Correct, Grendel, correlation does not imply causation.
What it does provide, where a causative mechanism is known, is evidence that either:
- a causative relationship exists between the two observed phenomena, or
- both of the observed phenomena are influenced by another external phenomena.
In more plain English, that means that if you can provide a reason for one thing causing another, correlation suggests that it may be right, unless both things are being influenced by something else (such as US music quality and oil production, which have both been influenced by an increase in international trade).
The mechanism for warming occurs at the sub-atomic level: certain substances block more low frequency light than high frequency light, hence the energy differential is retained in the atmosphere as heat. Now if quantities of those certain substances are increasing in our atmosphere, then theoretically, warming will occur. It might be prudent then to see if warming has occurred over the same period that the substances (GHGs) have been increasing. And lo and behold ...
The correlation between GHG levels and global temperatures therefore provides evidence that either:
- increasing levels of GHGs cause higher global temperatures,
- increased global temperatures cause increased level of GHGs, or
- both global temperatures and GHG levels are influenced by a common external phenomena.
Unless,
- an uncanny coincidence has occurred.
Perhaps the following graph would have been more appropriate: