People send me stuff. Strange, what could be so wrong or threatening about this story headline that it simply had to “vanish” without so much as a correction or a note as to why? Fortunately the Internet has a memory. This screencap below is from Google cache:
But if you go to Paul Hudson’s BBC blog right now…
http://www.bbc.co.uk/blogs/paulhudson/
or to the original URL:
http://www.bbc.co.uk/blogs/paulhudson/2010/12/coldest-december-day-on-record.shtml
You won’t find this headline. Instead, you’ll find this one:
Note the post time of 17:14 matches on the two stories, suggesting this has been a headline rewrite. Though, the article says:
So its been the equal coldest December day in the Vale of York since records began in 1932.
and
Scampton in Lincolnshire has experienced its coldest December day on record with minus 5.5C. Records here go back to the early 1950’s.
So, with those sentences, the headline would still be valid.
The story headline still exists in Google search findings, which is how I located the cached copy.
Note the headline was originally on the BBC home page according to the top link. I guess somebody didn’t like the original headline and decided it must be changed. h/t to WUWT reader “Pingo” for noticing.
On a related note, have a look at this headline right below the newly revised BBC story:
Ummm, sadly no. A “dead heat” is defined as:
dead heat
a race in which two or more contestants reach the finish line at exactly the same time; tie
The problem is that we have not yet reached “the finish line” for 2010, and compared to 1998, 2010 certainly doesn’t look like a “tie” to me. Here’s Dr. Roy Spencer’s UAH plot with some lines in purple I added comparing peaks of 1998 and 2010, and comparing the 13 month running average peaks of 1998 and 2010 in red:
Clearly, the peak temperatures between 1998 and 2010 are significantly different as shown by the gap in the two magenta lines.
But, what Mr. Hudson is focusing on with his “dead heat” statement is the red 13 month running average line, which “could” be said to be in a tie with 1998 at this moment. There’s only one problem, which becomes clear when we magnify Dr. Spencer’s UAH graph:
As indicated by my red arrow in the magnified view above, the 13 month running average stops in June, 2010, and the months of July, August, September, October, and November apparently aren’t included in it.
If they were, the red plot line would extend to the end of the graph. What’s comical about all this is that the 13 month running average was added by Dr. Spencer in response to complaints that the 25 month running average he had been using “hid the increase”. Read his explanation here: Is Spencer Hiding the Increase? We Report, You Decide
Since the temperature continue to drop, when we do finally get the completed 13 month running average for 2010 that includes all temperatures for 2010, Mr. Hudson will discover that red peak in 2010 will have dropped, and is nowhere close to a “dead heat” when the finish line is actually reached.
But since everyone wants to “close out 2010 early”, so as to help those partiers down in Cancun reach some sort of consensus and action, such stories claiming 2010 will be equal to or warmer than 1998, or the “hottest year on record” or in the “top three hottest years on record” seem to be the overreaching norm for media these days.
Since BBC is interested in correcting headlines, I’m sure our UK readers will want to point out this error to the BBC so that they can change it right away.
Place your bets now.
For those that might want to run their own plots to compare, here’s the actual UAH data for 1998:
ANNUAL CYCLE BASED ON 79001-98365 12-MON RUNNING MEAN YEAR MON GLOBAL NH SH TRPC NO.DAYS GLOBAL NH SH TRPC DAYS 1998 1 0.582 0.612 0.552 1.097 31. 0.103 0.149 0.056 0.213 365. 1998 2 0.753 0.857 0.649 1.291 28. 0.160 0.211 0.109 0.330 365. 1998 3 0.528 0.655 0.401 1.025 31. 0.207 0.263 0.152 0.442 365. 1998 4 0.770 1.014 0.525 1.059 30. 0.287 0.358 0.217 0.563 365. 1998 5 0.645 0.685 0.606 0.885 31. 0.347 0.419 0.274 0.653 365. 1998 6 0.562 0.635 0.490 0.536 30. 0.394 0.469 0.318 0.702 365. 1998 7 0.510 0.659 0.362 0.442 31. 0.430 0.511 0.348 0.706 365. 1998 8 0.518 0.544 0.492 0.447 31. 0.465 0.539 0.392 0.715 365. 1998 9 0.458 0.571 0.345 0.312 30. 0.495 0.563 0.427 0.708 365. 1998 10 0.416 0.519 0.312 0.339 31. 0.519 0.592 0.445 0.711 365. 1998 11 0.192 0.272 0.113 0.130 30. 0.519 0.606 0.431 0.688 365. 1998 12 0.277 0.416 0.138 0.073 31. 0.516 0.618 0.414 0.632 365.
