Previous post about the “climate myth” that the “IPCC is alarmist” was about an actual statement from Dr. Roy Spencer that was contorted into something that was not recognizable as his statement anymore. In that post, I made the remark that fabricating arguments of the other side seems some sort of a habit of the “skeptical”science team. I have seen them doing the same thing several times before.
This week I bumped into yet another example in which the skepticalscience team, who are clearly alarmist, made up the “arguments” of the other side. That example is called the “escalator” and can be found in the right sidebar of their website.
For those who didn’t see the graph before, it is a moving gif depicting two scenarios. The first scenario depicts how the, ahem, “contrarians” see the graph with global surface temperatures since 1970:
The moving gif slowly iterates through every blue trendline and at the end the blue lines disappear. Then these are replaced by one single red line showing how, ahem, “realists” see the same graph:
Basically, those “contrarians” are so short-sighted that they see a series of cooling trends, while not recognizing the overall upward trend. Skepticalscience explain it as those contrarians don’t know the “difference between short-term noise and long-term signal” and “inappropriately cherrypick short time periods that show a cooling trend”. This:
simply because the endpoints are carefully chosen and the trend is dominated by short-term noise in the data.
At that time, I was already rolling over the floor laughing…
Those endpoints are indeed “carefully chosen”, let there be no doubt about that! But not by those “contrarians”. Most of those blue trendlines didn’t make any sense when it comes to their claims. Those blue lines were clearly chosen by the skepticalscience team themselves for maximum visual effect.
That is easy to see. First consider the source of this data:
The data (green) are NASA GISS monthly global surface temperature anomaly data from January 1970 through December 2014
Which “contrarian” in his right mind would use NASA GISS data to describe the warming of the last fourty years?!?!? Aren’t they the ones who doubt the validity of this dataset, because of the many adjustments made, the non-climatic biases, the uneven spatial coverage and so on? As far as I know, they prefer the satellite record because it is actually global, includes the whole area in which the atmosphere would warm, is less sensitive for non-climatic interference like UHI and so on. This satellite data basically shows a graph which is rather flat over the last two decades.
If they would use a surface dataset, then the NASA GISS dataset would probably the very last one to use…
Secondly, there are so many of those blue lines all over length of the graph (my emphasis):
with linear trends for the short time periods Jan 1970-Oct 1977, Apr 1977-Dec 1986, Apr 1987-Oct 1996, Aug 1997-Dec 2002, Jan 2003-Jun 2012, and Jul 2012-Feb 2014 (blue), and also showing the far more reliable linear trend for the full time period (red).
That are six cooling trends over a 45 year period. Which “contrarians” make such weird choices?
The only trendline that made sense was the first one. It goes from 1970 to 1977, which was at the end of the new ice age scare. A cooling trend would make some sense there.
The three following trend lines are rather strange. There are way too many of them. The period of say 1975 to 2000 is considered an upward trend, why even use three steps during that period for what is seen as one upward movement? And wasn’t the end of the 1990s considered by those “contrarians” as the beginning of something called … the “pause”? So how does the trend line from 1997 → 2002 aligns with that notion?
The last two blue trend lines (from 2003 → 2012 and from 2012 → 2014): again too many of them. Why not use just one trendline from January 2003 → December 2014? Now that period is broken up into two trend lines: a big one from January 2003 → June 2012 and a tiny one from July 2012 → February 2014. That last 1.5 year trend at the very end doesn’t make much sense.
More, it ends in February 2014, but the NASA GISS data goes until December 2014. Even if we would assume, for the sake of the argument, that there is a good reason for a separate 2012 → 2014 trend, then why not use the full length from July 2012 until December 2014?
There is a very good reason for that. Look a bit closer at the end of the animation. When that last tiny trend line is drawn (1), there is a little pause in the animation. Then the data between February 2014 → December 2014 slowly starts to appear (2 and 3) and directly after that, the blue trend lines disappear while the red linear trendline from 1970 → 2014 is drawn (4). At this point the monthly data between February 2014 → December 2014 gets its full green color.
They apparently came into problems at the end of the graph. On the one hand, imagine how much the upward trend of this tiny blue line would be if they had made that trendline until December 2014. On the other hand, if they had used one trendline from 2003 → 2014, then that would probably be slightly increasing trend. Both would ruin their little theory that those “contrarians” see the graph as a series of cooling trends and they have to explain that this is not the data the “contrarians” tend to use.
The 1.5 year trend at the end of the graph seem to be some kind of compromise. Speaking of “carefully chosen”.
Stopping the blue line in February and the red only in December also has the advantage that it makes this red line a bit more steep.
But, but, would you say, this graph was probably never intended to be an accurate representation of how “contrarians” see the warming of the last 45 years. It is probably used to illustrate the different views of “contrarians” versus “realists” of the same data.
