Climate change: European scale
Global warming until now
The global (land and ocean) average temperature in the decade 2006-2015 was 0.83°C to 0.89°C higher than pre‑industrial levels (1850–1899 average). This makes it the warmest decade on record. 15 of the 16 warmest years on record have occurred since 2000, and 2015 was the warmest year on record - around 1°C warmer than the pre-industrial period. Over the decade 2006-2015, the rate of change in global average surface temperature was between 0.10 and 0.24 °C per decade (124).
According to a recent study there is no statistical evidence for a ‘slowdown period’ with a significantly reduced rate of warming in the period 1998-2014 or an acceleration of global warming in recent years 2014, 2015 and 2016. Rather, the data on global surface temperatures from 1970 to 2016 are fully consistent with a steady global warming trend since the 1970s, superimposed with random, stationary, short-term variability (141). Whether there was a hiatus or slowdown at some point is still debated, with some arguing strongly for and others saying it lacks scientific basis; the hiatus has not changed our projections of the overall magnitude of climate change or the emission reductions that are required to address it (142).
Standard data sets of worldwide surface temperatures may underestimate recent global warming. A recent analysis (66) indicates that the planet has warmed most where scientists are watching least. A new method (kriging) was used to fill in missing temperatures over sea ice. The results show a slowdown in warming that’s half as big as previously thought: temperatures in recent years go up, and temperatures around 1998 go down.
Recent extreme events
Following an extreme climate or weather event, the standard response from scientists has typically been that global warming does not “cause” any single event in a deterministic sense, but it can make some of them more likely to occur or more intense when they do. Because of advances in the relatively young science of extreme event attribution, however, it is now possible in some cases to provide quantitative information about how climate change may have impacted the probability or intensity of an individual event (143).
The occurrence of any individual extreme event, by itself, does not prove or disprove that the climate is changing. Nevertheless, event attribution studies seek to calculate how much human-induced climate change has affected an individual event’s magnitude or probability of occurrence (144). Confidence in attribution analyses of specific extreme events is highest for extreme heat and cold events, followed by hydrological drought and heavy precipitation. There is little or no confidence in the attribution of severe convective storms and extratropical cyclones (143).
Specific individual extreme weather events cannot be attributed to climate change. It has been shown, however, that the overall probability of climate change having an effect on extreme events can be quantified by using a global aggregation of events data (81). From such an analysis it was concluded that today about 75% of the moderate daily hot extremes over land are attributable to global warming. This high fraction is due to the fact that an upward shift of the temperature distribution rapidly increases the chances of temperatures in the upper tail of the undisturbed distribution. For moderate daily precipitation extremes it was concluded that 18% of these events are attributable to global warming, and this fraction is forecast to rise to about 40% when warming reaches 2 °C relative to pre-industrial temperatures (81).
Extreme events in previous decades
Over the past decade, a new field of science called “extreme event attribution” has emerged, which allows for answering the question: did climate change play a role in this specific extreme event? Many attribution studies of recent hot extremes find a dominant contribution of anthropogenic climate change to their probability of occurrence. For example, human-induced climate change increased the likelihood of the record hot Australian summer of 2012/2013 by at least five times (99) and the record warm Central England temperature of 2014 by at least 13 times (100).
It might be expected that record warm events in previous decades may also be attributable to human-induced climate change. This was studied for observational data sets of large parts of the world, back to 1900 (98). It turns out that the effect of climate change on extreme events extends back several decades. An example is the record- breaking hot summer of 1997/1998 in Australia. The study was carried out for both record-breaking hot years and record-breaking hot summers. The results for both hot years and hot summers are similar: the likelihood of record hot events due to anthropogenic influences has increased. Many past record hot summers have a substantial fraction of their probability of occurrence attributable to anthropogenic climate change, including high profile events such as the 2003 hot European summer.
Whether or not an anthropogenic influence can be discriminated from interannual temperature variability depends on the size of the anthropogenic effect with respect to the size of variability. Interannual variability reduces with an increase of the spatial scale of a study. Thus, for time series representing larger regions of the world the imprint of an anthropogenic influence extends further back as interannual variability is reduced allowing the anthropogenic signal to appear earlier (101). The largest spatial scale is a global series of observations: all of the last 16 record hot years in this global series, starting as early as 1937, have a fraction of their probability of occurrence attributable to the anthropogenic influence on the climate (98).
Another interesting feature is the fact that in the middle of the twentieth century the likelihood of hot years reduces in the Northern Hemisphere. This effect is related to anthropogenic aerosols (102) reducing the likelihood of hot years. The effect of anthropogenic greenhouse gases supersedes that of anthropogenic aerosols in later decades leading to an increase in the likelihood of hot years related to the total anthropogenic influence (103). In Australia, where the anthropogenic aerosol effect is weaker, the emergence of an anthropogenic influence, due to greenhouse gas emissions, is earlier (98).
After the 1980s, there are indications that warming in the Mediterranean region has been larger than at global scale; before the 1980s this cannot be concluded, climate variability is too large (179).
Global warming in the 21st century
IPCC 5th Assessment Report
For the Fifth Assessment Report of the IPCC a different type of climate change scenarios has been defined than for the previous IPCC reports: the so-called Representative Concentration Pathways (RCPs) (73). More information on these scenarios is presented below under the heading 'Climate change scenarios'. Scenarios that are often used for climate projections are the low RCP2.6, the moderate RCP4.5 and the high-end RCP8.5 scenario.
According to the IPCC the projected change in annual global mean surface air temperature (°C) with respect to the present day (1986–2005) is between 1.0 (RCP2.6) and 2.0 °C (RCP8.5) by the mid-21st century and between 1.0 (RCP2.6) and 3.7 °C (RCP8.5) by the end of the 21st century. These projections are based on tens of models. See the table below (104).
