Predictions of winter 2005/2006 in the UK & Europe

By Andrew Masterman 18 November 2005

The late Professor H.H. Lamb was one of the greatest climatologists of the 20th century and he wrote a paper entitled, "Types and spells of weather around the year in the British Isles: Annual trends, seasonal structure of the year & singularities (Lamb, 1950). In this paper, Lamb remarked that "The occurrence of occasional long spells of weather, marked by persistence of one or another easily recognised and definable type is a well-known feature of our climate" and that "such spells commonly determine the prevailing character of a particular season". A more concise version of these remarks is "Weather patterns within a season tend to repeat themselves". What the Useful Info winter & summer forecasts try to do is to predict what these recurrent weather patterns might be and from these predictions make some tentative forecasts about parameters such as temperature and rainfall.

North Atlantic Oscillation (NAO) Forecast

The NAO was discussed in detail in last winter's forecast and a number of factors which dictate NAO variations were identified. Before discussing the status of individual causal factors in 2005, a number of interesting features of the NAO in different periods can be seen in Figure 1. The green line (NAO data from 1900-2004) represents the long term average change in the NAO through the year and shows a peak in winter and a secondary peak in August. The red line is the average for the 1989-2004 period which has been characterised by persistent above average temperatures all year round and winters with strongly positive NAO (ie persistent westerly winds) and a notable feature of this period is that Septembers & Octobers had anomalously negative NAO. In contrast, the blue and yellow lines represent two cold periods with negative NAO winters (ie. persistent northerly & easterly winds), 1963-69 & 1979-87 respectively, and both these periods had below average NAO in August and above average NAO in September & October but in November & December in the case of 1963-69, the NAO dropped below average. Data for January to September 2005 is also presented (turquoise line) and are clearly more similar to the two colder periods (1963-69 & 1979-87) than the warm period (1989-2004) with February having well below average NAO and September above average although January & August values are more characteristic of the warm 1989-2004 period. In October 2005, the NAO was positive in the first half & negative in the second half (see graph).

Figure 1 Monthly values of NAO in 2005 relative to longer periods: 1900-2004; 1963-1969 (a cold period with negative NAO winters), 1979-1987 (a cold period with negative NAO winters) and 1989-2004 (a warm period with positive NAO winters). Data from the Climatic Research Unit.

Now some factors which dictate variations in the winter NAO are discussed:

ATLANTIC SEA SURFACE TEMP (SST) ANOMALIES: these are shown in Figure 2 which compares global SSTs observed on 12 November 2005 with data a year and two years ago. Overall, it appears that globally SSTs have cooled, notably in the North Pacific and Indian Oceans but also to some extent in the North Atlantic. A feature of the North Atlantic in all three years is the so called Atlantic tripole of anomalies: the combination of warm anomalies in the tropical and northern parts and negative/average anomalies in between favours a negative NAO winter. Rodwell, Rowell & Folland ( 1999) identified positive SST anomalies off the eastern seaboard of the USA and in the Greenland Sea north of Iceland, and negative SST anomalies south of Greenland and east to Iceland as favouring positive NAO winters while the opposite anomalies favour negative NAO winters . As of 12 November, the Atlantic off the eastern seaboard of the US is close to average (neutral signal) while the Greenland Sea north of Iceland shows cold anomalies and the Atlantic south of Greenland & east to Iceland is warmer than average: these two characteristics favour a negative NAO winter. Another paper emphasises SST anomalies south and east of Newfoundland as the critical region of the North Atlantic: negative SST anomalies there favouring negative NAO and positive anomalies positive NAO (Ratcliffe & Murray, 1970). Data for the 12 November 2005 show mostly average SSTs in this region so a somewhat neutral signal is present here. So overall, there is definite signal from Atlantic SSTs for a negative NAO winter.

