Predictions of summer 2005 in the UK & Europe

Useful Info Forecast for the UK & Europe Summer 2005

Andrew Masterman 28 May 2005

A number of factors which were used to make the summer forecast are described below:

Jan to March North Atlantic Oscillation (NAO)

Although the NAO in the first half of winter 2004/2005 was strongly positive, in the latter half, it was strongly negative as can be seen in this chart of January-March NAO. And this is relevant to summer 2005 as January-March NAO accounts for 25% of the variation in the summer C index for the UK (r = -0.373, P<0.01). The C index is a measure of anticyclonicity across the UK and warmer more anticyclonic summers tend to occur after Jan-March periods with positive NAO (strong westerlies) and wetter, cooler and more cyclonic summers tend to follow January to March periods with negative NAO (weak westerlies). So the January-March NAO in 2005 favours a cooler cyclonic summer. This relationship owes to the tendency of the phase of the NAO to persist from late winter through the summer months and indeed from year to year. And the summer NAO and the summer C index are also negatively and significantly correlated (r = -0.408; P<0.01) indicating that cool unsettled cyclonic conditions tend to prevail across the UK when the NAO is negative in summer months and the this correlation is somewhat stronger for the July-August period (r = -0.498; P<0.001) than for the summer as a whole.

Figure 1 shows the relationship between the January-March NAO and the following summer C index for 1951-2004. During the 1960s when winters were cold with strongly negative NAO, the summers were cool and unsettled (high C index) while conditions switched to milder westerly winters in the 1970s with warmer more anticyclonic summers including the two hot summers of 1975 & 1976. From 1979 to 1987, the UK climate flipped back to colder winters again and mostly cool unsettled summers although the summers of 1983 & 1984 were good. Since 1989, most winters in the January-March period have had highly positive NAO which brought mild winters and summers from 1989 to 1995 were remarkably warm and anticyclonic confirming the usefulness of January-March NAO as a predictor for summer weather. From 1996 to 2004, January-March NAO has shown a trend to more negative NAO and the summers have become more unsettled and cyclonic with the notable exception of 2003. So this trend to more cyclonic and unsettled summers needs to be considered in the summer forecast for 2005.

Figure 1 Relationship between the January-March North Atlantic Oscillation (NAO) and the summer C index 1951-2004 (r = -0.373; P<0.01)

Recent trends in the NAO continue to show a strong bias to negative NAO at times which has brought several weeks of below average temperatures to the UK in the middle part of May in addition to the sustained period of below average temperatures from mid-February to mid-March. So the signs are there that the NAO will be negative at times during the summer months resulting in cool cyclonic spells across the UK.

Position of African Inter-Tropical Convergence Zone (ITCZ)

In 2003, it was the unusually far northward position of the ITCZ across Africa which was a major contributor to the heatwave summer across Europe as it pushed the belt of sub-tropical high pressure northwards up into Europe bringing anomalous heat with it. At the time of writing of this forecast during the last week of May 2005, the ITCZ was 1 degree latitude south of its average position for the time of year and this position represents a movement south compared to recent weeks. Given that the northward extent of the ITCZ generally roughly coincides with the summer solstice and the solstice is only three weeks away now, it is fair to say that the African ITCZ currently does not indicate anomalous heat building across Europe in 2005.

Atlantic Sea Surface Temperature (SST) Anomalies

Figure 2 shows current global SST anomalies and several features should be noted:

the marked tripole of anomalies in the western North Atlantic: warm anomalies in the tropical and northern part and cold anomalies sandwiched in between. This favours negative NAO and probably explains why there is currently a strong bias for negative NAO patterns in the atmosphere and why a continued bias for negative NAO is likely to persist.
average SST anomalies around and to the south-west of the UK which have cooled from warmer than average owing to the higher incidence of cold spells in recent months owing to negative NAO patterns. Warm anomalies to the SW of the UK not only warm air arriving in the UK from the prevailing wind direction but also favour a strong Azores high pressure which is characteristic of warm anticyclonic summers in the UK. Warm SSTs around the UK also favour high pressure near/over the UK so the current average SST anomalies favour neither a strong Azores high, nor high pressure near/over the UK.
colder than average SST off the SE coast of the US: a strong ridge off the SE coast of the US is favoured by warm SST anomalies here and it supports a strong ridge across western Europe. In this case, the cold anomalies here do not favour a strong high pressure in western Europe

Figure 2 Global Temperature SST Anomalies for the week ending 21 May 2005 relative to 1971-2000 averages.