and 2010:
ANNUAL CYCLE BASED ON 79001-98365 12-MON RUNNING MEAN YEAR MON GLOBAL NH SH TRPC NO.DAYS GLOBAL NH SH TRPC DAYS 2010 1 0.648 0.860 0.436 0.681 31. 0.313 0.363 0.263 0.286 365. 2010 2 0.603 0.720 0.486 0.791 28. 0.340 0.375 0.306 0.351 365. 2010 3 0.653 0.850 0.455 0.726 31. 0.380 0.419 0.340 0.426 365. 2010 4 0.501 0.799 0.203 0.633 30. 0.408 0.459 0.356 0.477 365. 2010 5 0.534 0.775 0.292 0.708 31. 0.441 0.511 0.371 0.542 365. 2010 6 0.436 0.550 0.323 0.476 30. 0.473 0.558 0.389 0.573 365. 2010 7 0.489 0.635 0.342 0.420 31. 0.479 0.596 0.361 0.565 365. 2010 8 0.511 0.674 0.347 0.364 31. 0.501 0.633 0.369 0.562 365. 2010 9 0.603 0.555 0.650 0.285 30. 0.509 0.630 0.389 0.536 365. 2010 10 0.419 0.365 0.473 0.152 31. 0.514 0.633 0.396 0.517 365.
Source: http://vortex.nsstc.uah.edu/public/msu/t2lt/tltglhmam_5.3
Update: Dr. Roy Spencer uses the term “dead heat” in the posting here.
Nov. 2010 UAH Global Temperature Update: +0.38 deg. C
Which I’ll have to say, I didn’t read, since that day I was attending to my wife in the hospital. Charles the Moderator posted the story for me (on WUWT). So this is where Mr. Hudson got the term, and I’m in error in assuming it was his term.
My point about the year not being finished, and average line in 2010 dropping with time, and the comparison of absolute peaks remains valid though.
1998 1 0.582 0.612 0.552 1.096 31. 0.103 0.149 0.056 0.213 365. 1998 2 0.753 0.857 0.649 1.291 28. 0.160 0.211 0.109 0.330 365. 1998 3 0.528 0.655 0.401 1.025 31. 0.208 0.263 0.152 0.442 365. 1998 4 0.770 1.014 0.525 1.059 30. 0.288 0.358 0.218 0.563 365. 1998 5 0.645 0.685 0.606 0.885 31. 0.347 0.419 0.275 0.653 365. 1998 6 0.562 0.634 0.490 0.536 30. 0.394 0.469 0.319 0.702 365. 1998 7 0.510 0.659 0.362 0.442 31. 0.430 0.511 0.348 0.706 365. 1998 8 0.513 0.555 0.470 0.456 31. 0.465 0.540 0.390 0.715 365. 1998 9 0.432 0.564 0.300 0.284 30. 0.493 0.564 0.422 0.706 365. 1998 10 0.394 0.512 0.276 0.324 31. 0.514 0.592 0.437 0.708 365. 1998 11 0.190 0.265 0.116 0.134 30. 0.514 0.605 0.423 0.685 365. 1998 12 0.289 0.415 0.164 0.086 31. 0.512 0.617 0.408 0.631 365.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
I’m afraid rewriting content on their website without notice is just normal form for the BBC – see eg this (not climate-related) rather hilarious episode.
I think I calculated that if Dec is ~.39C anomaly, it will truly be a dead heat with 1998….
Winter Dec to Feb inclusive in Britain and
Europe will be exceptionally cold and snowy –
like hell frozen over at times – with much of
England, Germany, Benelux and N France
suffering one of the coldest winters for over 100
years. It is expected that two of the three months
Dec, Jan & Feb are likely to be in the three coldest
for a 100 years (eg using Central England Temperatures).
