16 November 2013

Data massage of the highest order

This is a rebuttal to:

I thought Al Gore knew a thing or two about massage, but these guys are good. This is how to make a 15 year warming hiatus disappear: massage the data.

Using satellite data to fill in those gaps, the study found that global surface temperatures have been warming 2 1/2 times faster over the past 16 years than previously believed, calling into dispute the widely-reported global warming "pause,"

And this is data massage of the highest order. Kevin Cowtan and Robert G. Way have explained that the missing warming is in places we can't measure it, specifically at the earth's poles where there are no thermometers!

Here's the diagram:
Temperature data from the U.K. Met Office (thin lines) compared to a global reconstruction (thick lines).

The thick black line is the new fudge factor with the computer generated "reconstruction" from climate scientists Cowtan and Way.
The weird thing is that the missing heat is so selective, being present at the end of the 15 year warming hiatus around the year 2013 but not at the beginning of it around the year 1998. The thick (massaged) line is pretty much neck and neck with the thin HadCrut original unadjusted line except near the years 1998 and 2013. Strangely the deviation is downward near 1998 but upward near 2013.
The magical selective temp adjustment of Cowtan and Way.
How can such adjustments be justified other than to say it fits the global warming narrative, specifically of rising CO2 during that time? I guess the pay-walled paper spends a lot of time justifying such adjustments.
Now I rebut a couple of other points on the Wunderground article by Terrell Johnson:
First, the HadCrut data is described as unadjusted for the purposes of this paper, but it  has been through a series of adjustments that are not uncontroversial. So, the HadCrut data is not really raw as characterised on the abstract for Cowtan and Way.

Second, Weather Underground's Dr Jeff Masters said:

"You shouldn’t just focus on periods of 10 or 15 years when you look at climate change," he added. "Climate is defined as a period of 30 years or longer, and we expect to see ups and downs in 10- or 15-year periods. So we shouldn’t pay too much attention to slowdowns or speed-ups during these relatively short periods of time."
Well, two can play at that game. Here's the CO2 rise from 1960s to today:

 And here's t
he global average temperature according to HadCrut from 1940's:

OK, so according to Masters 10 - 15 years is insignificant while 30 years is  significant. For the period from 1940 to 1979 I see no correlation between temperature and CO2, and that's longer than 30 years. For the period from 1998 to 2013 I also see no correlation between temp and rising CO2. Then from 1979 to 1998 there's a mild correlation. These latter two periods are 19 years and 15 years, so neither are long enough to be considered significant according to Masters.

So...there's as much or more lack of correlation than correlation, so that means we don't know with any confidence whether the period of warming from 1979 to 1998 was caused by rising CO2 or whether it was a coincidence.

But now we have this new study to convince us that selective warming that fits the global warming narrative is happening in regions we can't even measure with thermometers! Man this global warming movement is getting desperate!

Update: oops fixed second last paragraph that was incoherent drivel. Sorry about that! No more drunk posting!


Update 19 Oct 2013:

This warming reconstruction has already been debunked by actual observation which shows no warming trend in the Arctic:


No comments:

Post a Comment