It is relevant to mention the NAO during last winter in which the first half was dominated by positive NAO and the latter half by negative NAO. Atlantic SSTs in November 2004 (Figure 2) weren't as favourable for negative NAO as they are this year as the positive anomalies in the Greenland and Barents Seas favour positive NAO and these were probably a significant factor causing the positive NAO period from late December to mid-January when deep lows developed in the Greenland/Nordic seas and pushed Atlantic warmth right across Europe and deep into Siberia. This hypothesis is supported by The Early Twentieth-Century Warming in the Arctic—A Possible Mechanism which showed that the marked warming of the Arctic from the 1920s through to the 1940s was caused by reduced arctic sea ice extent triggering a positive feedback in the form of enhanced cyclogenesis in the Barents Sea region during winters which maintained a feed of warm Atlantic air into this region. Later in the winter of 2004/2005, the other signals which favoured negative NAO managed to overwhelm the effect of the warm SSTs in the Greenland & Barents Seas and strongly negative NAO predominated. So the cool anomalies in the Greenland & Barents Seas this year are possibly very significant as an early winter like last year is unlikely.

STRATOSPHERIC VOLCANIC DUST: there have been no major volcanic eruptions this year and the latest Weekly Volcanic Activity Report shows no major eruptions so stratospheric volcanic dust does not favour a positive NAO winter.

SOLAR ACTIVITY/SUNSPOTS: as we are heading towards a sunspot minimum in 2007/2008, sunspot counts are relatively low but currently there is a large sunspot, 822 which may direct some M-class flares toward Earth around 20th November (source: Spaceweather). However, a sustained period of high sunspot activity is unlikely so this factor does not favour a positive NAO winter.

The UK Met Office NAO forecast is for a negative NAO winter and this forecast is based on Atlantic SSTs in May. And the University College London NAO forecast is that negative NAO is three times more likely than positive NAO.

So Atlantic SSTs favour negative NAO and stratospheric volcanic dust and sunspots do not favour positive NAO and the two NAO forecasts mentioned above both favour negative NAO. Other data to support a prediction of negative NAO in winter 2005/2006 is

the Jan-Sep 2005 data shown in Figure 1 which is more similar to the average trend of cold periods such as 1963-1969 than a warm period 1989-2004.
the clear trend to negative Jan-Mar NAO in recent years (see chart)

Now the NAO is not easy to predict so the above cannot be taken as definite proof that the NAO in winter 2005/2006 will overall be negative to neutral rather than positive which has been the norm post-1988 but there are strong indications that it is likely to average neutral to negative and this would mean a cold winter across Europe but not necessarily the UK.

Other Northern Hemisphere Teleconnections

In last winters forecast, the Pacific North American teleconnection (PNA) was shown to have some associations with the weather patterns across the UK. Table 1 shows the results of the analyses in which winters 1950/1951 to 2003/2004 were allocated to groups defined by the PNA across North America: strongly negative (< -3.00), moderately positive (1.00 to 2.99) and strongly positive (>2.99) and the weather patterns across the UK in each of these groups was assessed using PSC indices.

P is the index of progressiveness and is positive when westerly winds predominate and is negative when non-westerly winds predominate
S is the index of southerliness and is positive when southerly winds predominate and negative when northerly winds predominate
C is the index of cyclonicity and is positive when cyclonic weather patterns predominate and negative when anticyclonic weather patterns predominate

The focus of the analyses was the effects of positive PNA which occurs when a strong ridge occurs across western North America while a strong trough sits across the east of North America and this pattern tends to get mirrored across Europe with a ridge off the coast of western Europe and a trough in eastern Europe.

Table 1 Number of PSC quintiles for each Group defining the level of PNA in three categories: low (quintile 1 & 2); average (quintile 3); & high (quintiles 4 & 5).

