2005 Will Probably be Warmest on Record
Nilmat writes "A Washington Post Article notes that 2005 will probably have the highest mean global temperature of any year since the advent of systematic temperature records. At the moment, the mean temperature is about 0.75 degrees C above the global mean from 1950 to 1990, approximately .04 degrees higher than 1998, the year of the previous record. Only something dramatic, such as a major volcanic eruption, could cause enough cooling to miss setting a new record."
No, correlation is not causation. But when you have correlation and the most accurate models imply causation, you definitely have to think hard about what you're doing. The fact that global warming was predicted by the models before the data could be taken further suggest that it's not simply alarmist readings of the data.
Science is hard; in many fields it's impossible to prove causation completely. But when you have a theory, and the theory holds up to all the available data, you act as if the theory were true and make decisions based on that. You don't over-react as long as there are competing theories that imply otherwise, but this is one more piece of data to suggest that global warming is very real and quite possibly man-made.
The "quite possibly" means that we shouldn't over-react; as you say, the correlation need not imply causation. But as the burden of evidence falls on the side of man-made global warming, it becomes increasingly dangerous to rely on "Yeah, but are you really, utterly, totally, completely sure?" arguments against action.