It’s time to get this show on the road as the day has come for me to release the Metcast seasonal forecast for Winter 2018-19. Plenty of conflicting factors for this season but hopefully I’ll be able to make a reasonable compromise based on all my research I have done. I will be explaining everything as easily as I can too just in case you are not sure what I mean at points here in the forecast. A lot of these things I will have already reiterated many times so do not be afraid to skip ahead to the punch line if you wish to. I apologise for any rambling on in parts as you know, I can’t resist adding in a bit of history here and there!

Long range models

To start off, let’s discuss the long range or seasonal model output. The vast majority of the known long range models were pointing towards the occurrence of northern blocking up over the Arctic, some more extensive than others with a particular emphasis over Greenland. Models like the UKMO Glosea5 and JMA went especially bullish on Greenland blocking in the second half of the Winter around February. These models in their November updates as a result anticipated the chance of a very cold to even severely cold February if we were to go solely off of their 500mb height anomalies and or mean sea level pressure anomalies. Using Hanna’s Greenland Blocking Index (GBI), similar Februaries to the extent of blocking shown on these models include 1947, 1901, 2010, 1902, 1895, 1942, 1960, 1855, 1900 and 1965. Several of these Februaries were either notably snowy (for example, 1947), cold (for example 1895 and 1947) or anticyclonic (for example, 1965 which later became cold and snowy too into March). Models were fairly mixed on the early part of the Winter but they generally pointed towards a rather zonal December with relatively mild and wet conditions especially the ECM. January looks anticyclonic but a bit split on the temperature aspect. If the high pressure aligns itself over top of the country, just to the north or to the east then it’s likely to be a chilly January with clear skies leading to frost and fog. If the high is more aligned to the south then it will be a mild month. February holds the greatest potential for blocking (and therefore cold conditions) on the models with quite a strong signal for three months away. One would normally think the first month (December) would hold the highest confidence of what is likely to happen but in contrast, it is actually the most uncertain of this forecast. Models like the DWD and the CFSv2 which show zonal and wet conditions cannot be discounted as their solutions are certainly possible given factors like the QBO, El Niño (especially early in the Winter) and the Atlantic sea surface temperatures.

Atlantic sea surface temperature anomalies

The Atlantic sea surface temperature (SST) anomalies continue to show this strange profile of cold-warm-cold-warm in the North Atlantic with the cold anomalies forming the shape of a horseshoe which has been the case since Spring generally. This is not favourable for a negative North Atlantic Oscillation (NAO) going strictly by the SSTs alone as they would have to be warmer around Greenland to do so. In case you do not know, the NAO is an index showing the difference in mean sea level pressure between the Icelandic Low and Azores High, two elements which make up the normal zonal pattern for western Europe. A positive NAO involves both of these elements being strong and normally translate to mild, wet conditions in Winter time whilst a negative NAO involves both of these elements either being weak or sometimes, non-existent as the pattern can completely reverse, which usually translate to cold conditions in Winter. This is not always the case with either side of the NAO index as there have been instances of cold conditions with positive NAO just as much as mild conditions with negative NAO. Every Winter since and including 2013-14 has had a positive NAO which has been one of the reasons why Winter hasn’t been particularly cold recently and sometimes, record mild or wet as such happened in 2013-14 and 2015-16. Summer has been the opposite for most years since 2007 with every Summer since bar 2013 and now 2018 having a negative NAO. In fact, Summer 2018 was record breakingly positive. Will Winter 2018-19 see a similar flip to negative NAO? Not very likely given the Atlantic sea surface temperature anomalies though not impossible. I did an analogue on Winters following very positive NAO Summers and in fact, the signal was strong for an Atlantic driven Winter. Therefore from this point of view, there is an increased chance of a mild, wet Winter.

The stratosphere

One thing that could be conducive to a negative NAO regardless of the sea surface temperatures is a sudden stratospheric warming (SSW) event. A sudden stratospheric warming event is where observed temperatures within the stratosphere rise by a few tens of degrees within a short period of time thus why it’s so-called “sudden”, and the zonal winds in the stratosphere reverse to east compared to how they normally travel in a westerly flow. The impacts of these sudden stratospheric warming events include the risk of intense northern blocking over the Arctic and an increased chance of easterly winds for the UK. There is no guarantee with them as it depends on the positioning of the high pressure and also the type of SSW. These impacts are felt in the troposphere around 9 days minimum after the initial SSW event. Sudden stratospheric warming events are normally favoured around the transition from easterly to westerly QBO as such happened in the Winters of 2012-13 and 1984-85. This does not necessarily mean we’ll see one this Winter as it is based on history and SSW events do not follow patterns. The CFS model has kept toying with the idea of a SSW sometime in December this Winter and the GFS ensembles recently have shown some instances of minor warming events through December similar in style to December 2017. With this, it’s more of a wait and see than what you can necessarily predict. If one does end up occurring, you could say there is the risk of a potent cold spell at some point during the season.