There will be some milder periods in which West
Russia may also be briefly milder. Ireland, especially
West Ireland will have some mild periods when the
rest of Europe remains bitterly cold.
http://www.weatheraction.com/docs/WANews10No37.pdf
Your criticism of the UAH 13 month average being out of date is unfortunately incorrect. It is calculated over the 13 month period November 2009 to November 2010, but it is plotted at the central point, i.e. May 2009 (not June as you state). This is in keeping with the traditional (some would say more correct) practice of plotting the moving average at the central point, rather than at the end of the period, hence the use of a 13 month average, rather than a more useful 12 month average. Personally I think that this practice is now outmoded and more likely to cause confusion, as in this case. On the other hand, you are correct in stating that some sources are jumping the gun in their comparisons of 2010 with 1998. For example the U.K. Met. Office has recently published a news release in which they claim that the “preliminary” anomaly for 2010 is 0.52c, “equal to the record breaking 1998”, while only using figures up to October. However, even if the anomalies for November and December are the same as October, the final anomaly for 2010 would be only 0.48c, compared to the Met. Office figure of 0.517c and the C.R.U. figure of 0.548c for 1998. The reason why the C.R.U. figure is different (higher) than the Met. Office figure is that it is calculated using the monthly averages, not the more complex method used by the Met. Office, so there are actually two different annual figures for 1998 (and most years) despite the fact that the monthly figures are identical.
Sorry, I forgot to post this link to the Met. Office news release:
http://www.metoffice.gov.uk/corporate/pressoffice/2010/pr20101202b.html
Deborah says:
December 7, 2010 at 12:20 am
I wondered that too – certainly at disforth and topcliffe temps have been in the minus double figures or minus teens. But it seems the reported temps are averages?
Let’s see…
How many temperature reporting station were there is 1998?
And how many are their now?
If there are far less now, then their is an error in how many actual record fell.
Nylo says: {December 6, 2010 at 8:55 pm}
“So 12 years later, we may perhaps break the record by a couple of hundredths of a degree? And that’s “alarming”? Then what will the record be by 2100, a tenth of a degree more than now? OMG.”
The “record” that was “broken” was the anomaly using the base period of 1979 -1998.
Take it with a grain of salt.
From the Met Office New blog – Cold records tumbling…
Did Paul mention -20.4 C
The coldest temperature overnight was -20.4 °C at Braemar in Aberdeenshire
http://metofficenews.wordpress.com/2010/12/03/loca…
Overnight Records 2/3 December Previous record
-17.9 °C LEEMING 07-01-1970 (-15.0°C)
-17.3 °C LINTON ON OUSE 14-02-1991 (-14.2°C)
-17.5 °C CHURCH FENTON 14-02-1991 (-15.2°C)
-10.4 °C INVERBERVIE NO.2 29-12-1995 (-8.1°C)
-12.8 °C SPADEADAM 04-03-2001 (-12.7°C)
-12.6 °C DISHFORTH AIRFIELD 21-12-1963 (-10.0°C)
-11.7 °C YEOVILTON 29-12-1964 (-10.0°C)
– 7.8 °C SHOEBURYNESS 15-12-1991 (-5.8°C)
– 6.0 °C ST BEES HEAD NO.2 28-12-1995 (-5.8°C)
– 7.5 °C HERSTMONCEUX 31-12-1996 (-5.3°C)
– 9.1 °C KENLEY AIRFIELD 20-12-1999 (-6.5°C)
– 5.2 °C LANGDON BAY 20-12-1999 (-4.5°C)
– 6.0 °C SOLENT 20-12-1999 (-5.0°C)
-14.7 °C SHAP 31-12-2003 (-11.4°C)
– 5.3 °C CARDINHAM, BODMIN 28-12-2005 (-4.5°C)
Met Office News Blog
Deborah says:
December 7, 2010 at 12:20 am
“I live in East Yorkshire where Paul Hudson gives the local forecast. Reports from a couple of colleagues tell me about day time temperatures of -12 degrees Celsius on their way to York.”
