Winter 2012 / 2013 Part 2

[Excitable_Weatherwatcher]

there is some very interesting weather this month. the teps are up and down and there is a chance of thunderstorms.

[johnholmes]

Indeed already had 2 hours on the white stuff this morning, amazing what you get in parts of Yorkshire!

[cheeky_monkey]

You know I personally feel we should discuss previous winters on this thread rather than the coming winter. This is because its too early for anyone to make a forecast. Maybe in another 4 weeks time should we return to discussing this coming winter.

I thought that is why there is a history section on this forum to discuss things last past winters etc etc

[Barry12]

• Penzance will have very little snow.
  

picture 6

http://news.bbc.co.u...000/9251626.stm

Edited October 1, 2012 by Barry95

[Weather-history]

Michael Fish couldn't believe this midday temperature for the middle of February, it was that remarkable that Peter Cockcroft wanted to make cameo appearances during the broadcast. The dreaded Bartlett high although I thought it was a lovely spell when the sun came out.

[Bishop Brennan]

Michael Fish couldn't believe this midday temperature for the middle of February, it was that remarkable that Peter Cockcroft wanted to make cameo appearances during the broadcast. The dreaded Bartlett high although I thought it was a lovely spell when the sun came out.

The Scots must have been pretty brassed off that day - heavy rain & strong winds whilst the English were basking under the warm sun!

Bish

[chionomaniac]

Michael Fish couldn't believe this midday temperature for the middle of February, it was that remarkable that Peter Cockcroft wanted to make cameo appearances during the broadcast. The dreaded Bartlett high although I thought it was a lovely spell when the sun came out.

UUUURRRGGHHH!!!

[snowfluff]

UUUURRRGGHHH!!!

Yes, perhaps a green tie would have been better

[garbagebags]

The Scots must have been pretty brassed off that day - heavy rain & strong winds whilst the English were basking under the warm sun!

Bish

Oh we are used to it

[BLAST FROM THE PAST]

Been working on my thoughts for winter 12/13. Particularly looking at this 'new' thing re low ice and southerly jet pattern. Well in 0708 winter we certainly didn't have a southerly jet pattern, although we were perilously close to a major Siberian HP blast which at last knockings sank SE away from us. PDO and pertubation cycles were same then as now, record low ice then as now....so what's different this year?

Food for thought

BFTP

[SMU]

Evening All-

Nice blog from Matt- very good - to support people learning about the winter factors.

I wanted to put some commentary on what I think she be presented in a winter outlook - this I think correlates to our current level of understanding-

By this I mean we have a lot of teleconnection data coming online & each year another 'fluid' dynamic seems to enter the fray.- I hear a lot of people saying that its the new 'fad'-

Last few years have been the stratosphere & its impacts, this year is arctic ice, in years prior to that was ENSO & the Jet in the Mix- & that old chestnut- the even larger teapot-

Bottom line is this-

Our sample size & dataset is to thin to create a representative analogue.- To create and predict a good match to cover the following-

Arctic Ice,

PDO,

QBO,

ENSO,

Solar flux,

Stratosphere,

Ozone levels

SST

I think we would need at least 500 years worth of consistent reporting data,- even if we rationalise each Variable to a low level tangable rating of Strong, Neutral & Weak signal I bet on all of the above with the data we have you might get one match that had a strong Signal for the winter forecast,

Also with forecasting its about attributing 'weighting' to the signals & their apparent influence to the OVERALL hemispheric pattern-

Im sure with enough data the scientific community could complete an algorithm whereby each signal could be fed in & generate a 'type' of outcome. The current CFS is what we got but sadly its very limited on capability & data capture -

anyway-where we are in 2012 that 500 year dataset doesnt exist so we have to go on the SMALL dataset we have & try to decipher what that means in the mix of other factors-

it ALL boils down to 2 things- which is especially important to the 55 N line around the mid lattitudes-

* Net overall projected anomaly of the mean zonal Wind.- IE- are we facing blocking scenarios or zonal conditions-

* Focal points of where the rossby waves are likely to show the greatest wavelength against the norm & in relashon to that where the zones of strong CAA & WAA are going to be - as these areas will have the biggest +VE & -VE temp anomalies V the norm.

What could & should a Winter forecast offer-

If you are reading about specific events around specific dates then at worst its a hopecast, at best a forecaster has reviewed the historic datelines around winter - in terms of dates that generally have colder periods & so on-

Forget dates & events- We dont have that capability yet to forecast.

