lol, I’ll give you that. But if that’s true then we can all jump on them for not getting internal variability right.

I just read your comment about the 75% weighting of trend in the peninsula. It did have an error but by the same method when corrected I get about 50% weighting. Nic made a nice post by adjusting trend but I don’t think that gives us a true weighting either because the regression can (and of course does) compensate using other station information.

When I ran the B matrix version of the calculation I found about a 50% weighting of the infilling of all stations based on the peninsula information. This doesn’t give us the reconstruction final weight because the satellite PC’s need to be separated out. I think what I took away from it all was that the peninsula weight was overstated in all reconstructions so far but not as badly as my first calculations showed.

Nic also made a good point that with regard to trend, most of the negative weighted stations subtract trend from the total whereas my initial calculations assumed the stations were positive contributors. A negative station with similar information to other stations would remove trend from the reconstruction.

Any way I think about it, the true weighting is difficult to express since the weighting changes as more or less data is available in the individual rows (months) of the matrix.

If you are wondering how over weighted the peninsula is, you can see it visually in the difference between area weighted and EM plots at this post.

http://noconsensus.wordpress.c.....#more-4448

One thing I hope to try this week is to re-weight the stations in Ryan’s EM algorithm according to area to insure that high concentrations of surface measurements (as in the peninsula) don’t over weight information in the reconstruction.

Maybe not, but at least now if things do stay cooler, they can say it doesn’t change things, just a pause, and point out how they’ve already posted about it.

Just for clarity, can we tell with certainty that RC is:

a) predicting a pause in Global Warming? or
b) projecting a pause in Global Warming? or
c) something else?

I ask seriously because as a Denier, I can only interpret this as a small defeat for the AGW Movement. The books are still cooking. Why turn down the heat? It’s not like we are currently experiencing a mass exodus from the AGW ranks. Or are we?

Andrew

Hadcrut3 global temps peaked at +0.364C in 1878 (versus last month’s Hadcrut3 at +0.400C). HadSST2 peaked at +0.264C in 1878 (versus last month at +0.500C).

So, there are cycles up and cycles down which is more clear with a longer-term perspective [and don't forget, they have been playing around with these historical indices (SSTs and land surface measurements) and past temperatures were almost certainly higher than is being recorded in the databases now.]

If there’s a lag, I sure don’t see any pattern. Something isn’t right.

It depends whose OHC data you refer to:
http://i44.tinypic.com/5uizit.png

The graph is from my post on the data presented by Levitus et al:
http://bobtisdale.blogspot.com.....et-al.html

Regards

But, unlike the HADSST data, the 2003 and 2005 peaks in the OI.v2 version are higher than June 2009.

An SST anomaly map:
http://i26.tinypic.com/23tl3xg.jpg

The two biggest contributors to the increase last month were the North Pacific with a rise of 0.24 deg C and the North Atlantic with a rise of 0.19 deg C.

From my monthly update for June 2009:
http://bobtisdale.blogspot.com.....pdate.html

On this graph, I only showed the line as a connect the dots. Adam pointed out we may be near comparable points in terms of the MEI index, so this might be a rough guide to comparing this SST level to previous ones. (Or not.)