longer-than-usual loss of statistical significance at the 2?/95% probability level (around 18-20 years, depending on the dataset, compared to the 10-15 years required for statistical significance most of the time since 1980

Since 1974 the average duration of years in GISS that achieve statistically significant warming was 14-15 years. Two 19 year periods (1977-95 inclusive and 1979-97 inclusive) each had just significant warming and two 18 years periods (1978-95 inclusive and 1980-97 inclusive) each had just significant warming. The commonest duration was 14 years.

So 19 years is not an unprecedented duration needed for significance. I’d say we’d have to wait until it becomes 20 years before we could say, yes, that is pretty unusual.

https://dl.dropboxusercontent.com/u/2007207/Screen%20Shot%202014-06-23%20at%204.04.14%20PM.png

I’d like to see it, but I don’t think it exists.

Seriously, this has been a somewhat longer than usual “pause”/”deceleration” in the surface temperature warming trend than any we’ve seen since the epic pause circa ~1944-1978. It’s the resultant longer-than-usual loss of statistical significance at the 2?/95% probability level (around 18-20 years, depending on the dataset, compared to the 10-15 years required for statistical significance most of the time since 1980) that’s created a climate confusion opportunity for those so inclined and a teaching moment for those who know better about the constantly shifting balance sheet of natural and anthropogenic forcings, the complexities of short-term internal climate variability and how we need to look everywhere – air, oceans, cryopshere – to figure out if the planet is warming consistently with AGW theory and current estimate of climate sensitivity.