Lookin' for Correlation

-- by Horatio Algeranon

(after a song made famous by Johnny Lee -- "Lookin' for Love" )

I've spent a lifetime lookin' for correlation
Error bars and downward trends, only aggravation
Playing a fools game, hopin' to win
Telling those sweet lies and losin' again.

I was lookin' for correlation in all the wrong places
Lookin' for correlation in too many cases
Searching the skies, lookin' for traces
Of what.. I'm dreamin' of...
Hopin' to find a trend, hand-in-glover
God bless the day I discover
An "R = 1" corre-la-tion.

When Global Warmin' was alone, no correlation in sight
I did everything I could to make it come out right
Don't know where it started or where it might end
I turn to a "t-test", just like a friend

I was lookin' for correlation in all the wrong places
Lookin' for correlation in too many cases
Searchin' the skies, lookin' for traces
Of what.. I'm dreamin' of...
Hopin' to find a trend, hand-in-glover
God bless the day I discover
An "R = 1" corre-la-tion

TSI** came a'knockin' at my back door..
TSI is everything I've been lookin' for..

No more lookin' for correlation in all the wrong places
Lookin' for correlation in too many cases
Searchin' the skies, lookin' for traces
Of what.. I'm dreamin' of...
Hopin' to find a trend, hand-in-glover
God bless the day I discover
An "R = 1" corre-la-tion

**IMPORTANT NOTE: Causrelations may also be produced
simply by replacing "TSI" above with other acronyms:
PDO, ENSO, AMO, etc.

"Correlation, (often measured as a correlation coefficient ["R"]) , indicates the strength and direction of a linear relationship between two random variables" (from Wikipedia) . Two things that are correlated tend to vary together: when one increases, the other either also increases (positive correlation) or decreases (negative correlation).

Some jump at an apparent correlation between recent global warming and anything but atmospheric CO2 increases. They are keen to entertain even the most outlandish claim supported by little or no evidence, provided it "disproves" the link between human-caused CO2 increases and recent climate change. Cosmic rays, TSI (total solar irradiance), climate on Mars, (phases of the moon? UFO's?) -- you name it.

Alas, as a very wise man once said: "If you look for something long and hard enough, you will find it." Whether it means anything is another matter entirely. (Hackneyed Cliche Warning!!!) If I let (an endless supply of) monkeys bang at a keyboard long (and hard) enough, they will eventually reproduce "Hamlet" word for word. Would that mean my monkeys were geniuses?

************

The Glossary of Statistical Terms defines a "trend" as

"A long-term movement in an ordered series, say a time series, which may be regarded, together with the oscillation and random component, as generating the observed values."

A trend is usually shown with a straight line (obtained with "linear regression") and is an overall indication of what is happening. For example, an upward "trend" in stock prices over several decades is shown with an upward sloping line that indicates that stock prices have gone up over the long term -- even though the price may fluctuate up or down from one year to the next (due to "noise"). Similarly, global average temperature may fluctuate (or remain flat) over the short term, but the trend over the past few decades (and even century) is upward as shown in this IPCC graphic:

The graph shows several different trend lines (with different slopes) for the different time spans. Global warming has been most pronounced (about 0.18C per decade) over the past few decades.

An "R = 1" correlation is a "perfect positive correlation" between two things ("variables", in the statistics lingo). It means that when one thing changes (eg, increases), another changes in perfect sync. Such perfect correlation never occurs with real data. Ever-present "noise" reduces the correlation coefficient "R" to a value less than 1.

A "t-test**" is a statistical test than can be used to determine the "significance" of a correlation.

**Special Note: There is also a much less well-known kind of "T"-test that tests whether a particular blog post (about "monkeys typing Shakespeare", for instance) is any good. If you get an email like this

It's a stupid meaningless thought experiment that I
don't even think is true. I know, it's easy to argue
the other side, but it's just a cliché that make
Tee want to puke.

from "T" (or "Tee") you know you have failed this second T-test. Alas, Horatio did...and did.

TSI = "Total solar irradiance describes the radiant energy
emitted by the sun over all wavelengths that falls each
second on 1 square meter outside the earth's atmosphere"
-- NOAA

Some claim that the global warming over the past few
decades is due to an increase in TSI, but as this graphic 
from the Max Planck Institute shows, TSI has remained
essentially flat over that time period, while the
global average temperature has gone up significantly.
(solar output in blue, global temperature in red).

