Guest post by Steven Goddard
25 September 2008
The Met Office forecast for the coming winter suggests it is, once again, likely to be milder than average.
Seasonal forecasts from the Met Office are used by many agencies across government, private and third sectors to help their long-term planning.
The meteorological winter is over, and the official results are in :
The UK had its coldest winter for 13 years, bucking a recent trend of mild temperatures, the Met Office has said.
The average mean temperature across December, January and February was 3.1C – the lowest since the winter beginning in 1995, which averaged 2.5C.
Peter Stott, of the Met Office, said despite this year’s chill, the trend to milder, wetter winters would continue.
He said snow and frost would become less of a feature in the future.
….
The Met Office added that global warming had prevented this winter from being even colder.
2009 is expected to be one of the top-five warmest years on record, despite continued cooling of huge areas of the tropical Pacific Ocean, a phenomenon known as La Niña.
2007 is likely to be the warmest year on record globally, beating the current record set in 1998, say climate-change experts at the Met Office.
Cognative dissonance or just plain faith? Note: I didn’t write the “R” word!
John Philip (23:29:04) :
“In fact Over the nine years, 2000-2008, since the Met Office has issued forecasts of annual global temperature the mean value of the forecast error is just 0.06 °C.”
“Or, to put it another way … Bullseye!”
Steven Goddard (05:51:36) wrote:
John Philip,
“I calculated the standard deviation of monthly UAH temperature data from 2000-2009, and it came out to 0.15. Thus a chimp forecasting the same mean value every month would come close to the error you claimed for the Met Office. Very impressive.”
These posts needed repeating.
If a person’s income were tied to AGW, this cliche’ would fit: “Facts don’t lie, but _____ figure!”
Thank you Steven for pointing out the standard deviation.
markm
Anthony, I know this is off topic but I’ve always wondered what role plate tetonics may play in climate change, especially CO 2 . I’m not a scientists, but an avid reader of anything related to climate and the earth in general. It would seem logical that all the earth movements and subsequent venting would have some effects even if it’s trapped in the oceans until something helped release it. I would appreciate any thoughts or comments.
Well I live in Hong Kong, and February was warm. It was 5.1 degrees Fahrenheit warmer than normal.
http://www.hko.gov.hk/wxinfo/pastwx/mws200902.htm
I didn’t realize the UK determined global weather.
Gee, that’s a surprise…not!
FYI: Great Lakes Ice:
Weekly Ice Cover compared to average
http://ice-glaces.ec.gc.ca/Ice_Can/GL/CVCSWCTGL.gif
Historical Date Ice Coverage
http://ice-glaces.ec.gc.ca/Ice_Can/GL/CVCHDCTGL.gif
NOAA’s Winter Forecast issued in Oct 2008 didn’t verify too well either. If I remember correctly, NOAA had the middle third of the US above normal for temps, below for precip, with the upper Plains being well above normal. NOAA MRF team offered little explaination other than a continuation of weak La Nina conditions.
As it turned out, a La Nina did develope,which induced a persistant 4-Corners High until Feb 2009. This induced drought conditions into Cali. The deep polar air masses which plagued Alasaka and NW Canada advected one after another into the Plains and Eastern Third of the US and Canada. Below normal temps reigned there while the Rockies and Texas remained high and dry. A weakening of the ridge out West (possibly caused by the MJO) allowed more zonal flow (and rainfall) to return to the Rockies.
David Archibald (05:48:41) :
Oulu has not broken its uptrend and F 10.7 is now 69 and perhaps headed back down to 65 over the next three months.
http://www.leif.org/research/oulu.png clearly shows the downtrend. And why should F10.7 head down when over the past three months it has headed up?
Yes, fail enough time, you’ll end up getting it right sometime in the future. Unless we are in a 30 year cycle or worse, they will be dead before it warms up again.
David Archibald. I agree. High energy prices, over time, will solve themselves, regardless of government intervention. I believe it was Mark Twain that said” Everyone complains about the weather, but nobody does anything about it” lol
John Philip (04:52:27) :
“…in the period in question the anomaly has varied between 0.27 and 0.47 so a mean forecast error of 0.06C is pretty good…”
Actually it is very bad. When I read meteorology I was taught the way to evaluate a forecast was to check it against the “persistence forecast”, i e “it will be the same as before”. If your forecast did not do better than that it was useless. Using the “persistence forecast” from 2000 to 2008 (“it will be the same temperature as last year”) gives a mean forecast error of 0.0535 C, so an error of 0.06 C is worse than useless.
Predictions are not meant to be remembered. They are only meant to be headlined and marveled at for a few days, then forgotten.
I must say that you show an amazing amount of chutzpah to hold climate officials to performance standards. It simply is not done, old boy.
Climate institutions are the new monastic centers of world spirituality. They should be revered, not questioned.
Philip Bratby et al:-)
Absolutely right! I am wrong. I was racking my brains trying to remember who it was & I was convinced it was Ian M. Sorry to Michael F. I couldn’t for the life of me remember his name, my wife & I went thro’ everyone we could think of, got half a dozen names, even my father-in-law couldn’t remember exactly who it was & I perhaps should have checked the Met Office site but time was agin me! AND yes it has to have been 87 not 88 because it occurred at least 2 years before we moved down to the south-west. It’s my age.
