Monday, October 22, 2018

Thoughts on Climate Change -- part 8: A Tale of Two Graphs


I will set before you two different graphs representing the relation between CO2 levels and global temperatures: number 1:

Global temperature vs. CO2 concentration


[added 12-22-18]: Since the graph presented above tells us little about 21st century temperatures, I've decided to add this one, which represents the so-called "hiatus" from 1998 through 2015 much more clearly. (The spike we see for 2016 is due to an especially strong El Nino and does not reflect long-term temperature trends):


number 2:



(from a blog post titled  Does CO2 correlate with temperature? by one Robert Grumbine.)

When I "eyeball" the first graph it seems evident, as I've contended in the past, that there is NO correlation between CO2 levels and global temps. during the entire 100 year period between 1880 and 1980. Nor do I see any sign of correlation between  1998 and the present. I DO see a correlation between ca. 1980 and 1998, but that represents only 20 years out of the last 138.

Graph number 2, a scattergram in which CO2 levels are plotted directly against temperatures, presents a radically different picture, where the two appear to be very strongly correlated. How is this possible?

(According to Grumbine "It's awfully hard to look at this and say that there's no correlation between CO2 and temperature.")

I scratched my head over this for some time before I realized that the two graphs represent more or less the same relationship, only presented in very different ways. Only when we attach dates to that second graph does it become possible to see that both are essentially the same, only with the time scale distorted in the second. After all, while graph no. 2 is a scattergram and graph no. 1 is not, essentially the same set of temperature data is represented in the vertical axis of both. For example, the year 1940, when temperatures peaked after a long increase, corresponds roughly with a CO2 level of 310 parts per million, as represented in graph no. 2, which shows temperature peaking at the same point; the year 1980, when temperatures began to rise dramatically, corresponds roughly with a CO2 level of 335 ppm. where the temperature begins to rise in graph no. 2 as well.

No sooner do we begin to recognize the relationship between the two graphs does it become apparent that there is in fact NO correlation in graph no. 2 until it gets close to the 335 level, around 1980. Similarly, we see no correlation after 370 ppm, the level reached around 1998, the beginning of the well-known "hiatus," as represented in graph no. 1.

The reason graph no. 2 appears to depict a correlation is due to the time distortion produced by the fact that CO2 levels shot up so rapidly from ca. 1980 to ca. 1998. [Correction: the previous phrase should read: "temperatures shot up so rapidly from ca. 1980 to ca. 1998."] Thus, unlike graph no. 1, which presents a more or less accurate picture of climate history since 1880, graph no. 2 distorts that history to emphasize the relatively brief 20 year period when both CO2 levels and temperatures were increasing at the same time.


When both graphs are examined critically it becomes clear that there is no evidence whatsoever of a correlation between CO2 levels and global temperatures, aside from the brief 20 year period at the end of the previous century, which all but rules out the possibility of a causal relationship, thus placing the entire human-caused "climate change" meme in serious doubt. 

Anyone following the analysis presented above should better understand why I'm so skeptical when it comes to the excessive reliance on statistical methodologies when attempting to evaluate scientific evidence. As a wise man once said, "If you torture the data long enough it will confess to anything."

 
Add to Technorati Favorites