The Seasonal Forecast Thread

[summer blizzard]

Good to hear roger and with a probable el Nino developing anything is possible. Think I already am liking this years weather a lot better in comparison to last year as it feels a lot mote how it should be.

http://www.liveweatherblogs.com/index.php?option=com_community&view=groups&task=viewdiscussion&groupid=8&topicid=2639&Itemid=179

Neutral most likely however so long as we are the warm side of neutral, with a +QBO it may be good enough.

I remain optimistic for summer.

[VillagePlank]

Nov: Fore 8.26C, Actl 6.8C, Err: +1.46C

Dec: Fore 4.40C, Actl 4.8C, Err: -0.40C

Jan: Fore 2.74C, Actl 3.5C, Err: -0.76C

My Feb forecast is 3.68C

If Feb CET does turn out to be 3.0C then I'm -0.68C out, which gives a total absolute error of 0.61C/month. Not too bad for my first, serious, effort, I don't think. Also, it's clear that my method pathologically underestimates the temperature, so, for instance, if I simply added 0.5C to the forecast for each month my average absolute monthly error would only be 0.18C which implies that somewhere I have a hidden constant as 0.18C is very likely to be within the margin of noise.

I'll wait until March to do a full analysis. Original methods summary, here.

Edited February 26, 2013 by Boar Wrinklestorm

[VillagePlank]

Still no comments? I recall that everyone said that all the forecasts are bust. Mine isn't, much. Anyone have anything to say? I think JH said it all ....

[VillagePlank]

lol

[johnholmes]

Still no comments? I recall that everyone said that all the forecasts are bust. Mine isn't, much. Anyone have anything to say? I think JH said it all ....

go on tell me what I said BW, I certainly cannot remember, probably just as well.

You seem pleased with your result but if 1C is okay that is equivalent to about 30% or is my maths as bad as my memory?

[cheeky_monkey]

Still no comments? I recall that everyone said that all the forecasts are bust. Mine isn't, much. Anyone have anything to say? I think JH said it all ....

i was going to make a comment that if you take the winter as a whole and take the mean averages of your forecast values and actuals then as a season your were only -0.15c out...there how is that?

[Iceberg]

um....to comment or not to comment that is the question.

How about a nice general comment that's not about any particular forecast but maybe a more general thought.

LRFs fail not due to hardwork or dedication(unfortunately hardwork and dedication dont make things right), but simply because any ability to forecast the major building blocks for the hemisphere, doesn't mean that you get the weather right for a country like the UK. This is where i completely agree with Pete even when the major blocks are known only 5-10 days away it takes super computers (not your home PC) to try and predict the weather for the uk and even then they get it wrong.

I really dont see how any LRF can play with anything but vague probabilities 1-2 months away. When LRF's try and predict specific events x weeks away...( I'll leave that for others to finish).

[A Winter's Tale]

AWT'S SPRING FORECAST 2013

MARCH...overall quite close to average, with a settled start, followed by transient and quite changeable weather with temperatures overall at around average. Temperatures 0.5C below to 0.5C above average. CET 7.0C. Rainfall 70% in south and east and 85-90% in north and west. Possibly some wintry interludes during the middle third of the month, much of the wintry precipitation confinded to higher ground in Scotland. Sunshine 0-5% above average.

APRIL...overall close to average, but a month of contrasts with some cold outbreaks at times and some warmer spells developing (especially during the second half of the month). Most of the times, quite normal showery April weather with less frontal activity and some settled spells. Temperatures 0.25 below to 0.5 above average. CET 8.6C. Rainfall 60% in south east and 75-85% in north west. Some wintry showers on high ground in north west at times aswell as some hail and thunder. Sunshine 5-10% above average.

MAY...overall warmer than average with some pleasant, settled and warm periods, combined with shorter spells of more unsettled, showery and cool spells with occasional frontal activity and with a threat of thundery outbreaks. CET 12.5C. Rainfall 60% in south east (some thundery downpours) and 75% in north west. Temperatures 0 to 1.0C above average. Sunshine 5-15% above average.

MARCH - MORE DETAILED FORECAST

First week: Quite chilly and settled to begin with and some cloudy periods for some. Turning slightly milder later with some rain, especially in western areas.
Second week: Staying changeable with fluctuating spells of high pressure influencing from the south and low pressure to the north. Mostly mild, chilly on some clear nights.
Third week: A mostly unsettled period with temperatures closer to average or quite chilly. Some showery spells, changeable conditions but quite bright at times
Final week: Turning more settled, possibly cool and cloudy to begin with before becoming brighter later, some frost at night.