Table. Projected change in annual global mean surface air temperature (°C) by the mid- and end of the 21st century relative to present day (1986–2005) for RCP2.6 (32 models), RCP4.5 (42 models) and RCP8.5 (39 models) as well as the 5 - 95 % ranges from the models’ distribution (numbers between brackets) according to the IPCC 5th Assessment Report (105).
|2046 - 2065||1.0 (0.4 - 1.6)||1.4 (0.9 - 2.0)||2.0 (1.4 - 2.6)|
|2081 - 2100||1.0 (0.3 - 1.7)||1.8 (1.1 - 2.6)||3.7 (2.6 - 4.8)|
In a special report in 2019, the IPCC has presented updates of the projections of global warming, and used different time slices and a different reference (the period 1850-1900) (213):
Table. Projected global mean surface temperature change relative to 1980 - 1900 for two time periods under fpur RCPs.
mean (likely range)
|2031 - 2050||1.6 (1.1 - 2.0)||1.7 (1.3 - 2.2)||1.6 (1.2 - 2.0)||2.0 (1.5 - 2.4)|
|2081 - 2100||1.6 (0.9 - 2.4)||2.5 (1.7 - 3.3)||2.9 (2.0 - 3.8)||4.3 (3.2 - 5.4)|
Several studies have projected climate change beyond 2100 based on the so-called extended concentration pathways (ECPs). The central estimates for global mean temperature increase by 2200, relative to pre-industrial levels, are between 1.3°C for ECP2.6 and 7.1°C for ECP8.5 (131).
IPCC 4th Assessment Report
In the previous Fourth Assessment Report somewhat different results on projected global warming were reporting. One of the reasons for these different results when compared with the more recent Fifth Assessment Report is the fact that different climate change scenarios were used.
According to the Fourth Assessment Report the projected increase in this century globally is between 1.8 and 4.0°C (best estimate), and is considered likely (66% probability) to be between 1.1 and 6.4°C for the six IPCC climate change (SRES) scenarios and multiple climate models, comparing the 2080–2100 average with the 1980–1999 average. These scenarios assume that no additional policies to limit greenhouse gas emissions are implemented (2). The range results from the uncertainties in future socio‑economic development and in climate models.
Global heat wave changes in the 21st century
When will hot summers be so widespread that half of the population regularly experiences summers that are hotter than any summer of the past? Much sooner than most people probably expect, according to a recent study (114). Under an intermediate (RCP4.5) scenario of climate change, more than half of the world’s population is projected to experience a summer that’s hotter than any summer of the past in 1 out of 2 years within 20 years, and nearly the entire world population just after 2050. In fact, in just over a decade such a hot summer in 1 out of 2 years may already be a fact for half of the world’s population under a high-end scenario (RCP8.5) of climate change (114). These results were obtained from a probabilistic approach that combines information from observations on temperatures in the past (the period 1950 - 2012) and future climate model projections for the period 2013 - 2069. For this study global land surface was divided into 26 regions.
In most of the world a large part of the observed temperature changes in the past can be attributed to anthropogenic forcing. It is estimated that hot summers in many regions of the world are now ten times more likely to occur than would have been the case if greenhouse gas and aerosol concentrations had stayed at preindustrial levels (114). Human impact is largest in the Mediterranean and the Sahara region. These results agree with previous studies (115).
The study on the likelihood of hot summers for different parts of the world suggests that the Mediterranean, Sahara, large parts of Asia and the Western US and Canada will be among the first regions for which hot summers will become the norm (i.e. occur on average every other year), and that this will occur within the next 1 - 2 decades. After 2035 very few summers in the Mediterranean will be colder than the heat wave summer of 2003. These early dates are consistent with the strong summer temperature responses to climate change in these regions as documented in the IPCC 4th assessment report (AR4) (114).
Global heat waves and the Paris agreement (2015)
According to the Paris agreement of 2015, global warming should be kept below 2 °C above pre-industrial levels, and preferably be limited to 1.5 °C. What if we do succeed in reaching these goals, what would future heat waves look like? This was studied with a very high-resolution global model. The model results were used to calculate an index for wave magnitude that takes into account both heat wave duration and intensity. The higher this index, the more extreme the heat wave. Three types of heat waves were distinguished: severe, extreme and exceptional heat waves. These heat wave types more ore less agree with the heat waves of the Balkans (2007), France (2003) and Russia (2010), respectively.
1.5 °C global warming
Even at 1.5 °C global warming, a significant increase in heat wave magnitude is expected over Africa, South America, and Southeast Asia. Deadly heat-related climatic conditions are expected to increase over most of the tropical developing countries. Heat waves will become not only more intense, but also more frequent (165).
Typical European summers at a global warming level of 1.5°C are those of 2003 and 2017. The 2003 summer temperatures in Europe will be commonplace around the 2040s (173).
2 °C global warming
At 2 °C global warming, most of the tropical countries will face severe heat waves at least once every five years. The increase from 1.5 °C to 2 °C global warming will double the frequency of extreme heat waves over most of the globe, including most of the tropical countries, the continental United States and the Mediterranean countries. Exceptional heat waves, with magnitude similar or higher than that of Russia 2010, are expected to occur especially over regions particularly vulnerable to climate change, such as Algeria, the Horn of Africa and the Arabian Gulf (165). These regions have been identified as hotspots for critical future human habitability because of extreme temperatures (166). In fact, exceptional heat waves will occur in some areas where these heat waves do not occurunder 1.5 °C warming. In particular, 10% of the land over East Africa and Southeast Asia will be affected by exceptional heat waves at least once every 20 years (165).
Summe rtemperatures in Europe in 2010 are comparable to the average summer in a 2°C warmer world. Under 2°C of global warming, every other European summer month would be warmer than the warmest summer month on record in current climate conditions (174).
