From: Michael Mann To: Tom Wigley Subject: Re: help Date: Tue, 22 Sep 2009 11:00:28 -0400 Cc: Phil Jones , Caspar Ammann Hey Tom, thanks for checking w/ me on this. Re, Moberg: Yes, in fact we (me, Phil, Tim, Keith, Caspar, etc.) submitted a comment to Nature about the problem w/ the variance scaling used by Moberg. It can easily be shown to inflate the low- frequency variance in synthetic experiments. I've attached both the original comment (which they judged to be too technical to merit publication) and also a J. Climate paper where we discussed the same result (see Figure 5 and associated discussion). Re, Von Storch et al. Yes, the paper you have in mind is Osborn et al Climate Dynamics '06. I only seem to have the preprint though (attached), please let me know if I can be of any further help w/ an of this, mike p.s. you can delete the U.Va email address--haven't been there for 4 years! On Sep 22, 2009, at 10:31 AM, Tom Wigley wrote: > Dear all, > > (Apologies Mike for email address confusion -- one of them will > get you I hope.) > > I need some help to finish a report I've had to write for EPRI -- > which is due in a few days. Hence the questions below ... > > (1) The Moberg paper (2005 Nature) is used by the skeptics as evidence > that most of recent warming could still be natural. Has anyone > published a critique/criticism of this? It seems to me take this > work is fundamentally flawed. First, variance scaling is crap > statistics as it produces results with far less explained variance > than normal least-squares regression. Second, the paper seems to > have no independent validation. Third, what happens if one just takes > his low-frequency (numbered in his Fig. 1) points and calculates > the area average? Surely this will have much greater variability > than the full global mean? (If no-one has done this please let me > know -- I can do it very easily myself.) But perhaps his scaling > method circumvents this "problem"? > > (2) What is the paper of Caspar's (with Doug Nychka) that shows > that McIntyre is wrong? Are there other papers I should see/cite > in this regard? > > (3) What are the papers that explain what is wrong with the von > Storch ECHO simulation? I think Tim Osborn did something on this. > > Many thanks for your help, > Tom. > -- Michael E. Mann Professor Director, Earth System Science Center (ESSC) Department of Meteorology Phone: (814) 863-4075 503 Walker Building FAX: (814) 865-3663 The Pennsylvania State University email: mann@psu.edu University Park, PA 16802-5013 website: http://www.meteo.psu.edu/~mann/Mann/index.html "Dire Predictions" book site: http://www.essc.psu.edu/essc_web/news/DirePredictions/index.html Hey Tom, thanks for checking w/ me on this. Re, Moberg: Yes, in fact we (me, Phil, Tim, Keith, Caspar, etc.) submitted a comment to Nature about the problem w/ the variance scaling used by Moberg. It can easily be shown to inflate the low-frequency variance in synthetic experiments. I've attached both the original comment (which they judged to be too technical to merit publication) and also a J. Climate paper where we discussed the same result (see Figure 5 and associated discussion). Re, Von Storch et al. Yes, the paper you have in mind is Osborn et al Climate Dynamics '06. I only seem to have the preprint though (attached), please let me know if I can be of any further help w/ an of this, mike p.s. you can delete the U.Va email address--haven't been there for 4 years! On Sep 22, 2009, at 10:31 AM, Tom Wigley wrote: Dear all, (Apologies Mike for email address confusion -- one of them will get you I hope.) I need some help to finish a report I've had to write for EPRI -- which is due in a few days. Hence the questions below ... (1) The Moberg paper (2005 Nature) is used by the skeptics as evidence that most of recent warming could still be natural. Has anyone published a critique/criticism of this? It seems to me take this work is fundamentally flawed. First, variance scaling is crap statistics as it produces results with far less explained variance than normal least-squares regression. Second, the paper seems to have no independent validation. Third, what happens if one just takes his low-frequency (numbered in his Fig. 1) points and calculates the area average? Surely this will have much greater variability than the full global mean? (If no-one has done this please let me know -- I can do it very easily myself.) But perhaps his scaling method circumvents this "problem"? (2) What is the paper of Caspar's (with Doug Nychka) that shows that McIntyre is wrong? Are there other papers I should see/cite in this regard? (3) What are the papers that explain what is wrong with the von Storch ECHO simulation? I think Tim Osborn did something on this. Many thanks for your help, Tom. -- Michael E. Mann Professor Director, Earth System Science Center (ESSC) Department of Meteorology Phone: (814) 863-4075 503 Walker Building FAX: (814) 865-3663 The Pennsylvania State University email: [1]mann@psu.edu University Park, PA 16802-5013 website: [2]http://www.meteo.psu.edu/~mann/Mann/index.html "Dire Predictions" book site: [3]http://www.essc.psu.edu/essc_web/news/DirePredictions/index.html Attachment Converted: "c:\eudora\attach\MRWA-JClimate05.pdf" Attachment Converted: "c:\eudora\attach\62811_0_merged_1109271201.pdf" Attachment Converted: "c:\eudora\attach\osbornetalClimDynInPress06.pdf" References Visible links 1. mailto:mann@psu.edu 2. http://www.meteo.psu.edu/~mann/Mann/index.html 3. http://www.essc.psu.edu/essc_web/news/DirePredictions/index.html Hidden links: 4. http://www.met.psu.edu/dept/faculty/mann.htm