2011-01-21 12:05:17Another rejected blog post by Frans Dijkstra
John Cook


Frans emailed me another guest blog post for SkS - the opening and closing lines make it seem real SkS material, don't you think?

Was 2010 the hottest year on record?

A very popular alarmist argument in the last weeks of 2010 was, that 2010 was becoming the hottest year on record. Many newspapers and broadcasting programmes made headlines of this claim. But is it true?

The answer is: ‘that depends on the data set you use!’ Let’s examine the facts. There are 5 datasets for the average global temperature: surface measurements HadCRUT, GISS, NOAA and satellite measurements UAH and RSS. Maybe there are more, but these are most cited. From all datasets except HadCRUT all data for 2010 are available. For HadCRUT the data are available until November 2010. The data for December 2010 can be estimated as the average of the five preceding December months.

Before 2010 in most datasets 1998 was the hottest year on record: HadCRUT, UAH and RSS. On the other hand NOAA and GISS had 2005 as the hottest year. So we can test the claim that 2010 was the hottest year on record by comparing the anomalies for 2010 and 1998 in all data sets. If 2010 does not pass that test, it will certainly not pass the test for being hotter than 2005.

The following graph shows the data:

The datasets differ significantly. NOAA and GISS show that 2010 is hotter than 1998, whereas UAH, RSS and HadCRUT show that 1998 was hotter.

The average of all stations shows that 1998 was 0.0004 degrees Celsius hotter than 2010.

So, when we consider all evidence, 1998 was a little bit hotter than 2010, but only just. Needless to say, that such a small difference is not statistically significant.

The claim that 2010 was the hottest year on record must be rejected, unless we want to cherry pick by limiting our analysis to the GISS data, as a Dutch newspaper did on 21 December 2010.


One more comment can be made on this issue: there is a significant difference between the  GISS data and the other data sets. The average of the difference between 2010 and 1998 in the data sets except GISS is -0.033 °C with a standard deviation of 0.038 °C. GISS is 0.161 °C higher. This is more than 4 times the standard deviation of the four other sets. On pure statistical reasons the GISS data could be rejected as an outlier, unless there are other arguments why GISS could be different.

The difference between GISS and other data sets is sometimes explained as ‘GISS includes the polar regions, which the other datasets do not’. This is in the first place not quite true, and moreover, this cannot explain the magnitude of the difference. GISS has no more data from the polar regions than the other data sets.  GISS only fills in the gaps for missing stations by extrapolating the data from the closest existing stations. The other data sets do not include fabricated missing data. The UAH satellite data do include the polar regions, and do not deviate like GISS.

But even if the GISS extrapolation procedure were correct, the results seem far too high. The polar regions (above 70 °C North and South) cover only 6% of the surface of the globe. If the polar regions must explain a rise in global average temperatures of 0.161 °C, the average temperature of the two polar regions should have risen 2.7 °C between 1998 and 2010. This is not true, as can be seen in the available datasets.

So, my advice for climate alarmists: ‘if you want your claims to be reliable, do not use GISS data!’



My emailed response to Frans:


Hi Frans,

Actually, there was a recent post by Robert Way that examines this very question, is 2010 the hottest year on record:


However, Robert goes a little deeper – he includes 5 other datasets in his analysis. And crucially, many of these analyses do cover the globe but using an independent method to GISS to calculate polar temperatures – they integrate ship, buoy, surface and satellite data into a synthesised temperature record (note – the purely satellite records, UAH and RSS don’t cover the whole globe). These analyses confirm that the Arctic is warming faster than the rest of the globe – thus to take the partial records’ (HadCRUT, UAH, RSS) approach of assuming the global average over polar regions is to underestimate the warming trend.




2011-01-21 13:55:57Another reason for rejection
James Wight

My problem with this submission is not just the word "alarmist" or that it contradicts Robert's post. It doesn't make a lot of sense to compare 2010 to one year, especially an outlier like 1998. It could mean GISS overestimated the temperature in 2010, or underestimated it in 1998, or both, or it might mean the other records tended to overstimate 1998. It makes more sense to compare 2010 to the same baseline, and when you do that GISS agrees pretty well with the other records.
2011-01-24 04:58:51


He has the UAH and GISS (at least, I checked only these two) numbers wrong. As for the latter, he read the wrong column.

A two years comparison is meaningless. And why 1998 which exagerate the difference btween the datasets?

He does not say anything about uncertainty and, above all, he does not refer to the offical news releases but to an artcle on a dutch newspaper on Decembre 21st. I can't read dutch but I bet the article was talking about the meteorological year.

Not true that UAH covers polar regions and not true that GISS fabricated data.

Final remark: not worth publishing on the SkS Journal, try Energy & Environment  :)

2011-01-24 06:22:59agreed
Dana Nuccitelli

Yes Robert's analysis was far better, even setting the "alarmist" comments aside.  Looked at more data sets, analyzed the trends, didn't make bogus claims about UAH and GISS, etc.

But seriously, asking SkS to publish an article repeatedly using the term "alarmist"?  Not too bright!

2011-01-24 10:17:38Asking SkS to publish an article repeatedly using the term "alarmist"
John Cook

Gotta admire the guy's chutzpah. He never replies to my rejection emails though - just sends me another guest post. Guess he's building up to a "SkS is censoring me" meme.
2011-01-24 21:34:23

It would be interesting to see if you could get him to incrementally change the article(s) to the point that it becomes acceptable.
2011-01-24 22:18:41


John, be carefull, you know how easily deniers can misrepresent what you say. I'm afraid he's collecting your responses and sooner or later will come out with something "unpleasant".

2011-01-24 23:45:27My responses
John Cook

I have been fairly careful (I hope). Generally when I email a skeptic, I assume any or all of what I say can be used against me in a kangaroo court of law. In this case, the fact that Frans is not even replying says to me he is storing up my emails for some upcoming public tirade. So I'm writing as if they'll be broadcast publicly at some time. But I feel I have to respond in good faith too. It's a fine line to walk.
2011-01-25 01:17:14

I think you're doing fine so far. Who knows, he may eventually learn something.
2011-01-25 03:44:46


It is weird that he tries to get a guest-post published on SkS. If it were just about getting his stuff published, why doesn't he do so on his own blog or send it off to others (WUWT comes to mind)? He should realise that wordings like "alarmists" don't stand much of a chance getting published here - is he perhaps adding them on purpose? Likewise with taking jabs at GISS. If he isn't totally clueless, he should have realised by now that his writing is not up to SkS-standards/quality.....

I may be interpreting way too much into this, but is it possible that he isn't working on his own but is just a front for others who want to discredit SkS any way they can?


2011-01-25 04:04:28


I think it would be worthwhile pointing out to him that labeling a large portion of his potential audience as "alarmists" is not an effective way of reaching them.

Really, just the same way we ask SkS posters to tone down when they use terms like "'denialist" on planned SkS postings.

Maybe he aspires to being accepted by SkS. Maybe we can "train" him to be a little bit reasonable.

2011-01-25 09:40:25Training a skeptic
John Cook


If he ever replied to my replies, perhaps a dialogue and progress might be possible. But the fact that he never replies speaks volumes.

The 'alarmist' rhetoric speaks of his intractability too.

I'm not going to expend too much effort on a lost cause - time and energy can be used more productively in many other places - but I will send polite, hopefully informative replies.