Impact on population
- Severe heat waves (like Balkans 2007): At 1.5 °C global warming, 13.8% of the world population will be exposed to severe heat waves at least once every 5 years, and around half of the world population at least once every 20 years. These fractions strongly increase to 36.9% (every 5 years) and 70.9% (every 20 years), respectively, when global warming rises from 1.5 °C to 2 °C (165).
- Extreme heat waves (like France 2003): The number of people exposed to extreme heat waves every 20 years will increase from 9.0% to 28.2% when global warming increases from 1.5 °C to 2 °C, corresponding to a difference of around 1.4 billion people.
- Exceptional heat waves (like Russia 2010): At 2 °C global warming, 8.3% of the world population may be hit by exceptional heat waves at least once every 50 years. This corresponds to around 452 million people more than in a 1.5 °C world. These people are mainly located in developing countries such as the Horn of Africa, the area of the Gulf of Guinea, Indonesia and the coastal regions of South-America from Venezuela to Brazil.
Compared with 2 °C warming, limiting global warming to 1.5 °C will result in around 1.7 billion fewer people frequently (every 5 years) exposed to severe heat waves, and 420 million and 65 million fewer people frequently exposed to extreme heat waves and exceptional heat waves, respectively.
This study shows that implementing ambitious mitigation strategies to limit warming below 2 °C or even to 1.5 °C will drastically reduce exposure to the most severe impact of temperature related extreme events in terms of intensity and frequency of extreme heat waves. What’s more, it will drastically reduce the probability of occurrence of exceptional heat waves, with magnitude similar of higher than that occurred in Russia 2010 (165).
Summer monthly mean temperatures at 2°C of global warming are projected to become around 1°C higher over Europe than at 1.5°C of warming (175). On a daily basis, the differences in extreme temperatures will probably be even larger. At 2°C global warming extreme maximum daily temperatures can be reached that are up to 1.5°C higher than maximum temperatures at 1.5°C global warming, according to a recent model study on summer temperature variability at these levels of global warming (174). This is due to the fact that the variability in summer temperatures is higher at 2°C global warming compared with 1.5°C. As a result, the 10% most extreme summer maximum temperatures in a 2°C warmer world cannot be reached when global warming is restricted at 1.5°C. 10% may not seem that much, but these events would correspond to the most extreme and severe heat waves, the ones with the most critical consequences (174).
Air temperature changes in Europe until now
Between 2006 and 2015, the average annual temperature over the European land area increased by 1.45°C to 1.59°C, relative to the pre-industrial period. This increase is larger than that for global average temperature and makes it the warmest decade on record (125). Moreover, 2014 and 2015 were the joint warmest years in Europe since instrumental records began. Anthropogenic climate change made these temperature records 35-80 times more likely (126). Moreover, climate reconstructions show that summer temperatures in Europe in recent decades are the warmest in at least 2000 years and that they lie significantly outside the range of natural variability (127). The relatively rapid warming trend since the 1980s is most clearly evident in the summer (124). Nocturnal temperatures have increased more than daytime values (14).
Geographically, particularly significant warming has been observed in the past 50 years over the Iberian Peninsula, in central and north‑eastern Europe and in mountainous regions (3,4). In the past 30 years, warming was strongest over Scandinavia, especially in winter, whereas the Iberian Peninsula warmed in summer (146). High-temperature extremes like hot days, tropical nights, and heat waves have become more frequent, while low-temperature extremes (e.g. cold spells, frost days) have become less frequent (2,5).
Urbanization explains only a very small part of warming in Europe: only 0.0026°C/decade of the annual-averaged pan-European temperature trend of 0.179°C/decade over the period 1960–2012 (76).
There is growing evidence that the rate of warming is amplified with elevation, such that high-mountain environments experience more rapid changes in temperature than environments at lower elevations (77). Elevation-dependent warming can accelerate the rate of change in mountain ecosystems, cryospheric systems, hydrological regimes and biodiversity. More rapid changes in high-mountain climates would have consequences far beyond the immediate mountain regions, as mountains are ‘water towers’ and the major source of water for large populations in lower elevation regions (78). Besides, mountains provide habitat for many of the world’s rare and endangered species and are therefore sensitive to environmental change (77).
Elevation-dependent warming may be caused by a number of mechanisms (77). One of them is the snow–albedo feedback. The surface absorption of incoming solar radiation increases around the retreating snowline, causing enhanced warming at that elevation (79). In the Swiss Alps, the daily mean 2-m temperature of a spring day without snow cover is 0.4 °C higher than one with snow cover (80) (mean value for 1961–2012). The current snowline, which varies in elevation across different mountain ranges, is expected to retreat to higher elevations as the overall climate system warms. Other mechanisms linked to elevation-dependent warming involve clouds, water vapour and aerosols (77).
The IPCC concluded in 2019 that mountain surface air temperature observations in the European Alps show warming over recent decades at an average rate of 0.3°C per decade, with a likely range of ± 0.2°C, thereby outpacing the global warming rate 0.2 ± 0.1 °C per decade, (214). In the European Alps, warming has been found to be more pronounced in summer and spring (215)
Anthropogenic influence on record‐breaking temperatures trends
From the 1980s onwards, a change in the record evolution is observed: the number of upper temperature records increases and the number of lower temperature records decreases (96). This agrees with previous studies (97). Daily minimum and maximum record-breaking temperatures tend to occur less and more often, respectively, than in a stationary climate (96). This may point at a changing climate. However, over the second half of the twentieth century (observations over the period 1956–2005) no statistically significant changes in the annual and seasonal evolutions of the upper and lower records over Europe can be detected yet. According to climate model simulations, the upper and lower record evolutions at the end of the twentieth century are still in the range of the model’s internal variability (96).
