Sunday, August 07, 2016

Eli Does the Sou Thing

With Tom Karl's retirement the folk over at the Watt's Shop are busy with the time machine, coming up with the idea that Tom Peterson is going to be the follow on.  Peterson, of course retired about a year ago himself.  Zeke pointed this out to only moderate grumbling and Tony Willard (cause he is a backwards kind of guy) did some strikethrough action.  Bunnies can read all about it at Sou's

Now some, ok Eli to be sure, have the lack of sense to go back to the original, and indeed great fun to be had, among which Roger Sr. still grumbling that TomP really didn't respect him in detail.  Roger, of course, runs the forgettery about his dissing all and sundry, but the treasure is the last comment from Pat Michaels.  To be more to the point the last line in the last comment, but just to help, here is the whole thing which has some interesting implications:
pat michaels August 4, 2016 at 5:40 pm
Hey, Tom, I’m pretty sure you’re reading these comments.  Some are a bit ad-hom. But I have important questions for you. 
You were the science chief for the first (2000) National Assessment. I found that the two models used, the CCM2 and the Met Office one, resulted in a larger residual error AFTER applied to 10-year running means of the global temperature average than what was in the raw data. In other words, they added noise to signal. 
You wrote back that indeed I was correct But you went forward. This is EXACTLY like a physician prescribing a treatment that he or she knows will cause more harm than good. It’s unethical, pure and simple. What you did was noticed and created a great distrust of anything out of NCDC. 
Why did you do it? Was it worth it? 
Pat M. 
I saved your job in 2000. You were on a hit list and I had you taken off because I thought you were a straight shooter. Seven months later what is detailed above happened.
Hmm. Hit list, 2000. . . got Bob Watson but appears to have missed Jim Hansen too, so this is probably something that pat had a hand in putting together in his dream of getting a real job, but who has the Emails?

An imaginative bunny might even think about who or what was on the drafting team.  The link to the incoming Bushies was almost certainly through the Marshall Institute and ran through Dick Cheney via Star Wars links.  Fred Seitz and Jastrow were still alive but if Pat Michaels was involved, then it is not a bridge too far to believe that a few august members of the American Association of State Climatologists of which Pat Michaels was an unusual member were also involved.  Perhaps Eli would enter the name of R A Pielke Sr.   Be good bunnies and drop the guy a line, he don't listen to Eli anymore ont he Tweets.  Other names that occur are Roger Srs. buddies George Taylor and David Legates.

20 comments:

Fernando Leanme said...

How much does the job pay?

Anonymous said...

So is Pat Michaels openly admitting that there was some kind of hit list to try and get climate scientists fired? Are there any more details?

Anonymous said...

I am not a statistician, but:

You were the science chief for the first (2000) National Assessment. I found that the two models used, the CCM2 and the Met Office one, resulted in a larger residual error AFTER applied to 10-year running means of the global temperature average than what was in the raw data. In other words, they added noise to signal.

Why wouldn't the difference between HadCM2 model runs and the 10-year running mean of global temperatures produce greater residuals than as seen in the difference between model runs and the raw temperature data? A 10-year running smooth removes all natural variations and can also produce strange artifacts at the very end of the smooth (especially here as 1998 was a whopper of an El Niño).

EliRabett said...

Well yes, this is an admission that the Denial Nomenklatura had a hit list, and they probably still have an updated one that they are passing on to Lamarrrrr Smith and Senator Snowball.

More to the point, the Michaels has let the cat out of the bag about their cabal, the existence of which will be very useful to reference when anybunny starts about Климат ворота, which clearly was run out of Russia.

Start tweeting folks

Anonymous said...

What possible role or position did Michaels have in 2000 that he could have 'saved' Tom Karl's job, and 'had him taken off a hit list'?

Is this just self-aggrandizing BS or is Michaels inadvertently revealing some true corruption and politicization of climate science, but on the side of the contrarians who are always crying wolf?

THE CLIMATE WARS said...

The link to the incoming Bushies was almost certainly through the Marshall Institute and ran through Dick Cheney via Star Wars links.

A very imaginative bunny . IMHO 'twas the demise of the evil empire that did in Marshall's orginal Star Wars incarnation- O'Keefe's arrival from the American Petroleum Institute turned it into just another a post-cold war PR shop. I recall Fred being kind of old by then- he was born in 1911.

EliRabett said...

Michaels was (and is a nothing) but he had two levers. First was as the Virginia State Climatologist (not really, start here then go here ) he was linked to state climatologists in the US, and that meant to folk like RPSr, so he had allies. Second he had a consulting firm (which employed Chip Knappenberger) New Hope Environmental Services, which provided testimony and beards on payment for a bunch of coal companies amongst other things. Thire, there were always links to CATO and those types. Right now about all he has left is CATO AFAEK. Eli will leave mag the fun of googling

caerbannog said...


Start tweeting folks...

Done.

THE CLIMATE WARS said...

AFAIK Michaels & his K-Street pals invented State Climatologists.

Though vital resume fodder for the op-ed wars, the invented ofice is about as much an existential necessity for ships of state as State Flowers , Reptiles or Aromatherapists

Unknown said...

Regarding Pat Michaels' charge against Tom Karl, Pat made up a test for the models suggesting that to be useful they had to be able to predict the running 5-year average variability through the 20th century; he found that they did not have any skill at doing this so coined the clever phrase that climate models are no more than random number generators. Given the test he chose which has nothing really to do with climate (defined by NOAA as a 30-year average), the models did exactly as expected in that the 5-year variability is dominated by ENSO (with a bit of volcanic response thrown in) and these are not predictable influences. For climate change, per the detection-attribution analyses that are done, one wants to use the slow but accumulating forcing of GHGs and other long-term forcing and look at the multi-decade response. Those of us involved in the meteorological aspects of the National Assessment wrote a peer-reviewed BAMS article to explain all of this (MacCracken, M. C., E. Barron, D. Easterling, B. Felzer, and T. Karl, 2003: Climate change scenarios for the U. S. National Assessment, Bulletin of the American Meteorological Society, 84, 1711-1723). That Pat still fails to understand this is pretty pathetic. Tom Karl has done a very commendable job and deserves our gratitude. Mike MacCracken

Anonymous said...

