Bishop Hill The Met Office's cold winter forecast
Jan 13, 2011 Climate: MetOffice Paul Hudson has produced a copy of the Met Office's previously unseen "cold winter forecast" - the one that was sent to the cabinet office but was not released to the public.
As Hudson notes, this is supposed to represent the average for December-February, so it wouldn't be expected to show the severity of the weather we experienced before Christmas. I'm not clear on whether those are temperatures or anomalies at the bottom.
So it seems as though there was a forecast of cold weather. Assuming that this was what was handed to the Cabinet Office, we then have the mysterious question as to why the Met Office's website at the time was suggesting that it would be warm.
Something to do with Cancun perhaps?
Autonomous Mind's post on the subject is well worth a look too.
Reader Comments (28)
Probably suppressed so that hoon Huhne could go on his jaunt in the sun with a publicly clear conscience.
Why didn't Paul Hudson show that "cold winter forecast" in his blogs on the 1st or 5th October if he had it at the time? He could have shown the two charts together then, as Autonomous Mind does now.
Curiouser and curiouser.
The graphic must show the forecast anomaly because an average temperature of -0.5C in northern Norway and +0.5 in Libya would be extremely newsworthy.
Assuming the scale is anomoly from average and the average is 0 it hardly suggests the MET was forcasting that there would be a Dec as actually happened, and it seems to miss out that the North got hit more than the South as its showing the further North you go the nearer to 0.
Following your comment on my post BH, I will look again at my last section. Thanks for your input.
The question mark here that needs clarification is about the profile map above. Do we know this was the forecast given to the Cabinet Office, because all reports on the topic suggest the CO was told to expect an extremely cold start to the winter - which this profile map does not bear out.
Hudson himself says that it doesn’t by definition give any clue as to the severity of the weather.
@AM
I posted that very question to Paul Hudson in the comments, awaiting responce.
"previously unseen"
They've been busy colouring it in...
Having read Autonomous Mind's piece and the comments, I wonder if there's another book for the Bishop in this? The Loneliness of the Long-Distance Forecaster, perhaps..?
I can see CAGW pawprints/mindset/delusion all over this. Whatever happens is what was predicted!
It doesn't clear things up. It actually makes it worse.
1. Mean temperatures across the UK in December were 5C below the average, the coldest December on record.
2. Quarmbey reported, "The Met Office gave ‘early indications of the onset of a cold spell from late November’ at the end of October."
3. National Grid Winter Outlook reported, "The Met Office have now ceased formal publication of their long term winter weather forecast, however their website continues to provide some long term analyses. For the period of December through to February the data presented suggests a higher probability of above average temperatures rather than below average temperatures."
4. Roger Harabin reported, "In October the forecaster privately warned the Government – with whom it has a contract – that Britain was likely to face an extremely cold winter. It kept the prediction secret, however, after facing severe criticism over the accuracy of its long-term forecasts."
5. Paul Hudson reported, "The colour map below shows the actual forecast that I obtained at the time and wrote about. It's a Met Office winter temperature profile, and there can be no doubt that it does show that the UK and Europe could expect a cold winter. This should put an end to the ongoing discussion as to whether the Met Office forecasted a cold winter or not. It is worth stressing that this is an average temperature profile across winter - December, January and February. It suggests that the winter would be cold, but it doesn't by definition give any clue as to the severity of the weather that we experienced in December - nor would it since seasonal forecasts are just that - an average for the season."
The Met Office forecast map provided to Paul Hudson shows that UK would experience on average a winter that would be 1C below normal. This forecast makes no mention of the onset of a cold spell from late November as Quarmby reported, nor highlights what Roger Harrabin reported that the Met Office did predict an extremely cold winter. As Paul Hudson states that seasonal forecasts are just that - an average for the season.
Also the Met Office forecast map provided to Paul Hudson contradicts the long term analyses provided by the Met Office on its website, analyses which which the National Grid uses to produce it's Winter Outlook reports.
It is beginning to look like that we are now dealing with THREE Met Office winter forecasts.
1. A winter analysis that predicted a 60-80% chance of a warm winter.
2. A winter forecast that predicted an average of 1C below normal winter temperatures.
3. A winter forecast that predicted an extremely cold winter with the onset of the extreme cold from late November.
It is this third forecast that both Harrabin and Quarmby alluded to that still remains a mystery.
