Here's yet another nail in the coffin of CO2 climate change culpability. Unfortunately, our political masters seem intent on hamstringing the developed world's economies while letting China, India, Brazil and other economies off the hook at the upcoming Bali round of climate talks.
The Climate Faithful regularly question the credentials of those presenting research contrary to the so called 'consensus' that human created CO2 is the primary source of the current warming trend. What these people miss is that climate science is fundamentally a mathematical/ statistical analysis of climate data. Therefore, qualifications in those disciplines are mandatory in order to determine whether claims of correlation can be supported.
If school curricula were up to me then besides more rigorously testing reading, writing and arithmetic I'd include fundamental statistics and economics. With an understanding of those two latter disciplines people would be less easily fooled by trumped up claims presented as science such as CO2 as the cause of global warming or be sucked into the socialist ideas promoted by environmental groups, universities and sections of the mainstream media.
One often hears the phrase "correlation is not causation". To many people, when they see a graph in which two variables seem to track together they assume a relationship. It's a natural instinct and has been the subject of many studies. However, the graph could be the result of a deliberate falsification of the data. Statistical analysis is the method used to identify this falsification and it is an extremely important part of the validation process. Imagine where we'd be if we couldn't determine whether claims of mineral finds or financial market performance could not be verified.
At this point, it's worth reminding people what correlation means and what an r-squared statistical test achieves.
Correlation: In probability theory and statistics, correlation, also called correlation coefficient, indicates the strength and direction of a linear relationship between two random variables. In general statistical usage, correlation or co-relation refers to the departure of two variables from independence. In this broad sense there are several coefficients, measuring the degree of correlation, adapted to the nature of data.
Note that I'll use R2 to represent r-squared, as I don't have a superscript font available.
R-squared: In statistics, the coefficient of determination R2 is the proportion of variability in a data set that is accounted for by a statistical model...R2 is a statistic that will give some information about the goodness of fit of a model. In regression, the R2 coefficient of determination is a statistical measure of how well the regression line approximates the real data points. An R2 of 1.0 indicates that the regression line perfectly fits the data.
In summary, an R2 of 1 indicates 100% correlation and 0% indicates no correlation.
If I stopped now and took a survey of what the general population predicted the correlation between CO2 and temperature was over the 20th century then what would be the most popular range? I suspect it would be somewhere between 80% and 95%.
The following paper is from retired meteorologist Joseph D'Aleo, whose bio reads:
JOSEPH S. D'ALEO has 30 years experience in professional meteorology. He has BS and MS degrees in meteorology from the University of Wisconsin and did doctoral studies in meteorology at New York University. He taught meteorology at the college level for over eight years and was a cofounder and the first director of meteorology at the cable TV Weather Channel, a position he held for seven years. He joined Weather Services International (WSI) in 1989, where he was a marketing manager and chief meteorologist. Currently, he is senior editor (also known as "Dr. Dewpoint") for WSI's Intellicast Web site. D'Aleo is a Certified Consultant Meteorologist and was elected a Fellow of the American Meteorological Society. He has authored, presented, and published numerous papers focused on advanced applications enabled by new technologies and how research into ENSO and other atmospheric and oceanic phenomena has made possible skillful seasonal forecasts.
When I talk about the low correlation between CO2 and the 20th century temperature record it's because of analysis such as this.
US Temperatures and Climate Factors since 1895
By Joseph D’Aleo, CCM
The now familiar plot of the US climate network since 1895 shows a cyclical pattern with a rise from 1895 to a peak near 1930 and decline into the 1970s and then another rise with an apparent peak around 2000. Note the minor warming from the peak in 1930 to the peak in 2000.
The short term fluctuations are driven by factors such as ENSO and volcanic eruptions. The longer term cycles are mainly driven by cycles in the sun and oceans although changes in the last half century have been increasingly blamed on anthropogenic factors.
Let’s look at the three factors mentioned and how well they correlate with the observed temperatures.
USHCN AND CARBON DIOXIDE
I first took the CDIAC annual mean carbon dioxide estimates since 1895 and plotted that against the annual USHCN. For a correlation, I did a 11 year smoothing to eliminate any effects of the 11 year solar cycle and to make it consistent with the other correlations. I got an R2 of 0.29. Without that smoothing the correlation was 0.149.
