virtualsphere

The two I remember well are snow showers in 94/95 and the very cold new year in 01/02. Agree recent new years have been very bland in comparison with the 90s - 2010 period. The one I remember in the last decade was 2013-14, being woken by thunder overnight and then our chimney cap being blown away by a gale on new year's morning! I wonder if we'll end up with your option c this winter? Looks promising at this point, a bit like 12/13 but running a few weeks earlier.

5.1C and 103mm please

There's some 10mb charts for 1963 in this paper - although it's behind a paywall, you can currently create a free account to read a limited number of articles per month. JSTOR: Access Check WWW.JSTOR.ORG JSTOR is a digital library of academic journals, books, and primary sources. As for 1947 I notice that this paper looking at historical SSW onset dates through reanalysis doesn't give anything for winter 1946-47 (Dec '47 is mentioned though): https://www.mdpi.com/2073-4433/10/9/519/pdf

Always a good sign when the vortex resembles a ninja turtle.

The one that surprises me there is 2012. I don't remember it being that cold: I recall there was an easterly modelled which never materialised then a wet and windy spell which must have been after 16th looking at your data.

I don't think I can quote a post from the old thread but I like the ECM animation posted by @IDO here: https://www.netweather.tv/forum/topic/94764-model-output-discussion-winter-has-arrived/?do=findComment&comment=4313159 Ignore our locale and take a look at the deep purples of the vortex and their direction of travel in this animation, note that at the end around 16-17 Dec the lobe that develops to the west of Greenland over N Canada is actually spinning back west towards Alaska - for me a positive trend away from any potential zonal outlook further down the line.

A 10.4C difference in the CET for the same month within 5 years between 2010 and 2015 is quite something. To put that in perspective, if we had a July that was 10.4C warmer than the coldest in the series, we'd end up with 23.8C! There is a similar gap between the warmest and coldest Januaries but December 2015 was awful here though: the wettest month in the last decade by quite some margin and zero dry days. Much preferred 2010!

As others have pointed out, worth remembering that where we are going into December doesn't always represent how the winter turns out overall: However the trough dropping through / close to the UK that's currently being modelled is a positive sign looking at precursors to some recent cold spells: The big difference in the first three charts from 1988, 1989 and Dec 2013 is the angry secondary lobe of the vortex building over NE America. In the last three charts this isn't present, or is further west over North America with a smaller main vortex. The positive thing about the current modelling to me is that upstream the vortex looks more like the latter examples:

Here's a pic of the power network in our area taken in a nearby village in Jan 2013, which is also when we last had any meaningful settling snow. Any outages are usually minor and are repaired within an hour. The last few I remember didn't coincide with severe weather. Wind gusts of 40mph are fairly frequent, 50+ are rarer but it does happen. Our chimney blew down in Jan 2014 but the power remained on! Not sure, but it does seem better recently. I'm thinking of intermittent very short power outages which were fairly common in icy weather around 12 or more years ago. Looking back more recently we didn't experience any in the cold winters of 2010 and 2013 so I'm assuming it might simply be some rogue equipment which has since been fixed.

Conversely another wet and very dull one here. I don't think we'll beat October for rainfall but we will come fairly close. At the moment I've only recorded 3 dry days in Nov compared to 5 in Oct: not out of the ordinary for this time of the year in wet west Wales but we did manage 10 last Nov and even 9 in Nov 2018 despite recording an impressive 290mm rainfall that month! Overall here in 2020 we're now past the annual rainfall total for the whole years in 2013, 2016, 2017 and 2018. Unless Dec is very dry we'll probably end up wetter than last year also. I'm looking forward to a pattern change later this week and a drier interlude!

Well at some point that huge gap between 1974 and 2015 will be breached, so let's go with 8.5C and 94mm. And hope I'm wrong

The Feb 2007 event was completely unexpected as it was forecast to quickly turn to rain. It still counts as the deepest single fall I've experienced here. Only other ones I can think of are Nov 2005 which wasn't completely unexpected but the amount of lying snow took my by surprise, and Jan 2013 when the snow also lasted much longer than forecast. BBC NEWS | Wales | Road gridlock after surprise snow NEWS.BBC.CO.UK

This made me think - here in SW Wales most of the power network is on overhead cables rather than overground and considering the strength of the winds we get in this area it is remarkably robust. Most of our power cuts haven't been weather related though we used to get them more frequently in cold weather.

