Did all these models or any of them forsee the cold winters of 2010 and 2011?

If they did I don't remember although I do remember the MetO Model clearly predicting a mild winter in 2010, this coming after their BBQ summer prediction of 2009.

Not saying that this winter will be even average never mind cold, just making the point that models rarely predict cold winters and we have had 4 in the past 9 years.

Andy

Totally agree with the point you make about long-range forecasts and their accuracy or lack of it, Andy. As Matty H once said about LRFs, "Just LOL".

---

 

That was true of both the 80's and 90's but I'm not so sure it was the case at the end of the noughties. Winter 2007/08 was mild overall but not exceptionally so from what I recall, but the following two winters were both colder than average, notably so in the case of 2009/10. Winter 2008/09, from what I remember, also brought the first notable snowfalls to parts of southern England since February 1991 if I'm not mistaken.

 

I wouldn't bother using data to rubbish Ian's claims; he has his agenda and he'll jolly well stick to it.

 

---

AccuWeather winter forecast for the UK and Europe

https://www.accuweather.com/en/weather-news/europe-winter-forecast-uk-to-germany-on-alert-for-wind-storms-rain-to-ease-drought-wildfires-in-spain-and-portugal/70003011

 

That's not very encouraging, is it?

---

 

 

That's not very encouraging, is it?

They might actually get it right for once, their last two winter forecasts for this side of the world have been total crap.

 

---

Interesting forecast from Accuweather. I've had a few journalists asking me directly about my winter forecast in recent days. I assume they're having problems finding anyone who is prepared to call cold this year. In my case it's simply because I always wait to see how the second half of October and November pan out. Issuing now would make me more £s but I don't consider it worthwhile switching the TWO schedule.  

---

Interesting forecast from Accuweather. I've had a few journalists asking me directly about my winter forecast in recent days. I assume they're having problems finding anyone who is prepared to call cold this year. In my case it's simply because I always wait to see how the second half of October and November pan out. Issuing now would make me more £s but I don't consider it worthwhile switching the TWO schedule.  

One thing I noticed while following the detailed explanations of the teleconnections involved in long range forecasts both here and over on Netweather last winter is that there seemed to be so many potential variables that I'm not entirely convinced of the point.

It seemed to be that present conditions would give rise to such an outcome in 3/4 weeks time. Fair enough but it usually seemed that a week later those present conditions,  or at least one of those factors driving those present conditions had changed so considerably as to totally alter the outlook.

No disrespect to any long range forecasters because I really admire the knowledge and work that goes in to those forecasts and I really enjoy reading them. Are there just too many variables to make anything other than a very broad brush forecast for a wide area?

I'm not saying I'm right btw, just how it seemed to me.

 

 

I wouldn't bother using data to rubbish Ian's claims; he has his agenda and he'll jolly well stick to it.

 

Yeah, you're right Saint. Trying to reason with Ian is a total waste of time; that first became obvious to me about a decade ago.

When one has an agenda and continually posts comments which attempt to further that agenda, rationality goes out of the window completely and contructive debate with them becomes pointless.

 

---

 

One thing I noticed while following the detailed explanations of the teleconnections involved in long range forecasts both here and over on Netweather last winter is that there seemed to be so many potential variables that I'm not entirely convinced of the point.

It seemed to be that present conditions would give rise to such an outcome in 3/4 weeks time. Fair enough but it usually seemed that a week later those present conditions,  or at least one of those factors driving those present conditions had changed so considerably as to totally alter the outlook.

No disrespect to any long range forecasters because I really admire the knowledge and work that goes in to those forecasts and I really enjoy reading them. Are there just too many variables to make anything other than a very broad brush forecast for a wide area?

I'm not saying I'm right btw, just how it seemed to me.

I think the best we will ever do is identify conditions which raise or lower the probability of deviations from your standard weather over a 3-month period. 

The problem is how few of those there seem to be. Being at the bottom of a solar cycle, or better still, just coming out of one, seems to increase the chance of a cold winter. So does an easterly quasi-biennial oscillation. But neither condition is any sort of guarantee, and you can get memorable winter spells without either. So in neither case can you actually make a forecast in the sense of a chart with dates and symbols.

All the other factors seem far more problematic to me. For one thing, it's a question of how they react with one another which makes it all so complicated and open to variation. And then there's the corollary that, as you say, the interactions result in those factors suddenly flipping the other way a few weeks down the line.

There ought to be some worth in studying sea-surface temperature anomalies, simply because bodies of water take a fair length of time to lose heat over the winter. But by the end of the winter even those anomalies could have flipped, as atmospheric patterns above, and ocean circulations below, do their stuff.

It doesn't help that we have so few years of data to work from. At the most about 150 separate years of global measurements, and a chunk of that, from before the satellite era, is guessed, I mean estimated. Anyone who's ever tried to identify patterns or analogies, or define what constitutes "average conditions", from past results in any field will tell you that 150 is a hopelessly small sample to work from. 

Give us another 1,000 winters and then we might be able to start drawing firm conclusions about how seasons will pan out in advance.

---

 

I think the best we will ever do is identify conditions which raise or lower the probability of deviations from your standard weather over a 3-month period. 

 

Bang on the money. That's the mantra to repeat whenever looking at these kind of seasonal forecasts. The Met Office always present them in this probabilistic fashion.

