Perhaps 50 years ago it would have been fair to say that policies to reduce CO2 emissions were based on the Precautionary Principle, because our climate science was still not very advanced and we could not say for sure how much damage CO2 was actually doing to the world's climate. We were pretty sure it was doing some damage, but we didn't have good models to say how much.

But today, it is very clearly the scientific method at work. Indeed, our climate models are astonishingly precise, given the complexity of the system involved; but because we have very good mathematical models and very fast computers to run them on, and we are constantly collecting new data in exquisite (some would say excruciating) detail from land-based, sea-based, airborne, and even satellite-based equipment. The current state of the art is not so much forecasting the climate itself as it is forecasting human behavior (obviously even more complicated) in order to estimate what the impact of various climate policies will be. It is based on these models that we know we need a dramatic reduction in CO2 emissions by the middle of this century or else the world will suffer grave outcomes that could kill tens of millions of people. (I'm sure that seems a bit macabre, but it's important to know---a lot is at stake here.)

Indeed, the real challenge is neither to forecast the climate nor to predict the outcomes of human behavior---it's to convince policymakers and the public to take the proper action. That part we still haven't figured out yet.