Interesting. Outside my house in NE derbyshire my car indicated -9.5c. As I drove downhill into derby road the reading continued to drop down to -16c. These figures are a lot lower than the bbc forecast low, which was -7c and a lot lower than the so called record at dishforth, which is further north. Sounds mighty suspicious to me.
I don’t see what the big deal is. The change in headline didn’t change the facts in the story, and wasn’t significant as a change in wording. It is much ado about nothing.
Regarding the second half of the story, the fact is, that the peak of the 13 month average temperature, is as high in 2010 as a previous peak was in 1998. This is the case despite the fact that 1998 saw an extremely strong El Nino, and was about halfway up the solar cycle, while 2010 saw the start of a La Nina cycle and is still at the bottom of a solar cycle, and has been there for nearly 2 years.. So 2010 is surprisingly warm for a year where the solar cycle and El Nino cycle are both in a phase that promotes cooler surface temperatures.
Hudson’s blog points out these important facts.
People who are suspicious about why the 13 month moving average is being used should ask Roy Spencer, who is the principal researcher responsible for the UAH data about why he did this. Roy is a prominent global warming skeptic.
Not to worry, Canadian smoothing alone will cover any spread.
I live just outside Tockwith in the Vale of York.
My dad ‘s car also has one of those thermometers.
At 9am on Monday he gave me a lift to Wetherby.
The diesel vehicle was slow to start and it recorded -15 as the temperature.
When we got to Wetherby where it recorded -13 as the outside temperature.
He says when he later drove into the centre of Tockwith village, as opposed to outside the village where we live, the temperature dropped to -16c.
If it weren’t for comparatively high moisture content aloft I’m thinking this would be a very average year, global tempwise.
Where I sit it’s shaping up to be below norm(and sadly I’m below weight) this winter.
O/T a bit, but I found this part of our wonderful metoffice forecast for NE England on their website just now:
This Evening and Tonight:
Temperatures will fall away rapidly this evening with another severe widespread frost likely. Towns and cities will fall to minus 4 or 5 Celsius but in the countryside and over high ground temperatures will be much lower. Minimum temperature -10 °C
Is this – or is this not – a confirmation of a significant UHI effect? Taking it at face value, the UHI could be as much as 5degC – but that probably covers the mentioned high ground (hereabouts that would only be perhaps 500m) and yet amazingly, parts of the Vale of York (perhaps max 50m in height?) will be around minus 15degC tonight!
Kev-in-UK ,
Under certain conditions, cold air, which is heavier, rolls down into the valleys and produces lower than average temperatures. I believe, for example that the agricultural college at Houghall, is subject to that phenomenon. On the other hand, I agree about the UHI. I know that they are taken into account, to an exent, in the global temperature figures, but is that just when the temperatures sensors are actually in urban areas, or is any consideration given to “radiation” from the UHI into the countryside?
From the UK Sunday Times newspaper, 05 December:
No respite from grip of the snow queen
“We have now reached -18C, or below, in five calendar months this year somewhere in the UK – January, February, March, November and December. This has never happened before in records dating back to the 90th century.”
[Section 1, p37, Weather. Isobel Lang]
Paul Hudson has updated the blog entry with an update:
UPDATE at 3pm
I can now confirm that Leeming in North Yorkshire had its coldest December day on record yesterday at minus 6.6C.
Last night Scampton in Lincolnshire recorded its coldest December night on record, with minus 15.6C. This beats the previous record which was set….last week!
ENDS
UPDATE at 7am
Some will have noticed last night I changed some of the contents of this blog. The observer at Dishforth realised we had confused daytime maximum temperature records between 0900hrs and 2100hrs with the records that apply to the 24 hour time period ending 9am this morning, and until that time period is up we will not know if the sites originally mentioned have established new records.
ENDS
They can’t continue to “hide the decline”. I believe we had MAXIMUM temps of minus 12 C in Glasgow today. Disappear THAT.