	Number of months
	__________________________________________________________________________

	Group A			Group B			Group C			Groups B + C

	P	S	C	P	S	C	P	S	C	P	S	C
_______________________________________________________________________________________
1- 2 (Low)	18	9	10	16	10	11	10	12	4	26	22	15
3 (Average)	2	7	10	1	4	6	6	6	10	7	10	16
4-5 (High)	13	17	13	7	10	7	8	6	10	15	16	17
________________________________________________________________________________________

Group A: winters where the PNA was below -3; Group B: winters with PNA between 1.00 & 2.99; Group C: winters with PNA >2.99

Table 1 supports this by showing a bias to northerliness (low S index) in Group C/strongly positive PNA owing to the ridge to the west of Europe while Group A/strongly negative PNA shows a bias to southerlies (high S index) which is consistent with a trough to the west of Europe and high pressure in eastern Europe. Pooling of Groups B & C shows that positive PNA is also associated with blocking/weak westerlies across Europe but there was no apparent bias for the C index which measures anticyclonicity. This was a surprising result as you would expect a ridge off western Europe to be associated with anticyclonicity across the UK. This surprising result led to two different types of the positive PNA pattern across Europe being identified: a cyclonic/west PNA in which the whole pattern is shifted west with the ridge in mid-Atlantic and the trough further west across western/central Europe and an anticyclonic/east PNA in which the UK is firmly under the influence of the ridge in the west and the trough is a long way east in eastern Europe. The importance of this distinction between two types of positive PNA pattern across Europe was seen during last winter. While the UK experienced a cold spell from mid-February to mid-March (see Bournemouth), eastern Europe had a cold spell lasting two to three weeks longer from around the 20 January (see Budapest) and this selective cold arose because initially the pattern was of an anticyclonic/east PNA bringing mild Atlantic air across the UK over the top of the high pressure in the west and it wasn't until the pattern evolved into a cyclonic/west PNA with the high pressure moving north-westwards to Iceland/Greenland that cold air was able to come westwards and affect the UK.

So what is the forecast weather pattern across North America in the coming winter? The NOAA Winter Forecast 2005/2006 predicts a warmer than average winter for central & western regions and equal chances of warm or cold for eastern states. The Accuweather Winter Forecast 2005/2006 predicts colder than average temperatures across the eastern third of the US, near normal in the central third and above average in the western third. The feature of a cold eastern US relative to the western states is indicative of a positive PNA pattern with a ridge in the west and trough in the east but a persistent positive PNA is not expected. Periods with a strong trough in the Pacific north-west are expected which would be associated with warmer weather right across the US so the PNA is likely to oscillate between positive & negative. This is likely to lead to oscillations downstream in the atmospheric patterns across the Atlantic and Europe. So the positive PNA pattern across North America is expected to be a feature of the winter as was the case in the last few winters so a positive PNA pattern across Europe is again likely to be a recurring feature but will it be in the form of an anticyclonic/east PNA or a cyclonic/west PNA?

During the preparation of this winters forecast, two further relevant factors have been identified: the level of the Atlantic hurricane season and the phase of the Southern Oscillation Index (SOI). Table 2 compares the PSC quintiles of winter months following below average Atlantic hurricane seasons (1-3 hurricanes) and above average hurricane seasons (>7 hurricanes) and the interesting result is that winters across the UK tend to be more cyclonic after above average Atlantic hurricane seasons (X² = 6.21, df 2, P < 0.05). The tests for the P & S indices were non-significant. This weakly significant result has some relevance to the anticyclonic/east PNA vs cyclonic/west PNA patterns as it suggests that the combination of a positive PNA pattern across North America and a preceding active Atlantic hurricane season would favour a cyclonic/west PNA pattern across Europe meaning a colder winter across the UK. From a meteorological perspective, an active Atlantic hurricane season is consistent with negative NAO in the following winter as it indicates low pressure in the tropical Atlantic and to balance this you need higher pressure further north in the Greenland/Iceland region. As already discussed, for cold weather to affect the UK, the trough in Europe must be further west across western/central Europe and an active hurricane season may be a useful predictor that this may occur. For example, the 2004 Atlantic hurricane season was an active one and in mid-February 2005, a cyclonic/west PNA pattern did develop giving a month-long cold spell in western Europe although at other times in the winter, an anticyclonic/east PNA prevailed.

