My Photo
Mobilise this Blog

Google

InvestigateDaily

INVESTIGATEMAGAZINE.TV

Kiwiblog

New Zealand Conservative

InvestigatePodcast

AmCam News Tips

  • Have you got mobile camera pix of breaking news, or a first-hand account you've written?
    email Investigate now on publicity [at] investigatemagazine.com and we'll get you online
Blog powered by Typepad

« A private life | Main | Like shooting fish in a barrel »

Comments

CM

Gosh, what a complete surprise that you picked out 1998 and 2002 as starting points! I fell off my chair in shock.
Not.

If we're able to be that subective; when I got up this morning it was colder than it is now. That's unarguable. So we're obviously in a clear warming mode, and mine is a more recent data set than yours. There's been a clear 'climate shift' (your words) since this morning.

Why don't you plot since 1979, if you don't trust the data before that? That would seem to be a more objective starting point.
E.g. Colin has plotted all four data series from that year here
http://tiny.cc/rDuAC
As he notes in his full explanation, they are in quite remarkable agreement.

It's also cherry-picking to point out that there have been some record lows, without ALSO pointing out that there have also been record highs. Which is why the 17 warmest years have all occurred in the last 20 years.

Oh, and my name isn't Colin. It IS interesting that you just assumed it was and posted it as fact though.....

Thomas Everth

For those who want to see the temperature trends and have a play with the data yourself online I recommend:

http://hot-topic.co.nz/keep-out-of-the-kitchen/

A great page showing where its at. It shows the bigger picture, something you will miss when looking at the graphs presented up here by Ian.

Thomas Everth

Oh, and thanks Ian for telling us about WoodForTrees.org
This is the website from which you constructed your cheery picked graphs.
Its actually a great website and I would encourage the inquisitive among your readers to go there themselves and play with the actual data.

Here is a sample (temp data before smoothing out the noise of short term fluctuations)
http://www.woodfortrees.org/plot/gistemp/from:1900/mean:1

and then this: (averaged over 50 data points, the noise is removed and the trend is visible.. now you can see the wood, not just the trees....)

http://www.woodfortrees.org/plot/gistemp/from:1900/mean:50

The reader can copy and paste the URLS into your browser.

And I congratulate Ian to the apt title of his blog entry....

CM

Ian, could you explain why you've suddenly decided to use offsets for the HadCRUT and GISTEMP data?

Also, what does the trend look like if you start at 1997, or 1999, or 2000? If they all look the same, it wouldn't seem as much like cherry-picking. Would you agree?

CM

Also Ian, could you please expand a bit more about how Wood for Trees is an "accepted alarmist site". Specifically, how are C++ software tools inherently biased towards showing that warming is occurring or is man-made?

"It's not the place of this Web site (or anyone else) to tell you the answers, even if I could! This is just a tool to help you dig into the data to help you form your own opinions. Whatever you decide the most important thing is that you learned what the issues in analysis are and how to test your ideas against real data."

http://www.woodfortrees.org/

Thomas Everth

For Ian's defense, offsets are sometimes useful when directly comparing data from different sources and won't change the trend.

But Ian did not read the notes at woodfortrees.org before posting his foot in mouth blog post here.

I recommend reading these:

http://www.woodfortrees.org/notes

And also to scroll down over the various graphs that the author presents himself there.

In regards to short term trends (last ten years) it is probably known to Ian that the Sun is currently at the end of a low in Irradiation (11 year Sun Spot Cycle) which has made its mark on dampening GW over the past years. The Sun is now coming out of that Cycle again and we should see a resumption of the temp along the long term GW trends.

Ocean Temps in June and July 2009 broke all time records already. (NOAA)

CM

Yeah I was really trying to get him to admit that by himself (and perhaps credit Colin or that website for providing him with the offsets which he's now using).

I also notice he says "Yes, that's right, the temperature sets most preferred by warming alarmists HAVE NOT BEEN PEER REVIEWED",

which isn't quite the same as

"Outside of Mann and Schmidt's clique of tame 'peers', neither the Hadley nor that GIS data has been independently peer-reviewed in any proper sense of the word."

which is what he said here
http://tinyurl.com/ygs9uhv

Not quite the same thing. I would have thought being precise was quite important for any reputable journalist. Otherwise you run the risk of being found out.

