While running across another one of those “global warming is a scam” posts this morning, I came across a most interesting claim in the that post’s comments.
“JcT” writes:
This is the real reason for “Global Warming” Notice how the dates match up perfectly!
And “J.L. Ohlinger” responds:
Come on now JcT… that would be down right… well… logical…
The page to which JcT links contains a picture and a caption:
Explanation: Every eleven years, our Sun goes through a solar cycle. A complete solar cycle has now been imaged by the sun-orbiting SOHO spacecraft, celebrating the 12th anniversary of its launch yesterday. A solar cycle is caused by the changing magnetic field of the Sun, and varies from solar maximum, when sunspot, coronal mass ejection, and flare phenomena are most frequent, to solar minimum, when such activity is relatively infrequent. Solar minimums occurred in 1996 and 2007, while the last solar maximum occurred in 2001. This picture is composed of a SOHO image of the Sun in extreme ultraviolet light for each year of the last solar cycle, with images picked to illustrate the relative activity of the Sun.
Does the sunspot cycle explain global warming? Is it really true that the “dates match up perfectly”?
Let’s look at global temperature data, direct measurements of which are available through NOAA for the completed years of 1880-2009. Let’s take a closer look at 1996-2007, the years in the SOHO sunspot image that JcT says “match up perfectly” with global warming variation. Here’s the pattern in global temperature:
1996 is the 18th warmest year on record
1997 is the 11th warmest year on record
1998 is the 4th warmest year on record
1999 is the 16th warmest year on record
2000 is the 15th warmest year on record
2001 is the 9th warmest year on record
2002 is the 5th warmest year on record
2003 is the 6th warmest year on record
2004 is the 8th warmest year on record
2005 is the 1st warmest year on record
2006 is the 7th warmest year on record
2007 is the 3rd warmest year on record
That’s not a primarily cyclical pattern like the 11 year recurring sunspot pattern. It’s a pattern in which the overwhelming tendency is to warmth, no matter where the sun sits in its sunspot cycle. Each one of the years from 1996-2007 is one of the 20 warmest years on record. 2008 is #10 and 2009 is #2, in case you’re curious, and not one of the warmest 20 years comes before 1980.
If you mined the data for not just this but previous sunspot cycles, you might find some interesting secondary effect of sunspot cycles on temperature. But sunspot cycles just can’t explain why the warmest years are all recent, or why not one of the coldest 20 years on record comes before 1920.
Another interpretation might be that annual sunspot numbers (source) are somehow correlated with temperature. But run a multiple regression on the global temperature data and you’ll find that global temperature data better fit a simple pattern of increasing as years progress (p value < 0.001) than they do by sunspot numbers (p value 0.53). Annual sunspot counts only predict about 4% of the variation in global temperature from 1880-2009. A simple linear yearly trend predicts 75% of the variation in global temperature from 1880-2009.
Over at Free Republic, there was great chatter in December of 2009 about how with the low sunspot activity it should be really, really cold now. Then it turned out that the winter of 2009-2010 was the second warmest Winter on record. Spring 2010 was the warmest Spring on record. That thread went quiet as global temperature data rolled in. The sunspot pattern just isn’t working out.
