The climate of Europe exhibits large differences from west (maritime) to east (continental) and from north (Arctic) to south (Mediterranean). The climatic effects of the distribution of land and ocean are further complicated by numerous high mountain ranges, which act as physical barriers to atmospheric circulation and often introduce large precipitation gradients within small regions (e.g., Frei and Sch�r, 1998). There is no synoptic consistency to the behavior of the European atmosphere across such a heterogeneous climatic domain.

Europe possesses long instrumental data records. The central England temperature series, for example, commences in 1659 (Jones and Hulme, 1997). Reconstruction of regional climates in Europe with proxy data sets identifies the magnitudes of natural temperature variability that have occurred on even longer multi-century time scales: 1,400 years of summer temperatures from tree growth in Fennoscandia (Briffa et al., 1990), 460 years of monthly temperature and precipitation patterns from hydrological and biological evidence in central Europe (Pfister, 1992), and 500 years of annual temperatures from ice cores in Greenland (Fischer et al., 1998). Direct sea-level measurements are available from the 18th century. There have been several attempts at deciphering the remote past, based on proxy data (e.g., Velichko et al., 1998; Zelikson et al., 1998).

13.1.3.1. Temperature

Most of Europe has experienced increases in surface air temperature during the 20th century that, averaged across the continent, amount to about 0.8�C in annual temperature (see Figure 13-1) (ECSN, 1995; Beniston et al., 1998; EEA, 1998). This warming has been largest over northwestern Russia and the Iberian peninsula (Nicholls et al., 1996; Onate and Pou, 1996) and stronger in winter than in summer (Maugeri and Nanni, 1998; Brunetti et al., 2000). An exception is Fennoscandia, which has recorded cooling in mean maximum and mean minimum temperature during 1910�1995 in winter but warming in summer (Tuomenvirta et al., 1998). The past decade in Europe (1990�1999) has been the warmest in the instrumental record, annually and for winter. Increases in growing-season length also have been observed in Europe�for example, in western Russia (Jones and Briffa, 1995) and in Fennoscandia (Carter, 1998). Trends in the intensity of the growing season as measured by growing degree days, however, are more ambiguous in both of these studies. The evidence for longer growing seasons in Europe also is supported by phenological data collected in central Europe (Menzel and Fabian, 1999). These data point to increases of about 10 days in average growing-season length since the early 1960s. Other biological indicators of a changing growing season in Europe include poleward shifts of 35�240 km during this century in entire ranges of 34 different butterfly species (Parmesan et al., 1999) and earlier breeding times for several species of amphibians and migratory birds during the past few decades (Forchhammer et al., 1997).

Figure 13-1: European annual mean temperature anomalies over land, 1901�1995, for the region 35�N to 75�N and 30�W to 50�E, with respect to 1961�90 mean. 10-pt filter applied. Data from IPCC Data Distribution Centre.

13.1.3.2. Precipitation

Trends in annual precipitation differ between northern Europe (wetting) and southern Europe (drying), reflecting a wider hemispheric pattern of contrasting zonal-mean precipitation trends between high and low latitudes (Dai et al., 1997; Hulme et al., 1998). Precipitation over northern Europe has increased by 10�40% in the 20th century, whereas some parts of southern Europe have dried by as much as 20% (see Figure 13-2). Romero et al. (1999) show that the numbers of days with precipitation over the Spanish southern coast and the Pyrenees region have decreased by 50 and 30%, respectively, in 1964�1993. In Italy, total precipitation in the 20th century has decreased by about 5% in the north and by about 15% in the south (Buffoni et al., 1999; Brunetti et al., 2001). Analysis of moisture extremes over Europe, using the Palmer Drought Severity Index (PDSI�Briffa et al., 1994) showed strong decadal-scale variability in drought frequency; the 1940s and early 1950s experienced widespread and severe droughts�a pattern repeated in 1989 and 1990.