A lot of these are sensationalist forecasts & headline grabbers-

Interestingly in the last 10 years how many sensational winter months have we had( in terms of extreme cold temps V the norm- maybe 2 or 3 ( AKA Dec 2010 )- & yet every year we are subjected to forecasts of extreme weather-

Sorry people, we have a run rate of probably 2 in 30 months of severe 'events' but a forecast run rate of 1 in 3-

IE 1 extreme month from every winter forecast from all these people that appear- its laughable.

Where are we at -

If I had the time I would be dividing my forecast into 4 clear sections-

* data & Indices-

Aknowledgment of data used, assumptions made MUST be communicated & clearly indicated in a medium where its clear what type of winter you are expecting-

I wouldnt dwell on this info & spend to long on it as a chunk of this is just available data to cascade & present-

Importantly the most UP TO DATE data is required- therefor a forecast issued prior to November the 15th is not going to be as accurate to the degree as one issued on Nov the 30th as the rate of change that can occur in this time period can be a big influence over the winter- especially with respect to the stratosphere.

* Projected longwave pattern-

Get the 'MEAN' longwave pattern right & your 95% of the way there- Especially when forecasting 'non typical' weather types-

If you go for a flat longwave pattern with minimal waves in the jet you are essentially saying that the weather type & pattern is essentially the norm with minimal variance-

On the 500MB plot for the NH there would be minimal Height anomaly zones-

Vice versa ( & interesting in terms of all these sensational forecasts ) is forecasting waves in the jet for the season- & how that correlates to the UK-

the MAX stationary waves at any one time for the NH is usually 4, however across a season its very unlikely that ALL 4 will show as a strong positive anomaly- its usually just 2 teleconnecting Highs that appear on the plots.-

So outlining these shows hemispherically where the 2 biggest areas of below average will occur & of course where the 2 warmest areas will be-

Also you will have smaller zones of anomaly where the 500 height anomaly signal is weaker.

dividing these up into 3 months is TOUGH for Nov 15th-

I would say a 6 week anomaly & a 7th-12th week preliminary view would be the way forward on the 15th Nov- with an update on the first / second week of Jan for the second 6 weeks of winter.

* Seasonal Temp Probabililty :

I firmly believe that rather that going for irrational CET estimates is not the way to go- we should break the forecast down to degrees of probability of a particular range of temps-

For example-

With all the background signals & the projected longwave anomaly you then present a 'signal' of pattern- I would have a temp range like this

-3 to -1c anomaly ( strong COLD) -1 c to +1 ( Neutral ) +1 to +3 ( Strong Warm)

The strength of your signal dictates the probability scale - I think this is VERY similar to how the met office operate with percentiles. & the range above could be broken down again & have weak cold inserted & weak warm as well-

* Rain days V the Norm ---- NOT accumulated PPN V the Norm.

I think this is where my thinking is slightly different to most ( if I cast my brief flirtation with trading energy & weather) Its better to predict rain days V the norm as opposed to predict totals V the norm-

The reason I say this is the total rain for one area can be delivered in one short heavy burst of heavy rain in an amplified flow or consistently short amounts accumulating over a period of time from a semi zonal flow.

For me being a numbers man- statistically we must measure ave rain days per month-

if you are producing a forecast & a particular prevailing weather type it will then stack up that from this you can forecast the amount of rain days anomaly V the norm-

So for example - your insight says strong atlantic high is the form horse- Rain days will be reduced-

I would be dividing this up into some sort of range V the norm- with the neutral zone being average zonal conditions & the average number of rain days- then signal for blocked conditions correlates to negative days & highly zonal conditions = positive days.

In Summary-

I think I would need need to invest 40-50 hours to produce a semi authentic forecast these days, all based on probability,- the difficulty being that theres not enough data on the wildcards like the arctic ice & Stratosphere......

best regards

Steve

[Coast]

So are we still saying Steve it's just too early to even make a stab at things?

[Thundery wintry showers]

picture 6

http://news.bbc.co.u...000/9251626.stm

Looks like my Penzance prediction isn't guaranteed to come off then! The easterly outbreak of the 11th-14th January 1987 was the coldest of the twentieth century, but it does show that Penzance can get sizeable snowfalls given an exceptionally cold easterly coming off the English Channel- I do recall reports of thundersnow from the extreme south-west in late November/early December 2010 also. It does require an exceptionally cold easterly type though.

The Scots must have been pretty brassed off that day - heavy rain & strong winds whilst the English were basking under the warm sun!