USHCN AND SOLAR
The sun influences the climate in direct and indirect ways. A more active sun is a brighter slightly hotter sun and when the sun is hotter the earth is a little hotter. This small effect is magnified by other more indirect solar influences. When the sun is more active although its brightness (mainly visible light) only increases by 0.1%, the ultraviolet radiation increases by 6-8% and the even shorter wavelengths by a factor of two or more. These UV rays create and destroy ozone in the high atmosphere, both of which are exothermic effects and produce heat. Work by Labitzke and Shindell at NASA GISS have shown this to be important.
Shindell showed how this factor may have been responsible for the little ice age. When the sun is more active there are more flares and eruptive activity that causes rapid increases in the solar winds, causing ionization storms in the earth’s atmosphere with resultant heating. Also importantly an active sun causes the earth’s magnetic shield to diffuse more cosmic rays from reaching into our atmosphere. Since these rays have a low water cloud formation enhancing effect (recently confirmed in the laboratory), an active sun usually means less low clouds and thus warmer temperatures. In all these cases, a more active sun brings warming.
Scafetta and West (2007) have suggested that the total solar irradiance (TSI) is a good proxy for the total solar effect which may be responsible for at least 50% of the warming since 1900.
I took the TSI from Hoyt Schatten and compared to the USHCN data (smoothing the data for 11 years to eliminate the 11 year solar cycle. I found a correlation strength (R2) of 0.64 lagging the temperatures 3 years after the solar. (Wigley and others have suggested a lag up to 5 years may be appropriate).
USHCN AND OCEAN MULITDECADAL CYCLES
We know both the Pacific and Atlantic undergo multidecadal cycles the order of 50 to 70 years. In the Pacific this cycle is called the Pacific Decadal Oscillation. A warm Pacific )positive PDO Index) as we found from 1922 to 1947 and again 1977 to 1997 is usually accompanied by more El Ninos while a cool Pacific more La Ninas (in both cases a frequency difference of close to a factor of 2). Since El Ninos have been shown to lead to global warming and La Ninas global cooling, this may have an affect on annual mean temperature trends in North America.
A similar mulitidecadal cycle exists in the Atlantic known as the Atlantic Multidecadal Oscillation (AMO). When the Atlantic is in its warm mode there tends to be more tropical activity and on average above normal temperatures on an annual basis across the northern hemispheric continents.
Since the warm modes of the PDO and AMO both favor warming and their cold modes cooling, I though the sum of the two may provide a useful index of ocean induced warming for the hemisphere (and US). I standardized the two data bases and summed them and correlated with the USHCN data, again using a 11 point smoothing as with the CO2 and TSI.
This was the jackpot correlation with the highest value of R2 (0.86!!!).
Note this data set started in 1905 because the PDO and AMO was only available from 1900.
THE LAST DECADE
Last week we showed how global satellite temperatures were uncorrelated with the monthly CO2 levels over the last decade (R2 of just 0.07).
For USHCN and the CDIAC annual data, the correlation is even worse mover this time frame (0.05).
USHCN temperatures show a cyclical behavior over the past 112 years with peak warming about 1930 and 2000. The temperature trends correlate with a number of factors. We examined them here. We found the correlation strengths to be as follows
Clearly the US annul temperatures over the last century have correlated far better with cycles in the sun and oceans than carbon dioxide. Whatsmore, this correlation with carbon dioxide seems to be weakening further in the last decade.
Given the recent cooling of the Pacific and Atlantic and rapid decline in solar activity, we might anticipate given these correlations, temperatures to respond downwards shortly.
So there we have it. The correlation between CO2 and temperature over the last decade of the US temperature record is a completely irrelevant 5% while the 1895-2006 correlation is just 29%.
Naturally, the Climate Faithful will point to the whole world, not just the US, in order to support their claims. One wonders how it is possible for the US to respond differently to CO2 production than the rest of the world.
As I've pointed out previously - the fixation, for political purposes only, on CO2 as the cause of climate change will have severe negative consequences going forward. Firstly, policy makers will become much less trusting of any science purporting to support a change in public policy. Secondly, there are clearly issues with land-clearing that affect climate and we're spending too little time studying this problem. Finally, it's the developed world that has the economic capacity to assist poorer nations in the world up the quality of life ladder. By restricting these economies through a carbon tax it impacts our ability to help those at the bottom end of the ladder.
But isn't that the way of all green policies?