Those following the fishtale of m=11 smoothing to test the TAR projection will be interested to learn that Stefan Rahmstorf confirms what we, in the more “excitable” part of the blogosphere, suspected: He has abandoned m=11 year smoothing to test observations against models. He now uses m=15 years. JeanS shows the difference in trend projections back when Rahmstorf et al was published in 2007 and now, using both m=11 years and m-15 years: [caption id="attachment_5517" align="aligncenter" width="499" caption="Figure 1: Comparison of change due to smoothing."]
[/caption]
Read more »
Comments: 49