Here you are using a computer, which means that you use technology. Technology uses science. Science is theory, research is done to prove or disprove theory. If the research provides evidence that does not match the theory [it happens] the theory has to be modified [if appropriate] or discarded completely. A theory is tested by whether it can predict previously unseen behaviour. Naturally, research is done to look for the predicted behaviour. The theory is confirmed or confounded as a result. Science is an iterative process.
This is briefly how science works.
In this respect, climate science is no different from the science of semiconductors that underpins the electronics in the PC or Mac that you are using.
Climate models are part of this. Climate models are quite different from your expressed opinion.
The fact that models have difficulty in modelling ice-melt is because the mechanisms are still not properly understood. Currently the models underestimate melting by a factor of three or so. If it was merely a matter of a little tweaking here and there, then where would the problem be?
Model modification requires SCIENCE! This is why you are wrong!
Very briefly, climate models calculate the climate by incorporating the known behaviour of the various components of the atmosphere and hydrosphere.
I suggest you read the following, it is an introduction:
http://www.realclimate.org/index.php...ions/#more-442
The following includes more information about climate models.
Spencer Weart's "Discovery of Global Warming" (AIP)
http://www.aip.org/history/climate/index.html
All of these lead to peer-reviewed science.
The rest of your post was not worth commenting upon. No substance, just rant. Why not check your facts first!