But today he decided to try his hand with some graphs.
The settings for this foray in to the graphic world of data was a comment by Craig Emerson on Twitter with Dr Karl Kruszelnicki:
In response Andrew Bolt today wrote a blog post titled:
In the faint hope Emerson can be educatedAnd he opens with this line:
One man presents science on the ABC. The other is a senior minister in the government that’s imposed the carbon tax.
You would expect both not to be so ignorant about the very basic data on global warming:Now it’s a brave man who starts calling others ignorant especially one who is not exactly known for his superb research skills.
In order to prove to Emerson and Dr Karl that they were ignorant Bolt tried his hands at graphing some data.
Now look. I don’t want to get all “demarcation dispute” with Andrew on graphs. Sure they’ve become a bit of my thing. And sure I probably have a passion for graphs that rather goes beyond the healthy stage. But here’s the thing. When I use a graph I know that I have to be just as careful in how I present that graph as I do when I write words, because you can lie just as easily with a graph as you can with words, and I don’t want anyone accusing me of attempting to defraud the public by presenting data that is actually false or misleading.
Bolt presents 5 graphs and I will show them all screenshot just to prove I’m not making this up (because by the time you get to the end of this you’ll be sure I am).
First he presents a “graph” from the Met Office in the UK:
It’s not really a graph worth considering because where the temperature was in January 1997 and December 2012 doesn’t actually say much about the trend, but it does nicely convey to his readers that there has been no increase in temperatures. This graph he’s clearly lifted from somewhere – probably the Watts Up With That website, as that’s usually where he gets his climate change “scoops”.
But then Bolt really goes wrong, because instead of grabbing a graph from somewhere else, he astonishingly decides to do some work himself and produce his own graphs.
He went to the great website Wood for Trees which has pretty much all the data on global temperature recordings you could want. It is a neutral site. No one disputes the data it provides. And best of all it allows you to use that data to make graphs. You can do graphs according to any time period and using a raft of different measures.
Here for example is a graph I made using the GISTEMP data (the NASA guys) from 1972-2012:
I have screen capped it all to show you how these graphs are made. Series 1 – (the red line) is “GISTEMP LOTI global mean from 1972”. LOTI stands for “Land-Ocean-Temperature Index. Series 2 (the Green line) is the Linear Trend from 1972 of “GISTEMP LOTI global mean”.
As you can see it’s a nice upward trend.
Have a look also at the top left of the graph. It clearly labels the red and green lines.
OK. Now here’s the graphs Bolt used and his captions for them (and remember he prefaced these graphs by saying “In the foolish hope that these two men may reconsider their opinions after consulting the evidence”:
First up is UAH. This is the data from the USA’s National Oceanic and Atmospheric Administration (NOAA) and it measures “Tropospheric and lower stratospheric temperature data collected by NOAA's TIROS-N polar-orbiting satellites and adjusted for time-dependent biases by the Global Hydrology and Climate Center at the University of Alabama in Huntsville (UAH)”. OK that’s a mouthful but I want to be exact about what it is.
OK. Let’s have a look at this graph. Go up to the top left. What do you see? You see the red line is “uah/ from 1997”, which is what it should be, but the green line – the linear trend line – is from “rss/ from: 1997/ trend”.
That is, the trend line is not the trend of the red line. It is not the trend line of UAH data, it is the trend line of “rss data from 1997”.
What is rss? Well that is also done by NOAA but it is their “remote sensing system’s microwave sounding units”. It rather nicely has shown a flat trend since 1997.
But that is not what Bolt is saying this graph is suggesting.
What he is saying is that the green line shows the trend of UAH measurements has been flat since 1997 – it “confirms” it to use his exact word. .But what he actually has shown is two different measurements. And the measurement he has used as the trend line is from data that reveals a flat trend line. But what about the actual UAH temperatures? Have they been flat?
Let’s use the same data Bolt has used but instead of the “rss linear trend” let’s actually use the UAH data to make a trend – because that is what he is saying his graph displays:
Notice anything different? Like the fact the linear trend is now clearly heading up?
OK onto his next graph and here we at least see some honesty, because (what a surprise) it is the RSS graph! The trend certainly is flat – if anything it might even be going downwards.
I have no problems saying that, because I trust the data and I am not afraid of admitting what it shows – even if it contradicts the argument I might wish to make about climate change.
But after the one honest graph of the RSS he quickly moves on to the next bit of “evidence”. We now come to our old friend GISTEMP:
Notice anything about that top left corner? Yep the wonderful “rss from 1997” trend line.
Gee how did that get there?
Now anyone with mathematical knowledge slightly above that exhibited by the ape creatures of the Indus (thank you Black Adder) would look at that graph and think hang on, that green line isn’t anywhere near the middle of the red lines – which is usually the case for trends (scratch that, not usually – always). But not our Andy. He saw it, and slapped it on his page with a Huzzah you ignorant Craig Emerson, can’t you see even NASA knows temperatures have stopped warming!
