Friday, March 29, 2013

The Australian: How to Lie With(out) Statistics

This "haitus" business is really getting ridiculous.

Graham Lloyd at The Australian has now put it at 20 years:
DEBATE about the reality of a two-decade pause in global warming and what it means has made its way from the sceptical fringe to the mainstream.
This is flat-out wrong -- warming over the last 20 years has been about 0.3°C.
  • Over the last 20 years, GISS shows a warming trend of 0.15 ± 0.02 C/decade, or 0.30°C for two decades.
  • HadCRUT4 has it at 0.136 ± 0.014 C/decade, or 0.27°C for two decades.
Both are very statistically significantly different from zero.

Can anyone stretch this lie out to 25 years?

P.S.: Yesterday HadCRUT4 just reported their surface anomaly for February: the 9th-warmest February in 165 years. Some haitus. Even their 15-yr trend is 0.05 ± 0.02 C/decade, and rising fast as the 1997-98 El Nino falls out the back-end of the 15-yr interval, and it's for suckers.

8 comments:

gallopingcamel said...

There is no doubt that global temperatures have risen since 1850. It was very smart of the IPCC to pick a start date that served their political agenda.

In recent years Hansen made "Adjustments" to the GISS records to maintain his meme of the "Hottest Year On Record". Fortunately, some of us have the unadjusted GISS v2 temperature records which say otherwise.

As the period in question (1988 to 2013) lies entirely within the satellite record there is a sanity check available:
http://www.drroyspencer.com/2012/09/uah-global-temperature-update-for-august-2012-0-34-deg-c/

UAH says no statistically significant warming since 1998. Are they liars too?

David Appell said...

UAH says no statistically significant warming since 1998.

Wrong.
The linear trend for UAH LT since Jan 1998 is 0.06 +/- 0.03 C/decade (1-sigma), which is statistically signficiant warming at a 95% confidence level.

David Appell said...

Besides, Peter, even if the UAH LT data showed zero trend for 15 years, there are plenty of reasons why the signal of AGW would not rise above the noise in the climate system over such a relatively short period, and forcings such as aerosols and black carbons that may be causing cooling.

There is also the possibility that the UAH and RSS data have developed a cool bias and are underestimating the actual trend.

gallopingcamel said...

"There is also the possibility that the UAH and RSS data have developed a cool bias and are underestimating the actual trend."

That is indeed a possibility.

In contrast it is a certainty that the recent retrospective NASA GISS adjustments (2008 vs. 2012) have a warming bias. Something that prompted Richard Lindzen to note in February 2012:
"We may not be able to predict the future, but in climate ‘science,’ we also can’t predict the past."

gallopingcamel said...

David,
It turns out you were right about one thing. N&K's 154.3 K temperature for the Moon is their own calculation.

I had assumed that they used the observed value from the Diviner LRE.

Are you sticking with your 212 K equatorial average?

David Appell said...

Are you sticking with your 212 K equatorial average?

Of course -- that's what the theory shows, and it agrees with observations.

David Appell said...

Why is it a "certainty" that GISS has a warming bias?

David Appell said...

Peter: Thanks for recognizing N&K's true value for the lunar temperature.

Does this make you rethink their theory?

Calculating the temperature of body with no atmosphere is about as simple as it gets....