13.1.3.3. Extreme Events

Analyses of trends in extreme weather events in Europe generally have been limited to national studies, making it difficult to provide a continent-wide overview of changes in hot/cold day frequencies, precipitation intensities, or gale frequencies. Gruza et al. (1999) analyzed data over the whole of Russia, the western third of which falls into our definition of Europe, and found a slight increase over the 20th century in the Climate Extremes Index (CEI), which combines daily temperature, daily precipitation, and drought extremes. Analysis of 85 long-term maximum 1-day precipitation records in the Nordic countries indicates that there is a maximum in the 1930s and a tendency of increasing values during the 1980s and 1990s�decades with relatively high regional summer temperatures. In western Norway, the past 2 decades have been exceptional, with substantial increases in orographic precipitation during autumn, winter, and spring (F�rland et al., 1998). Elsewhere, daily precipitation intensities over the UK have increased in winter over recent decades (Osborn et al., 1999), although not in other seasons. This increase in UK winter precipitation intensities has been paralleled by a marked decrease in the frequency of cold winter days in the UK (Jones et al., 1999). Changes in storminess over the northeast Atlantic have been analyzed by Schmidt et al. (1998) and WASA (1998); they show that although storminess has increased in recent decades, storm intensities are no higher than they were early in the 20th century. Wave heights around the shores of northwest Europe also show large decadal variability, but no long-term trends emerge.

13.1.3.4. North Atlantic Oscillation

One important cause for interannual and perhaps interdecadal climate variability in Europe, particularly in winter, is the North Atlantic Oscillation (NAO). NAO is a measure of the strength of the westerly flow over the North Atlantic; records go back to the early 19th century (Jones et al., 1997). Over the past 4 decades, the NAO pattern gradually has altered from the most extreme and persistent negative phase in the 1960s to the most extreme positive phase during the late 1980s and early 1990s�a trend that has been responsible for relatively mild and wet winters during the latter period over much of northwest Europe (Hurrell and van Loon, 1997) and can help explain the observed narrowing in DTR (Tuomenvirta et al., 1998). Following its long period of amplification, the NAO index underwent a sharp decrease to a short-lived minimum in the winter of 1995�1996, with radical recognizable changes in the North Atlantic. It should be noted that the NAO appears to exert significant control on the export of ice and freshwater from the Arctic to the open Atlantic (WCRP, 1998). Since that temporary minimum, a recovery toward positive values of the NAO has been observed. NAO certainly is a more important influence on European climate than the El Ni�o-Southern Oscillation (ENSO), although several studies have explored the influence of ENSO on European precipitation variability, and links between NAO and ENSO are being sought. Evidence for such an influence is weak and shows phase differences through the continent (Fraedrich and M�ller, 1992; Fraedrich et al., 1992; Rodo et al., 1997; Price et al., 1998).

The response of human activities and the natural environment to current weather perturbations provides a guide to where critical sensitivities to future climate change may lie. Even if adaptation may modify the response, analysis of present-day sensitivities helps us understand the likely impact of future climate changes on our environment and lifestyle.

The principal sensitivities in Europe to current climate and weather conditions are to:

• Extreme seasons � Exceptionally hot, dry summers � Mild winters
• Short-duration hazards � Windstorm (possibly associated with tidal surges) � Heavy rain leading to river-valley flooding and flash floods
• Slow, long-term change � Coastal squeeze.

Sensitivities vary across Europe as a result of differences in climate and topography, as well as the socioeconomic environment, but all of Europe is sensitive to the foregoing anomalous conditions.

13.1.4.1. Extreme Seasons

13.1.4.1.1. Exceptionally hot, dry summers

Extremely hot and dry summers were experienced in 1995 throughout much of western Europe (Palutikof et al., 1997) and in 1992 in CEE (Schellnhuber et al., 1994). Summer drought conditions can have a devastating impact on the natural environment, particularly by reducing the availability of water for flora and fauna. Pollution levels may rise to high levels in the subsiding anticyclonic air, with add-on effects for human health in terms of increased numbers of asthma attacks and associated hospitalizations. High water temperatures often lead to algal blooms, which may make water bodies unusable for recreational purposes.