The justification of the Jan-Mar NAO and Atlantic SST anomalies as predictors of summer weather in the UK is shown below where some good and poor summers are compared:

July-Aug 1984 were warm summer months and followed a Jan-March period with strongly positive NAO (+4.26) and SST anomalies for May 1984 show plenty of warm SSTs to the west and south-west of the UK to support a strong Azores high and high pressure near the UK and warm anomalies off the SE coast of the US to support a western Europe high were also present.
the summer of 1985 was particularly unsettled with both June & August being much cooler than average. The Jan-Mar period had strongly negative NAO (-3.11) and Atlantic SST anomalies for May 1985 show average SSTs near and to the SW of the UK with warm anomalies west of the Azores while there were warm anomalies off the SE coast of the US
1986 was another cool and unsettled summer and the Jan-Mar period also had strongly negative NAO (-2.56). Atlantic SST anomalies in May 1986 like in May 1985 show average SSTs near and to the SW of the UK with warm anomalies to the west of the Azores but in this year, SSTs were average off the SE coast of the US
1989 was a warm anticyclonic summer and the Jan-Mar period had strongly positive NAO (+7.14). Atlantic SST anomalies for May 1989 show warm anomalies around and to the SW of the UK and mostly warm anomalies off the SE coast of the US
1990 was another warm anticyclonic summer and the Jan-Mar period had strongly positive NAO (+8.61). Atlantic SST anomalies for May 1990 show warm anomalies around and to the SW of the UK and mostly warm anomalies off the SE coast of the US
1995 was a very warm anticyclonic summer and the Jan-Mar period had a strongly positive NAO (+5.83). Atlantic SST anomalies for May 1995 show a huge area of warm anomalies to the SW of the UK and particularly near to the Azores. And SST anomalies off the SE of the US were also warm.
summer 2002 was a poor unsettled summer which led to severe flooding in parts of Europe. Jan-Mar NAO was strongly positive (+5.41) favouring a warm anticyclonic summer but Atlantic SST anomalies for 28 May 2002 show average SST anomalies to the SW of the UK although around the Azores themselves, they were warm. Probably the most significant factor which accounted for the unsettled weather was the region of strong thermal contrast running at about 40 N right across the Atlantic. Jetstreams high in the atmosphere tend to follow the line of strong thermal contrast in the oceans and are strengthened by them. In this case, the jetsream was directed into Europe at a latitude of 40 N directing a stream of Atlantic depressions into the heart of Europe which eventually led to severe flooding in August 2002.
2003 was the year of the amazing summer heatwave in Europe and Jan-Mar NAO was moderately positive (+2.57). Atlantic SST anomalies for 27 May 2003 show generally warm anomalies around and to the SW of the UK and importantly very warm anomalies around the Azores. Warm SST anomalies were also present off the SE coast of the US. The African ITCZ was anomalously far north in this summer which was an important factor in the hot weather
summer 2004 was quite unsettled and warm with both June & August having above average temperatures but in the first three weeks of July, the weather was cool and unsettled before turning hot at the end of July and the first ten days of August. The Jan-Mar NAO was neutral (+0.04) but Atlantic SST anomalies for 29 May 2004 show warm anomalies around and to the SW of the UK and off the SE coast of the US

From the above, it can be seen that the combination of Jan-Mar NAO and Atlantic SST anomalies generally provide useful guidance on the type of summer ahead. Of course, this method does not give useful guidance when the different predictors give conflicting guidance or when they have neutral values, therefore in some years, it may be difficult to come to a conclusion on the nature of the coming summer using these predictors.

On this occasion in 2005, all these predictors favour a summer biased towards cooler unsettled conitions rather than warm anticyclonic weather. One final predictor which is relevant to August weather is considered.

Atlantic Hurricane Forecast

Dr Gray is one of the leading Atlantic Hurricane forecasters and his latest forecast issued 1 April is "We foresee an above-average hurricane season for the Atlantic basin in 2005. Also, an above-average probability of U.S. major hurricane landfall is anticipated". Although the official start of the Hurricane season is 1 June, in practice, relatively few hurricanes occur prior to late July and the main season is August-September. An above-average hurricane season as in 2005 is relevant to the UK only if strong anticyclonic conditions are not dominating the UK summer. The remnants of ex-hurricanes may become absorbed by developing Atlantic depressions which then move across the UK and the ex-hurricanes add extra moisture and heat to the depression resulting in heavier rain and higher temperatures than expected from an Atlantic depression. Clearly if the UK is stuck under an anticyclone all August, then such depressions will never cross the UK but in this case where cyclonic conditions are expected (according to the predictors above), then during August-September, it is likely that depressions containing ex-hurricanes/tropical storms will affect the UK.

The General Weather Pattern:

the last few months have seen some quite long spells of negative NAO as discussed earlier which has given some long spells with below average temperatures (chart) as northerly/arctic airstreams have spread south. But there have also been some southerly spells which have brought some anomalous warmth: second half of March; first few days of May; 26-27 May. Such large swings in temperature brought about by changes in wind direction from north to south and vice versa are characteristic of meridional weather patterns which have prevailed in the northern hemisphere for several years now. In Britain, the southerly spells have mostly been longer than the northerly spells resulting in well-above average temperatures. But recently, northerly spells have been increasing in frequency as can be seen in Figure 3 which shows quintiles of PSC indices from January 2002 to February 2005 where high P index indicates above average incidence of westerlies, high S index represents above average incidence of southerlies and a high C index represents above average cyclonicity (below average anticyclonicity; PSC indices are defined here). It is evident that above average southerliness dominated 2002 & 2003 but in 2004, five months had above average northerliness and January & February 2005 both had above average northerliness. PSC data for Mar-May 2005 are not yet available but May 2005 has certainly continued the bias to northerliness but the northerliness of the first half of March 2005 was probably balanced by the southerliness of the second half. Figure 4 shows that UK temperatures have come down somewhat in the last year and the mean monthly anomaly for Jan to May 2005 is +1.1 C compared with rather higher anomalies of around 1.5 C during most of 2002 & 2003: nevertheless, they are still well above the long term average but this chart shows that temperatures are now falling well below average for long periods at times. The cooling effect of recent colder weather patterns can be seen Figure 2 which shows sea temperatures around the UK are currently close to the long term average. This forecast assumes this shift to a higher incidence of cold weather patterns relative to recent years will continue during summer 2005 giving an overall bias to above average northerliness (suggested by Figure 3) and above average cyclonicity as indicated by the predictors discussed above.

Figure 3 Quintiles of PSC indices January 2002 to February 2005

Figure 4 Monthly anomalies of mean Central England Temperature (CET) relative to 1961-90 average: January 1970 to April 2005

