Comments on: Models! Pt. I http://www.calacademy.org/blogs/climate/?p=31 The science behind a global issue Fri, 17 May 2013 18:20:59 +0000 hourly 1 http://wordpress.org/?v=3.5.1 By: peter http://www.calacademy.org/blogs/climate/?p=31&cpage=1#comment-275 peter Mon, 01 Sep 2008 20:32:45 +0000 http://www.calacademy.org/blogs/?p=31#comment-275 Weather prediction models are not of the same sort that are used for long-term climate prediction. Weather prediction is generally not useful at time scales beyond a week because weather is a chaotic system. This does not mean that it is random. Chaos is deterministic, but prediction is limited severely by limitations on the accuracy of model input. However, I think that we can agree that long-term climate is reasonably predictable; I predict little rainfall in the Sahara Desert in January 2009, but moderate to considerable rainfall in northern California during the same interval.

Weather is not climate, and climate is not weather, though the two are related. Doubled CO2 does not automatically lead to an increase of 3 C, but that is certainly within the range of predictions of many of the GCMs. All these models are carefully documented, Mr. McIntyre has the same access to them that I do! As for data access, which seems to be his bugbear, all U.S. government data kept by NOAA and NASA are available for download.

]]> By: Don http://www.calacademy.org/blogs/climate/?p=31&cpage=1#comment-273 Don Mon, 01 Sep 2008 20:00:32 +0000 http://www.calacademy.org/blogs/?p=31#comment-273 It seems that GCMs can predict tomorrow’s weather. They also get fairly close one week out. Beyond that, the GCMs are useless. But that does not answer the question about CO2 and temperature. There may be other calculations that can be done. Steve McIntyre has a request of climate scientists that has not been responded to as far as I know. Maybe you can write out a careful “engineering quality” exposition of how doubled CO2 leads to 3 deg C in which all the key parameterizations and feedbacks are written down and described. Can you do that?

]]>