One of the big issues in the Climate blog world is the way in which temperature proxies are used to guess/analyse past temperature. This is really an issue, but unless you dig into things you might never know.

One of the blogs I read on occasion had a great description of one of these issues. Using Multi Variate Regression to estimate past temperatures.

Go check out The Climate Skeptic article for a fairly easy to understand explanation. Follow the logic of the argument, and then give it a think. You might end up shaking your head and thinking “what the heck??”.