Anthropogenic influence on Europe's warm year 2014
Attribution analysis enables scientists to estimate to what extent climate change may have attributed to the (increasing) likelihood or intensity of an event. It is a tool to communicate the impacts of the changing climate to the public. It also helps to inform decision makers how specific events were impacted by the changing climate and what this means for their future. To those dealing with droughts, for instance, attribution analysis shows what ingredients went into any particular drought, how it evolved, and whether it could have been predicted. This may improve early warning for drought and informs decision makers why long-term planning should account for changing climate (117).
One of these extreme events is yearly average temperature in 2014. That year broke the record for the warmest year in Europe, in the observational record (118). 2014 was especially warm in central Europe and Scandinavia. How much did anthropogenic climate change alter the likelihood of this European record? This question was answered by studying historical data, and by making climate model calculations on probabilities of the event occurring in the actual world with climate change and in the world without human influence. From these results it was concluded ‘with high confidence’ that anthropogenic climate change has made Europe’s warm year of 2014 at least 500 times more likely (117).
This signal of human influence is so strong because the area for which historical records were broken is extremely large. When the same approach of attribution analysis is employed to a small part of Europe, the signal of a human influence on yearly average temperature is much smaller. This tells us that the yearly average temperatures of 2014 are not exceptionally high for just one country, for instance, and similar temperatures may have occurred without anthropogenic climate change. For Europe as a whole, however, the situation is completely different: the fact that these relatively high temperatures cover such a large area makes the situation of 2014 extreme and point at a significant effect of climate change (117). This conclusion, that the risk of record annual mean warmth in Europe as occurred in 2014, has been greatly increased by anthropogenic climate change, has been confirmed by previous research (119).
Heat wave changes in Europe until now
Heatwaves, similar to those which occurred in western Russia in 2010 and Europe in 2003, develop when persistent anticyclonic patterns, often referred to as ‘atmospheric blocking’, initiate a cascade of self-reinforcing, heat-accumulating physical processes (87). It is suggested that the observed increase in extreme summer heat over Europe is attributable to both increasing frequency of blocking circulations and changes in the surface energy balance (88).
European heat wave changes since the 1950s
European summer temperatures have increased far more rapidly than the trends for the Northern Hemisphere since the early 1950s (157). While mean summer temperature in the Northern Hemisphere has increased by 0.21 °C per decade between 1951 and 2015, the increase was 0.46 °C per decade for Europe.
This rise in mean European summer temperatures has contributed to the intensity of recent European heat waves, like the ones in 2003, 2010, and 2015. What if climate had remained as it was in the mid-20th century, in terms of the number of hot days? This was analysed for temperature records of 14 climate stations across Europe (156).
If Europe’s climate hadn’t changed since the 1950s, the number of hot days with maximum temperature over 30 °C would have been 10-25% less. The number of hot days over 35 °C would have been 25-50% less, according to this study. This serves to illustrate that climate change is not simply a matter of gradual shifts in mean values, but also of strong increases of the frequencies of extremes. As a result heat waves are becoming more extensive and prolonged, with stronger impacts on society (156).
The heat wave of 2003
Much of Europe was affected by a heat wave during the summer of 2003 (June, July and August). It is estimated that this was the hottest summer since at least 1500 (6). Seasonal temperatures were the highest on record in Germany, Switzerland, France and Spain. Average summer (June–August) temperatures were far above the long-term mean, by up to five standard deviations, implying that this was an extremely unlikely event under current climatic conditions (7).
Over the period 1880 to 2005 the length of summer heat waves over Western Europe has doubled and the frequency of hot days has almost tripled (8). The study adds evidence that heat waves, such as the devastating 2003 event in Western Europe are a likely sign of global warming.
Shortly after the 2003 heat wave, researchers estimated that it is very likely that human influence has more than doubled the risk of a regional scale heat wave of at least the 2003 severity (11). Recently, the same researchers presented a new, updated, analysis of the same heat wave. According to this recent study (75), the chance of extremely hot summers would have increased dramatically since the 2003 European heat wave. In this study, the chance of extremely hot summers such as the 2003 European heatwave has been studied based on an ensemble of models and four climate change scenarios. This was done for the same area that was hit by the 2003 heat wave, including the countries where heat-related mortality peaked (France, Germany and Italy). It was found that although summers as warm as 2003 are still rare, they are now much more likely to occur. Events that would occur twice a century in the early 2000s are now expected to occur twice a decade. For the more extreme heatwave of 2003, the return time would have reduced from thousands of years in the late twentieth century to about a hundred years in little over a decade. As climate warming continues, the frequency of extreme hot events is expected to increase further. All climate change scenarios used in the study (RCP2.6, RCP4.5, RCP6.0 and RCP8.5) indicate that by the 2040s a summer as hot as 2003 will be very common. In fact, the scenarios with the strongest anthropogenic influence (RCP6.0 and RCP8.5) suggest the 2003 summer will be deemed an extremely cold event by the end of the century (75).
Based on the evidence published in scientific literature so far (2012) there is now strong evidence for some types of extreme, notably heat waves and precipitation extremes, that link specific events or an increase in their numbers to the human influence on climate. For other types of extreme, such as storms, the available evidence is less conclusive, but based on observed trends and basic physical concepts it is nevertheless plausible to expect an increase (53).
The heat wave of 2010
During the 2001–2010 decade, 500-year-long records were likely broken over ~65% of Europe, including eastern Europe (2010), south-western central Europe (2003), the Balkans (2007), and Turkey (2001). The 2003 and 2010 summers were likely the warmest on record over ~25% of Europe, standing as major contributors to the current European map of the hottest summers (54).
In summer 2010, many cities in eastern Europe recorded extremely high values of daytime (for example, Moscow reached 38.2°C), nighttime (Kiev reached 25°C), and daily mean (Helsinki reached 26.1°C) temperatures. Other countries (Baltic countries, Belarus, Ukraine, and Kazakhstan) also experienced extreme temperatures that broke the summer records of the last 140 years at many temporal scales (54).