************

For country music buffs (which Horatio ain't, since he originally thought the song was by Waylon Jennings -- though Waylon did sing it as well [or "too", at least]), according to Wikipedia,

"Lookin' for Love" is a song made famous by country music singer Johnny Lee. The song was part of the soundtrack to that year's movie, Urban Cowboy. This iconic "love song" was written by two school teachers, and was actually written about a classroom of second grade children.

The Global Warming "Skeptic" (Part II)

--by Horatio Algeranon

The warming "skeptic" is an incurious creature,
"Denial" its most prominent feature.
It changes spots and stories too,
But you won't find it at the zoo.

Its "logic" is implacable,
Its "argument" most laughable.
It has a nervous tic(k),
That it can't seem to kick,

Which makes it say the dumbest things:
"Global Warming's a hoax!" and "Pigs have wings!"
It doesn't have a single clue,
But criticizes those who do.

It signs its posts "The Warming Skeptic,"
Quick, go and get the antiseptic,
And pour it in my open eyes,
That I may cleanse them of the lies.



Read The Global Warming "Skeptic" (Part I).

Clima(c)tic Feedback

--by Horatio Algeranon

Hendrix is to decibel-level,
As water vapor is to temp
One makes minds explode,
The other takes no hemp.

The fancy term is "feedback,"
Though some call it "distortion,"
But no matter what you call it,
It increases the proportion.

When CO2 goes up,
It warms the earth somewhat,
But warming boosts the vapor,
So temp goes up a lot!


For a wee bit more technical description of the effects of feedback on the earth's climate, read this excellent description by the "Master of Climate Clarity" -- "MCC Hammer" (aka, "Hansen's Bulldog" [aka "Tamino"] ).

Causrelations?

-- by Horatio Algeranon

Some claim correlation and imply causation,
Between air temperatures and TSI,
PDO + AMO,
Cosmic rays and ... my oh my.

This "skeptic" mumbo-jumbo,
Is making my head spin,
It's got little to do with science,
So the scientists just can't win.

According to NOAA "Total solar irradiance [TSI] describes the radiant energy emitted by the sun over all wavelengths that falls each second on 1 square meter outside the earth's atmosphere".

Some claim that the global warming over the past few decades is due to an increase in TSI, but as this graphic from the Max Planck Institute shows, TSI has remained essentially flat over that time period, while the global average temperature has gone up significantly (solar output in blue, global temperature in red).

According to Wikipedia

"The Pacific Decadal Oscillation (PDO) is a pattern of Pacific climate variability that shifts phases on at least inter-decadal time scale, usually about 20 to 30 years. The PDO is detected as warm or cool surface waters in the Pacific Ocean, north of 20° N. During a "warm", or "positive", phase, the west Pacific becomes cool and part of the eastern ocean warms; during a "cool" or "negative" phase, the opposite pattern occurs."

Atlantic Multidecadal Oscillation (AMO) is a similar pattern that manifests itself in sea surface temperature changes in the North Atlantic Ocean.

Some have implied that such phenomena are more likely than CO2 to have been responsible for the global temperature increase in recent decades, but the evidence for such claims is weak (to say the least) as explained in this post by Hansen's Bulldog.

It is a common error to believe and/or imply that "if two things are correlated statistically one must be the 'cause' of the other."

But, simply put, "correlation does not imply causation" -- ie, a correlation between two things does not mean that one of the two "caused" the other to happen.

First, even "strong" or "high" statistical correlation may not mean much (if anything), as show in this post by Hansen's Bulldog.

Second, even if the correlation is "real" (not just an analytical artifact), it might simply mean that each of the two things in question has been caused (or just "influenced") by a third thing -- and, possibly, that the two things (eg, "sea surface temperatures in the Pacific and US "Lower 48" surface air temperatures) have influenced one another through "teleconnection"**).

**According to the American Meteorological Society, "teleconnection" means

A linkage between weather changes occurring in widely separated regions of the globe.

A significant positive or negative correlation in the fluctuations of a field at widely separated points. Most commonly applied to variability on monthly and longer timescales, the name refers to the fact that such correlations suggest that information is propagating between the distant points through the atmosphere.