They can't even get the weather correct on a day-by-day basis and we're expect to believe them when they say what it's going to be like in 50 years.
Here's a challenge that I use:
Take any of the computer models being used to forecast climactic changes in the future and ^run it backwards^ to say, the Great Lakes region, May 30, 1650 AD.
Does it precisely (within +/- 0.05 C) predict the recorded temperature for that day?
(That large "POP!" you hear is the gorites head exploding.)