>>>http://science.nasa.gov/headlines/y2009/29may_noaaprediction.htm
>>
>>I especially like this quote:
>>
>>"It turns out that none of our models were totally correct," says Dean Pesnell of the Goddard Space Flight Center, NASA's lead representative on the panel. "The sun is behaving in an unexpected and very interesting way."
>>
>>That is the problem with trying to model natural, dynamic processes. It is very difficult to model chaos. I loved modelling because of the tweaking/tuning possibilities, but in the end, they are only models, not reality.
>
>I cannot agree with yours: "That is the problem with trying ", and "they are only models, not reality".
>It is not a problem, but the only way we can explore the nature: "on error think again and revise the approach".
>The more we learn on our failures, the closer we are to "reality". Example: compare weather forecast today and 50 years ago.
>
>My $0.02

Don't get me wrong, I'm not anti-model, I love modelling. It was just refershing to see an admission that modells are imperfect. Models are useful and they are the only way we can explore many natural processes. It is what I did for a living in my previous career. My main problem is with how model results are presented to the public. Most consumers of models (management, politicians, the public) always wanted the 'number', something easy to understand. They typically don't understand that the reported 'number' represents one possibility of a range of possibilities. Consumers we are usually not provided with any indication of how wide that range is, unless they are willing to dig into the source documents.