Signal v. Noise

Real Climate had a pretty good article a few weeks ago about Edward Lorenz, who recently died.  The article speaks of Lorenz's butterfly effect, atmospheric modeling, and chaotic systems.

The article is definitely worth reading, but I found the last two paragraphs very insightful:

...how can climate be predictable if weather is chaotic? The trick lies in the statistics. In those same models that demonstrate the extreme sensitivity to initial conditions, it turns out that the long term means and other moments are stable. This is equivalent to the 'butterfly' pattern seen in the figure above being statistically independent of how you started the calculation. The lobes and their relative position don't change if you run the model long enough. Climate change then is equivalent seeing how the structure changes, while not being too concerned about the specific trajectory you are on.

Another way of saying it is that for the climate problem, the weather (or the individual trajectory) is the noise. If you are trying to find the common signal that is a signature of a particular forcing then averaging over a number of simulations with different weather works rather well. (There is a long standing quote in science - "one person's noise is another person's signal" which is certainly apropos here. Climate modellers don't average over ensemble members because they think that weather isn't important, they do it because it gives robust estimates of the signal they are usually looking for.)