The only thing we can be sure of about the future is that it will be absolutely fantastic. So if what I say now seems to you to be very reasonable, then I have failed completely. Only if what I tell you appears absolutely unbelievable, have we any chance of visualizing the future as it really will happen.

- Arthur C. Clarke (h/t Brin)

Sunday, January 2, 2011

Garbled Reasoning at WUWT

Stop the presses, huh?

The WUWT folks have been making hay out of some completely garbled nonsense as usual. They habitually (at best) don't understand what they are looking at and proceed to interpret it to support their peculiar certainty that modern CO2 greenhouse effect somehow doesn't work, so we can spew as much of it as we like, and of course that was always guaranteed by liberty or something.

Anyway there's a particularly striking example up on an article by Don Easterbrook.

First, he shows a graph by Richard Alley, no less:



And then somehow jumps to:

So where do the 1934/1998/2010 warm years rank in the long-term list of warm years? Of the past 10,500 years, 9,100 were warmer than 1934/1998/2010. Thus, regardless of which year ( 1934, 1998, or 2010) turns out to be the warmest of the past century, that year will rank number 9,099 in the long-term list.


Well, this is quite silly. You aren't exactly comparing apples to oranges, but you are comparing apples to applesauce. 1998 and 2010 are contenders for warmest year in the instrumental record. As such, they are probably the warmest yearss in the last 2000 years or so, but certainly not in the past 8000 years. Mainstream climatology ha[[ily stipulates that the period from 8000-6000 years ago was warmer than today, but not by much.

Mentioning 1934 (a contender for warmest year in the USA) is just silly beyond belief.

But the global mean is about 15 C, and the graph is centered around -31 C, so what accounts for the 50 C (90 F) difference? Well, the graph refers to a signal called the delta-O18 temperature proxy. It is a regional temperature for the region of Greenland. What's more, this proxy is effectively a multidecadal running average, not an instantaneous temperature. And finally, the "present day" is geologists' jargon for 1950. The last 60 years don't show up in ice cores or in the record as shown.

So, to recapitulate:

2010 is the hottest year on record or close to it globally

2010 or 1998 is likely the hottest in thousands of years globally

6000 to 8000 years ago was a bit warmer than today. We will probably cross that threshold soon enough but we haven't yet.

By the end of the century, unless we get a grip we will be seeing the warmest years in the last tens of millions of years, but we are not there yet.

The graph measures multidecadal running averages regionally up to 1950 and cannot be directly compared to annual global averages.

It's nonsense.

I like it better than the recent nonsense in Forbes magazine because it only makes one uninformed point rather than a dozen. But it's noise. Noise, noise noise.

It is still pretend science and not real science. Somehow the political sector needs to learn to tell the difference. A tall order, but crucial.

Update: Gareth has more at Hot Topic.

27 comments:

Michael Tobis said...

Here's another stupid one:

Lack of Historical Pewrspective by Bishop Hill.

Identification of the lack of perspective is left as an exercise.

Lou Grinzo said...

I'm sure everyone here knows the saying from the legal profession: "If the law is on your side, argue the law. If precedent is on your side, argue precedent. If neither is on your side, pound the table and yell." It seems that since pounding a table and yelling don't work in cyberspace, WUWT and similar outlets take it to the next step -- just make up stuff.

WUWT makes my brain flinch.

Steve Bloom said...

It looks to me as if some labels have been added to that graph, so it's no longer exactly Alley's. Also, it seems to say that it ends at 95 years BP, not 1950, unless I'm reading it wrong.

Steve Bloom said...

On looking again I see that it identifies the present as 2000 AD, so if I'm right the record ends in 1905. The amount of warming at the end seems consistent with that. In any case, given that it's WUWT and Easterbrook, it's probably not good to be making any assumptions about the graph without comparing it to the original.

Stephen said...

Note a couple of other things:

First, what I like to call the "Excel rule" - if someone post a graph that was clearly made using Excel default settings, they probably have no idea what they're doing.

Second, he used years as a categorical axis instead of a numerical one, which brings me back to rule one . . .

Neven said...

How about this one by the inimitable Goddard, closest thing to a Poe in the world of lala?

Steve Bloom said...

Well, here's the Alley paper. It's not his graph at all. Possibly there's a universe somewhere where someone like Don Easterbrook would render Alley's data honestly, but I suspect it's not this universe.

guthrie said...

The Easterbrook post is so bad that I, who does not have a PhD but have recently gotten another MSc (In an unrelated subject), think my lecturers would give him 51% in an essay. Maybe. Or possibly 48%.
The reason being that the Greenland graphs before the one you show have delta O18 on them, but no temperature comparison. They all end by the 1960's or so, and give no idea of change in temperature, but do give the impression that there't not a lot happening. He's lying to the reader, in other words. It is totally illegitimate to use these graphs in any sort of attempt to compare them to "today".