I repeat a comment I made against another post before this one appeared:
Hudson's article is all spin. What he shows first (October 1) is that the US NOAA forecast for Europe was very accurate, but then it was forecasting this 9 months ago as well. We know what the Met Office were expecting in October for winter - milder than usual, with less than 5% chance of much colder than normal. The evidence is still on their website, but hard to find. Hudson's second piece (October 5) is about the European model:
"This would mean not especially cold temperatures, as the wind direction would not be from the North or East. So it's no surprise that the European model is expecting temperatures this winter to be close to normal."
So he has one post about the US model and one post about the European model. The Met Office hardly gets a look in.
So, here's the spin: what Hudson says on his January 12 piece is that combining the forecasts for the Met Office (slightly milder), the European (average) and the US models (much colder than normal), we end up with an average that is colder than normal. The "Winter Forecast 2010-11 with all predictors" is clearly not the Met Office's own forecast, but what it says, an average of a number of a number of model outputs from a number of national/international centres. The Met Office have put their logo on the average because they've presented the data average, but this is not THEIR forecast.
So, we are getting to the truth. When the Met Office briefed that it was going to be a colder than normal winter, this was their take on what the consensus was, of which they were a minority. What the Met Office really predicted from their own models was a milder than average winter, but this data was swamped by all the others that predicted normal or colder winter.
So they can talk with a forked tongue: they can say
We believe it will be a milder than average winter (if we trust our models: that's what they are telling us); or
We believe it could be a colder than average winter (if we take into account that we're in the minority and factor in what everyone else is saying).
Truth is, their models are seen to be a heap of junk.
Cease publication of forecasts.
Doesn't mean you can't get them with an FOI request does it?
ha ha
Scientist for Truth @ 10.31
"We know what the Met Office were expecting in October for winter - milder than usual, with less than 5% chance of much colder than normal. The evidence is still on their website, but hard to find."
It is not that hard to find. Just go to http://www.metoffice.gov.uk/science/specialist/seasonal/ensemble/ens_mean.html , then select "October 2010" in the "Issued" dropdown menu and the map appears.
Regards
Mike Post:
Thanks for your instructions. The red font heading text for the map is bizarre. It refers to "raw data". But the anomalies shown are for the future (as of October). There was no data for Nov/Dec/Jan when it was issued, much less 'raw data'. I think I've noticed this uniquely perverse use of 'data' by 'climate scientists' before. The other aspect of the heading that is weird is the final caveat: "this does not constitute a seasonal forecast for a given location". How can that possibly be true? The map shows a three-month forecast (in the layman's sense of that term) for the world. And the world is comprised of a set of regions. One cannot exhibit such a map and simultaneously deny that it makes claims about Europe, the US, Oz, etc.
Throwing up smoke screen like mad , this latest map merely provides fodder to argue over without actual given much information. It’s clear that realised to the public or not the MET still do long and midterm forecasting, therefore this information, no matter how it was or was not supplied to, exist.
All the MET office has to do is release the information on its forecasts covering this period into the public realm, it as accurate as they like to claim it’s hard to see why they don’t. Who sat on it , who decided not to tell the public etc are secondary to the basic question , did the MET office accurately predict the weather seen over this period?
I agree SFT
We have THREE Met Office forecasts
MO-1 A winter analysis that predicted a 60-80% chance of a warm winter
MO-2 The Met Office forecast given to Paul Hudson looks like an aggregate of the Met Office, ECMWF and NCEP forecasts predicting a colder than normal winter, basically 1C colder for the UK.
MO-3 The Met Office forecast reported by Quarmby and Harribin that predicted an extremely cold winter with the onset of the extreme cold from late November.
Take a look at this site.
http://www.met.reading.ac.uk/Data/CurrentWeather/wcd/blog/seasonal-forecasts/
This site reports;
"For Europe the NCEP/CFS is forecasting temperatures slightly below average earlier in the period (in particular over eastern Europe) but close to average by late winter and early spring."
The ECMWF’s seasonal forecasts will be now discussed. .......... The signal over Europe is weak but there is a hint of cooler than normal temperatures in particular over south-eastern Europe
"Slightly below average" by NCEP and "a hint of cooler than normal" by ECMWF does not add to an extremely cold winter reported by Harrabin nor a cold winter reported by Hudson.
Now if you combine these seasonal forecasts with the Met Office's own web site seasonal analyses you end up with an aggregate that could be at best termed as a prediction of normal winter temperatures.
