What's the problem with global climate models?

The thing about climate models is that they're based on the laws of physics. Some people argue that they're worthless, no better than guessing, but this neglects not only the fact that they've proven quite accurate so far, but the fact that we know the laws of physics quite well.

If something causes an energy imbalance on Earth, its temperature will change in response. If solar irradiance increases, the planet will warm. If the concentration of greenhouse gases in the atmosphere increases, the planet will warm. Granted it's difficult to predict exactly how much the planet will warm due to uncertainties like water vapor and clouds, as you mentioned. Nevertheless, we have a pretty good idea that if we double the amount of CO2 in the atmosphere, the planet will warm about 3 degrees Celsius in response. Not exactly 3 degrees, but not far from it. We know this not just from running models, but from observing how much the global temperature changed in the past in response to energy imbalances.

Then there's the fact that climate models were pretty darn accurate even 20 years ago, when they were much less sophisticated than they are now.

http://www.realclimate.org/index.php/archives/2007...

http://tamino.wordpress.com/2008/03/26/recent-clim...

When you take this all into consideration, the arguments against climate models aren't very convincing. But even if you throw the models out the window, the physics tells us that the planet will keep warming if we keep adding CO2 to the atmosphere.

*edit* oh geez I love Ben O's comments that hindcasting is a simple task. This is clearly somebody who's never worked with computer models. Just try fitting a chaotic set of data with a model using 30 degrees of freedom. It's incredibly impressive that the models hindcast the 20th century temperature changes so accurately. As I've said, they can't just tweak them however they want - the models have to be consistent with the laws of physics and the scientific data we've collected.

This is really the problem people have with climate models - they simply don't understand how they work.

That's very much like asking "what information do maps present". Maps are models of geography that can present may things on many scales, but leave out most of reality because it just detracts from what you are trying to understand - soil type doesn't matter much to folks who are trying to follow the interstate, for example, although it was vital knowledge to the builders of the system. A computer model of a physical system will include as much info as is necessary to understand the variables you're studying. Milankovic cycles are important to the long-term climate, for example, but are understood well enough that you don't have to put them into a model that is looking at 100-year detail. Many times a simple one-dimensional model will show important constraints on complex systems. Consider the joke about "you're an engineer if you've ever assumed a spherical horse".

Actually, 30 years ago the climate models were predicting global cooling. There is a bias in the research and the prediction with the data.

Climate change models are indeed an estimate which as more data becomes available become more accurate. This is why scientific organization have ranged estimates for their models like Arctic ice decline, sea level rise or temperature rise. That deniers keep complaining that models are not exactly accurate shows how little they understand what models are or how they are interpreted.

That some like Peter J talk of the "bad hurricane years of 2006, and 2007... did they hold that prediction for 2008?"

For a start these increases in hurricane activity are expected to happen over the next 50 years so any short time change is not really expected, but even so, what did happen in 2008

http://www.noaanews.noaa.gov/stories2008/20081126_...

Someone needs to work on their research skills.

With enough research and facts derived from the research anything can be proven. Besides there is little debate on whether there is global warming. Only some nuts out there deny that it is happening. The effect we have on global warming is what is debatable and how much we have accelerated this cycles warming vs past warming cycles or whether it really changed the cycle by much or just by a few tens of years. I believe NASA projects that the tipping point will happen within the next 100 years what is debatable is, would it have happen a thousand or a million years from now if it were not for the what we do or 150 years from now?

The goal of climate models is to determine the sensitivity of the atmosphere to CO2, and by inference, the associated heating of the atmosphere. There is a great deal of variation in the 20-odd IPCC models when it comes to sensitivity.

At nature.com, (a well known skeptical denier fringe site - NOT!), an April 30 article says: "Gauging how the planet will respond to rising emissions remains one of the biggest questions in climate science."

In the same article, climate physicist Reto Knutti says "There is a true climate sensitivity. We just don't know its true value."

"If scientists could nail the number for sensitivity exactly, it would give a much clearer view of how global warming will change the face of our planet."

http://www.nature.com/climate/2009/0905/full/clima...

In those models, assumptions are made about hundreds of interactions in the atmosphere, resulting in a large disparity (an order of magnitude). What if the assumptions were incorrect? What if the basic assumptions were reversed? Dr. Roy Spencer has published data that proposes that is the case. If Spencer's data is correct, and his data is run thru the IPCC models, the net change may well indicate that global temperature changes are within normal variations.

Both by publishing his data, and by personal invitation, Spencer has asked the MGW community to review his work and prove him wrong. To date, am unaware of any takers... Wonder why.... Hmm.. maybe suspect the MGW alarmists don't want to see the results.

I'm still waiting for the record warmth, and the really bad hurricane years of 2006, and 2007 ... did they hold that prediction for 2008?

Edit...

I'm sorry, is there some reason I should let people pretend they can predict the climate when they can't? Should I not comment that they've pretended they can predict the climate and they haven't been able to? Should we just listen to them and pretend their predictions are worth something when they aren't?

With all due respect sir, if they're making predictions, and the only reliable thing about their predictions is that they're wrong, why should we take them seriously?

Predicting the past is not difficult. For anyone with an understanding of mathematics - fitting a model to past data is an absolutely trivial exercise. The only way to check if a model works is to predict something before it happens.

Predictions of global warming keep getting postponed. You may believe they are accurate, but the unprecedented warming predicted 10 years ago has been replaced with cooling and who knows when we will see any warming, never mind unprecedented warming as we are in a cooling trend now.

Peter J, how about you create your own, more accurate, climate model that correctly explains why the "record warmth, and the really bad hurricane years of 2006, and 2007" didn't occur.

Or how about you provide sources for these claims. It'd be lovely to know who "they" are.

It'd also be great if you could explain why a climate model should've been able to predict the freak weather events of 2007 and 2008 (re: hurricanes).

It'd be great if you wrote something -- anything -- of actually substance besides "DURRRRR NUH UH".

With all due respect, sir. :)

First up is the basic assumption that is made concerning co2 functioning as a greenhouse gas. Every published test or experiment I have seen to date uses water first to produce the co2 and secondly to verify the presence of co2 in the sample by turning the water milky. So we have a heavy presence of water in an experiment that is going to be heated to 100 degrees and probably more. What happens in nature when you heat water, it vaporises forming water vapor or humidity. What is the most powerful and greatest concentration greenhouse gas, humidity or as some would say water vapor.

What single instrument is always missing from all of the published co2 experiments, a hygrometer to measure the humidity of the sample, why because the presence of hygrometers in the samples would reveal the variation in humidity between the saturated co2 sample and the drier ambient air sample. I have done these experiments both with and without the hygrometer and with samples having identical humidity levels and with unbalanced humidity levels.

In every case whichever sample had the highest humidity level always retained more heat, warmed faster and stayed warm longer. The only way co2 can ever be shown to be a greenhouse gas is by contaminating it with humidity. Remove the humidity and it is less a greenhouse gas than nitrogen which comprises about 80% of our planets atmosphere. Only water vapor can be scientifically shown to be a greenhouse gas because any other gas must have a high concentration of it in order to perform as such and this is why all the experiments were written procedurally to use lots of water vapor in them.