In that same post (a response to Colin) Ian also suggests that 2009 studies (this year, and multiple) "suggest UHI contamination can be as high as 0.1C PER DECADE". Ian, can you direct me to these. The only relevant recent study I can find is Jones et al, but that's just one, and it's 2008, and what you've suggested would be a misrepresentation.

http://www.agu.org/pubs/crossref/2008/2008JD009916.shtml

Thomas Everth

One last graph from WoodForTrees.org

http://tinyurl.com/yjhzqjw

Its a composite of four different temperature measurement series and their respective linear regressions from 1980 to present.

A good way to look at the bigger picture.

peter

Thomas and CM - I am sure we are all very grateful for your enlightenment.

In his "some mothers do have 'em" fashion, Ian has put a VERY appropriate title on this thread:

"Lies, Damned Lies and Statistics"

May this be alesson to all!

CM

Colin is the man really.
I mean the other Colin ;-)

Thomas Everth

Thanks Peter, yes, Ian's spiel with the data is a prime exercise as how to fudge the message by cherry picking your data.
Fortunately he pointed us to this website www.woodfortrees.org where everybody can have a go at playing with statistics and learn the tricks of fudging by careful selection of starting point and brief enough time scales to create graphs that produce the intended message.

The more the general public is educated by science sites like woodfortrees.org the less they can be manipulated by polemics.

I guess this was not the outcome Ian intended.

robk

You guys are quite funny, really :-D Give yourselves a good pat on the back...

Ian Wishart

I've been down country most of the day, and while I've been out the mice did play, I see.

Thomas, CM et all - you have utterly missed the point. I chose a ten year trend because that was the time frame Gareth claimed to have "debunked".

As the data above deals explicitly with that ten year period, it conclusively shows a deceleration in warming, which in ordinary English is known as a 'cooling trend'.

I think we can all agree it has not yet fallen below the 30 year anomaly average, but temperatures now (notwithstanding dodgy NOAA data) are cooler than they were eight or ten years ago.

It is not "cherry picking" to use the data in support of a precise point, and with respect you guys have manifested on this thread in my absence like a colony of noisy fruitbats.

CM, I used the offsets that CTG plugged into his graph over at HT so we could compare apples with apples (his graph vs mine). It wouldn't be much of a comparison if I didn't match his baselines.

The overarching point there is that the graphs prove my point about GIS being in a world of its own and Hadley not far behind.

Finally CM, you queried me on the UHI studies. The correctly cited studies are found in Air Con, Chapter 8. I'm sure you have a copy.

John Boy

I can't work out what the point of the argument is really unless it all points to a UN agenda that is unsavoury (as is most UN stuff I guess). There seems no tipping point where it goes pear shaped as the earth will manage itself within boundaries set by far more than CO2. It seems historically that warming is good, cooling is very bad. With the current stress on resources because of a bulging global population why would you want it colder? Eventually nature will balance us with what the world can realistically support and getting colder will do it a darn site quicker than warming. Why not just blatantly kill people off and help nature out. Hey, I think dooming billions to poverty will do that anyway. Good plan. No one will notice.

Thomas Everth

Betting on Fossil Fuels to carry us on into the future will surely doom the poor as resource constraints will rise the price of Oil and Coal to the point where the poor can no longer afford it. The spike of 2008 gave us a preview of that.

The current crop yields getting humanity through are only possible with massive irrigation projects and soil moisture reduction due to warming can quickly push large areas of productive land below the threshold where production is viable. You do not need 4Deg warming to create large problems for agriculture world wide.

And Ian, you graphs and your entire argument hinges on one data point: The exceptional year of 1998. In this year a massive El Nino event pushed the envelope. This is especially amplified in the UAH and RSS data.
So picking that year 1998 as the starting point of your argument allows you to make a point - moot as it is - as anybody can see who looks at the bigger picture.

Ian Wishart

And 1998 affects the 2002 start point how?

Crop yields have been boosted, by the way, by increasing atmospheric CO2 concentrations.

Thomas Everth

I beg to differ. The downside of warming on crop yields far outweigh any increase from CO2.

Just today this study was in the news:

"One of the world's most influential scientists has warned that climate change could devastate Africa, predicting an increase in catastrophic food shortages."


http://www.guardian.co.uk/environment/2009/oct/28/africa-climate-change-sir-gordon-conway

Ian Wishart

Peer reviewed?

I note he admits in the first instance that the main drivers of climate in Africa are not CO2 related

"The African climate is determined at the macro-level by three major processes
or drivers: tropical convection, the alternation of the monsoons, and
the El Niño-Southern Oscillation of the Pacific Ocean. The first two are local
processes that determine the regional and seasonal patterns of temperature
and rainfall. The last is more remote in its origin, but strongly influences
the year to year rainfall and temperature patterns in Africa. Despite the
importance of all three processes, we poorly understand how they interact
and how they are affected by climate change."