The 2010 event exceeded the 2003 episode in terms of amplitude and spatial extent. The maximum extension of areas experiencing record breaking temperatures in 2003 was ~1million km2, which is considerably lower than that of 2010 (more than 2 million km2) (54). The European mean 2010 summer was ~0.2°C warmer than the previous warmest summer of 2003 (55).
The severe 2010 heat wave in western Russia was influenced both by natural climate variability and anthropogenic climate change (90). Climate variability led to extremely low soil moisture content. Evapotranspiration was very low due to the very dry soil that cannot provide enough water to evaporate. As a result, little of the available surface net radiation was used for evaporation (and turned into latent heat flux) and most of it was turned into surface warming (sensible heat flux).
Looking at the 2010 heat wave, the dry soil moisture alone has increased the risk of a severe heat wave in western Russia sixfold, while climate change from 1960 to 2000 has approximately tripled it. The combined effect of climate change and the extremely low soil moisture yields a 13 times higher heat wave risk. Thus, internal climate variability causing the dry 2010 soil moisture conditions formed a necessary basis for the extreme heat wave (90).
Anthropogenic climate change increased the probability of occurrence of this heat wave. However, the magnitude of the heat wave was within the range of natural climate variability. Even though climate change had an influence, its contribution to this heat wave magnitude was small compared to that of natural variability (see also 91).
The heat wave of 2013
Anthropogenic forcing played a substantial part in Western Europe’s hot, dry summer in 2013 (71). North Atlantic sea surface temperatures were likely a factor in the large contrast with summer 2012. Averaged over Western Europe, the seasonal mean (June–August) anomaly in surface air temperature was 1.33°C above the mean over the period of 1964–1993, which is 3.2 standard deviations of the interannual variability. This magnitude of warming is slightly less but comparable with the previous hot summers in Western Europe, such as 2003 and 2010 for which summer mean surface air temperature anomalies were 1.46°C and 1.86°C respectively, corresponding to 3.5 and 4.5 standard deviations of the interannual variability (71).
The heat wave of 2015
Europe’s summer of 2015 was highly unusual, as persistent heat and dryness prevailed in large parts of the continent. In central and eastern Europe, a combination of record-low seasonal rainfall and record-high monthly July/August temperatures were observed over an area stretching from France to western Russia. The anomalous temperatures were caused by a sequence of four intense heat waves that struck the region from the end of June to early September (120). The 2015 summer heat intensity and drought conditions were especially extreme in the Central part of the continent, where a large range of partly century-old record values of minimum and maximum temperature indices have been broken, e.g. in Prague (since 1775) and Vienna (since 1855) (138).
Europe experienced the hottest August ever recorded (122), and the entire summer season ranked third after the unusual summers of persistent heat in 2003 and 2010 with hotspots in France and western Russia, respectively (54,90). On 1 July, London experienced its hottest July maximum temperature on record: 36.7°C. Paris recorded its second hottest day ever on 2 July, with a high temperature of 39.7°C, and Berlin experienced its highest temperature on record, 37.9°C, on 4 July (121).
Both an analysis of long-term observations (2001-2015) and climate model simulations show that human-induced climate change has contributed to the increase in the frequency and intensity of short-term heat waves and heat stress such as the central and eastern Europe 2015 event (120,121).
The heat wave of 2017
Under the on-going climate change, mega heat waves are becoming more frequent, intense, and longer (168), and this trend is expected to continue in the future. In addition, Europe’s summer season starts earlier, by 4 days per decade according to observations from 1979 to 2012 (169). As a result, mega heat waves may occur unusually early in the year when compared to the historical record. This was the case in 2017.
In June 2017 temperatures were extremely high across Western Europe. It was the hottest June in Spain in half a century (170), and in France, Switzerland and the Netherlands in 100 years (171). The June 2017 mega heat wave lasted two weeks and affected a wide area from the east Atlantic to western and central Europe. Perhaps the most prominent feature of this event is its advanced timing. In fact, it was the earliest mega heat wave in Europe since at least the midtwentieth century (167).
An atmospheric circulation that spread subtropical air from Africa over Western Europe caused the heat wave. As such, the mega heat wave didn’t have a climate change signature, it just occurred earlier than usual. The temperatures in Spain and Portugal were exceptionally high, however, which seems to be due to the changing climate (167). In fact, the June 2017 mega heat wave could be an actual manifestation of summers that are becoming longer and start earlier (169), in agreement with future projections of global warming (172).
Earlier green-up of vegetation amplifies frequency and intensity of spring heat waves
The earlier green-up of vegetation in Europe amplifies spring warming, especially the frequency and intensity of spring heat waves, according to a recent study. On the other hand, no effect of increased early-season transpiration and late-season dryness on summer heat waves was found, which is in contrast to previous studies (92).
The onset of green-up of plants has advanced in response to climate change. On the other hand, the presence of green land cover earlier in the season may also alter the seasonal climate through a number of effects, including changes in surface albedo and evaporation that affect soil moisture and surface energy (93). Thus, there may be a positive feedback of spring advancement of vegetation growth due to climate change on further amplifying global warming.
Evidence exists that an earlier green-up may enhance the likelihood of summer drought because soil water is depleted earlier in the growing season (94). For Europe, previous modeling and observation studies have suggested that an earlier onset of vegetation green-up and a prolonged period of increased evapotranspiration have enhanced recent summer heat waves by lowering soil moisture (95).
This effect could not be confirmed in a recent study, however: the canopy evapotranspiration is indeed increased in response to earlier green-up, but this is counteracted by a decrease of ground evaporation. There is therefore no propagation of a drier soil from spring to summer and consequently no impact on summer heat waves (92).