[Bobd29]

AWT'S SPRING FORECAST 2013

MARCH...overall quite close to average, with a settled start, followed by transient and quite changeable weather with temperatures overall at around average. Temperatures 0.5C below to 0.5C above average. CET 7.0C. Rainfall 70% in south and east and 85-90% in north and west. Possibly some wintry interludes during the middle third of the month, much of the wintry precipitation confinded to higher ground in Scotland. Sunshine 0-5% above average.

APRIL...overall close to average, but a month of contrasts with some cold outbreaks at times and some warmer spells developing (especially during the second half of the month). Most of the times, quite normal showery April weather with less frontal activity and some settled spells. Temperatures 0.25 below to 0.5 above average. CET 8.6C. Rainfall 60% in south east and 75-85% in north west. Some wintry showers on high ground in north west at times aswell as some hail and thunder. Sunshine 5-10% above average.

MAY...overall warmer than average with some pleasant, settled and warm periods, combined with shorter spells of more unsettled, showery and cool spells with occasional frontal activity and with a threat of thundery outbreaks. CET 12.5C. Rainfall 60% in south east (some thundery downpours) and 75% in north west. Temperatures 0 to 1.0C above average. Sunshine 5-15% above average.

MARCH - MORE DETAILED FORECAST

First week: Quite chilly and settled to begin with and some cloudy periods for some. Turning slightly milder later with some rain, especially in western areas.

Second week: Staying changeable with fluctuating spells of high pressure influencing from the south and low pressure to the north. Mostly mild, chilly on some clear nights.

Third week: A mostly unsettled period with temperatures closer to average or quite chilly. Some showery spells, changeable conditions but quite bright at times

Final week: Turning more settled, possibly cool and cloudy to begin with before becoming brighter later, some frost at night.

I'll take May thanks, like that one!

[Interitus]

Nov: Fore 8.26C, Actl 6.8C, Err: +1.46C

Dec: Fore 4.40C, Actl 4.8C, Err: -0.40C

Jan: Fore 2.74C, Actl 3.5C, Err: -0.76C

My Feb forecast is 3.68C

If Feb CET does turn out to be 3.0C then I'm -0.68C out, which gives a total absolute error of 0.61C/month. Not too bad for my first, serious, effort, I don't think. Also, it's clear that my method pathologically underestimates the temperature, so, for instance, if I simply added 0.5C to the forecast for each month my average absolute monthly error would only be 0.18C which implies that somewhere I have a hidden constant as 0.18C is very likely to be within the margin of noise.

I'll wait until March to do a full analysis. Original methods summary, here.

There appears to be a simple error in your calculations - if Feb is 3.0°C then you are +0.68 out, not -0.68. The absolute error margin stays the same, but idea of adding 0.5°C to your winter figure gives an average monthly error of 0.51°C instead of 0.18°C.

Still seems to be a creditable performance however as a measure of skill, purely using the 1961-90 climatology alone (and the assumed 3.0°C Feb) gives a lower average monthly error of 0.37°C

[Roger J Smith]

BW, your numerical forecasts did appear to improve on random chance -- as you probably know there are "tests of significance" but for a set of four or five monthly CET forecasts the sample is too small to draw very strong conclusions. Have you tried to predict trends on a finer scale than monthly, like half-monthly, weekly etc? That might establish a stronger test.

Do you predict any trends beside temperature (precip, pressure, wind direction etc)?

If you're using historical data and extracting cycles or trends, then you might want to examine your working hypothesis about the AGW signal. I am somewhat of a skeptic myself but I would acknowledge that past data have to be warmed up by about 0.5 C for whatever reason (in the immediate sense, the reason is probably higher ocean temperatures but that of course is postulated to be part of the AGW signal).

This past month managed to hide in plain sight as far as being a cold, blocked month -- the monthly CET will look rather bland years from now perhaps but I suspect that the very strong storm that hit the eastern U.S. managed to push into a blocking pattern with just enough energy to distort the circulation for 3-4 days and in a month as short as February that can make quite a difference in rank order finish. It wasn't really that big a weather event for the UK so much as in Ireland where it basically added another degree to their version of the CET by pushing the Atlantic back in for about 2-3 days.

[Methuselah]

It does seem that the winter was close to average, in pretty much every respect...

[mushymanrob]

i reckon that things will continue much the way it has this year... cold spells less cold spells alternating into may.

when we have had cold marches, the pattern seems to last until mid april at least.

