The first paragraph in an article from physics.org:
"Researchers from institutions including the Department of Energy's Oak Ridge National Laboratory have created a new method for statistically analyzing climate models that projects future conditions with more fidelity."
___________________________________
"We don't judge models individually," said Elias Massoud, a computational ecohydrologist at ORNL. "Instead, we see how they can be put together, using their combined information to get projections of the future."
A federal agency combines predictions of the future to arrive at a more accurate prediction than that of any one individual prediction. This will increase the "fidelity" of their own predictions. Since no one can currently assure us of the details of the climate future, that fidelity can't be measured at this time, regardless of the methodology used.
Certainly if a very large number of predictions are included in the study, perhaps all possible outcomes, a small number of them may be accurate. Those can't be determined until future conditions become present, or even better, past conditions. The ORNL study is a Bayesian one, using posterior data in a statistical format to arrive at an outcome. If it was that simple we could all be rich.
No comments:
Post a Comment