Bish

I remember having similar feelings in Tyne and Wear on the 9th-12th February 1998 when we were hearing about the warm sunshine that was extending up from the south-east, while stuck under dull, dry, mild conditions with a strong south-westerly wind. The sunshine extended further north on the 13th to cover Tyne and Wear, and I went over to Carlisle that day where it was also generally sunny, but Lancaster recorded just 0.2 hours of sunshine indicating that some parts of NW England stayed cloudy.

[SMU]

So are we still saying Steve it's just too early to even make a stab at things?

As of today the strong positives ( if you want a colder than ave weather type for winter ) Is the SST atlantic profile & the feedback loop from the arctic to the thermal gradient.

Easterly QBO & Negative PDO ( remember the 1962 esque analogue)

lowish solar output-

The 'to early to tells' are the strength of ENSO & the temp of the stratosphere-

As it stands I would be inclined to forecast a higher probability of a colder winter than a lower one-

However this can all be wiped out by a sudden variance in the above 2 indicies- which have the time to manifest before Winter

S

[chionomaniac]

Couldn't agree more with Steve,

Would also add late Autumn Eurasion snow cover into the mix.

[SMU]

Cheers

[gottolovethisweather]

Couldn't agree more with Steve,

Would also add late Autumn Eurasion snow cover into the mix.

Ok, we are two whole months away from proper winter but the following thread is still worth watching and learning from.

http://forum.netweat...00#entry2376270

Thanks again for your input, chaps.

Edited October 1, 2012 by gottolovethisweather

[SMU]

Just to add on above-

In a prelim I would box off the first 6 weeks of winter as-

a signal for weak cold weather in my mentioned range (poss -0.5 to -1.5) - which also has consistencies with this years CET pattern- & a signal for below average rain days-

The next update should be ideally Nov 01- where trending will be emerging from the other players--

S

Edited October 1, 2012 by Steve Murr

[feb1991blizzard]

Just like to thank people who have contributed to this thread, it has worked very well in my opinion because it has the knowledge and educational value of chionomaniac, Steve Murr, the forecast team and many more but also is relaxed in terms of moderation and therefore has great flow, thanks to Weather-History for his excellent videos as well.

[SP1986]

Reading this thread has tripled my scepticism.. not meaning to be a party pooper, but this is way way way too early.

However the historical information is as ever, very interesting!

I'm looking forward to winter whatever it throws up, as ever there should be something for everyone..! A few thunderstorms wouldn't go amiss as there has been a distinct lack of them this year here. As for snow, it would be nice, but I can take it or leave it, if some interesting weather prevails instead (ie winter storms etc etc)

[lorenzo]

ENSOitself is so finely balanced the 01 October update from CPC gives nothing away.

Models looking at Nino conditions through to Feb, CFSv2 counters this somewhat hovering at neutral to very weak Nino.

SSTs in CP have taken a marked fall back from the previous warming out in the last update, the last 4 weeks slide shows a cooling trend.

Another paper archived on the ECMWF Seminar pages looked at the influence of Central Pacific vs Eastern Based Pacific El Nino events on NH Winter temp profiles. So even within the ENSO should it be weak or building over November / December, there are analogs within the analogs ! Minefield !

Edited October 1, 2012 by lorenzo

[bobbydog]

Ok, we are two whole months away from proper winter but the following thread is still worth watching and learning from.

http://forum.netweat...00#entry2376270

Thanks again for your input, chaps.

don't forget though, its within these two months that the factors which will decide our winter weather, will reveal themselves...

some already are....

[gottolovethisweather]

Guys and gals, you might be interested in my interpretation of the coming couple of weeks below for locations due East, Northeast and Northwest.

It's all just a bit of fun and games, for the start of October really.

http://forum.netweat...20#entry2376377

Edited October 1, 2012 by gottolovethisweather

[cyclonic happiness]

Ok, we are two whole months away from proper winter but the following thread is still worth watching and learning from.

http://forum.netweat...00#entry2376270

Thanks again for your input, chaps.

I'm still convinced that following the developing snow and ice will be a great pointer to where the snow and cold is likely to come from and how much.

Its like a visible record or tally that is spread across the globe.

We need a white scandi and a green southwest russia for early cold.

Im so glad i started that thread, its been awesome watching it grow , thanks guys ;-)

[feb1991blizzard]

I disagree with this. Obviously like all forecasters he gets it wrong sometimes but I have been following TWO's forecasts for ten years and in my opinion he doesn't have a bad track record.

I agree, probably does not give as much detail as GP as to his reasoning but to be fair he has been pretty accurate the last 2 winters and went against all the evidence and was pretty much alone in forecasting a mild second half of feb but turned out to be correct.