Quasi-Biennial Oscillation (QBO)

The Quasi-Biennial Oscillation (QBO) is an index that reflects the variation of the zonal winds in the stratosphere above the equator. These winds travel in a belt around the planet and change direction approximately every 14 months. It is said to be among the most regular phenomena on earth but as 2016-17 showed, this is not always the case, when the QBO was meant to go into an easterly phase. There was an upward displacement of the westerly phase in early 2016 which cutoff the normal downward propagation of the easterly phase and 2016-17 was a second successful westerly QBO Winter. We are transitioning to the westerly phase of the QBO after being in the easterly QBO phase since mid-2017, which should increase the chances of a milder and wetter Winter but there have been cold Winters with westerly QBO (for example, 1978-79) before as much as mild Winters with easterly QBO (for example, 2007-08). There is a time lag before the westerly QBO makes full impact on the weather conditions and one thing that can break this impact before arrival is the occurrence of a SSW event but again no guarantee as described above.

October Eurasian snow cover advancement

Eurasian snow cover extent through October 2018 was slower than recent years but still relatively above average according to the snow advancement index by Cohen. Comparable years to October 2018 in terms of Eurasian snow cover going by the index include 2010-11, 2004-05, 2000-01, 1999-00, 1996-97, 1982-83, 1978-79 and 1973-74. My analogue of these years showed there is above average heights over and just to the south of Greenland with below average heights in the central North Atlantic into southern and central Europe – even into the east of Europe. This leaves us pulling in north to northeasterly winds. This would be a cold Winter if it were to verify with snow potential.

El Niño-Southern Oscillation

The El Niño-Southern Oscillation (ENSO) is an irregular index with variation in winds and sea surface temperatures over the equatorial Pacific Ocean. The warming phase is known as El Niño whilst the cooling phase is known as La Niña. Southern Oscillation is the accompanying atmospheric component coupled with the sea temperature change. These phases of ENSO have various impacts on weather patterns around the world but the impacts from it on Europe are unknown due to event to event variation. However, it has been said that El Niño increases the chances of cold later in the Winter whilst La Niña does the opposite of increasing the chances of cold earlier in the Winter. Examples of El Niño Winters doing this include 2014-15 (where February was the coldest of the Winter with a frosty spell during the first half) and 1977-78 (where February produced a severe cold spell with one of the worst blizzards on record to hit the West Country). An example of a La Niña Winter doing this includes 2010-11 (where November/December produced some of the coldest weather on record to occur so early in the Winter including the coldest December on record for most). However there are many odd ones out. For instance, you only have to go back to last year for La Niña and February 2018 was the coldest of the Winter. There has been talks of an El Niño Modoki for 2018-19 rather than a traditional El Niño event. Most of the time, warming occurs in ENSO regions 1+2 but sometimes, you can have another type of El Niño called a Modoki. With this type of El Niño, the warming is focused on the central part of the Equatorial Pacific i.e. ENSO region 3.4. The CFSv2 continues to forecast an El Niño for Winter 2018-19 with the model now showing the possibility of moderate El Niño conditions but this is kind of an outside chance at the moment as it will have been very late for such an ENSO event to occur and the model forecasts a slight drop off in early 2019 too into weak threshold. The overall reanalysis of El Niño Modoki Winters in the past show above average heights over Greenland with below average heights to the southeast of us drawing in an easterly or northeasterly flow and is a clear cold signal. Mind you, this is no guarantee as there were some El Niño Modoki Winters that were mild and or wet such as 1994-95 (the wettest on record for a good part of Ireland before 2013-14) but the majority were cold. The most recent example of an El Niño Modoki occurred in Winter 2009-10. Will we see a similar Winter for 2018-19? Very unlikely, that was the coldest Winter since 1978-79 after all but not impossible! As this will be an El Niño Winter, it is more likely to start off mild with an increased risk of cold later.