Table 2 Number of PSC quintiles in winter months following below average Atlantic hurricane seasons (1-3 hurricanes) and above average hurricane seasons (>7):
winters 1950/51 to 2003/04; hurricane data from Atlantic Oceanographic & Meteorological Laboratory

		Number of winter months with PSC quintiles
	________________________________________________________________________________________________

	P		S		C
	Below	Above	Below	Above	Below	Above
	______________________________		_____________________________		_______________________________

______________________________________________________________________________________________________________________

1-2 (Low)	11	20	7	16	8	11
3 (Average)	1	6	7	10	10	8
4-5 (High)	12	16	10	16	6	23
____________________________________________________________________________________________________________________

Table 3 compares PSC quintiles of months in winters 1950/51 to 2003/04 where the SOI was either negative (<-2.0), neutral (-2.0 to +2.0) or positive (>+2.0) and the only interesting feature is the tendency of neutral SOI winters to be blocked across the UK (low P index). A chi-squared test for the P indices displayed in Table 3 was non-significant but was weakly significant when P quintiles in categories 1-2 & 3 were pooled and compared with quintiles category 4-5 (X² = 6.01, df1, P < 0.05). An explanation of this result is attempted below.

Table 3 Number of PSC quintiles in winters with negative (<-2.0), neutral (-2.0 to +2.0) and positive (>+2.0) SOI:
winters 1950/51 to 2003/04; SOI data from NOAA

				Number of winter months with PSC quintiles
	_________________________________________________________________________________________________________

		P			S			C

	negative	neutral	positive	negative	neutral	positive	negative	neutral	positive
	______________________________			______________________________			________________________________

__________________________________________________________________________________________________________________________

1-2 (low)	24	37	13	27	28	20	17	23	9
3 (average)	6	12	7	9	19	10	18	21	11
4-5 (high)	27	20	19	21	22	9	22	25	19
___________________________________________________________________________________________________________________________

Last winter, there was a weak El Nino (slightly negative SOI) and every time there was pulse of negative SOI, the northern hemisphere atmospheric pattern changed significantly. The reason why this occurs is probably that the strengthening of the Pacific sub-tropical jetstream associated with each pulse of negative SOI injects a lot of heat from the tropics into the extra-tropics and this results in a significant alteration of weather patterns around the northern hemisphere. The effect of such pattern changes is large swings in weather and temperature throughout the winter as was the case last year. In contrast, positive SOI/La Nina is associated with a strong polar jetstream around the northern hemisphere which means strong westerlies across the Atlantic but neutral SOI (Table 3) is associated with weak westerlies across the Atlantic (low P index). And weak westerlies provide a window of opportunity for arctic air to come southwards and westwards across Europe and give the UK some cold spells provided other factors are favourable for negative NAO. This would appear to be the case this year as forecasts of SOI all favour neutral conditions for the coming winter (SOI/ENSO Forecasts). It is noteworthy that the five coldest winters in England since 1950 (defined by Central England Temperature CET) all occurred with neutral SOI: 1963; 1979; 1982; 1985; & 1986.

Table 4 below is reproduced from last winter's forecast (in which it was Table 5) and shows the eleven winters from 1950/51 to 2003/04 with positive PNA (>1.0) and negative NAO and it contains a high proportion (~30%) of cold easterly months including the infamous winter of 1962/63. Table 4 was used in last winter's forecast to tentatively predict a return of cold easterly type weather which had been absent from UK winters since the mid-1990s and this prediction proved accurate with the occurrence of a cold easterly spell in the final week of February 2005. Now if the eleven winters in Table 4 are reduced further using the criterion of neutral SOI (-2.0 to +2.0), then just three winters remain all of which had at least one cold easterly month (5/9 or 55%): 1963; 1977; & 1986. So neutral SOI does appear to favour a cyclonic/west PNA pattern across Europe in which the cold air occurs further west to affect the UK.

Table 4 PSC quintiles and CET anomalies (from 1961-90 averages) for the eleven winters from 1950/51 to 2003/04 with positive PNA (>1.0) and negative NAO


	PSC quintiles - T is CET Anomaly
	__________________________________________________________________
	December				January				February

	P	S	C	T	P	S	C	T	P	S	C	T
_________________________________________________________________________________________
1958	5	2	4	-0.2	2	2	5	-0.4	4	1	3	+0.9
1963	3	1	2	-2.9	1	1	1	-5.9	1	3	3	-4.5
1964	1	4	1	-2.1	1	4	1	-0.4	1	5	2	+0.7
1970	2	3	4	-1.4	1	5	3	-0.1	4	1	5	-0.9
1977	1	2	3	-2.7	1	1	5	-1.0	3	4	5	+1.4
1978	2	3	3	+1.4	3	2	5	-0.4	1	3	4	-1.0
1987	5	2	3	+1.5	1	4	1	-3.0	2	4	2	-0.2
1986	3	3	4	+1.6	3	1	5	-0.3	1	3	2	-4.9
1998	3	2	3	+1.1	3	4	3	+1.4	5	3	2	+3.5
2001	1	4	5	+1.1	1	4	4	-0.6	2	2	3	+0.6
2003	1	5	4	+0.9	2	1	3	+0.7	2	4	2	+0.1

Totals for each quintile group
1- 2 (low)	6	5	2		8	6	3		7	3	5
3 (average)	3	3	3		3	0	3		1	4	3
4-5 (high)	2	3	5		0	5	5		3	4	3

No of months with below average CET				5				9				5

							19/33 (58%) months with below average CET
_________________________________________________________________________________________

There is some folklore that autumn weather can foretell the weather of the coming winter. The most famous saying is "Ice in November to bear a duck, the rest of the winter will be slush and muck." and this in plain English means cold Novembers foretell a mild winter and vice versa. The truth of this saying is tested below (Table 5) by comparing the PSC indices of winter months following Novembers with below average S indices (S₁/S₂ indicating cold weather) and above average S indices (S₄/S₅ indicating mild weather). Table 2 of this study shows the S index to be highly correlated with temperature in autumn i.e warm autumn months have a high incidence of days with southerly winds. Table 5 shows that there was no association between the November S index and PSC indices in the following winter.

Table 5 Number of PSC quintiles in winters months following Novembers with below average (S₁/S₂) and above average (S₄/S₅) S indices
winters 1950/51 to 2003/04;

		Number of winter months with PSC quintiles
	____________________________________________________________________________________________________

	P		S		C

	Low	High	Low	High	Low	High
	_____________________________		________________________________		______________________________

______________________________________________________________________________________________________________________

1-2 (low)	29	28	32	24	21	15
3 (average)	16	5	19	11	22	20
4-5 (high)	27	21	21	19	29	19
______________________________________________________________________________________________________________________

No. of winters following: S₁/S₂ Novembers: 24 S₄/S₅ Novembers: 18

Table 6 shows the results of the same exercise for October S indices and again there was no association between the October S index and PSC indices in the following winter. Next winter, this exercise will be repeated for the P index and judging by Figure 1 which shows that the cold 1963-69 and 1979-87 periods were characterised by positive NAO in September & October (i.e above average westerlies), a significant result may be found.

Table 6 Number of PSC quintiles in winters months following Octobers with below average (S₁/S₂) and above average (S₄/S₅) S indices
winters 1950/51 to 2003/04;

		Number of winter months with PSC quintiles
	____________________________________________________________________________________________________

	P		S		C

	Low	High	Low	High	Low	High
	_____________________________		________________________________		______________________________

______________________________________________________________________________________________________________________

1-2 (low)	15	41	17	35	13	23
3 (average)	4	10	6	15	11	20
4-5 (high)	17	27	13	28	12	35
______________________________________________________________________________________________________________________

No. of winters following: S₁/S₂ Octobers: 12 S₄/S₅ Octobers: 26

Recent trends in PSC Indices

The logic of this part of the forecast is that what the atmosphere has been doing so far in 2005 in the region of the UK is likely to be of some relevance in the coming winter. This logic won't always be true as a significant pattern change may occur but as a method of detecting any trends in the weather patterns in the UK region, this technique has merit. Figure 3 shows PSC quintiles for each month from January 2002 to October 2005 and there are several trends which are obvious:

throughout this period, the P index has tended to be below average indicating weak westerlies across the Atlantic (41% were below average)
from January 2002 till July 2003, the S index tended to be above average indicating above average incidence of southerlies but since July 2003, 42% of months have had below average S index compared with 29% of months with above average S index indicating a bias to northerliness over the last two years
from 2002-2004 there was a bias to months with below average C index (44%) indicating anticyclonicity (31% with above average C index) whereas seven of the months Jan to Oct 2005 have had above average cyclonicity

This above average cyclonicity is perhaps surprising as rainfall has generally been below average (see 2005 rainfall graph) this year but what has been characteristic this year is oscillations between dry anticyclonic spells and wet cyclonic spells with the cyclonic spells predominating.

These trends in PSC indices are expected to continue during the winter 2005/2006 with below average westerliness, above average northerliness and a slight bias to above average cyclonicity as the oscillations between anticyclonic and cyclonic spells continue with the latter being slightly more predominant.

Note of caution: objective Lamb types since February 2005 are currently unavailable from the CRU website so the author subjectively identified Lamb types for each day from 1 March to 31 October 2005. This means a few of the PSC quintiles for this period may be incorrect

Figure 3 Quintiles of PSC indices for each month January 2002 to October 2005.

So now follows the second Useful Info winter forecast for the UK which is based on the various factors and trends discussed above.

The General Weather Pattern

although the NAO is considered difficult to predict, Atlantic SSTs are favourable for negative NAO at the moment, are more favourable than last year with respect to the Greenland & Barents Seas and the annual profile of NAO through this year (Figure 1) is more similar to cold years (1963-69 & 1979-87) than warm years (1989-2004). So the probability is that the NAO will be negative for more days than it will be positive during the coming winter. (High Confidence)
the expected neutral SOI in the coming winter favours weak westerlies across the Atlantic and increases the likelihood (relative to last year when SOI was somewhat negative) of cold easterly type spells when a positive PNA pattern prevails across North America
winter forecasts for the US again expect a cold winter in the east and warm in the west indicating a recurrent positive PNA pattern which is likely to be associated with northerly & easterly spells across the UK (High Confidence)
the occurrence of cyclonic spells this year, the neutral SOI and the above average hurricane season all suggest that when the positive PNA pattern occurs across Europe, it has more chance of being a cyclonic/west PNA rather than an anticyclonic/east PNA which has been characteristic of recent winters. This is an important point as it implies the UK would be included in the cold & snowy weather rather than it being confined to eastern Europe (Moderate Confiidence)
the weather patterns across North America are expected to oscillate markedly between cold and warm spells and this will no doubt have some effect downstream across the Atlantic & Europe but perhaps in Europe we can expect less volatility than in recent winters because of the expected neutral SOI (Moderate Confidence)
the probability of persistent mild westerly winds looks to be low as all factors appear to point to weak westerlies so such spells are likely to be brief (High Confidence)
the predicted pattern above suggests a cold winter for Europe and the coldest airstreams are likely to come further west and north than in recent years when SE Europe (Greece/Turkey) has received the coldest airmasses. Drought in Iberia is a big concern at the moment and from the water shortage point of view, strongly negative NAO is good news as it means lots of low pressure and precipitation in the Mediterranean.

The Weather

TEMPERATURE: with the occurrence of both northerly and easterly spells, some very cold weather is likely and with neutral SOI, less volatility in the pattern may occur than in recent years which means some long cold spells are possible (last winter managed a four week cold spell in the UK but 6-7 weeks in eastern Europe) but how cold it gets and how long the cold spells last will be dictated by random events which at this long range cannot be predicted. What can be said is that there is a real potential for cold spells this winter as was the case last year and the potential this year (from the point of view of the analyses above) appears to be greater.
75% Confidence Range for winter CET Anomaly (from 1961-90 average): -1.5 to +0.5 C (Moderate Confidence)
RAINFALL: the trend to increasing incidence of cyclonic spells this year and the neutral SOI suggests a cyclonic/west PNA pattern may be recurrent across Europe this winter which would mean a wetter winter than last year which was markedly anticyclonic. Central & eastern regions are therefore likely to have average or slightly above average rainfall while western regions below average rainfall, especially the north-west. (Moderate Confidence)
75% Confidence Range for % of 1961-90 average of England & Wales Rainfall: 90-120%
SNOWFALL: the trend of recent winters to increasing snowfall looks likely to continue. Northerly spells are likely to give some heavy snowfalls to northern Scotland and eastern & western coastal regions exposed to the north and easterly spells are likely to give some heavy snowfalls to eastern Scotland and central & eastern England
above average snowfall generally but average in some western regions (High Confidence)
GALES: weak westerlies and the expected negative NAO suggest the incidence of gales and severe gales is likely to be below average (Moderate Confidence)

Prospects for the Scottish Ski Centres

The prospects for winter 2005/2006 look very good with frequent northerly spells bringing regular snowfalls to the Scottish ski resorts and with perhaps a higher tendency for cyclonicity this season relative to last, snowfalls could be heavier. The likelihood of both northerly and easterly spells means some long cold, dry and sunny spells of weather ideal for skiing with not too many skier days lost to high winds. The early part of the season last year was ruined by a period of strong westerly winds from late December to mid-January but such a spell of weather is considered less likely this winter because SSTs anomalies in the Greenland & Barents Seas are negative currently compared with positive anomalies last year. And other factors do not appear to favour persistent mild westerlies.

The season last year was brought to an early and abrupt end by persistent warm southerlies from mid-March till the end of the month which quickly melted the unconsolidated snow pack. While it is certainly the case that negative NAO can lead to spells of warm southerlies when lows to the SW of the UK stall for a few days, the sustained period of warm southerlies last winter was unusual and is unlikely to be repeated this winter. With the potential for a strongly negative NAO this winter, the 2005/2006 season is likely to start much earlier and end much later than last year's short but sweet season and hopefully be the best since 2001 (Moderate Confidence).

Other Forecasts for the UK & Europe

Seasonal forecasting for Europe is considered very difficult so the Useful Info winter forecast or indeed any other should not be taken too seriously. A number of forecasts from professional and amateur forecasters are shown below. The consensus seems to be for a winter close to or slightly below normal but most predict some cold spells similar to or colder than those of the last few winters. Some of these forecasts are updated monthly and are likely to change but the comments below apply to the 18 November 2005 when this forecast was completed.

Met Office Winter Forecast colder than average winter for much of Europe including southern UK
Metcheck average to slightly above average temperatures across the UK
US National Weather Service unofficial seasonal outlooks temperature charts show a cold winter for much of Europe with the UK on the border of warm and cold anomalies
Net-Weather a slightly colder than average winter in the UK, with a cold January
International Research Institute for Climate & Society above normal over most of Europe including the UK
Ian Currie &nbs; a colder than average December but January & February near normal in the UK
Tom Presutti near or slightly below normal temperatures in the UK with a cold February
UKWeatherWorld slightly below average
The Weather Outlook slightly below average

References

Lamb, H. H. (1950). Types and spells of weather around the year in the British Isles: Annual trends, seasonal structure of the year & singularities. Quarterky Journal of the Royal Meteorological Society, 76, 393-427.

Ratcliffe, R.A.S & Murray, R. New lag associations between North Atlantic sea temperature and European pressure applied to long range weather forecasting. Q. J. R. Met. Soc., London, 96, 1970 pp 226-246.

Rodwell, M.J., Rowell, D.P. and Folland, C. K. 1999: Oceanic Forcing of the wintertime North Atlantic Oscillation and European climate. Nature, 398, 320-323.