Anyone know what his PhD is in? Media studies?

Tom Curtis said...

Stephen, it is, so far as I can see, based on Alley's graph found in the supplementary material:

http://www.ncdc.noaa.gov/paleo/pubs/alley2000/alley2000.html

http://www.ncdc.noaa.gov/paleo/pubs/alley2000/alley2000.gif

Of course, the graph used only plots the temperature up to 10,000 years before 1950, while Alley only plots the temperature from 10,000 years to 16,000 years before 1950 in the published material of his paper.

Anna Haynes said...

Tangential topic - re "delta O18" (and its non-global nature), is this the same as the "oxygen isotopes from ocean cores" axis that the Smithsonian Institution Human Origins Exhibit uses in the muddled "Our Changing Climate" circa-15-foot-long graph that greets visitors coming in the back way?
(The scale goes from "1.5" ("warmer") up to "5.5" ("cooler"), & the graph starts 10 million years ago)

Kooiti MASUDA said...

The data are made public at the Paleoclimate division of the National Climatic Data Center of NOAA [The entry of Alley (2000)]. I think that using it is a better practice than copying a figure from a published paper and add annotations, if a proper time axis is used.

There is a problem in the documentation at NCDC: I do not find explanation of what "present" means in "before present". I know that there is a convention about Carbon 14 ages where "present" is 1950 CE, but I am not sure whether the same convention applies here. (I have not read the text of Alley 2000 yet.) I

If "present" is 1950, the newest data point "95 years BP" is the year 1855.

Though I have not searched, there must be other studies which examined O-18 of the shallower part of ice is examined in the shallower part of ice in the same area, though probably not of the same core sample. So, a person with appropriate knowledge can extend the record with the same variable.

There was a similar case. The paper Robinson, Robinson and Soon (2007) used by the Oregon Petition had a plot of data of Keigwin (1996) [entry at NCDC Paleoclimate]. Its major problem is that the original record is local but Robinson et al. pretended that it was representative of the global situation. It had also a minor problem that "present" was 1950 in the data at NCDC (though careful reading of the published paper is necessary to know it), Robinson et al. assumed it to be 2000.

Same Ordinary Fool said...

This is the seventh in a WUWT series preparing its followers for a GISS-hottest 2010.

#7 "2010-where does it fit in the warmest year list?" Dec 28
#6 "Do we care if 2010 is the Warmest year in History?" Dec 25
#5 "How Germany's weather team views the hottest year ever" Dec18
#4 "Arctic Oscillation spoiling NASA GISS party" Dec 15
#3 "Tisdale K.O.'s GISS latest 'warmest-year nonsense'" Dec 12
#2 "Hansen feels the need to explain why GISS is high in the midst of frigid air" Dec 12
#1 "Warmest year ever? 2010: an Unexceptional El Nino Year" Dec 6

Michael Tobis said...

Gareth of Hot Topic is working on a more detailed posting on this particular fiasco. He writes "I checked with Richard Alley yesterday: GISP2's present is 1950, and
therefore the [most recent] data point in the temp series is 1855."

Kooiti MASUDA said...

Excuse me for failure in editing out duplicated words ("is examined in the shallower part of ice") in the previous comment of
mine.

NCDC Paleoclimate has a collection of GISP2 data, and under "Stable Isotopes", there is "GISP2 Stable Isotopes (Oxygen)". In that document, it is clear that "present" is 1950, and the newest data point is -36.88 yr BP, i.e. 36.88 years after present.

Kooiti MASUDA said...

I find that the "temperature" record of Alley (2000) is not a simple linear transformation of GISP2 oxygen 18 record that I mentioned. I should check the paper by Cuffey and Clow (1997), cited by Alley (2000), about how the values were derived, but I am sorry I do not have time to dig more.

I think this kind of complication should be resolved by messages from experts. But I am afraid those people who doubt integrity of experts easily interpret this kind of apparent inconsistency as evidence of fabrication.

Kooiti MASUDA said...

Anna Haynes, I have made a remark at a Google Group yesterday (responding to Timothy Chase), but I write again with some modification.

O-18 in H2O of ice sheets is a proxy of regional temperature at interannual time scales.
It is difficult to specify the scale of "region",
as fractionation of isotopes occurs during both evaporation process
from the sea surface and condensation process in the clouds.

O-18 of calcium carbonate in deep sea sediments (also
O-18 of oxygen molecules in bubbles in ice cores) represent, in the
time scale of glacial-interglacial cycles, global displacement of mass of water between ice sheets and the ocean. (There are other factors local sea water temperature etc., but these are considered relatively minor at the time scales of glacial cycles.) So they are also used as proxies of the global sea level. When there are more extensive ice sheets, the sea level is lower, and the concentration of heavier isotopes in the sea water is higher.