Ray says:
December 7, 2010 at 1:19 pm
not sure I can believe that about Houghall (I assume you mean the College?) – which I know well – Firstly, its next to Durham Uni, where i did my masters many years ago, and also where I lectured on a couple of occasions to the arborists. Its near to river level, but hardly a massive height difference from surrounding topography.
Of course, you are right to consider that cold air will ‘settle’ in valleys, but you should know (I hope) that the ‘Vale of York’, is more of a generic term being a very wide area rather than a small steep sided valley!
As for UHI adjustments – has anyone seen a RAW data temp series for somewhere like London (Greewich Observatory or the Royal Society) and thereafter the ‘homogenised’ station data e.g. as used by Hadcrut3? If they have, please let me know, cos I’d like to have a look at it – just to see how much UHI has been accounted for. However, I suspect you or I will never see this data – it’ll be some of the stuff Jones has lost! LOL
Veronica says:
December 7, 2010 at 2:16 pm
ah, but thats the point Veronica – you believe thats what you have had in Glasgow – that will almost certainly disappear after data corrections (ok, lets be posh and call it homogenisation!)! LOL
Kev-in-UK says:
“not sure I can believe that about Houghall (I assume you mean the College?) – which I know well – Firstly, its next to Durham Uni, where i did my masters many years ago, and also where I lectured on a couple of occasions to the arborists. Its near to river level, but hardly a massive height difference from surrounding topography.”
There was an item on “Look North”, I think at the end of last winter, on the subject, in which they said it was prone to such frosts. Not sure if I have a copy, but if I do, I will post any relevant quotes.
In relation to the Vale of York, I suppose it depends on precisely where the low temps. are expected.
eadler says:
December 7, 2010 at 9:13 am
I agree with some of your points.
Frankly, I don’t really see how ANY moving average calculation can be used in preference to another? I would assume that a 13mth average is used to account for slight fluctuations in seasons (early winters, later summers, etc) but I know diddly squat about stats. Why are 5yr or 11yr moving averages used? etc, etc… All averages have a smoothing effect of course, but I sometimes wonder why they are picked at all!.
I look at the headline warmest year rubbish by the metoffice et al – and cannot reconcile their claims with my local experience and knowledge – so I assume it is down to the averaging across the county, or country, or continent!
Based on the limited understanding I have; this means (excuse the pun) that we have a daily average (max-min) used to create a monthly average used/mixed/averaged with other averages to create a regional (gridded?) average used to create a NH average – which is then used to create a global average – have I got that right?
OK – so we then take that data and process it using an 11 yr moving average – yes? – so how many years of GOOD and RELIABLE data would we need to establish a realistic baseline average? Lets say it’s a 100 yr solar cycle (just for example) – we would need at least 120 yrs of (good, reliable?) data to ‘see’ that represented over our ‘long term’ plot. Do we have this? I think not!
and finally, adding in any unknown (as yet, and I am sure there are many – for flips sake, they cannot even agree on the effect of clouds after all this time!) reason for climate variation – I find it quite incredulous that the anthropogenic CO2 hypothesis can be upheld as the sole/primary cause of any measured (from the averages) warming. Perhaps CO2 does play an important role – but it is just impossible to prove how or why at this time.
I have taken a local stations data for the last 8 years and see absolutely no warming – if I knew how to put up the plot in a post I would – but the fact remains that locally, there is absolutely NO warming, probably a very slight cooling.
I just find it quite strange to stand back and reflect (without emotion or viewpoint) on what data we have and what we are trying to do with said data. When you consider the very basic processes used to obtain data and analyse it – and then consider the vast bio/ecosystem that constitutes planet Earth – surely our ‘data’ and knowledge pales into the level of ridicule!
Ray says:
December 7, 2010 at 3:29 pm
yeah, that would be interesting. I don’t know if the Observatory (perhaps a mile away) was used for met data, but if it was, it would be interesting to compare data from the two. Off the top of my head, I’d say the Obs would be maybe 50m higher than the college. Is Houghall a recording station? Hmm.. I might look into that one day.
Typo above, sorry. Should be:
“We have now reached -18C, or below, in five calendar months this year somewhere in the UK – January, February, March, November and December. This has never happened before in records dating back to the 19th century.”