So I stand/sit corrected. Thanks.
David C. Archibald;-)
Thanks for that too.
AtB
It seems that there is plenty of room for private forecasters. Private or even individual research it is (and it has been historically) by far better than any (if any) done by government employees, they are sure nobody will remove them from their jobs.
John Philip (23:29:04) :
“Hmmmm, let us not confuse weather with climate…as we are looking at their climate forecasting”
No John, they were forecasting the weather for this winter not the climate.
It’s quite difficult to find the full list of CET from 1659 onwards but after a while I found this on http://www.islandweather.org/php/Central_England_Temperaturelist.php?start=301
Divided into six pages, with 59 entries per page and an average annual temperature given for each page, cherry picking does not enter into the equation with such a random division. So we find:-
1659 -1718 average 8.79
1719 -1778 9.21
1779 -1838 9.11
1839 -1898 9.09
1899 -1958 9.41
1959 -2007 9.56 (last complete listed year)
What’s the fuss?
Well, although I know precious little about the UK Met Office, I feel we should look a bit closer at weather forecasting.
In general, NOAA does a pretty fair job forecasting short-term weather for an area restricted in size. While not perfect, their 12 to 48 hour forecasts for aerodromes is generally quite usable.
The problem is simply that as you enlarge the forecast area and/or extend the forecast time, the reliability falls rapidly.
Still, I suspect that seasonal forecasts are not much more than a WAG, and I doubt that they they are ever fully analyzed or subject to validation. Further I doubt these WAGS are handled much differently in the UK.
Having said that in defense of forecasting in general, I can’t resist asking how accurately forecast validation can be performed when we have so much difficulty just accurately determining a simple actual temperature at any given time or place without the after-the-fact help of NASA scientists?
Roger
Please see my post on this thread 6 38 05-my highly motivated and efficient team of dedicated climate researchers researched the CET in some detail.
Tonyb
Roger, it’s http://hadobs.metoffice.com/hadcet/cetml1659on.dat
Not ‘final score’ yet – the Met Office continues to get it wrong still!
” 25 February 2009 : Mild weather is expected to see out what remains of winter.”
Yet the current forecast is: “Wintry showers continuing. Frosty tonight.”
On the longer term, the Met Office continues with its deluded doublethink on global warming (as pointed out recently by Lucia). Their ‘Fact 2’ at http://www.metoffice.gov.uk/climatechange/guide/bigpicture/fact2.html is “Temperatures are continuing to rise” while their own graphs at http://hadobs.metoffice.com/hadcrut3/diagnostics/global/nh+sh/ show exactly the opposite.
“Global warming had prevented this winter from being even colder.”
I would have to say that this means that global warming is a blessing rather than a curse.
It is when it gets too cold that there is widespread death from famine.
OT but i need some help. Our local newspaper ran a column by an environmentalist and they wrote
“Once again, average global temperatures are rising, with 2008 temperatures placing it as one of the 10 warmest years on record. All of the record-setting years for global temperature have occurred since 1997.”
Can one of you provide a link to the data that most effectively refutes the above? I will be writing a letter to the editor and want to have the facts I need.
Thanks.
The last couple of years were mild? Compared to what?
Of course if one was to turn to Piers Corbyn of Weather Action you would have seen that his forecasts were correct. But then you see, he’s not a believer.
http://www.weatheraction.com/id7.html
Why are they taking so long to get the GLobal Temperature Anomaly number for February 2009? Could the value be inconvenient for their AGM agenda?
Steve – I guess we’ll just have to differ about the usefulness of comparing the mean forecast error to the standard deviation on its own as a method of evaluating forecasting skill. Can you point me to a statistical authority that recommends it? A SD of 0.1 (I get a smaller number btw) simply means that the difference between the mean and the actual value was 0.1 or less about 68% of the time and greater than 0.1 about 32% of the time so the Met Office’s actual average performance of a 0.06C sits well within that range, and is actually less than the uncertainty range in the global mean. It hardly seems likely that the Met Office would publicise the error value in their Press Releases if it demonstrated that their forecasts are useless.
I have more sympathy with tty’s point about the ‘persistence forecast’ however it seems to me that examining just 9 data points is not conclusive: the period in question is one of historically unusual stability (The Std Dev of this period is less than 25% that of the whole dataset) so a naive persistence forecast would perform well anyway. It would be more useful to see how well the forecast performs during an ENSO event.
cheers,
JP.
Re: Martin Kidds enquiries@metoffice.gov.uk (reply to james above)
I endorse the contribution by Phillip Bratby. This is a typical case of equivocation and obfuscation to justify the Party Line!
I swear, all employees of Government departments must undergo an operation to deprive them of independent, rational thinking. Come to think of it, BBC employees have the same problem.
The irony is that we poor mugs are required to support them at every level by taxes on our hard-earned income.