[damianslaw]

It still feels very much like mid winter, im in full winter mode right now. This year has started off in very similiar vein to 1996, 2001. Whilst the springs of both years were cold, v cold in case of may 96, the summers were preety décent. Another dry quite cold winter followed by à cold blocked march was

2006, and we know how that summer panned out. Mins going back to the 3 cold marchés in 85, 86, 87, we saw 3 preety chilly summers.

[snowking]

I have just begun to take a look towards summer 2013

Some early observations. QBO wise, the best analogue period is 1978. The main reason for this being that we have now been under an east QBO phase for 20 months, and are now heading into positive (westerly) territory. 1977/78 is the next best match period I can find for this at 19 months. 1978 has some weak ENSO correlation too - a generally neutral stance (though the difference I am expecting between 2013 and 1978 is a slightly negative index as opposed to the slightly positive index that characterised '78).

Funnily enough 1962 (mentioned already here in jest more than anything) offers perhaps the best ENSO match in terms of my expectations of weakly negative, with some minor QBO support in terms of it being an increasing westerly

There are a number of other analogue years I have picked out, and these have produced composites consistent with those created by 62 and 78 (based around ENSO and QBO correlation), and I will go into these at a later date.

The overall picture offers a mean sceuro trough:

With a resultant negative affect on temperatures:

Though rather surprisingly, the precipitation anomaly remains around average on the whole:

As a side note, based upon these composites, the best month of the year may end up being October.

So, there's some preliminary insight. I expect it will take me another month or so before I can offer up anything conclusive.

SK

Edited March 29, 2013 by snowking

[Rollo]

Thanks for the info Snowking.

[Alex]

My Summer Forecast 2013 By Alex:

it's been a pretty cool Spring this year which has been very similar to Spring 2006. Will the summer follow a 2006 type? The answer is possibly! We have very different Synoptics to the last few years - interestingly this cool unsettled end to May has coincided with the last stratospheric warming and the warming out of the current nearby trough is likely soon - how soon leaves us is questionable! However I expect this to happen by at least the first ten days in June which will allow the Azores high to ridge across the UK initially into western areas and then across all areas. It will probably settle over Scandinavia or mainland Europe for a while with a hot period likely towards the end of June ESP in southeastern areas. The rest of the summer although unlikely to be a 2006 repeat is likely to be the warmest and sunniest summer since the famous 06 summer! However there will be unsettled spells more especially in August. July is the one to watch with a drier than average month quite consistently progged by the CFS v2 model. The fly in the ointment to my predictions will be whether the upstream patterns can continue to thwart the possibility of la nina conditions developing. If an active MJO continues along with positive angular momentum - then I don't see this an issue, however August is most at risk from a la nina type developing which would in turn mean a more unsettled cooler month.

Overall I predict a decent summer but probably not a memorable one.

For stats I am looking at a CET of around +0.5c above avg for 1981-2010 (June, July and August). Rainfall I expect to be varied with thunderstorms likely at times - however over England and Wales 90-95% can be expected with sunshine probably ending up 105% of average.

Add me on twitter@alexbweather - for more of my weather views!

Edited May 23, 2013 by Alex

[Roger J Smith]

Here's my outlook for the summer of 2013:

 

Summer of 2013 -- warm, dry, sunny with brief interruptionsIn general, I believe there will be considerable improvement of the weather pattern ahead, and some chance of prolonged warm, dry spells. At the very least, I think this will be one of the better summer seasons in recent years, and there's some chance of it being an exceptional summer.I have to respect the strength of the recent cold signal, which although not overly wet has maintained rainfalls at normal values which makes the season seem rather wet since the combination has led to reduced "evapo-transpiration." This signal may not just disappear on the first or even second instance of more settled weather under higher pressures. It may even try to dig in and fight back. So I want to temper the outlook by saying most likely a combination of warm, dry spells and cooler, unsettled spells, although I could see theoretically how the warm, dry signal could lock in and repel any returns to cool, unsettled weather for more than a few days here and there. The compromise for my forecast would be to say that warm, dry spells will dominate and there will be relatively brief interruptions from unsettled, cooler intervals.No particular timing was suggested by the research, which shows a general trend to above normal temperatures from June to August and into parts of September, so more or less the entire summer season.Since we have this situation of signals for above normal temperatures but a lingering trace of the long-wave trough, it might be expected that rainfall, even if it averages below normal, could include one or two heavy falls in brief episodes of severe storm activity. This would be more likely in Wales, the Midlands, and parts of south central England, and for readers in Ireland, the inland northern half of the Republic as well as inland Northern Ireland. Drought is also possible given the strength of the warm signal at the foundation of this forecast, and that would be most likely to develop in various parts of England and Wales (including some locations prone to severe storms which alone would not be enough to overcome drought), and for Irish readers, the southern third of the country and possibly around Dublin.Anyway, that's the outlook, I hope that the actual results are not too far removed from this warm, sunny and relatively dry prediction. Just to place some numbers for later verification, I will say that June, July and August will each average 1.2 degrees above long-term normals, rainfall will be 70 to 90 per cent of normal, and sunshine will be 10-30 per cent above normal values