the trend to more cyclonic summers post 1997 (except 2003) in Figure 1 is fairly clear. This chart shows that from 1880-1970, July & August were on average amongst the most cyclonic months of the year (together with November & December) and it also shows the marked rise in the P index during June as the westerlies return after their spring minimum. This chart shows how anomalously anticyclonic Julys & Augusts have been from 1989-2002 and this chart shows the bias to southerliness from 1989-2002 relative to 1880-1970 averages. Recent Julys & Augusts have been unusually warm and anticyclonic : Julys & Augusts are normally quite mobile westerly and cyclonic months with a completely different feel to recent late summers. The official scientific explanation for these warmer anticyclonic Julys & Augusts is that Man's burning of fossil fuels is to blame by increasing atmospheric levels of carbon dioxide and causing changed weather patterns in Europe: these summers are an unforeseen benefit of burning fossil fuels! The UKCIP02 scenarios are the official climate predictions for the UK in the 21st Century and for summers, the recent trends to warmer and anticyclonic conditions is extrapolated such that by the end of the century under the high emissions scenario, summers will be 4 to 5 C warmer and 50% drier. However it is the author's opinion that natural variability is the cause of the anomalous climate in the UK since 1989 and that there are already signs that northern hemisphere patterns are shifting to a more negative NAO regime meaning colder winters and cooler more unsettled summers: the trend to increasingly negative winter NAO shown in Figure 1 is clear and has been accompanied by an increasing incidence of cold spells in the UK & Europe with snow in recent years and the post 1997 shift to more cyclonic summers is also apparent in Figure 1.
so a summer more like the 1880-1970 average is expected where a more mobile westerly pattern sets in by the end of June to give an unsettled cyclonic July-August period. There is no Atlantic SST profile like in 2002 which directed the jetstream a long way south at 40 N into the heart of Europe but the strong negative NAO signal suggests that Atlantic depressions will probably travel under the high in the Greenland region and go across the Atlantic at roughly the UK's 1atitude and then off into Scandinavia or the Low Countries. But without a favourable Atlantic SST profile strengthening the jetstream, there is no reason to suppose long periods of really wet and windy weather, just a more typical summer of two hot days and then a thunderstorm as a succession of Atlantic lows cross the Atlantic interspersed by weak ridges.
during August, remnants of ex-hurricanes/tropical storms are likely to get absorbed into some Atlantic depressions bringing a risk of some heavy rainfall totals. However, an August like last year is unlikely as the pattern across the North Atlantic was very static resulting in depressions stalling near the UK giving heavy rainfall and the four hurricane strikes on Florida.
negative NAO weather patterns sometimes lead to a stalled depression to the SW of the UK/Azores region which pump warm southerlies across western Europe and this occurred for a long period in the second half of March and briefly in early May and 26-27 May. Further spells like this are likely during the summer giving some hot spells but they are likely to be brief before cooler Atlantic air sweeps back in
the overall pattern for Europe therefore is a continuation of the May pattern for much of June with a bias to cool NWly winds with a risk of brief southerly blasts before an unsettled mobile cyclonic westerly pattern sets in across the north (UK/Scandinavia/Low countries) for July & August giving temperatures close to average here which implies that the Mediterranean should have a settled anticyclonic summer but with the mobile pattern across the top and the African ITCZ not unusually far north, temperatures should not become unusually hot. However, as Mediterranean SSTs are currently above average (Figure 2) after a warm May, a warmer than average summer seems likely in southern Europe and western/central Europe may get some short-lived hot southerly blasts as explained in the bullet point above

The Weather

TEMPERATURE: 75% confidence interval for CET: +0.5 to -0.5 from 1961-90 average. But northern Scotland likely to be cooler than average as negative NAO patterns push cooler arctic air south and warm airmasses from the south will probably fail to get so far north. June & July are likely to have temperatures around average while August may well continue the recent trend of warm Augusts as ex-hurricanes/tropical storms are likely to inject some anomalous warmth into Atlantic depressions.
RAINFALL: June is likely to have average rainfall and July & August slightly above average rainfall but overall, it is unlikely that summer 2005 will be as wet as last year which was the 35th wettest summer since 1766 (140% of average): summer rainfall % of average estimated to be 110-120

Other Forecasts

There is some talk on the UK weather forums of hot summer like 1995 or 2003 coming up which is quite different to the Useful Info Summer Forecast. Metcheck have gone for a slightly warmer than average summer while the latest European headline from the UK Met Office is for mixed signals for the UK (does this mean average or they haven't a clue?) with a slight bias to warmer than average in southern Britain. Ian Currie has forecast a hot summer and the International Research Institute for Climate Prediction also go for a warmer than average summer.