ENSO Signal For 2018/19
The NMME ensembles (above) are going for a weak El Nino
(magnitude peaking at approximately +1) but with some
uncertainty around the mean, and some weak hints at a
weakening of the El Nino towards spring. The ECMWF
ensembles go for a weaker El Nino (magnitude most likely
less than +1), but the Met Office ensembles point
towards a stronger El Nino, with the majority of
ensembles going for a magnitude of around +1.5. Thus, El
Nino looks set to be dominant this winter, possibly
comparable to the 2009/10 El Nino but unlikely to
approach the magnitude of El Nino seen in the winter of
Looking at the MEI based ENSO time series for previous
seasons, the closest analogue is the winter of 1964.
Other near approaches are 1978, 1987, 2005 and 2007.
The overall signal is for the winter of 2018/19 to be
more anticyclonic than average, with a reduced mean
westerly flow relative to normal. Breaking it down by
month, though, we see the traditional El Nino winter
progression from relatively westerly-dominated weather
early in the season, to blocked patterns and more
frequent easterly winds late in the season.
The respective Decembers were on average more westerly
than normal, with an anticyclonic tendency in the south.
The Januarys tended to be anticyclonic with the mean
sea-level pressure pattern being anticyclonic
north-westerly. Note that the above Januarys were a
mixed bag, however, containing the exceptionally mild
ones in 2005 and 2007 as well as the cold one in 1987
which was famous for its exceptionally cold easterly
incursion between the 10th and 15th. The Februarys had a
very marked tendency towards high latitude blocking,
especially between Greenland and Iceland, and an
anomalous easterly flow covering the British Isles. The
mean sea level pressure anomaly for the Februarys
(below) will be of interest to those who hope for cold,
snowy weather, and also matches up well with the Met
Office model ensemble mean projections for Jan/Feb/Mar
and Feb/Mar/Apr, which point towards high latitude
blocking late in the season.
Strong El Nino winters don't always end cold (indeed,
the winters of 1868/69 and 1997/98 both had
exceptionally warm Februarys) but winters with
relatively weak El Nino conditions appear to more
reliably see a reduced mean westerly flow by February.
Note that this doesn't always mean widespread snow for
the UK though - I still recall the second half of
February 2007, when much of Europe had a cold easterly
spell, but the cold air never made it across the North
Sea to the British Isles.
The QBO (Quasi-Biennial Oscillation) has been in an
easterly phase since the summer of 2017 but has moved
into a neutral phase during September and October. In
winter the easterly phase of the QBO is correlated with
a weak jet stream and more high latitude blocking than
usual, but the near-neutral present state of the QBO
means that it will probably have little bearing on the
winter of 2018/19.
The Madden-Julian Oscillation is currently forecast to
head through phase 1 and phase 2 during early December,
which means that there is an above average chance of
early December being dominated by low pressure and
relatively mild temperatures. A shift into phase 3 is
possible towards midmonth which increases the chances of
colder, drier, anticyclonic weather around mid-December.
If the MJO retains a strong signal, then late December
may be mild and westerly dominated with the MJO going
through phases 4 and 5. During January and February it
is impossible to forecast where the MJO will be at
particular times in the month, but phases 7, 8 and 1 are
the phases which tend to correspond to the increased
likelihood of high latitude blocking and cold and
potentially snowy weather over the British Isles.
Arctic Sea Ice
At this point in the year, sea ice was comparably
extensive in 2007 and 2013 (including the same regional
anomalies in the Arctic, with below average extent east
of Svalbard and in the Bering Sea). The winters of
2007/08 and 2013/14 were both mild, and 2013/14 was also
exceptionally wet, but 2007/08 contained an
exceptionally sunny February.
North Atlantic Oscillation and Arctic Oscillation
The North Atlantic Oscillation and Arctic Oscillation
are generally forecast to be near-neutral early in the
season, but with some disagreement between the models
and also the analogues from past years.
Late in the season, there is strong support for a
negative North Atlantic Oscillation and Arctic
Oscillation with a much reduced mean westerly flow
relative to normal over the British Isles and a
suggestion that easterly winds may be somewhat more
common than usual.
Temperatures and rainfall
The Met Office long-range model ensemble, which has had
a good track record at identifying the pressure patterns
for the summer and autumn of 2018, is suggesting that
temperatures over the UK are most likely to be near to
the long-term average despite the generally elevated
global temperatures. There is a signal for below average
pressure over central Europe and above average pressure
over Greenland, and the forecast for Jan-Feb-Mar and
Feb-Mar-Apr indicate an increasing chance of extensive
high latitude blocking and frequent easterly winds over
the British Isles late in the season, and a likelihood
of cyclonic weather being more dominant than usual early
in the season. Precipitation is forecast to be near to
slightly above average over much of the UK but below
average in north-west Scotland.
The Netweather long-range model is suggesting that mean
sea level pressure will be slightly below average over
the British Isles during December and January, with
January seeing more high latitude blocking than usual.
Britain is forecast to have slightly above normal sea
level pressure during February. The westerlies are
predicted to be slightly less dominant than usual.
The ECMWF long-range guidance is suggesting slightly
above pressure over the British Isles, with the anomaly
centre to the north, which indicates a slightly reduced
mean westerly flow relative to normal. Like the Met
Office, the ECMWF model suggests that near average
temperatures are most likely for the British Isles
during the winter quarter, though perhaps a little above
in the north, with an increased likelihood of extensive
high latitude blocking late in the season. The ECMWF
model has no clear precipitation signal for Britain. As
in recent winters, anomalous warmth is forecast for the
The JAMSTEC model outlook has a colder than average
winter over the British Isles, and below average
precipitation in the north-west and above average
precipitation in the south-east, suggesting a strong
likelihood of blocked weather patterns and a reduced
westerly flow relative to normal.
The CANSIPS model is suggesting that December will be
more cyclonic than usual with southerly winds more
frequent than usual, followed by a January with reduced
westerly flow relative to normal and above average
pressure over Greenland, and then a much reduced
westerly flow relative to normal for February.
Overall, the signals for December are quite mixed, but
the most likely outcome is that the month will be
unsettled and mild early and late, possibly with a
colder blocked spell around mid-month, and probably
wetter than average overall for the British Isles.
For January there is strong support from the analogues
for colder, more settled conditions and a mean westerly
flow somewhat veered from normal, pointing to
northerlies being more frequent than normal, although
two of the closest ENSO analogues are 2005 and 2007
which had very mild Januarys with rampant westerlies.
For February there is strong support for plenty of high
latitude blocking and easterly winds blowing more
frequently than normal, as has historically been very
common in winters with ENSO in a relatively weak El Nino
state, and both the analogues and the major long range
forecast models are in strong agreement on this. In
particular, any sudden stratospheric warming event in
January or February, especially later in the season,
would be particularly likely to give us a significant
cold spell given that the probabilities are already
weighted towards high latitude blocking for February.