Sunday, June 16, 2013

A Record for the Pause?

In what may be a record for the claimed length of The Pause in global surface temperatures, a columnist in South Africa named David Gleason writes,
Between 1995 and last year carbon dioxide levels in the atmosphere rose 10%, but global temperatures not at all.
This is complete nonsense. HadCRUT4 shows the globe warming 0.17°C (0.31°F) since January 1995, a number that is very statistically significant(*).

How that equals zero, I have no idea.

By the way, the David Rose Hole still exists, barely. There is still a 9-month period from 5/1997 to 1/1998 where there is warming-til-today, but whose statistical significance falls below 95% -- all the way to 89%. (*Again, no autocorrelation.)

(*) This conclusion doesn't includes autocorrelation. Slope uncertainties that take autocorrelation into account are nearly impossible to calculate using Excel, if you want to calculate them for a range of dates, because you need to create an array of the residuals for each interval. And I just haven't gotten around to writing a Python or R script to do it. Using the SkS trend calculator, I find that the HadCRUT4 trend since Jan-1995 is 0.095 ± 0.109 °C/decade (2σ), which is not quite statistically signficant warming at the 95% confidence level, though it is at the 92% level -- the large uncertainty simply showing that it's usually difficult to make statistically signficant conclusions about climatologically short time intervals (as I wrote about in the Sidebar here).

No comments: