I see that the media have started their annual cycle of hyping up the coldest winter in 50 years
is coming except this time it's the Beast from the East returning. I find this year's release relatively interesting (although as always frustrating) as it's a bit different to other years. Recently, a study was released from University College London by Saunders, Lea and Smallwood titled "Long-Range Forecast for the North Atlantic Oscillation and UK Weather in January-February 2020". I will highlight some parts of it with my own research and discuss my opinions on the forecast.
Firstly, they state the following quote:
There is an 87% chance the NAO will be less than 1981-2010 mean and a 65% chance the CET will be colder than the 1981-2010 mean. Examination of data between 1953 and 2019 shows that nine of the ten years where these predictor fields (referring to solar and stratosphere cyclic signals) had the same sign and similar magnitude to that in summer 2019 were followed in January-February by a negative NAO and by a CET colder than climatology.
87% chance of a -NAO to occur during January/February 2020 according to their teleconnections (which they go onto mention as monthly solar 10.7 cm flux data and monthly QBO data) based on Summer 2019. That's a very high chance I must say but keep in mind, that includes a 13% chance of -NAO not occurring which is still a significant chance in weather world in my opinion. They also state a 65% chance of the Central England Temperature (CET) being colder than the 1981-2010 which again is quite a high chance but that includes 35% chance of not being cold. I suspect this takes modern warming into account somewhat as well as -NAO winters not all being colder than average.
When I did an analogue on +NAO summers last year with winters that followed, there was a very strong signal for the +NAO to continue into the winter. These summers included but not limited to 1983, 1979, 1967, 1994, 1976, 1972, 1996 and 2013. A much different signal this year with -NAO summers which include but again not limited to 2012, 1958, 2011, 2015, 2016, 2008, 1993, 2009, 1998, 1980, 1954, 2014, 1957, 2010 and 1962. One thing to notice here is the frequency of anomalous -NAO summers in the 21st century so far - feedback from arctic amplification perhaps. There is definitely more of a signal for a -NAO winter with a -NAO summer but it's not clear cut either as if anything is in weather! I'll go more in-depth with this at a latter point as a NAO update is in the works.
The following table shows years with similar magnitude of solar flux and 50hPa QBO to Summer 2019 and the January/February NAO/CET values for those years. Looks to me that there is good correlation with these results and only 1975 was not cold (1966 average although even that had a cold Nov/Dec period). Also looks like that all the years had a westerly QBO in the April to June period preceding the winter which makes sense with 2019.
If you'd like to read the study for yourself, here's the link: http://discovery.ucl.ac.uk/10080518/...%282019%29.pdf
I know I shouldn't be saying this but there is always some people believing the crap. There is nothing to suggest that the Beast from the East will be returning. An event like 2018 was very, very rare and only comparable with the likes of 1947 and 1982 or in terms of cold, 1956, 1987 or 1991. The study is merely suggesting the risk of colder than average temperatures in January/February 2020 with a -NAO.