Exotic adventures in global data to unfind “the Pause”, by Karl in 2015

UPDATED: Ross McKitrick’s PDF file has some minor changes.

The Pause has been unfound, not with new data, but with new adjustments in one odd dataset.

The awkward “Pause” in global temperatures shows up in every major dataset. It’s the reality that conflicts with nearly every major climate model. But it’s there in the Hadley records of land surface and ocean, it shows up in the Hadley sea surface measurements, it’s there in NCDC, GISS, and of course in the satellite data of RSS, and UAH, and it shows up in the best data we have on the ocean, the ARGO buoys. It’s quite the challenge to unfind it!

 

To find global warming in the last 15 years, we need to ignore all that and use sea surface data blended from boats randomly trekking through shipping lanes with buckets and from ocean buoys (and that’s not ARGO buoys). But even that isn’t enough, that original data needs to be adjusted, and where sea ice gets in the way, gap-filled from sparse land data (as you would right?). Then we need to accept a lower-than-usual significance test, and carefully cherry pick the time periods to blend the past rapid warming with past cooling, so we can say we’ve found the holy grail, a quasi significant slight “warming trend” in data adjusted with a wildly uncertain correction estimate. And Professor Matthew England likes this kind of science.

Ross McKitrick points out that to get the new NOAA sea surface data they added 0.12 °C to the buoy readings, to make them more like the ship data. That magic number came from Kennedy et al. (2011) where  the uncertainty was reported as (wait for it) 0.12 ± 1.7°C. (Which is like saying there is definitely one apple here, give or take 17 apples. So this is what 95% certainty looks like?). Worse, that uber-extremely-uncertain-number was supposed to be used to adjust the ship data down so it was closer to the buoys.  The authors felt the buoys were more accurate than bucket-from-ships. Even Karl et al paradoxically agree (have cake, eat cake), saying that because the buoy data is better, it should be weighted higher. In this fashion, the best data can get adjusted the most, to make it more like the bad data, then it can count for more. This is Double-Good-Science!  Thanks to Ross McKitricks comments on Karl et al 2015 for details (1K PDF).

Here’s how this dubious chain of uncertainty and assumptions plays out in the media:

Professor Matthew England, Chief Investigator at the ARC Centre of Excellence for Climate System Science at UNSW, said the findings were no surprise.

“There’s nothing all that new in this paper and nothing that surprises me,” Professor England said.

“The bottom line is that multiple data sets and multiple lines of evidence have shown that global warming hasn’t stalled at all’”

Who is the denier then? The person who recognises the obvious results in all the major data sets, or the person who is “not surprised” when a contrived adjusted concoction of not-good data produces a result he expected? (Well really, who at this late stage is really surprised by the recent try-on?)

This is so desperate even other fans of the global warming theory are not buying it. They have some standards. They know the global temperature pause is real:

Dr Peter Stott, Head of Climate Monitoring and Attribution at the Britain’s Met Office Hadley Centre said…

“The slowdown hasn’t gone away, however,” and  “The results of this study still show the warming trend over the past 15 years has been slower than previous 15 year periods.”

Prof Tim Osborn, Professor of Climate Science at the University of East Anglia, said he would caution against dismissing the slowdown in surface warming on the basis of this study, nor to downplay the role of natural decadal variability for short-term trends in climate.

“There are other datasets that still support a slowdown over some recent period of time, and there are intriguing geographical patterns such as cooling in large parts of the Pacific Ocean that were used to support explanations for the warming slowdown,” Professor Osborn said…  (h.t to Waxing Gibberish).

[Reuters] “Some other experts said however the idea of a hiatus was still valid, since warming had probably slowed this century if compared to fast rates in the 1980s and 1990s.

“It is curious that a comparison with these decades was not included in this new study,” said Richard Allan, a professor of climate science at the University of Reading. (h/t Barry Woods)

In the endless quest to make reality fit the models, the adjustments to data have become their own parody. What is “evidence” anymore? Evidence schmevidence.

Seven steps to unfind the Pause?

Anthony Watts, Bob Tisdale, Judith Curry,  Pat Michaels, Richard Lindzen, Paul C. Knappenberger, and the GWPF and Fred Singer have already taken it apart.

@NOAA ‘s desperate new paper: Is there no global warming ‘hiatus’ after all? –  WUWT

Reports Of The Death Of The Global Warming Pause Are Greatly Exaggerated – GWPF

1. Use a weak test: p < 0.1, not the usual significance. Bit desperate.
2. Use assumptions and apply large adjustments to sea surface data. Don’t use the best dataset for sea-surface temps (ARGO). Ignore the satellites.
3. Create Arctic sea surface temperatures by extending data from land measurements. The ocean there is covered in ice. There aren’t many land measurements to go from. What could possibly go wrong?
4. Don’t mention the eighties or nineties. The fastest recent global warming occurred in the 1980’s and1990’s. Obviously the 2000’s are red-light bad news for the alarm-us camp, because that was when CO2 emissions increased dramatically but the warming slowed. The worst possible thing is to compare those decadal trends to the previous ones.
5. Cherry pick the time frames! Karl et al carefully choose a long trend — all the way back to the 1950s —  in order to find a weak long term warming trend that the recent decade can outdo. Back in the 50s and 60s, the world was cooling, so wrapping in and averaging the long cooling and then warming cycle, they can come up with a small warming trend number.
6. To find warming in the sea surface during the pause, it helps to adjust the late 90’s sea surface temps down and the recent measurements up, thus increasing the trend in the last 15 years, but not affecting the trend across the whole period. Check.
7. Ignore contradictions like: why can’t we find a hot spot? If the surface warmed more than we thought, the upper troposphere should have warmed even faster. This means the missing hot spot is more missing than before.

Anthony Watts and Bob Tisdale have some eye watering graphs on temperature adjustments. Get a load of this:

 

This is a fairly standard form of adjustment — cool the past, warm the present. Smooth the bumps.

From Anthony Watts we can see how USA thermometers were all biased to warm until WWII, then biased too cold. (Luckily someone at NOAA knew what the thermometers of 1915 should have read, and “fixed” them decades later. Scientists were pretty stupid in the early 20th Century, eh? Silly old Einstein.)

The desperate lengths of the Team to find some recent warming says a lot about how important the pause is.

If you can’t explain the pause, you can’t explain the cause.

(h.t  Hockeyschtick. I’m shamelessly copying their motto.)

REFERENCE

T.R. Karl; A. Arguez; B. Huang; J.H. Lawrimore; M.J. Menne; T.C. Peterson; R.S. Vose; H.-M. Zhang  (2015) “Possible artifacts of data biases in the recent global surface warming hiatus,” by at National Oceanographic and Atmospheric Administration (NOAA) in Asheville, NC; J.R. McMahon at LMI in McLean, VA.