Comments on: A Mixed Effects Perspective on MMH10

By: Weekend Roundup – August 14, 2010 « The Policy Lass

Weekend Roundup – August 14, 2010 « The Policy Lass — Sat, 14 Aug 2010 18:32:02 +0000

[...] at my favorite climate skeptic blog, CA, Steve McIntyre has several posts about his recent co-authored paper reviewing Santer 08′s analysis of the climate [...]

By: Christy actualiza la disparidad modelos – realidad « PlazaMoyua.org

Christy actualiza la disparidad modelos – realidad « PlazaMoyua.org — Sat, 14 Aug 2010 09:45:55 +0000

[...] Más McIntyre: http://climateaudit.org/2010/08/11/within-group-and-between-group-variance/ [...]

By: Climate Blog and News Recap: 2010 08 13 « The Whiteboard

Climate Blog and News Recap: 2010 08 13 « The Whiteboard — Fri, 13 Aug 2010 15:41:41 +0000

[...] McIntyre expands on the paper. [...]

By: Steve McIntyre

Steve McIntyre — Fri, 13 Aug 2010 15:15:21 +0000

there was a mistake in the collation of trends that I used in this post -which was (confusingly) supposed to help clarify things rather than confuse things. I used trends calculated up to 2099 instead of 2009, which ended up making the within-group standard deviations too narrow. I’ll rework this and re-post. However, I’m going to Italy next Tuesday and am working on a presentation so the re-post will have to wait. This has also been a distraction from the MMH issues so let’s focus on them.

So let’s stick to MMH10. Sorry about that.

By: Ron Broberg

Ron Broberg — Fri, 13 Aug 2010 14:51:05 +0000

Looking closer to the code I think I've got it and summarize below. ----------- MMH10 looked at mean and variance of the trend of the multi-model-means (inter-model or across-models) and compared that with the mean and variance of various observational data sets. Chad and I are looking at the mean and variance of the trends of individual models and comparing that with the mean and variance of various observational data sets. McIntyre in this post is looking at the mean and variance between the trends of multiple runs of a single model (intra-model mean or within-group mean). This shows that the modeled trends output are relatively insensitive to the perturbations of the intial conditions used in the CMIP4 runs. ----- Is that a fair summary?

By: Bernie

Bernie — Fri, 13 Aug 2010 14:44:11 +0000

Given the presenting policy issue of CO2 emissions and the agreed upon nature of CO2 as a GHG, why wouldn’t the null hypothesis be whatever the temperature trend is without any feedbacks, positive or negative, since that is what the models are leargely designed to incorporate?

By: Gekko

Gekko — Fri, 13 Aug 2010 14:30:44 +0000

I think it is more nuanced than that. confidence intervals overlap, but in 1 case of the 4 (RSS/UAH LT and MT), the confidence limit does not encompass zero (indicating significant positive trend) while the others do not.

By: Ruth

Ruth — Fri, 13 Aug 2010 14:10:22 +0000

I am not clear why ECHAM5 has a mean trend of 0.446 in your graph (and R script), and 0.227 in Table 1 in the paper. The other series agree much more closely (though not exactly). Can you explain why the means are not identical?

By: Nick Stokes

Nick Stokes — Fri, 13 Aug 2010 13:55:04 +0000

Re: TAC (Aug 13 08:26), According to MMH10, UAH and RSS also have significantly different trends.

By: Bernie

Bernie — Fri, 13 Aug 2010 13:40:51 +0000

Steve:
Have you addressed the issues around the ensemble of models elsewhere? If so can you point to a link? I don’t want to distract from the substance of this post.