On the other hand, the same study did show the amplification of spring warming in Europe, and especially heat waves, due to the earlier onset of green-up of plants. This was related to a decrease in low- and middle-layer clouds and associated increases of downward short wave and net radiation. The earlier green-up has substantial impacts on daytime temperature, especially heat waves, but makes little difference on nighttime temperature because the reduction of low- and middle-level clouds and water vapor in the atmosphere enables a higher outgoing long wave flux at night which offsets any impact of higher daytime temperatures (92).
Changes in hottest and coldes extremes in Europe until now
Hot days now about 2.3 ◦C warmer than in 1950
On average, across Europe, the number of days with extreme heat and heat stress has more than tripled from 1950 to 2018: from less than 2 days to more than 6 days per year. This was concluded from observations during the period 1950-2018 at weather stations all across Europe. In this study, a day is extremely hot when its maximum daytime temperature exceeds the limit of the 1% hottest days over the whole period 1950-2018 (12).
In addition to an increase in the number of extremely hot days, heat intensity on a hot day has increased as well. Across Europe, extremely hot days are now about 2.3 °C warmer than in 1950. This is an increase of 0.33 °C per decade, larger than the average global warming of about 0.2 °C per decade (210). Especially in Central Europe hot days are now much hotter than half a century ago: hot extremes have warmed about 50% more than the corresponding summer mean temperatures. In the rest of Europe, the trends of temperature increase for hot days and mean summer temperatures are similar (12).
The observed trends for hot extremes are remarkably consistent with climate model projections (211)
Cold days now more than 3 °C warmer than in 1950
Likewise, a day is extremely cold when its minimum daytime temperature falls below the limit of the 1% coldest days in 1950-2018. The annual number of extremely cold days and nights has decreased by a factor of 2-3 on average across Europe from 1950 to 2018: from more than 5 around 1950 to around 2 days/year in 2018. Cold extremes are now more than 3 °C warmer than in 1950: an average increase of 0.49 °C per decade. In Central and Northern Europe the warming of cold extremes is substantially higher than the corresponding winter mean warming and about 2.5 times larger than the average global warming (12).
The observed trends for cold extremes are remarkably consistent with climate model projections (cold extremes: 212).
Results from other studies
Another study has been carried out over the period 1880 to 2005. It was concluded that both the mean and the variance of daily maximum temperature have increased significantly (8).The highest trends in daily maximum temperature variability have been found in central Western Europe. The researchers suggest that their conclusions contribute to growing evidence that Western Europe's climate has become more extreme and confirm a previously hypothesized increase in the variance of daily summer temperatures since the 19th century. The study adds evidence that heat waves, such as the devastating 2003 event in Western Europe are a likely sign of global warming. According to these researchers current estimates of changes in June-August extreme daily temperatures (e.g. 9) have hitherto been conservative in their estimate of changes in extremes over the last 126 years. Over recent decades the annual number of warm extremes has increased significantly as a result of an asymmetric pronounced warming of the upper tails of the temperature distribution (10,62).
Range minimum - maximum temperature
Cold extremes are warming more rapidly than hot extremes, accprding to observations during the period 1950-2018 at weather stations all across Europe. Where hot extremes have increased by 0.33 °C per decade, the trend for cold extremes is 0.49 °C per decade (12). A 50% difference!
any impact studies assume that climate change results in changing daily minimum and maximum temperatures by the same amount and in the same direction. However, on local scales natural variability still plays a dominant role and can overshadow the global warming signal (86). For Europe this variability was quantified for the period 1950-2013 by quantifying trends in seasonal daily minimum temperature (defined as the 5th percentile of the daily temperature distribution for a season) and seasonal daily maximum temperature (95th percentile) (86).
The results show that especially the hot extremes show significant increasing trends over central and eastern Europe, and significant decreasing trends mainly over Norway, south-eastern Europe, and Turkey. The cold extremes, on the other hand, experienced warming trends over Scandinavia and eastern Europe and cooling trends over south-eastern Europe, mainly over Turkey. The extreme temperature range, thus, changes differently for different parts of Europe: it widens, for instance for Madrid (warm extremes increase and cold extremes decrease) whereas it reduces for Trondheim (Norway) (a large upward trend in the cold extremes and a decrease of warm extremes) (86).
Precipitation changes in Europe until now
Trends in extreme precipitation events
The number of days with very heavy precipitation over Europe has increased on average by about 45% in observations in the period 1981-2013 compared to 1951-1980. The more intense the heavy rainfall event, the higher the relative increase in frequency. The findings imply that what was a 1-in-1000 day heavy rainfall in 1951-1980 occurred about 45% more often in the 1981-2013 period. Models seem to underestimate the rate of change in daily heavy precipitation compared to observations (123).
There is compelling evidence for statistically significant trends, during the last century, for increasing frequency and total precipitation of rainfall extreme events over North and Western Europe (extreme events being defined as the 90th and 95th percentiles of daily rainfall in the October to March season) (85,113). Extreme daily precipitation started to increase in southern Europe before the year 2000 (178).
Similar results were shown for an analysis of extreme precipitation trends in Europe during December 1950 to February 2008 (145). In this study an extreme precipitation event was defined as a daily amount of precipitation exceeding the 95th percentile of daily rainfall in a certain month at a certain location over the period 1961-1990. Thus, the threshold value equivalent to the 95th percentile was calculated separately for each month and weather station across Europe, allowing for the temporal and seasonal variability in precipitation. Overall, the analysis showed significant upward trends of the number of extreme precipitation events in Europe in all seasons except summer. The results also point at distinct differences between the north and south of Europe, however: an increase of number of extreme precipitation events in the northern part of the continent and a decrease in the south. The prominent seasonality and strong regional variations detected in extreme precipitation trends are caused by the complicated impacts of both local (e.g. topography) and macro- and mesoscale (atmospheric circulation) factors that trigger precipitation formation and cause variability (145).
