use models constantly for risk assessment (not nearly as complex as these), and a statistician - George Box- says it best:  "All models are wrong, but some models are useful."  GCMs are useful for investigating what happens to climate globally if we alter some basic model inputs (like CO2 concentration, incoming solar radiation, atmospheric aerosols for example).  It can't predict the weather on a particular time of day in a particular city and was not designed to. To arrive at a conclusion that a model is not useful because it can't arrive at a "correct" answer to a question for which it was not designed nor intended to answer is just silly. Its exactly the same logic that draws the conclusion  that a car is useless because it won't fly you the moon. The argument is just flawed and you inaccurately assume there is 1 model and nobody has bothered checking them.

In reality, there are many models. Scientists look at the information from all of them, cross check, figure out what they are good at and bad at, compare answers, and evaluate the results to decide what is the big picture is.  The big picture is human activity is causing global warming and climate change.

Answer 2

Notice that the notion that the models are correct is all talk, without backing things up.  How about in 1990, they predicted that temperatures would be so much by today, and this is what they are.  They just make a claim that they are accurate and they want us to believe them.

Here is an interesting letter from NASA to James Watt that further show that the models are wrong:

    Thank you for your interest in the AIRS CO2 data product.

    We are still in the validation phase in developing this new product.     It will be part of the Version 6 data release, but for now those of us     working on it are intensively validating our results using in situ     measurements by aircraft and upward looking fourier transform IR     spectrometers (TCCON network and others).

    The AIRS CO2 product is for the mid-troposphere. For quite some time     it was accepted theory that CO2 in the free troposphere is     “well-mixed”, i.e., the difference that might be seen at that altitude     would be a fraction of a part per million (ppmv). Models, which     ingest surface fluxes from known sources, have long predicted a smooth   (small)variation with latitude, with steadily diminishing CO2 as you     move farther South. We have a “two-planet” planet - land in the     Northern Hemisphere and ocean in the Southern Hemisphere. Synoptic     weather in the NH can be seen to control the distribution of CO2 in the free troposphere. The SH large-scale action is mostly zonal.

    Since our results are at variance with what is commonly accepted by he scientific community, we must work especially hard to validate them.

    We have just had a paper accepted by Geophysical Research Letters that will be published in 6-8 weeks, and are preparing a validation paper.

    We have global CO2 retrievals (day and night, over ocean and land, for  clear and cloudy scenes) spanning the time period from Sept 2002 to the present. Those data will be released as we satisfactorily  validate them.

  

    Ã¢Â€Â”——-

    AIRS Team

Answer 3

No.

One guy (an engineer) thinks the global models are bad,  because they can't predict the weather in 8 specific cities?  Thousands think they're not, in trems of predicting the average climate of the Earth.   You'll forgive me if I don't find the one guy more persuasive?

So what if they don't predict local weather in 8 specific cities?  They do predict global average climate, very well.

Note also that various models all give similar responses.  They also give similar responses, even if parameters in them are varied within reasonable limits.  In scientific terms the models are "robust".    Here are just a few of the models, from all over the world.

Table 4. AOGCMs in the Intercomparison

Originating Group Country           Model

BCCR Norway                         BCCR-BCM2.0

CCCma Canada                        CGCM3.1(T47/T63)

CCSR/NIES/FRCGC Japan               MIROC3.2(medres/hires)

CNRM France                         CNRM-CM3

GFDL USA                            GFDL-CM2.0/2.1

GISS USA                            GISS-EH/ER

INM Russia                          INM-CM3.0

IPSL France                         IPSL-CM4

LASG/IAP China                      FGOALS-g1.0

MIUB/METRI/KMA Germany/Korea        ECHO-G

MPIfM Germany                       ECHAM5/MPI-OM

MRI Japan                           MRI-CGCM2.3.2

NCAR USA                            CCSM3

NCAR USA                            PCM

UKMO UK                             HadCM3

UKMO UK                             HadGEM1

This is not persuasive.  There's a reason this study is not front page news, and it's not the "liberal media conspiracy".

EDIT - I don't think these guys listen to moveon.org much.  You might want to listen to them:

"Former Republican House Speaker Newt Gingrich challenged fellow conservatives to stop resisting scientific evidence of global warming"

"National Review (the most prestigious conservative magazine) published a cover story calling on conservatives to shake off denial and get into the climate policy debate"

"Pat Robertson (very conservative Christian leader) 'It is getting hotter and the ice caps are melting and there is a build up of carbon dioxide in the air.  We really need to do something on fossil fuels.”

"I believe there is now more than enough evidence of climate change to warrant an immediate and comprehensive - but considered - response. Anyone who disagrees is, in my view, still in denial."

Ford Motor Company CEO William Clay Ford, Jr.

"The science of global warming is clear. We know enough to act now. We must act now."

James Rogers, CEO of Charlotte-based Duke Energy.

What's political denial is accepting this one guy over thousands of trained climatologists and every major scientific organization.

Answer 4

One model out of many.  Nothing to see here.

Bob's answer was not vulgar, was reasoned and contained no personal attacks.  Yet, you attacked him.  Perhaps you should look up the meaning of the word hypocrite.

Answer 5

I think this is an interesting study on the applicability of downscaling of certain GCMs to particular cities. Most GCMs are fit to past climate data in the mean, so we know they'll get that right and nothing in this study invalidates the broad conclusions of a warming climate, just the accuracy of regional assessments. I think downscaling is one of the most difficult problems for forecasting future climate and it's not at all clear that downscaling to single points in different climatic regions (as done in this article) is valid.  It would have been more useful to do 5 or 10 times as many locations in each region to avoid just the sort of fluctuations that they say invalidates other studies.

Note that many regional downscalings do hindcast past regional climate correctly, so while this paper seems to question pointwise downscalings I don't think it can invalidate the concept of regional downscaling in general.

It's a very good question though.

EDIT: While it's a good question, characterizing believers in AGW as "vulgar" is insulting. You might also want to question the accuracy of your sources if you truly believe that "wisenheimer" is an old German word--it is 20th century American slang.

Answer 6

looking at your other answers, it's fairly obvious you deny global warming.

and it would appear that you're fishing for support for your position.

consider that all of the honest science indicates that global warming is real.  eg.

NAS, NOAA, NSF, NASA, EPA, MIT, UCLA all agree. AGW (Anthropogenic Global Warming) is a serious problem.

http://www.nasonline.org/site/PageServer

"May 19, 2008: The National Academies have released the 2008 edition of "Understanding and Responding to Climate Change," a free booklet designed to give the public a comprehensive and easy-to-read analysis of findings and recommendations from our reports on climate change."

http://dels.nas.edu/dels/rpt_briefs/clim...  <== here's a good description.

http://www.funnyweather.org/  <== this is a more lighthearted link.

http://lwf.ncdc.noaa.gov/oa/climate/glob...

http://www.ncdc.noaa.gov/paleo/globalwar...

http://www.exploratorium.edu/climate/  <== not regulated by the government.

http://www.nsf.gov/news/news_summ.jsp?cn...

http://www.nsf.gov/news/news_summ.jsp?cn...

Answer 7

(eric c), makes a valid point. during the 1990 period climate sensitivity was a 3-6C toleration. Basically a 50/50 chance of accuracy. As with most technology, things improve exponentially with time. There are areas that don't reflect true comprehension or fidelity but within 8-10 years many environmental factors/studies along with super-main frames will allow for more precise analytical data.