Not disputing this, but thought I'd have a look at the data in relation to difference between ENSO regions 1+2 and 3.4 in Nina winters and what that means for the CET. In this chart anything below the horizontal 0 line has a greater negative anomaly in the 3.4 region for Dec - Feb; anything above has a greater negative anomaly in the 1+2 region. The further below / above the line, the greater the difference. We should see CP La Ninas below the line and there are still quite a few colder examples, including in more recent years 2010-11 and 2008-09. There are also some very warm ones of course like 1988-89, and the extremes do trend warmer, but it shows that anything's still possible in this setup.

The only monthly values below -2C in the 3.4 region I can see are Jan 1890, Feb 1890, Nov 1973, Dec 1973, Jan 1974, Oct 1988, Nov 1988.

You may well be right re. strong, will leave that to the experts The coldest Nina winters especially more recently are indeed in the 0 to -1 range - but there are actually very few examples below -1.5 and they are pretty evenly spread across the CET range from 1916/17 at a chilly 1.5C to 1988/89 at 6.5C. There's only 3 winters which average Oct - Dec periods with negative ENSO anomalies lower than -1.5 in both 3.4 and 1+2 regions, 1916/17, 1955/56 and 2007/08 and these return CETs of 1.5, 2.9 and 5.6, so that's 2 out of 3 that are sub 3C. There's a lot of other factors at play, no doubt, but looking at these has made me more confident that we can still get cold weather with strong(ish) La Nina than if it was strong El Nino - there are no examples of a winter CET below 3.5C following a positive anomaly greater than 1.5 in either 1+2 or 3.4 regions for the entire dataset (and there are more of them)! Of course worth mentioning this is averaged over 3 months which is likely to reduce the strength of any individual monthly anomalies.

Strong Nina might not be so bad for a colder winter when you look further back in the data. Here are some more visualisations using the historical datasets @SqueakheartLW posted above, and when you go back to 1870 there are quite a few examples of Oct-Dec sub -1 anomaly Ninas with sub 3C CETs for the following winter. First up, Nino 3.4 Oct - Dec average anomaly plotted against following winter CET with the shape indicating 3.4 anomaly (triangle positive, circle negative): And now Nino 1+2 Oct - Dec average anomaly, this time with the shape indicating 1+2 anomaly (again triangle positive, circle negative). Looks like we really don't want this to be strongly positive for colder winters!

That was probably me, I was trying to plot the ENSO data against the CET as mentioned in this post but the dataset I used only went back to 1950. I'll take a look when I get a minute, it will be interesting to use a longer period to see how the results come back, especially given some of the mention of 1917 and 1956. I did also look at the different ENSO regions but couldn't really find an effective way to plot this. Something to file for now and come back to! Talking of strong Nina events, this report gives some background and mentions briefly 1917 & 1955-56 although the main focus is on the effects of the 2010 Nina in Australia.

It will be interesting to see the regional variation in rainfall between the first and second halves of the month. Here in the west, the rainfall was fairly normal for the time of year at the start of the month but places further east and south seemed to get much more of a soaking. After the drier mid-month spell, this last week has turned very wet here and the models look to dump the heavier rainfall in the next few days over Wales and the NW too. A very wet month for many regions but delivered from two distinct spells with the impact in different areas.

While you'll almost certainly be right that we won't see cold like Dec 2010 this winter, it's worth remembering that someone might have said the same thing back in 1981 (coldest Dec since 1890) and it only took 29 years rather than 100. Similarly from a 'warm' perspective, it was unlikely April 2007's 11.2C CET would be exceeded as soon as 2011 even in a warming climate, especially since the May & June records have stood since the 1800s, but these things do seem to happen! We're certainly not overdue a sub-zero December but we're probably overdue a relatively cool anticyclonic one as you say. I have fond memories of some calm and frosty December weather in the 2000s - e.g. 2001, 2005, 2008, 2009 - early winter seems to have been much more unsettled since 2011 in general.

