The new assessment of the likely ranges now relies on a larger number of climate models of increasing complexity and realism, as well as new information regarding the nature of feedbacks from the carbon cycle and constraints on climate response from observations.
they are used to make cars, and some aeroplanes have been completely designed in them. and virtual wind tunnel tests too have proven sucessful.
its a useful thing computer modelling. but not perfect.
i'm afraid i have no experience of climate modelling. but i've done a smidgen in cosmology.
there are always simplications and the question if some effect is due to the crudeness of the model.
the same sort of concerns no doubt exist in climate modelling. but its not a worthless thing to do. much is understood, and the only way to do a forcast is to simulate. if you know whats going on. the forcasts should be reasonably good.
I presume the models have been well argued over and picked appart by the scientific community. and will continue to be.
I have no idea but im having doubts on their validity the more I look into it...
That was my early impression. I now realise that at the moment they are primarily a tool designed to assist the researcher in understanding and refining theories. However, their modelling of the energy balance of the earth has been well tested (which is the key in assessing global temperature changes). They represent the climate pretty well, and as more and more factors are added they get better. There is active research in seasonal and decadal prediction allowing probabilistic assessments to be made of regional climate changes. Their ability to accurately predict the impacts of volcanic eruptions could be very helpful, say, in modelling the impacts of pollution being produced in the east asian economies.
climate models were exposed for being less than accurate by the chap (corben) whos predictions of weather by using the sun as a reference were more accurate than the met office this is fact.
