Amid all the brouhaha about Ted Cruz’s insistence that the globe isn’t warming, based on his using satellite data for the lower troposphere (not Earth’s surface), insisting that it’s “the best we’ve got” (it isn’t), and ignoring absolutely all the other evidence (which isn’t just powerful, it’s overwhelming), we haven’t paid enough attention to the fact that Cruz, and Monckton, and Christy, and Spencer, and Curry are cherry picking. Not just a little — they are cherry cherry cherry picking.
They all start their “no warming” plots with the giant el Niño in 1997-1998. They pick that time because it shows what they want it to show — which is the textbook definition of cherry-picking.
What does cherry-picking do for you? It enables you to distort the real trend by picking out a trend you like, whether it’s really there or not.
Allow me to demonstrate.
Let’s use the RSS data for TLT (lower-troposphere temperature) — the one Ted Cruz and the rest of the cherry gang like to use. We’ll ignore the fact that it doesn’t agree with balloon–borne thermometers, instead it shows a drift. We’ll ignore the fact that it’s not measuring conditions at Earth’s surface. We’ll ignore the fact that it doesn’t actually measure temperature. We’ll ignore the fact that there are more adjustments to the satellite data than to the surface data. We’ll just take it at face value.
It shows this:
This is the graph Cruz and the cherry gang don’t show, because it includes the entire time span. They’d rather you didn’t see the whole time span — that would interfere with their cherry-picking.
Notice that if you plot a straight line through the entire time span (which I’ve done), last year’s value is dead on the money. That kind of throws a monkey wrench into the idea that it stopped warming. In fact none of the values is outside the two-standard-deviations limits above and below the linear trend line — except 1998. The monster el Niño. Where the cherry gang likes to start.
Here are the residuals from the full linear fit:
If it stopped warming, those residuals can’t be following a flat line. But they sure look like they do. Even so, we should test that idea, to see whether or not we can find evidence that it stopped warming.
So … let’s cherry-pick.
We’ll start with 1998. We’ll fit a straight line to the residuals from the full-time-span straight line fit, to see whether or not the slope isn’t flat — if it stopped warming, the post-1998 slope of residuals can’t be flat, it has to be less than the whole slope. We get this:
The slope is downward! A statistical test shows a t-value of -2.3!! That’s statistically significant at 95% confidence!!! OMG!!!!!!!
But wait … there’s more.
If you allow yourself to cherry-pick, then the “usual” t-value doesn’t follow the same statistics. It can get a lot more extreme, for no reason other than randomness, because there are so many possible choices to start your cherry-pick. But how extreme can it get? Can it get to -2.3?
We can answer that question with a little game called “Monte Carlo.” It’s a workhorse of statistical analysis: you use a random-number generator to create test series with no change, and test to see how much change you can find by allowing yourself to cherry-pick. I insisted that the cherry-pick had to leave at least seven years’ data before and after, and I ran 1,000 of those Monte Carlo simulations to get a good handle on how extreme the t-value can get just from random fluctuation, with no “stopped warming.” I even looked only for extreme downward slopes — after all, that’s what the cherry gang wants.
Here’s how the observed cherry-picked t-value (in red) fits in with the Monte Carlo distribution of cherry-picked t-values:
Rigorous statistics shows clearly, that it’s ridiculously easy to get such extreme values when you allow yourself to cherry-pick. Done right, the statistics does not confirm “stopped warming” at the 95% confidence level. It doesn’t even make the 90% confidence level.
The only reason that Cruz, and Monckton, and Christy, and Spencer, and Curry are able to point at satellite data and say “no warming” is that they cherry-pick.
We should definitely call them what they are. That includes two things: “denier” and “the cherry gang.”
Cherry cherry cherry. LOL!
If you like what you see, feel free to donate at Peaseblossom’s Closet.