---

I have been wondering why most of the seasonal models are going for a strongly NAO+ winter, what is driving that.

I suppose the thing is, even with none of the large scale drivers favouring a mild winter i.e. no strong nino or nina, no WQBO, no strong solar activity etc., it's still possible for us to get a mild winter! Just the probabilites and shades of grey shift about.

---

 

 

I wouldn't bother using data to rubbish Ian's claims; he has his agenda and he'll jolly well stick to it.

 

 

Well I always prefer the stats, if you take the winter months of 87/88,88/89,97/98,98/99,07/08 & 08/09 then you have 18 winter months - 16 were above average, many crazily so. 

 

---

 

One thing I noticed while following the detailed explanations of the teleconnections involved in long range forecasts both here and over on Netweather last winter is that there seemed to be so many potential variables that I'm not entirely convinced of the point.

It seemed to be that present conditions would give rise to such an outcome in 3/4 weeks time. Fair enough but it usually seemed that a week later those present conditions,  or at least one of those factors driving those present conditions had changed so considerably as to totally alter the outlook.

No disrespect to any long range forecasters because I really admire the knowledge and work that goes in to those forecasts and I really enjoy reading them. Are there just too many variables to make anything other than a very broad brush forecast for a wide area?

I'm not saying I'm right btw, just how it seemed to me.

Last winter certainly looked difficult and I didn't have the time or the confidence in the variables to issue one last year. 15/16 was a penalty kick, especially when taken in the context of the modern era.

I don't know if the same guys are having a go at the Net Weather LRF, it's been a disaster for them now for a number of years, unfortunately the pressure is on whoever does it to at least provide something 'interesting'.

 

---

Being a lurker here is certainly educational and at times very entertaining!!

From what's been posted on here so far I'd say the elephant in the room is the AMO.

The propounded higher occurrences of milder winters in more recent times is compared against, what might be considered by some, a run of untypical colder ones in the 80's. The timing of the perceived 'switch' does seem to have some correlation with what has been published regarding the AMO - all interesting stuff.

The Stolen Snowman view for what it's worth is that we will indeed see a run of colder winters again one day - but not just yet. 

 

---

 

 

Well I always prefer the stats, if you take the winter months of 87/88,88/89,97/98,98/99,07/08 & 08/09 then you have 18 winter months - 16 were above average, many crazily so. 

 

Can you quote the CET figures for all three months from each of those winters Ian to prove this?

I do remember February '09 turning much milder after a cold and wintry opening, but I doubt it would have been enough to have seen it finish significantly above average.

---

Further to the above, I seem to recall that Ian, in his 2009/09 winter forecast, predicted a mega-mild January for 2009 and forecasted a pretty mild winter overall for that season. The prediction of an exceptionally mild January failed to materialise and that winter overall turned out, if I'm not mistaken, to be the first slightly colder than average winter for quite a few years. And no-one needs me to remind them of the events of the following two winters!

---

 

They might actually get it right for once, their last two winter forecasts for this side of the world have been total crap.

 

 

Well it must be then quite easy to forecast this winter and just predict opposite of what they are forecasting.

In any case, you can still have a mild winter with above average temperatures and get a few very cold and snowy shots which would leave you feeling we had a nice and snowy winter after all these years. Therefore, I would not read much into these forecasts.

---

 

 

Well it must be then quite easy to forecast this winter and just predict opposite of what they are forecasting.

In any case, you can still have a mild winter with above average temperatures and get a few very cold and snowy shots which would leave you feeling we had a nice and snowy winter after all these years. Therefore, I would not read much into these forecasts.

Yes true - last year was similar they said Stormy'

When winter 2016/17 was drier than average cross the UK. - also ntoe it says stormy first? does this mean December or what happens after the first?

This accuweather forecast set up for the UK looks typical of El Nino type pattern. So despite the easterly QBO and , solar minimum, fewer sunspots, increased Atlantic hurricane activity as well as a potential La Nina...I do wonder what it will take to get cold wintry weather in winter!?

---

 

 

Well I always prefer the stats, if you take the winter months of 87/88,88/89,97/98,98/99,07/08 & 08/09 then you have 18 winter months - 16 were above average, many crazily so. 

 

To me, the only winter month that you can say that was crazily above average was December 2015.

Others may have been well above average but they did fall within the bounds of past records.

December 2015 was 1.6C above anything that was ever recorded for December. 

And that's an interesting point, with all these mild winter months, only once has a CET record fallen to a winter month and that was December 2015 and this is very recent.

We've had two spring months: April 07 and 11, two summer months: July 06 and Aug 95, four autumn months: Sep 06, Oct 95 and 01, Nov 94, all breaking CET records.

-------------------------------------------------------------------

 

 

 

 

---

 

Can you quote the CET figures for all three months from each of those winters Ian to prove this?

I do remember February '09 turning much milder after a cold and wintry opening, but I doubt it would have been enough to have seen it finish significantly above average.

 

David - 

 

87/88   5.91,5.44,4.88

88/89   7.42,6.16,5.92

97/98   5.84,5.2,7.4

98/99   5.45,5.46,5.4

07/08   5.06,6.72,5.39

08/09   3.59,3.16,4.19

From the above, I would class above 5.5 as 'very mild' which covers 7 of the above months. with many others close to that figure. Outside of the modern era going back well into the 20th century, such months are a rarity rather than being commonplace.

---