The number of extreme precipitation events has increased over most of the European land area, linked to warming and increases of atmospheric water vapour. For Europe as a whole, also the intensity of extreme precipitation such as heavy rain has increased in the past 30 years, even for areas with a decrease in mean precipitation, such as Central Europe and the Mediterranean (17). The proportion of Europe that has experienced extreme and/or moderate meteorological drought conditions did not change significantly during the 20th century (1).There is evidence for Europe and the United States that the relative increase in precipitation extremes is larger than the increase in mean precipitation, and this is manifested as an increasing contribution of heavy events to total precipitation (18).
Over approximately two-thirds of the Northern Hemisphere land area, greenhouse gases have contributed to the observed intensification of annual maxima of daily and five-daily precipitation amounts during the second half of the twentieth century (51).
The change in return period of rare precipitation events that occur on average once in 5, 10 and 20 years in the 1950s and 1960s have been analyzed using extreme value theory (63). This was done for 1-day and 5-day precipitation amounts over the time period 1951–2010 in Northern and Southern Europe in all four seasons. 1-day events are indicative of extreme showers which are important for flash floods on local scales. 5-day events are indicative of wet periods which may result in high water levels in larger scale river basins. Over the period 1951–2010, Northern Europe shows a wetting trend in winter, spring and autumn, and a drying trend in summer. Southern Europe shows a drying trend in winter, spring and summer, and a wetting trend in autumn. In Northern Europe, the picture for the changes in extreme precipitation is approximately the same as that for the trend in total precipitation amount. In Southern Europe the 20-year 1-day and 5-day events stay about the same in winter, but become slightly wetter in other seasons, although the regional trend in total precipitation amount in winter and summer indicates drying. Despite considerable decadal variability, the results of this study indicate that 1-day and 5-day precipitation events that occurred on average once in 5, 10 and 20 years in the 1950s and 1960s generally became more common during the period 1951–2010. For all regions, seasons and return periods, the median reduction in return period between 1951-1970 and 1991-2010 is about 21% with variations between a decrease of about 2% and 58% (63).
Trends in annual and seasonal precipitation volumes
Annual precipitation since 1960 shows an increasing trend of up to 70 mm per decade in northeastern and northwestern Europe, and a decrease of up to 90 mm per decade in some parts of southern Europe. At mid-latitudes no significant changes in annual precipitation have been observed. Mean summer precipitation has significantly decreased by up to 20 mm per decade in most of southern Europe, while significant increases of up to 18 mm per decade have been recorded in parts of northern Europe (146).
Annual precipitation trends in the 20th century showed an increase in Northern Europe (10–40%) and a decrease in some parts of Southern Europe (up to 20%). Mean winter precipitation has increased in most of Western and Northern Europe (20 to 40%), whereas Southern Europe and parts of Central Europe were characterized by drier winters (1). Drying has been observed in the Mediterranean and Eastern Europe and no clear trends have been observed in Western Europe (15).
Mean winter (December–February) precipitation is increasing 20–40% in most of Western and Northern Europe, because western circulation was stronger in winter. Conversely, Southern Europe and parts of Central Europe were characterized by a drier winter. Trends in spring and autumn were not significant (16). The changing precipitation regimes in Europe since 1950, reflected in fewer winters with wet conditions in southern Europe but more in northern Europe, coincide with changes in storm track activity over the central/eastern North Atlantic toward the northern British Isles (159).
More extremely strong fronts over Europe
High-impact, small-scale weather events, such as hailstorms, wind extremes, or intense precipitation, are frequently associated with the passage of large-scale fronts (134). These fronts are related to strong transitions in temperature and humidity across the front. The stronger temperature and humidity transitions across the fronts, the stronger the fronts are. Extreme weather is more likely linked to strong than weak fronts (133).
Scientists found an increase in the number of extremely strong, large fronts (with a minimum length of 500 km) over Europe over the period 1979-2014 (133). The increase is most pronounced after the year 2000, and refers to the summer (June-August) and autumn (September-November) seasons. The scientists argue that increasing atmospheric humidity primarily drives this trend in the number of strong and extremely strong fronts. Their findings are consistent with the fact that most extreme fronts are identified during the summer, when atmospheric humidity is higher than in the other seasons. Besides, their findings also agree with the observed increase in humidity over Europe in the considered period (135).
These extremely strong fronts can occur at various places over Europe: mountainous regions such as the western Alps, the Dinaric Alps, the Scandinavian mountains in south-western Norway, and the Cantabrian mountains in north-western Spain, as well as the west coast of Ireland and also many parts of France, northern Germany, Denmark, and the Baltic Sea (133).
The upward trend in the number of strong fronts is potentially one driving agent behind an increase in the number of extreme precipitation events over Europe (136). Indeed, several studies suggest an increase in the likelihood of extreme summer precipitation events for Europe (137).
Hail events changes in Europe until now
Hail events are among the most costly weather-related extreme events in several European regions, causing substantial damage to crops, vehicles, buildings and other infrastructure (146). For example, three hailstorm events in Germany in July and August 2013 caused around EUR 4.2 billion of combined damages to buildings, crops, vehicles, solar panels, greenhouses and other infrastructure (147). The number of hail events is highest in mountainous areas and pre-Alpine regions. Since 1951, increasing hail trends have been noted in southern France and Austria, and decreasing (but not statistically significant) trends have been noted in parts of Eastern Europe (146).
Snow cover changes in Europe until now
Snow cover extent in the northern hemisphere has declined significantly over the past 90 years, with most of the reductions occurring since 1980. Over the period 1967-2015, snow cover extent in the northern hemisphere has decreased by 7 % on average in March and April and by 47 % in June; the observed reductions in Europe are even larger, at 13 % for March and April and 76 % for June (146).