Crickets.

opit said...

Heck Rab. Where is the info showing that temperature readings for scattered locations on land lead to statistically useful averages on a water ( especially oceans )planet with storms and clouds : i.e. change of state of water mechanisms as well as convection mechanics, thermal inversions, thermoclines in the oceans, ocean currents, and on and on. I have yet to be convinced there is such a thing as a meaningful global mean temperature to be ascertained from land surface readings of some precision only in the last few decades.Everybody talks about cycles such as el Nino ( PDO ) being important, but it does not seem to shake the faith in alleged 'consensus' ( nor help to define exactly what such a consensus might be about ) Way too much Poisoning the Well Argumentation going on for easy dissing the Strawman Meme. The skepticism part is easy to define for those who care to listen : there is no way to show that the alarmist turkeys have any credibility. Lack of proof of futurism's warnings is inherent in the traditional header : Speculation.

EliRabett said...

Opit: Sea water temperature has been measured from ships and floats (look up Argo) as long as there have been surface stations and maybe longer thanks to navies and commercial sailing ships.

Second what is measured are anomalies, changes in temperature at locations on both land and sea.

Your ignorance is primarily your problem but we do have a small responsibility to educate you. Unfortunately we also expect that you are not exactly going to listen.

opit said...

What have also been measured by the Royal Navy is ocean currents. Can you educate me on how one can get a variation of less than a degree per century down to hundredths of a degree for variation in a global temperature for periods when no such accuracy was available ? How about ocean temperatures past the thermocline as they contribute to a baseline temperature which is also not available ? That's not a matter of not listening - but of paying attention to what is being ignored.

EliRabett said...

Same way you get high resolution from low resolution A/D converters, with the technique of oversampling and decimation used to increase the precision of a analog to digital converter

http://www.atmel.com/images/doc8003.pdf

Kevin O'Neill had a fine example of how this works

"As a metrologist, I'm surprised anyone would use metrology as an argument against averaging. One must take a series of readings and average if only to know the short-term repeatability to calculate uncertainties. And of course anyone with half a brain, metrologist or not, quickly understands that averaging adds precision.

Perhaps the mental stumbling block is that averaging readings from one device adds precision - not accuracy, but averaging multiple devices adds both precision and accuracy.

I once performed a simple experiment where I showed co-workers that I could get more accurate results from twenty-five 6 1/2 digit voltmeters than from one 8 1/2 digit voltmeter - even though the 8 1/2 digit voltmeter has a presumed accuracy 50 times better than the 6 1/2 digit voltmeters. I did 'cheat' a little by using statistical bootstrapping to increase the effective sample size from 25 to 1000. I would have to go back and find the final results, but the reduction in error was approximately from 85 ppm for a single 6 1/2 digit voltmeter to low single digit ppm error after bootstrapping."


Your problem is indeed the things that you know for sure and are wrong.

guthrie said...

Tamino had a good post on the question of how you got more accurate results from many measurements but I can't find it.

Bernard J. said...

What Michael MacCracken said:

"Given the test he chose which has nothing really to do with climate (defined by NOAA as a 30-year average), the models did exactly as expected in that the 5-year variability is dominated by ENSO (with a bit of volcanic response thrown in) and these are not predictable influences. For climate change, per the detection-attribution analyses that are done, one wants to use the slow but accumulating forcing of GHGs and other long-term forcing and look at the multi-decade response."

It seems to me that Pat Michaels is either incompetent in his job, or knew that he was using inappropriate metrics and therefore knowingly misbehaved professionally.

And the two alternatives are not mutually exclusive...

If anyone has a justification for Michaels' "test" I'd be interested to hear it.

Bernard J. said...

Opit opined:

"I have yet to be convinced there is such a thing as a meaningful global mean temperature to be ascertained from land surface readings of some precision only in the last few decades."

Others have already smacked you about for your lack of understanding of metrology and statistical accuracy (due to random measurement variation), but there's another point that bears emphasising, and that is that where there are consistent systematic biases in measurement, these biases have little to no effect in monitoring changes to the dependent variable in a time series.

If one's boat is sinking and one is measuring the depth of the water to the top of one's seat rather than to the gunwale, one will have a bias in the estimation of the depth of the encroaching water. One is still able, though, to accurately monitor the fact that the depth of water in one's boat is increasing, and that one's boat is therefore sinking.

It's time for you to stop pissing in humanity's boat from your rock of ignorance, and start helping with the bailing out.

opit said...

Whoa ! My ignorance of meteorology is what is causing me to misunderstand that averaging of ignorance adds accuracy ? Wherefore this then ? https://insideclimatenews.org/news/20120507/television-meteorologists-climate-change-skeptics-weather-global-warming-john-coleman-james-span-joseph-daleo Now if you wanted to say bias was reduced I would get that. But we are now talking data nebulous as quarks used to posit the dangers of further change when it is questionable that we have even reached the baseline accuracy of single entry readings.

Kevin O'Neill said...

opit writes:"Whoa ! My ignorance of meteorology is what is causing me to misunderstand that averaging of ignorance adds accuracy ....."

Meteorology? Try looking up 'metrology' -- a metrologist is not a weatherman.

As for 'averaging of ignorance adds accuracy' I'll just let that pass. Why bother?