But, if we look at variations in time scales shorter than
a millennium, the ocean cannot be considered uniform,
and the signals may be regional rather than global.

thingsbreak said...

Mainstream climatology ha[[ily stipulates that the period from 8000-6000 years ago was warmer than today, but not by much.
...
6000 to 8000 years ago was a bit warmer than today. We will probably cross that threshold soon enough but we haven't yet.

Perhaps I'm mistaken, but I was under the impression that this only holds true in summer and only in the northern hemisphere. If that's the point being made, some clarification might be helpful.

Michael Tobis said...

I seem to be out of date on this question, at least according to that page. I wish the NOAA page had some references. Anyway, thanks for the correction.

I'll patch the text tomorrow.

thingsbreak said...

@MT:

Don't take that link at face value, necessarily. It was the quickest I could find. I think there is more nuance to the position, in terms of a strong Holocene climate/thermal optimum outside of the NH extratropics. But IIRC, like the MWP/MCA, it was something that had been provisionally suspected to be a globally, temporally coherent period that turned out to be far more complicated.

Peter H said...

Does this mean the famous Petit et.al. 1999 graph only applies to the area around Antarctica?

Michael Tobis said...

In its fine details, it is a local measure. In its large scale structure, the agreement with other global evidence including Greenland ice cores over the first cycle indicates that the global temperature is represented.

Peter H said...

"In its fine details, it is a local measure. In its large scale structure, the agreement with other global evidence including Greenland ice cores over the first cycle indicates that the global temperature is represented. "

What I would guess. I agree with Easterbrook being up to no good, but doesn't something like the same comment apply to the Greenland graph? So that E. is dodgy but you can, um, draw global conclusions from the graph if not the ones he does but the ones Gareth does?

Michael Tobis said...

Well, it's reasonable to parse these as a global signal plus a regional signal. Since the large, slow excursions are very much in synchrony at opposite poles, you are asking to attribute a lot to coincidence.

Indeed, if you are leaning toward dismissing all of it, you have a triple coincidence to explain, because the CO2 signal is global.

Additionally, there are other proxies telling the same story.

Peter H said...

"...but you can, um, draw global conclusions from the graph if not the ones he does but the ones Gareth does?"

Doh.

but you can, um, draw global conclusions from the graph if not the ones he does and you can draw the conclusions Gareth does.

Peter H said...

"Well, it's reasonable to parse these as a global signal plus a regional signal. Since the large, slow excursions are very much in synchrony at opposite poles, you are asking to attribute a lot to coincidence.

Indeed, if you are leaning toward dismissing all of it, you have a triple coincidence to explain, because the CO2 signal is global.

Additionally, there are other proxies telling the same story."

Oh, I agree, I don't in any way dismiss it. It just struck me that the response of a thoughtful WUWT sceptic (humm) might be 'so the Petit graph is local?' and I wanted the answer. Thanks :)

NewYork said...

If the data ends in 1905 as this implies

ftp://ftp.ncdc.noaa.gov/pub/data/paleo/icecore/greenland/summit/gisp2/isotopes/gisp2_temp_accum_alley2000.txt

It seems we'd want to add roughly 2 C to the end of the graph. Anomaly of last decade compared to 1900-1909 average...

http://data.giss.nasa.gov/cgi-bin/gistemp/do_nmap.py?year_last=2010&month_last=11&sat=4&sst=1&type=anoms&mean_gen=0112&year1=2000&year2=2009&base1=1900&base2=1909&radius=1200&pol=reg

Yeah there's probably a better apples to apples comparision but it's a start.

What Easterbrook leaves out is indefensible.

thingsbreak said...

@PeterH:
Does this mean the famous Petit et.al. 1999 graph only applies to the area around Antarctica?

If you look at the temp swings, they're 9-12°C from glacial maxima to minima, so in a sense, yes. The global mean change is more like 5-6°C, and the Antarctic values reflect polar amplification.

If someone were to claim that Petit et al. 1999 demonstrated that the global mean change was ~12°C, he or she would be wrong.

Note the language in the abstract: "Atmospheric concentrations of carbon dioxide and methane correlate well with Antarctic air-temperature throughout the record."

And later: "The overall amplitude of the glacial–interglacial temperature change is ~8°C for DTI (inversion level) and ~12°C for DTS, the temperature at the surface (Fig. 3). Broad features of this record are thought to be of large geographical significance (Antarctica and part of the Southern Hemisphere), at least qualitatively.When examined in detail, however, the Vostok record may differ from coastal28 sites in East Antarctica and perhaps from West Antarctica as well."