Well OK, let’s put in the GISTEMP data to get a linear trend and see what pops up:
Hmmm. Now I may be seeing things all crooked after spending two weeks watching Channel 7’s coverage of the tennis, but that green trend line looks to be going up.
OK now he goes back to the UK Met Office and their “HADCRUT” data:
Don’t you love the nonchalant “ditto” like he has this whole climate change thing sewn up. And don’t you also like how when you look at the top left corner once again he is using the RSS data to represent the trend for different data. Once again even a basic knowledge of maths would have you realising that that green line could in no way be the trend of the red line.
But not Andrew Bolt. He saw that, saw nothing wrong with it. He put it on his blog and thinks QED.
Note as well that Bolt has used “HADCRUT3”, which has now been superseded by HADCRUT4.
OK what is the real HADCRUT3 trend
Again we see a trend line that is going up. Now sure it is only just going up and I wrote about this on The Drum a couple weeks back. Although it was only just going up I still noted it, even though I was trying to show that climate change is causing temperatures to rise. I did that because I am honest with data and am not trying to pretend the facts are one thing when they are something quite different.
Now Bolt used HADCRUT3. Why didn’t he use the more recent version, HADCRUT4? Maybe because then the trend would look like this:
You can’t even pretend that trend is flat. Also it clearly shows that even by the UK data, 1998 was not the hottest year.
I guess Bolt just wasn’t aware of that data.
Bolt then quotes NASA head climate honcho James Hansen:
An admission from the godfather of global warming, NASA’s James Hansen:
The five-year mean global temperature has been flat for the last decade, which we interpret as a combination of natural variability and a slow down in the growth rate of net climate forcing.Well OK, but Emerson and Dr Karl was talking since 1998 which is rather a tad more than 5 years (I’m not 100% sure on this, but I sure as hell hope Bolt realises that 2012 minus 1998 is not 5). So Hanson’s quote would be worth citing if Dr Karl had said “the world has warmed in the past 5 years”.
Incidentally here’s the GISTEMP for the past 5 years
Yep it is flat.
But here’s the thing. Who gives a damn? No one tries to measure changes in the climate in just 5 year blocks because it is completely meaningless. Here’s what
As you can see some went down, some went flat, some went up. But the trend since 1972 is obvious.
Now look, Bolt normally does such cherry picking. It’s intellectually dishonest, but maybe he actually does believe 5 year trends are important.
But when it comes to representing trends on a graph and suggesting that they show something which they do not, well that is not just intellectually dishonest; it is completely so.
But here’s the kicker he ends his graphs with a quote from Phil Jones regarding temperatures from 1995 to 2009. I’m not sure why Bolt is not concerned about the past 3 years, but oh well, let’s stick with what he thinks is irrefutable evidence.
The line Bolt quotes is this:
I also calculated the trend for the period 1995 to 2009. This trend (0.12C per decade) is positive, but not significant at the 95% significance level.Rather hilariously he got this line from the great blog Skeptical Science (he even links to it!), which has the quote in full:
BBC: Do you agree that from 1995 to the present there has been no statistically-significant global warmingBolt for some reason didn’t graph 1995-2009, so I’ll do it for him, using HADCRU4 as that is Phil Jones’ data:
Phil Jones: Yes, but only just. I also calculated the trend for the period 1995 to 2009. This trend (0.12C per decade) is positive, but not significant at the 95% significance level. The positive trend is quite close to the significance level. Achieving statistical significance in scientific terms is much more likely for longer periods, and much less likely for shorter periods.
BBC: How confident are you that warming has taken place and that humans are mainly responsible?
Phil Jones: I'm 100% confident that the climate has warmed. As to the second question, I would go along with IPCC Chapter 9 - there's evidence that most of the warming since the 1950s is due to human activity.
Bolt ends his diatribe of misleading information and context-slaughtered quotes by writing:
Do Dr Karl and Emerson truly dispute this evidence? Are they so in denial?
PS: an apology would be nice.It takes a hell of a lot of front to demand an apology after misleading the public and displaying an utter ignorance of statistics.
Andrew Bolt doesn’t deserves an apology, he deserves censure.
Bolt has now changed his post stating:
(APOLOGIES - I originally and inadvertently used two graphics with the wrong trend line. This HADCRUT graphic is now corrected. The other graphic was of GISTEMP, one of the data sets showing a mild warming.)That’s not quite accurate either. As you can see from above he “inadvertently” used three graphs with the wrong trend line, not two.
He puts up the correct HADCRUT graph – still using HADCRUT3 not HADCRUT4. Rather oddly however he doesn’t put the the GISTEMP graph on this new updated blog, he just links to it. And he completely ignores the UAH graph!
I note that on some browsers the graphs aren’t coming up. I use Chrome and it seems fine on them, but not on explorer (although you can right click and “show graphs”) and on Firefox it’s just the link. I don’t know why – as I put them up the way I do all my other graphs in the past. Will try and sort it out.