Despite this massive uncertainty, he then expresses certainty:

"What we can be sure of is that global warming—expressed, for example,
through higher sea and land surface temperatures—is affecting the outcomes,
increasing the incidence and severity of the droughts, floods and
other extreme weather events that these drivers produce."

Hopeless. Read Air Con, you'll learn far more about climate change than from articles and papers like this one.

CM

“As the data above deals explicitly with that ten year period…..”

Ah but Ian, you didn't choose a ten year period. You’ve carefully picked an 11.8 year time frame (1998 to mid 2009 inclusive) and a 7.8 year time frame (2002 to mid 2009 inclusive).

Unfortunately for the cherry-pickers out there like yourself and Garth George, we've now gone past the point where you can pretend to pick a round number like 10 years (when in fact the real reason is simply to ensure 1998 is included).

Even if you discount 2009, on the grounds that it’s not a complete data set (although I know you reject that notion because you and others still have your panties in a wad over Hadley changing their data presentation), then the 10 complete years would be 99, 00, 01, 02, 03, 04, 05, 06, 07 and 08. THAT would the data set for the last ten years. But of course, if you plug that into Wood For Trees, it shows a warming trend. Or, if you want to include a 10 year period ending now, you need to start at October 1999. Which shows a warming trend. If you start in 1997 you get a warming trend. If you start in 2000 you get a warming trend.

So now, in order to still include 1998, which you ALWAYS need to in order to get a desired result (which is the very definition of cherry-picking of course) you’ll have to try and make some convoluted argument that 12 or 13 or 14 years is more appropriate. Cherry-picking becomes even more obvious. Hey I know, maybe you can pretend that you now accept that 10 years is too short, and say you’re just bending over backwards to try and play by the rules imposed on you? I’m sure you’re well aware that anyone can argue absolutely anything if they’re determined enough, and if they have no problems working backwards from the desired result. Which is what you’ve clearly done here, so that clueless people like Garth George can point to and use to continue to misinform those who are unfortunate enough to read his column.

As Colin said about his gadget: The point of my toy is to show that it is meaningless to pick two points on the graph, draw a line between them and say “there is a cooling trend”. You’ve proven that in no uncertain terms. As everyone else has said, the title is very apt (although clearly not in the way you meant it).

Notwithstanding that, a “deceleration in warming” is not a “cooling trend” in any language (ordinary or otherwise). The only reason anyone would use that term would be to deliberately mislead. Which is intellectually dishonest.

“I think we can all agree it has not yet fallen below the 30 year anomaly average, but temperatures now (notwithstanding dodgy NOAA data) are cooler than they were eight or ten years ago.”

You mean 11.8 or 7.8 years ago? ;-)

“It is not "cherry picking" to use the data in support of a precise point”

Why don’t you use the CORRECT precise data then, as opposed to starting at 1998 or 2002, neither of which are 10 years ago. If being precise is the key, why don’t you show us what the trends look like starting either 10 years ago from October 2009, or from prior to 2009. And if you’re ‘precisely’ looking at a trend over that period, then how are recent ‘record breaking cold snaps’ even remotely relevant?

“CM, I used the offsets that CTG plugged into his graph over at HT so we could compare apples with apples (his graph vs mine). It wouldn't be much of a comparison if I didn't match his baselines.”

You sure are a slippery character Ian! As Colin has already explained: “The baseline for GISTEMP is the average from 1951 to 1980, whereas the HadCRUT baseline uses 1961-1990, so the baseline figure for HadCRUT is about 0.09°C warmer than GISTEMP’s baseline. So although HadCRUT spends a lot more time below its average than GISTEMP does, you have to realise – as Wishart apparently does not – that they are different averages. If you want to compare GISTEMP and HadCRUT on the same graph, you need to offset the GISTEMP figures so that they refer to the same baseline.”

I.e. to compare apples to apples in terms of the GISTEMP and HadCRUT data, you need to apply the offset.

But then this, from you, back at HT suggests you didn’t realise an offset needed to be applied to compare apples with apples:
“If you go back to your own gadget and compare the HadCRUT plot with the GISTEMP, you'll see GISS plotting significantly above the average for most of the 20th century, while Hadley plots below it for the same period.”
http://tinyurl.com/yflz88t

CM

It's Craig by the way. Cheers for the correction (not that it really mattered). ;-)

The comments to this entry are closed.