There is something seriously amiss here.
The only way the Met Office could have produced the forecast map given to Paul Hudson, an aggregate of forecasts, that would also allow Roger Harrabin to state the Met Office did forecast an extremely cold winter is for the Met Office to actually have predicted an extremely cold winter.
Again this mysterious and secret THIRD Met Office forecast completely contradicts the Met Office's own web site analyses.
How did the Met Office predict an extremely cold UK and European winter?
Mike Post "It is not that hard to find. Just go to http://www.metoffice.gov.uk/science/specialist/seasonal/ensemble/ens_mean.html , then select "October 2010" in the "Issued" dropdown menu and the map appears."
I wasn't suggesting it's hard for me to find - I've no problem finding them. But is is not easy for the 'general public' to find. The MO deliberately buried it (they have said so themselves) away from where most users will browse so that Joe Public doesn't stumble across it.
Mac "How did the Met Office predict an extremely cold UK and European winter?"
Well, Paul Hudson is pointing to the aggregate chart isn't he, and who knows how many have been aggregated? - it could be a lot more than the US, the European and the MO. By October practically everyone and his dog (EXCEPT the MO) were forecasting a colder than average winter, so if "all predictors" are, say, 10 weather offices then the MO's data is diluted by 90%, and practically disappears. Bear in mind that the MO was predicting warmer than normal only by about 0.5 degC, so throw that in the mix with lots of others showing significant cooling and it doesn't play much part in the overall aggregate.
SFT
It is possible that other Met Office's were predicting a colder European winter.
Going back to the web site below indentifies two more.
http://www.met.reading.ac.uk/Data/CurrentWeather/wcd/blog/seasonal-forecasts/
"They (Japanese) suggest that in this coming winter temperatures are likely to be below the climatological average across western Europe"
and
"It is important to stress that one cannot look only at three independent seasonal forecasts to make predictions. The forecast issued by the Beijing Climate Center (available here), for example, shows that it is more likely for temperatures to be above than below average over north-western Europe in the coming months.
So we have the American, Europeans and Japanese predicting a cooler (not colder) than normal winter. The Chinese predicting an above average winter. Whilst the Met Office own seasonal analysis on their web site predicting a warmer than normal winter with a 60-80% probabilty.
Any Met Office forecast based on an aggregate on these 5 forecasts points to normal winter temperatures, not warmer and certainly not colder..... unless that is if the Met Office did actually predict an extremely cold winter, an outlier that may have produced a possible aggregate forecast map given to Paul Hudson.
Again, there is something mysterious about all of this. The Met Office and the various BBC stories do not run straight.
It is time the Met Office released their actual secret forecast and explained how they produced it.
Perhaps the Met Office should be allowed to issue two weather forecasts per season... as long they are both clearly marked:
1) WInter forecast - raw
2) Winter forecast - ideologically adjusted
"The Met Office gave ‘early indications of the onset of a cold spell from late November’ at the end of October."
I think I could safely predict that for 2011, 2012, 2013, 2014, 2015 etc.
Its called "Winter" isn't it?
On a more serious note, looking at the met Powerpoint slide shows even that got it wrong. Surely the coldest anomalies were in the North and West of the UK, not the South and the East? They also got the amplitude of the cold anomaly completely wrong.
This looks like a classic AGW "have your cake and eat it" argument. These are not the droids you are looking for...
Scientist for Truth @ 1.04
"But is is not easy for the 'general public' to find. The MO deliberately buried it (they have said so themselves) away from where most users will browse so that Joe Public doesn't stumble across it."
Point taken!
Regards
One thing's for sure. The weather is showing symptoms of a serious psychological disorder in refusing to acknowledge the science of catastrophic global warming.
These behavioural problems prove beyond doubt that the weather is in denial.
It does appear that the Met Office does need a new computer to predict the weather and a new Met Office to go with it.
Jezzus -- anyone can be right if you make a number of predictions covering from melting-lead heat to absolute zero cold and then you get to chose which is right after the fact.
Cartoon for Josh -- Met Office forecaster rifling through a stack of predictions and pulling one out saying "See, we got it right! Here's it right here!"
Peter S
I like the notion of the weather being in denial on climate change.
Peter S
These behavioural problems prove beyond doubt that the weather is in denial.
Most excellent!
I think we could have an umbrella slogan here gentlemen:
"The problem with the weather is it's always in denial!"