Experimental Long Range Forecast for SW England. December 2017 issue.

Published 18 December 2017.

1. Influences.

Changes in sea temperature.


Apart from a little cooling, relative to “normal”, near the North of Scotland Atlantic anomalies continue to be above normal. The tropical area is forecast to remain slightly above normal as shown in the UK Met Office plot below (right). The La Nina in the Pacific is well established, with a large area of below normal sea temperatures in the eastern Pacific south of the equator. The La Nina is forecast to slowly decline during the Spring and early Summer as shown by the NMME plot below (left).


IRI published a statistical relationship between La Nina and UK rainfall and this is shown below for January to May, however, as the La Nina becomes weak so the statistical relationship may decline.


IRI probabilities of Wet /Normal /Dry seasons based on possible links to La Nina

The plots above suggest that SW England has slightly higher probabilities of Normal to Dry conditions as we move into Spring.

Another statistical relationship which may play a part in the remainder of the Winter and Spring seasons is the Quasi-Biennial Oscillation (QBO). The QBO is linked to conditions over Western Europe during late autumn and early winter through an influence on the phase of the North Atlantic Oscillation (NAO) at the surface. An easterly phase of the QBO tends to moderately increase the chances of a negative phase of the NAO, which in turn increases the chances of below-average temperatures.”


QBO monitoring data (shown above) indicates that the easterly (negative) phase of the QBO remains above 50hPa.

The North Atlantic Oscillation index (shown by a 500hPa index in the image below) continues to oscillate between positive and negative phases with ensemble predictions of remaining in a positive phase through to the end of December. This maintains a signal for a  milder/wetter winter season to develop.


The North Atlantic Oscillation can also be affected by the Madden-Julian Oscillation. In an early issue of the contingency forecast (15 December 2017), the Met Office states that “the Madden-Julian Oscillation is an area of enhanced thundery activity that moves eastwards through the tropics over a period of several weeks. It has recently been active and is currently in a location that frequently leads to a negative phase of the North Atlantic Oscillation about 10 days later. The negative phase of the NAO brings colder-than average conditions to the UK, implying an increased chance of colder than-average temperatures in early January.” For more information see Met Office NAO information.

The next statistical relationship to consider is the development of a sudden stratospheric warming. Should this occur there can be an impact on the Jet stream causing it to weaken and allowing High pressure to develop over Europe. This could bring colder temperatures further west and possibly into the UK.

The stratospheric polar vortex has been displaced eastwards over recent weeks by a series of warming events across the N Pacific and into Canada, as can be seen in the temperature plot for the North Pole.


It is not clear whether there will be a full disruption of the polar vortex, though this remains a risk, possibly leading to an upset in the model forecasts. Several of the seasonal models are able to handle changes in the stratosphere.


ECMWF 30hPa chart 17 December and forecast for 27 December 2017

Analysis and Numerical (model) output.

A: Recent Climatology.


Temperatures across the SW of England have been near normal and for many areas it has been a fairly dry period. Recent reservoir levels remain above normal for the time of year, but further dry weather may impact the storage levels for the winter as a whole.  River flows in the region for the Autumn season show western areas above normal and eastern areas below, although November data shows reduced river flow in more areas.


The  full summary is available from the Hydrological Summary  –  November 2017 summary PDF 

An experimental product from the Global Flood Awareness System suggests the flood risk areas for the next 4 months based on ECMWF seasonal forecast.


Hint at below normal river flows in Somerset and high flows in the Exe estuary, the latter looking odd. Compared to the previous forecast there is a lower probability of flooding in some Wales and N England.

B: Upper Troposphere:


CFS2  200hPa monthly mean Jan Feb March 2017 and anomalies (lower)


CFS2 200hPa monthly mean contour March April May 2017 and (lower) anomalies

CFS2 forecast for 200hPa suggest above normal heights across the UK but still with a signal for enhanced jet across the Atlantic in February.

C: Lower Troposphere:

Comparing the CFS2 mean of 10 days forecasts (top row of each pair temperature anomaly, lower row indication for above or below normal rainfall (reds are above).


Jan Feb March 2017 temperature and precipitation anomalies


March April May Temperature and rainfall anomalies

The main theme from the above is for above normal temperatures and a wet period early in 2018 but a drier Spring with temperatures nearer normal, for at least some of the months.