 

I could mention anecdotally that the current atmospheric set-up is not very different from this stage of 1976 and I have the feeling that very strong ridge-building could be ahead to bring possible heat wave conditions at times in later June or July.

[tadpole]

Here's my outlook for the summer of 2013:

 

Summer of 2013 -- warm, dry, sunny with brief interruptionsIn general, I believe there will be considerable improvement of the weather pattern ahead, and some chance of prolonged warm, dry spells. At the very least, I think this will be one of the better summer seasons in recent years, and there's some chance of it being an exceptional summer. 

 

Gosh I hope you're right. I'm sick to the back teeth of this cold gloomy weather. 

[weather_it_will]

This summer will be average at best in terms of warmth.

 

It is clear that our summers are now owned by the energy companies.

 

If there were good summers and mild winters in the UK, the energy companies would be looking for takeover bids. It's not going to happen.

 

Have you ever wondered why the shops are heavily discounting barbecue equipment in May?  It's because they know it's going to be a rubbish summer.

[Purga]

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

"Met Office 3-month Outlook

Period: June – August 2013 Issue date: 24.05.13

The forecast presented here is for June and the average of the June-July-August period for the United Kingdom as a whole.

This forecast is based on information from observations, several numerical models and expert judgement.

 

 

SUMMARY - TEMPERATURE:

For June, below-average UK-mean temperatures are more likely than above-average. For June-July-August as a whole,

near-to-below-average temperatures are most probable. Sea surface temperatures of the surrounding seas are below-average,

increasing the probability of below-average temperatures across the UK.

Overall, the probability that the UK-mean temperature for June-July-August will fall into the warmest of our five categories is

15% and the probability

that it will fall into

the coldest of our five categories is 20% (the 1981-2010 probability for each of these

 

SUMMARY - PRECIPITATION:

For June, there is a large degree of uncertainty, but on balance above-average rainfall is more likely than below-average. For the

June-July-August period as a whole above-average rainfall is also more probable than below-average rainfall. There is less

confidence in the signal emerging in the forecast for rainfall than for temperature.

The probability that UK precipitation for June-July-August will fall into the driest of our five categories is around 15% and the

probability that it will fall into the wettest category is around 20% (the 1981-2010 probability for each of these categories is 20%)"

 

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

 

There is still a 'probability' that it will turn out warm and dry but the odds and expert opinion appears to be against it.

 

I've posted in this thread as it's the most appropriate and relevant one and kept it out of the more 'popular' ones due to the over emotional reactions usually encountered.

 

Edited June 4, 2013 by Paul

[summer blizzard]

Is the Netweather forecast out today?

[summer blizzard]

Attempting a more experimental forecast method...

 

QBO (Positive)

 

2008

2004

1999

1997

1995

1990

1985

 

MEI (Neutral-positive)

 

2012

2005

2004

2002

2001

1995

1994

1991

1990

1980

1979

1969

1965

1957

1953

 

PDO (Neutral)

 

1989

1959

1957

 

GLAMM (Neutral-positive)

 

2004

1990

1988

1987

1981

1978

 

April-May

 

1990

2004

 

1957

1995

 

 

abb

baa

baa

baa

aba

aba

 

June: 50% chance above/below average (range is 13.6C-15.3C)

July: 50% chance above/below average (range is 15.8C-16.9C)

August: 84% chance above average (range is 17.6C-18.0C)

Edited June 15, 2013 by summer blizzard

[snowking]

I have just started putting together some new analogue years for the preliminary Autumn forecast - the two years that still carry over from the previous years are 1955 and 1962

 

You may remember me mentioning from the analogue years for the summer that there was a very strong signal that October could end up being the best month of the year. Well, heres the latest result....

 

 

SK

[summer blizzard]

June CET was 13.6C which is within my range so pleased. Not so sure that i can rescue July though.