Sir Ivor,
The relationship between levels of co2 and temperature seem to be at
the crux of the problem. I have seen one graph, and one graph only, so
I am not sure how much reliance to put on it, that showed the
exponential relationship. The curve has rolled over so far that a
doubling of the co2 would be hardly detectable in temperature change.
From memory, it required something like a 1000 times increase in co2
to get less than one degree rise.
I wish I could remember where I saw it.
However it might explain the statement that all of Australia's saving
of co2 emissions will result in a temperature .0004 deg C lower than
it otherwise would be.

We have a much more serious problem than AGW ahead of us.
Unless we can stop the government from hiding our energy crisis from
us we will plough headlong into a liquid fuels shortage.
Not even the oil companies closing refineries seems to have woken them
up from their, "Don't talk to me about that" attitude.

When they even go to the trouble of suppressing reports from their
own organisations you know we have a problem.
Now there is a real worry !

cohenite, you astound me more with each post you make. Your knowledge of the scientific publication process is woeful. The manuscript version is what gets submitted. It is long because it is double spaced and all the figures are on separate pages. The final printed version will be much shorter. GRL is a peer reviewed journal. http://www.agu.org/journals/gl/submissions.shtml The designation “Letters” in the title indicates papers need to be short, not that they are correspondence.

I think you need to go back and read your posts. This one in particular http://forum.onlineopinion.com.au/thread.asp?article=14089#243852

So Figure 3 is now an addendum? When did that happen? The manuscript you linked to on ArXiv has Figure 3 in the body of the paper. It presence is even indicated in the last line of abstract. Are you reading the same Stockwell and Cox manuscript I am reading? It certainly appears not.

I picked on Figure 3 because it is complete rubbish and is a clear example of the rubbish in the rest of the paper. Do you agree with me that the figure is complete rubbish?

I know the difference between a Chow test and a linear regression, do you? Do you know the Type 1 errors that are likely to arise in the use of a Chow test and how to manage for them? Has it not occurred to you that my scepticism about the use of the Chow test in this circumstance is because there is no a prior reason to use such a test on a data set like this?

And you astound me when you say this:

"my scepticism about the use of the Chow test in this circumstance is because there is no a prior reason to use such a test on a data set like this".

In doing so you dismiss the bulk of the field of meteorology that studies DISCRETE changes in atmospheric regimes, ENSO transitions between El Nino and La Nina, PDO, drought and flood and other discrete climate changes. You also dismiss the oceanography concerned with the effects of sudden changes in ocean currents, such as the shutdown of the Gulf Stream.

You also reject large parts of econometrics that have shown the folly of assuming that all the parameters of a model are always constant. The Chow test relaxes that assumption by allowing some of the parameters to change their value. In a statistical model of a part of a complex system (which a fit to surface temperature is), it’s a risky assumption that all the parameters remain constant.

Heat going alternately into the upper 700m of the ocean, then into the lower 700-2000m could produce the kind of alternation of trend around a general rising trend that Fig 3 models. The Stockwell paper attempts to model such a process.

How is that rubbish Agronomist?

You mention Type 1 errors Agronomist. The Chow Test was employed in Stockwell to ascertain the validity of breaks in data to a confidence level of 95%. In a linear trend there can be breaks contrary to the trend within the overall linear trend; those breaks are not breaks in the linear trend because before and after them the linear trend continues.

If a break is determined by the Chow to the 95% CL it means the trends before and after were different and not linear. A linear model cannot show this.

The point can be illustrated by a typical linear model, the running mean. A series of numbers: 11, 12, 13, 14, 15, 16, 15, 14, 13, 12, 11; is a sequence of values. The first 6 have an average of 13.5; the 2nd 6, an average of 14.16; the 3rd 6 14.5; 4th 6, 14.5; the trend is increasing but the values are declining so every trend after the 1st is misleading.

The Chow would pick the break in trend at 16 and isolate the 2 different trends before and after the break.

What Type 1 errors has the Chow Test caused to “arise” Agronomist? Are you saying the effect of AGW is linear?

Agronomist; I am waiting for your response to my reasonable comments; the general gist and tenor of your's to date have been both patronising and ill-informed. I had assumed you had some scientific capacity to justify your condescension, which I don't object to, as long as it is accompanied by some insight.

It seems I will have to put you into the same basket as the other time wasters and trolls on the issue of AGW.

Some people say "look at me, look at me, look at me" ... others just have a real life and live it.