D: Comment:  Seasonal temperature and rainfall forecast data for the Autumn have tended to over forecast above normal temperature and rainfall. There has been some indications of below normal rain in southern areas but these have not been consistent. Forecasts of below normal temperature seem to be rare in the model output. So far in December temperatures have been near or slightly below average and the rainfall about half of the total for a normal month but this can easily change. Verification summary.

2. Forecast.

Remainder of Winter ( 2018  January and February) :

There remains a strong indication for a milder than average January and February in the SW of England  This does not rule out some colder spells, probably with North or NW winds rather than easterly, but suggests that the cold spells will be short lived.

Rainfall is uncertain month to month but the suggestion that mobile westerly patterns will come to dominate the weather patterns suggests above normal  rainfall especially in Cornwall and to the west of the moors. Elsewhere near normal or even slightly below normal rainfall seems possible, perhaps with January being the wetter month. The snow risk is reduced compared to average with little if any snow for lower ground but some snow for the moors.

Due to the statistical relationships, outlined in part 1, there is a moderate risk that colder and drier weather may develop making this forecast summary uncertain.

Winter Climate: 1981-2010 average temperature values for lowland areas 7°C in the West and 5°C in the East. Rainfall;  January typically wetter than February. 1981-2010 Autumn average 300-400 mm lowlands but 200-300 mm areas to E of Dartmoor. Snow climatology less than 5 days lying snow over lowland areas 5 to 10 hills, say hills around 200 metre elevation – one in three years have no lying snow.

Spring 2018 (March April and May):

Temperature probably slightly above normal for the season but possible colder periods in March and again in May. Rainfall near normal in the west, elsewhere drier than average.

Spring climate: 1981-2010 average mean temperature 9 or 10°C but a few degrees cooler over the moors. Roughly Mar 7 or 8°C Apr 8 or 9°C  May 11 or 12°C. Average 1981 to 2010 rain 200 to 300mm lowest values in E Devon and over parts of Somerset.

Summer 2018  (June July August). Limited model data:

Temperatures near normal, perhaps a cooler start and a warmer end to summer compared to average. Uncertain rainfall but perhaps above normal in June and near normal later.

Summer climate: 1981 to 2010 average daily mean temperature 14 or 15°C in many areas to 16 or 17°C in main urban areas also locally east of the moors and more widely in Somerset. Maximum temperatures average 19 to over 21°C in similar areas. July often warmer than August. Average 1981 to 2010 rain 300mm over the moors, typically 200 to 250mm in many coastal and eastern areas. June often drier than July and July drier than August.

3. Caution.

Experimental Long Range Forecasts do not have a good success rate. The data used for the above forecast summary and brief verification of previous Long Range Forecast summaries can be seen at   http://www.weatherservice.co.uk

The attempt at a Regional Forecast for SW England aims to test whether such a forecast of temperature and rainfall variation from average can be made using numerical model data available on the internet. The forecast should not be used for any other purpose. A brief verification summary for the UK and Eire is routinely published at http://www.weather-info.co.uk/wxsvc/Verification.html or Click here for the Teignmouth and Dawlish summary

4. References.

SST anomaly NOAA Remote Sens. 2014, 6(11), 11579-11606; doi:10.3390/rs61111579

IRI statistics: Mason, S.J. and L. Goddard, 2001: Probabilistic precipitation anomalies associated with ENSO. Bull. Amer. Meteor. Soc., 82, 619-638.

UK climate details see: http://www.metoffice.gov.uk/public/weather/climate/

NMME information:   http://journals.ametsoc.org/doi/abs/10.1175/BAMS-D-12-00050.1

CFS2  info

GLOFAS Acknowledgement: Data were provided by the Global Flood Awareness System – GloFAS (http://www.globalfloods.eu/) of the European Commission Joint Research Centre and the European Centre for Medium-Range Weather Forecasts.
Reference: Alfieri, L., Burek, P., Dutra, E., Krzeminski, B., Muraro, D., Thielen, J., and Pappenberger, F.: GloFAS – global ensemble streamflow forecasting and flood early warning, Hydrol. Earth Syst. Sci., 17, 1161-1175, doi:10.5194/hess-17-1161-2013, 2013.

‘Copernicus Products’  as listed in the C3S or CAMS Service Product Specification or any other items available through an ECMWF Copernicus  http://climate.copernicus.eu

Stratospheric Diagnostics from Free University Berlin