Figure 13-2: Trends in annual (top), winter (middle), and summer (bottom) precipitation expressed as % per century and calculated on 2.5� grid. Black circles denote wetting and white circles denote drying.Magnitude of trend is related to circle size. Shaded trends are significant at 90% (data from New et al., 1999).

Agriculture and water supply are the two economic sectors that may be most severely affected by exceptional heat and drought in summer. Large proportions of the working populations in Mediterranean countries are engaged in agriculture, and even in �normal� years water shortages may be a limiting factor. These countries, therefore, appear to be more vulnerable to exceptional summer weather. An increased frequency of hot, dry summers is likely to reduce tree growth and affect timber yield and quality. Reduced tree vigor often favors outbreaks of pests and pathogens, and hot and dry conditions trigger wildfire.

13.1.4.1.2. Mild winters

Milder winters exert a major impact on the natural environment. Overwintering of species is more successful. However, failure to kill off pests and diseases that prey on wildlife, as well as failure to cull weaker members of the species—which then compete for food in the following springs and summers—in the end may be counterproductive. Cannell and Pitcairn (1993) show that during two mild winters in the UK, various life events occurred earlier—for example, insects and other animals moved about and fed more actively, causing widespread aphid damage on oak and mite damage on lime. Milder winters can affect fulfillment of chilling requirements in some plant species, which would have an impact on the formation of leaf and flower buds (e.g., Murray et al., 1989). This could lead to regeneration failure; some susceptible species are common across much of Europe today (e.g., Picea abies) (Sykes and Prentice, 1996).

The skiing industry in Europe is vulnerable to variations in snowfall�that is, too little or too much snow. Many European ski resorts have suffered severe shortfalls in earnings when snow cover was not sufficiently deep during Christmas and winter school holidays. If temperatures remain below freezing, a more active hydrological cycle might lead to more episodes of heavy snowfall; the winter of 1998�1999 is a good proxy for this situation. Although the number of cold spells generally has decreased over the past few decades, cold spells still have a range of impacts on agriculture, settlements, the built environment, and transport.

13.1.4.2. Short-Duration Hazards

Severe windstorms affected Europe in the 1990s. The January 1990 storm (Daria) caused insured losses of about US$5.7 billion (at 1997 prices) and 95 deaths; in the following month, the storm of 26 February (Vivian) caused a further US$3.9 billion in insured losses (Swiss Re, 1998) and 64 deaths. The UK experienced a severe windstorm at Christmas 1997, causing 13 deaths and losses of US$500 million (Swiss Re, 1998).

The major impact of windstorm on the natural environment is on trees, woodland, and forest. It is estimated that 15 million trees (5 months� production of coniferous wood and 2 years� production of broadleaf timber production�see Quine, 1988) were blown down in the UK during the October 1987 storm.

Windstorm associated with tidal surge is a particularly deadly combination for low-lying coastal areas, as epitomized by the event of 1953. Severe northerly winds combined with an exceptionally high spring tide caused overtopping in The Netherlands and much of East Anglia, resulting in more than 2,000 deaths. Since then the policy has been to keep the sea out at all costs, and in fact several storm surges of similar size have occurred with no loss of life.

Several devastating river floods, of high severity, occurred in Europe in the 1990s (Kundzewicz and Takeuchi, 1999). Among the flood events to hit the headlines was the Odra flood in the summer of 1997, during which historic flow records were broken. The deluge caused more than 100 deaths and economic damage in excess of US$5 billion (Kundzewicz et al., 1999).

13.1.4.3. Coastal Squeeze

Coastal squeeze occurs when coastal habitats are �squeezed� between rising sea level and fixed hard defenses (Bijlsma et al., 1996). It is occurring in most northwest European countries, and it already is a coastal management issue (Rigg et al., 1997). It is considered in more detail in Sections 13.2.1.3 and 13.3.5.

IPCC Homepage