Unfortunately no computer model prediction 100 years into the future has ever been tested. That's the essence of science: propose a theory and then back it up by testing its predictions.

Models are getting good now at predicting the past, i.e., they can correctly predict 1980s climatology from earlier data without depending on too many "free parameters". But of course the experimenter already knows what the "right" answer is in this case. To gain credibility models need to predict, say, 2012 climatology and then have it come true (and do it more than once to avoid luck).

Climate prediction is just a very hard problem. Historic data is untrustworthy (in the case of tropical cyclones "historic" means any data earlier than 1980(!) may be questionable) and nobody knows what cycles dominate over which time scales. The computers are good enough but we just don't know what to put into them.

It would be very nice to be able to plan and prepare for a warmer or colder world in the future but there are no credible (i.e. a history of tested predictions *that came true*) forecasts of that sort.

James_Van_Artsdalen makes an excellent point above that the predictions need to be tested for and demonstrate accuracy, before they should be given much creedance. The short term models keep getting tested and refined when there are inaccuracies, and thereby have gotten better.

Long term models, the data is still out.

We as a people like everything now, we are into instant gratification. Our modern lives are built on this now idea. We want food instantly (fast food and micro waved food) we want the answers to any question answered NOW! We want fast internet so the pages load NOW! We are so into instant everything that we forget (or don’t care) that many things take time to get the answers or to get the right answers.

We see these reports on what it’s going to be like in hundreds of years and are unwilling to wait for the results to be proven right or wrong. I don’t see it as a good thing to act on a prediction that has absolutely no base in facts and that is politically motivated. I do think it’s a good idea to work for cleaner energy production, but not so much as to harm our living standards as seems to be the desire with the global warming proponents right now.

Hi James_Van_Artsdalen




The science is pretty well understood (the physics and thermodynamics of the atmosphere i.e how the amount of CO2 will tend to warm a volume of Nitrogen and Oxygen when illuminated with IR radiation etc), finite element analysis modelling is well understood, even solving complex non-linear differential equations is well understood.

What the BBC experiment did was to start the computer model of the global climate in 1920s using the limited real world data set of conditions in the 1920s of the atmosphere as a starting point of the simulations. The variability of the starting condition parameters was assigned to each of the computers within the distrubution. Each computer within the distribution then ran its own individual climate prediction. Hundreds of thousands of climate predictions were computed each with a slightly different set of starting conditions. The recorded data of atmospheric conditions from the 1920s would have of course been subject to recording error.

There was not just one climate model, which was selected, there were over 250,000 models. Each climate model was then grouped and averaged statistically and the results produced an overall trend.

What is interesting are the results, the slight cooling in the 1970s is seen. The results do seem to follow the temperature trends of the global climate from the 1920s up to the present day with remarkable accuracy to the recorded climate data since the 1920s. In effect the model has been tested in accordance to the principles of science. The theoretical modelling used in distributed climate modelling exercise is being compared to the hundred of thousands of atmospheric recordings that have been made since the 1920s until the present day.

http://www.bbc.co.uk/sn/climateexperiment/theresult/graph1.shtml

It's called the "solar constant", which turns out not to be constant. Except that the variability over the last 50 years doesn't seem enough to amount to anything. Except .. the last "Little Ice Age" seems to have coincided with the "Mauder Minimum", a period when there were 3-4 orders of magnitude fewer sunspots than we see now. Coincidence? Important? That's the kind of thing people still do their dissertation on...

If you flip a coin how good are you at predicting which side comes up? At best your a bit better than 50% accurate.

Now flip that same coin a million times. I can't tell you with any degree of accuracy what any one flip will come out. But I can give you extremely good odds that over a million flips very close to 50% will end up heads. And the more flips you make the closer to the ideal the numbers will get.

In statistics it is known as the 'rule of large numbers'. Which is why even the best meteorologists often can't predict what the weather will be tomorrow but any competent climatologist with a statistical background can, with very high confidence, predict global warming based on the preponderance of the evidence available recording sources back thousands of years. When you get a large number of data points that record things over a very long time the trends become startlingly clear.