Monday, 28 April 2008

The Mysterious Vanishing Ice Age

On 23 April 2008, The Australian published an article, "Sorry to ruin the fun, but an ice age cometh" by Phil Chapman, geophysicist and astronautical engineer, who was also the first Australian to be sent into space by NASA.

Possibly he's still out there, because he tells us:
All four agencies that track Earth’s temperature (the Hadley Climate Research Unit in Britain, the NASA Goddard Institute for Space Studies in New York, the Christy group at the University of Alabama, and Remote Sensing Systems Inc in California) report that it cooled by about 0.7C in 2007. This is the fastest temperature change in the instrumental record and it puts us back where we were in 1930. If the temperature does not soon recover, we will have to conclude that global warming is over.

I haven't tracked down all four of the data sets he mentions but two should be sufficient since he claims they are all consistent.

Let's look at Hadley: HadCRUT3 is their most commonly quoted data set. What does it show (each year is represented in two rows: the first has the year, 12 monthly values and ends with an annual value, the second gives percent coverage for each monthly measure)? The global average for 2006 was 0.422° (measured as an "anomaly" versus the average over the years 1961-90). The number for 2007 was ... wait for it ... 0.402° -- a drop of 0.02°. Only a factor of 3.5 out.

What of NASA's GISS data? Their data is slightly different. Their anomalies are relative to the average over 1951 to 1980, and their data gathering and analysis also results in slightly different numbers. But this is good: if the rival data sets are consistent, it adds confidence to the credibility of the data. The numbers we want are in the column "AnnMean J-D". For 2006: 0.54° and for 2006: 0.57° for a difference of 0.03° -- in this case an increase rather than a decrease.

Whether we look at either of these figures separately or look at them together, the overall conclusion is the same: 2007 was much the same on average as 2006. If you look at NASA's graph of temperatures, you will see that these variations are well within the error bars (green vertical lines).

So where does the 0.7°C drop come from? I have to guess since he doesn't supply his working. The GISS data has a drop of about this order between the monthly average for January 2007 (0.86°) and the monthly average for January 2008 (0.12°). HadCRUT3 shows a similar but smaller effect between the Decembers of the two years (a drop from 0.536° to 0.201°). However, short term swings are common. One on this scale is unusual, but it is not uncommon for a comparable month to differ by a lot more from its predecessor than yearlong averages.

A fair number of people posted comments on the original article pointing the error out. So what did The Australian do? They made the version with comments inconspicuous (no longer linked from the main opinion pages) but kept a version without comments in a more conspicuous location.

Then, to drive the point home, the inimitable (fortunately) Christopher Pearson, the same one who incorrectly quoted the Pope as being in the anti-global warming camp, followed up with an article on 26 April 2008 "A cool idea to warm to" repeating the error. Pearson quotes Keynes: "when the facts change, I change my mind."

How about a couple more quotes for you, Chris?

Senator Daniel Patrick Moynihan:
You are entitled to your opinion. But you are not entitled to your own facts.
Or how about Richard Feynman:
Reality must take precedence over public relations, for nature cannot be fooled.

How long will we have to wait for another mea culpa? Three weeks like last time? Or is it OK to be loose with the facts unless you offend the Catholic Church?

Follow-up: The Australian on 29 April 2008 published another article, "Warming trend has not been reversed" by David Karoly, professor in the University of Melbourne School of Earth Sciences, correcting the many errors in the original article. Not bad: only 6 days after the original. The Pope will feel slighted.

More data: Karoly refers to the World Data Centre for Solar Terrestrial Physics at the National Geophysical Data Centre in Boulder, Colorado as the main source on sunspot numbers; if you want to check his 2008 data, here it is.


Anonymous said...

This one is doing the rounds in the deniosphere, of course.

I think you made an error: surely you meant Jan 07 to Jan 08? After all, some parts of the world have just emerged from a cold winter, following a mild one last year. This was ripe for cherry-picking by the usual suspects.

Jan 07 - Jan 08
HADCRUT3 0.632 - 0.056 = 0.576
GISTEMP 0.86 - 0.12 = 0.74

Looks like GISTEMP, showing the greater drop, was chosen.

To claim that this shows a year-on-year drop of this magnitude is either gross incompetence or simple bare-faced lying.

Philip Machanick said...

truesceptic: thanks, corrected.

The Australian today has letters on Karoly's response including one claiming his sunspot numbers were per month, not per day. Look up the link I gave ("More data"). You will see that Karoly is correct. It doesn't surprise me too much that some people don't understand how natural variability could mask a trend of 0.02 to 0.03 degrees per year -- but the inability to comprehend what a simple average means? You'd think that someone telling a highly regarded professor they're wrong would at least bone up on their high school arithmetic.

I have a question for the self-proclaimed AGW sceptics: why do you not vigorously correct errors in attacks on the "consensus"? It is you who have the most to lose from being associated with buffoons.