At last, I’ll discuss my final analogues. The years that cropped up most in my analogues for this Winter are 1941-42, 1963-64, 1979-80, 1994-95, 1996-97, 2010-11, 1901-02, 1940-41, 1968-69, 1987-88, 1995-96, 2001-02, 2008-09 and 1990-91. As to be expected, this batch of years is mixed in terms of their Winter conditions. 1901-02 had an unremarkable December and January whilst February was quite cold. 1940-41 and 1941-42 were cold Winters with some potent cold spells particularly in January of both years. 1963-64 was exceptionally dry and quiet. 1968-69 had a White Christmas and a severely cold February with a mild January despite the NAO and AO being negative then; it was also an extremely cold Winter for parts of Eurasia. 1979-80 was generally on the mild side with a rather wet December though January was a bit cooler. 1987-88 was a very mild and wet Winter and the start of the cluster of three notoriously mild Winters. 1990-91 was changeable with a fairly cool but unsettled December with a snowstorm for the midlands, an anticyclonic January after a stormy start and a cold February with a very cold spell near the start of the month. In terms of the methodology given, 1990-91 provides the most similarities to 2018-19. 1995-96 had an extremely cold spell at the end of December whilst most of January was very mild and dull and February was cold, wet and snowy but sunny; a bit of an unusual combination. 1996-97 was a front loaded Winter with a cold spell during the second half of December into the first part of January including some ice days but the rest of January tended to be milder and February was a wet, mild, stormy month. 2001-02 had the sunniest December on record with mainly dry and cool conditions due to frost later on whilst January and February tended to be stormy and mild. 2008-09 had a cold spell at the end of December and start of January with severe frost but spells of sunshine rather than snow along with a very snowy spell at the beginning of February but became milder in the second half. 2010-11 contained two severe and unprecedented cold spells at the start of December (continuing on from end of November) and again from mid-December to Boxing Day. When you put all these years together, you get well above average heights right over the top of Greenland with a trough of below average heights just to the west of Iberia. In this scenario, the mean wind direction is easterly and with cold air coming in from the east across the relatively warm North Sea, there would be instances of lake-effect snow. The intensity would depend on the potency of the cold air and the exact state of the sea surface temperatures.

Let’s take it month by month (in terms of what my analogues showed):

December – Blocking just to the northwest of Ireland with below average heights to the south and east of Europe. Looks to me as it would be a cold but anticyclonic month with us just on the periphery of easterly winds bringing cold air from Siberia.

January – Absolute monstrous blocking stretching from Greenland to Scandinavia with a trough of below average heights in the Atlantic going through the Mediterranean regions. This would be an extremely cold January I think with a lot of easterly winds.

February – The blocking reduces from the extent of January’s but still to be found with below average heights further northwards over top of us rather than to the south of us. This to me still looks cold with the blocking but maybe not as straightforward as January. It would be more of a knife edge between very cold and snowy or very wet and mild.


What is the final verdict from all of what I described above? What do I think is likely? Well, the signs from my analogues and long range models are very compelling aren’t they for a cold Winter. However, I just don’t think it will be as straightforward as some would interpret from these. I don’t think this will be a classic cold Winter. Rather, I think a changeable Winter is more likely similar in vein to 2017-18 especially earlier in the season with more of a lean towards a risk of prolonged cold later in the season.

December – A close to average month in terms of temperature if a little bit on the milder side. Above average precipitation is more favoured than below average despite my analogue showing an anticyclonic month on the cards, due to a more zonal flow earlier on and there is a possibility of this continuing through the month. However, if there is frequent high pressure to the east and or south, the south has a higher chance of drier than average conditions. This would be a reverse if the wind would turn easterly with the north tending to be drier.

January – A close to average month again with no large deviation though slightly leaning towards the colder side due to frost or blocking. Precipitation likely to be on the drier side.

February – Holds the greatest prospects for cold currently with the models clearly showing a backloaded Winter for 2018-19. I think February 2019 will be a cold month but not remarkable in the historical record with a deviation somewhere between 0.5 to 1.0c below average. Precipitation likely to be relatively above average due to volatility in the weather conditions as the Atlantic battles against blocking to the north.

So that’s the Metcast seasonal forecast for Winter 2018-19. Overall, an unremarkable Winter being signalled going by the forecasts which is referenced to climatology aspects but that does not mean, it won’t be an interesting Winter. 2017-18 was similar to what I’ve described above and look how much interest that garnered even pre-the Beast from the East. Remember, this is only just for fun as long range forecasting is unreliable and not to be taken as gospel!

We’ll be on live chat (blue button at the bottom right hand side of this page, however when no agents are available it will not be visible) if you have any immediate questions or you can leave a comment below…

Thanks for reading!

Sean Bruen is a forecaster for Metcast (and Snow Watch). His main interests are historical and long range weather. He LOVES snow (his Twitter account is @SnowbieWx, go figure!) and his favourite season is Winter.