I can understand this reasoning, but I would rather doubt that, even when a “contrarian” would be tempted to give his/hers interpretation of the NASA GISS data, that the result would then resemble an escalator with six distinct “cooling” steps. That is an alarmist interpretation.
At this point those red and blue trendlines are rather distracting and that makes it rather difficult to visualize another trend than the ones that are shown. So, to explore other possible trends, I first removed the red trendline using an image editor program in order to have a blank canvas. This is the result:
Just look at the last 12 – 13 years of this blank graph. At first glance, there doesn’t seem to be much of a trend going on there.
I can now draw another possible trendline (for illustrative purposes only):
I am not claiming that this trendline is more correct than the red or blue trendlines that the skepticalscience team came up with, but this is also a possible trendline and it even makes the “pause” visible. I think it leans more to how “contrarians” would view the warming of the last 45 years than those daft escalator steps drawn by skepticalscience.
Why would someone make such an effort to fabricate the arguments of someone else? If the science is that clear, then just debate the science and get it over with. Yet, that is not how it is done here. In those three cases (the IPCC being cautious, table 1 Cook & Lewandowsky and the escalator), the alarmist side of the debate represented the arguments of the other side AND explained them AND subsequently debunked that strawman.
The answer is of course: control. By controlling the arguments, one is controlling the debate. This has advantages. It is easier to debunk a weaker and/or a framed version of the argument rather than an actual argument. It avoids an actual debate that could give the impression that there is still something to debate about and it could shed doubt when it get lost, especially those who are declaring high certainties.
Another advantage is that framing the argument makes the opponent’s argument look ridiculous and therefor can win the sympathy of the bystanders for the own case. It will make the own side look better. They have seen through the arguments of the opposition, making their own arguments superior over those of the opposition.
But then again, if the science is so clear and those “contrarians” are so very wrong, why even the need to misrepresent the other side?
Update on the origin of the escalator graph
The concept of the escalator graph was first proposed by Bob Lacatena in a comment on skepticalscience. He apparently was inspired by a java applet made by Colin Sharples. I couldn’t find the applet ietself, it was on the (currently defunct) website of Colin Sharples, but here is a description of it (following the link by Bob Lacatena):
Hot Topic reader Colin Sharples (aka CTG) is a whizz with Java programming. Prompted by a discussion on climate trends at ReaClimate, he took NASA’s GISS global temperature dataset (to be precise, it’s the GISTEMP series for the climatological year (Dec-Nov) and he’s estimated a figure for 2009 by completing the year by calculating the average anomalies for the last few months of the year over the last five years) and the Hadley Centre’s HadCRUTv3 series and produced an interactive graphic that shows how changing the length of the period you select for trend calculations affects the trends you see.
Apparently, the interactive applet used the NASA GISS dataset and had a slider to change the trend period. The shorter the selected trend, the more cooling periods popped up.
These weird steps indeed most probably didn’t come from a contrarian, but originated from/were inspired by playing with the slider of this java applet. It was then later adapted by skepticalscience to a 1970s – 2014 period and presented as “the view of contrarians on global warming”.
Trust, yet verify 
Trying to make one’s opponents look more stupid than you are is a classic piece of misdirection. It becomes your opinions against the opponents, which in Cook and Lewandowsky’s case is supporters of the academic scientific consensus against those of bloggers. Yet they are reversing what most academic subjects try to achieve. That is to go beyond established opinion to look at the actual real world evidence in support of hypotheses. This principle is as true in history and theology as in physics. Cook and Lewandowsky’s view of a linear warming trend over the last few decades is in line with others in the climate community who attacked David Whitehouse of The GWPF for daring to point out in November that global average land temperatures had plummeted by 1C at the end of El Nino. Your rough trend line of slowing temperature rise post 2000 needs to be compared to the theory. The central estimate is that a doubling of CO2 levels will lead to a 3C rise in global average temperatures. I looked at the period from 1950 and calculated what would happen if the temperature rise was in the year following the CO2 rise.
The rate of rise in temperatures should be accelerating particularly after 2000. Convert this to linear trends with the same end year, and the slope should be increasing. Therefore the trend in C/decade from 2000 to 2016 should be more than 1990 to 2016, which in turn should be more than 1980 to 2016 and so on. But look at the actual data trends (I used HADCRUT4, Cook uses GISTEMP) and the shorter period trends are lower than the longer period trends. As an addition, the warming from CO2 should be less than the actual warming if CS=3, whilst it is less. But whilst CS could be 1.5 to 2.0, the fact that warming stopped when theory says it should have accelerated is a clear divergence between theory and evidence. This suggests that the rise in greenhouse gases does not account for all the recent warming either.
LikeLike