Thursday, May 01, 2008

When is a Projection Not a Prediction?   [Chris Horner]

I can't let the patent absurdity of Andrew Revkin's Dot Earth blog item from last evening go unremarked. Revkin's spin on the forthcoming Nature report is that they are predicting cooling now because they wanted to make their models more realistic. Titled "Moving from Projections to Predictions on Climate," the piece speaks for itself.

Scientists are forecasting a temporary cooling, from shifts in the oceans including the Pacific Decadal Oscillation, which has cool and warm phases shown in this graph. (Credit: Joint Inst. for the Study of the Atmosphere and Ocean/NOAA & Univ. of Washington). . . .

The researchers, writing in the journal Nature, stress that this is a preliminary attempt to shift climate models toward becoming a forecasting tool, mainly by tweaking them with real-world data (in this case ocean temperatures) as they churn through their simulations.

What a novel idea, "tweaking" climate models with "real-world data" so that they can forecast actual climate phenomena. What does that tell you about what the climate models were designed to do previously?

05/01 04:00 PMShare