8.2C and 80mm please

How about this from 1978? Not suggesting we'll get a repeat of 78/19, but plenty of time for things to change at this stage.

Posted this in the ENSO thread but may be of interest here. This shows the mean winter CET for all years since 1951 (on the horizontal axis) and mean ENSO ONI Oceanic Niño Index data for the winter (on the vertical axis). Symbols indicate QBO data averaged for the winter months, circles are EQBO, squares are WQBO, filled are stronger values, unfilled are weaker, and an asterisk represents very weak / neutral values. The year shown is for January / February, so '2010' represents winter 2009/10. Taking 2010 as a bit of an outlier, this suggests the closer to ENSO neutral the better if it's a cold winter you are after. QBO also varies, with west and east QBO both associated with warm and cold winters. Obviously this takes a rather broad brush approach and the interplay between these variables and others are likely to be a key factor as Mike mentions above. Solar is an interesting one: certainly recent cold periods have generally been associated with low solar activity - 1995/6, 2009, 2010, 2018 - also further back 1963 and 1986,87 - but some other memorable cold periods have been closer to solar max, e.g. 1947 (shown below), 1979, 1991 and 2013. The research I've read suggests it's actually the descent from solar max that is least associated with cold winter weather in NW Europe, so now we are out of the descending period we should have a greater chance of cold weather in the coming winters than in most recent years between 2013 and 2018. Others have commented that it's not just the broader solar cycle activity but actually whether immediate activity is strengthening or weakening that makes a difference - March 2013 is an interesting example in this respect as a sharp period of reduced activity in between two peaks - but I've not seen much research about this. It does look like colder weather could be favoured this winter with our current solar position being similar to Autumn 2009, and westerly QBO isn't necessarily a problem (1986, 1991 both WQBO for example). One thing to keep an eye on is how strong the La Nina becomes, if this remains on the weak side then certainly a chance of a cold winter but as we saw last year even with some favourable background signals this isn't a certainty.

If there's a similar temperature series for Canada to the CET, it should be pretty straightforward to drop it into the same template - let me know and I can give it a go? I have downloaded the PDO data so that's something I'm hoping to add to the mix, at first glance the very coldest winters in the UK (CET below around 2.8C) look fairly close to neutral and then a fairly random spread.

I've been teaching myself data visualisation in Tableau for work, so thought I'd also have a go at plotting some CET data against ENSO / QBO data. Here's a chart showing the mean winter CET for all years since 1951 (on the horizontal axis) and mean ENSO ONI data for the winter (on the vertical axis). The symbols indicate QBO data averaged for the winter months, circles are EQBO, squares are WQBO, filled are stronger values, unfilled are weaker, and an asterisk represents very weak / neutral values. The year shown is for January / February, so '2010' represents winter 2009/10. It's a bit crude as it covers the whole winter and so doesn't indicate (for example) whether a winter is front-loaded or back-loaded for cold, and I could only find the ONI data for ENSO which covers the Nino 3.4 region, so this doesn't indicate the impact of Central or Eastern Pacific La Nina / Modoki El Nino for example. A couple of interesting observations from the chart: (1) aside from 2009-10 which is a bit of an outlier, it seems that coldest winters tend to be clustered closer to ENSO neutral, and while it doesn't seem to make much difference whether it is El Nino or La Nina, some of the warmest winters have been associated with the strongest events: e.g. 2015/16 and 1988/89. (2) the QBO doesn't seem to make a great deal of difference to the outcome, with both EQBO and WQBO of varying magnitudes at either end of the scale, at least when averaged over the winter here. I'm still learning so it would be interesting to know if there are any other datasets that would be better to use, or for example whether using previous autumn or OND / NDJ values for ENSO / QBO would be more helpful if there is likely to be a lag in the effect on UK weather.