Data from satellite monitoring from 1966 to 2005 show that monthly snow-cover extent in the northern hemisphere is decreasing by 1.3% per decade. Snow-cover trends in the mountain regions of Europe vary considerably with region and altitude. Recent declines in snow cover have been documented in the mountains of Switzerland (19), Slovakia (20), and in the Spanish ski-resorts in the Sierra Nevada and the Pyrenees (21), but no change was observed in Bulgaria over the period 1931–2000 (22). Declines, when observed, were largest at lower elevations. Lowland areas of central Europe are characterized by recent reductions in annual snow-cover duration of about 1 day/year (23). At Abisko in subarctic Sweden, increases in snow depth have been recorded since 1913 (24), and trends towards greater maximum snow depth but shorter snow season have been noted in Finland (25).
Mean and maximum winter snow depths
Mean and maximum winter snow depths are decreasing over Europe except for the coldest regions. This decrease is widespread and has accelerated since the 1980s in parallel with stronger global warming. This was concluded from daily snow depth observations over the period 1951-2017 in a large number of European countries (180). In this study, mean winter snow depth is defined as the average daily snow depth during the winter months December, January and February. Maximum winter snow depth is defined as the 95th percentile of daily snow depth values in this winter period.
The decrease in mean snow depth is stronger than the decrease in maximum snow depth, and this contrast has become stronger in the last few decades. Averaged over Europe, the trends in maximum snow depth are about one tenth weaker than the mean snow depth trends. The average of decreasing trends in mean snow depth for the studied weather stations is −12%/decade. The average of decreasing trends in maximum snow depth is −11%/decade.
The steep decrease in mean snow depth can be related to a combination of decreases in mean snowfall, less days with snowfall, and stronger snowmelt which (combined) lead to shorter periods with snow cover. The results agree with the expectation that in all but the coldest areas in Europe the fraction of below-zero temperatures will decrease, thereby decreasing the fraction of winter precipitation that falls as snow and increasing the amount of snowmelt (181).
For the coldest areas and high latitudes in Europe, an increase in mean and maximum snow depths is observed. Here the increase in precipitation, probably associated with the warming climate, dominates any snowmelt signal in the mean snow depth. For these areas, where temperatures in winter are still below the freezing point, future projections indicate that increases in total precipitation outweigh the effects of warmer temperatures, thereby increasing snow depth (182).
Ice cover changes in Europe until now
An analysis of long (more than 150 year) ice records from lakes and rivers throughout the northern hemisphere indicated that for a 100 year period, ice cover has been occurring on average 5.7 ± 2.4 days later (± 95% confidence interval), while ice break-up has been occurring on average 6.3 ± 1.6 days earlier, implying an overall decrease in the duration of ice cover at a mean rate of 12 days per 100 years. These results do not appear to change with latitude, or between North America and Eurasia, or between rivers and lakes (26).
At the Hungarian section of the river Danube, the date of the first ice appearance has shifted 19-29 days later over the 1876-2011 period, while the date of the final ice disappearance has shifted 18–23 days earlier (153). In Lake Kallavesi (eastern Finland), the freezing date has shifted to 15 days later in the period 1833-2011, while the break-up date has shifted to 12 days earlier in the period 1822-2011 (154). The ice break-up date in the Lake St. Moritz (Swiss Alps) has shifted to 15–20 days earlier since 1832 (155).
Sea temperature changes in Europe until now
On a global scale, ocean surface temperatures have increased, on average, by around 0.7°C since the beginning of the 20th century (2) and 0.17°C between 1969 and 2008 from the surface to a 700 m depth; the temperature increase in the Atlantic Ocean during 1969 and 2008 was 0.30°C (57). Observational records of ocean heat content show that ocean warming is accelerating (184).
For each of the most recent decades, the surface of the Mediterranean Sea has warmed by around 0.4°C (186).
Wind climate changes in Europe until now
Storminess in Europe has shown considerable variation over the past century, but with no clear long-term trend (27,146). The past 60 years have been characterised by an increase in European storminess, with a strong increase from the 1960s to the 1990s and a decline thereafter. The extent to which this is part of natural variability or related to anthropogenic climate change is not known and needs further investigation. Long-term changes in European storminess are not yet clear as studies taken as a whole are inconclusive, with sometimes conflicting results (70).
Evaluating high tide levels along the North Sea in the past century showed clear changes in mean levels (related to sea-level rise) but no storm-related variations (27). Similarly, in the northern Adriatic Sea the trends for high sea levels and the subsequent occurrence of storm surges cannot be associated with any trends in storminess (28).
Storm tracks associated with extratropical cyclones have moved polewards over the past 25 years (before 2012) (52). A northward shift in mean storm track position since about 1950 is consistent in studies on wind climate in Europe over the last decades (111). This northeast shift together with the trend pattern of decreasing cyclone activity for southern mid- latitudes and increasing trends north of 55 - 60°N after around 1950 seems consistent with scenario simulations to 2100 under increasing greenhouse gas concentrations (112).
Impacts on glaciers until now
A general loss of glacier mass since the beginning of the measurements has occurred in all European glacier regions, except some glaciers in Norway. The Alps have lost roughly 50 % of their ice mass since 1900 (148). Norwegian coastal glaciers were expanding and gaining mass up to the end of the 1990s owing to increased winter snowfall on the North Atlantic Coast; now these glaciers are also retreating (149).
The centennial retreat of European glaciers is attributed primarily to increased summer temperatures. However, changes in winter precipitation, reduced glacier albedo due to the lack of summer snowfall and various other feedback processes, such as the increasing debris cover on the glacier, can influence the behaviour of glaciers, in particular on regional and decadal scales (146).
Data on the longest and most continuous series for six glaciers in the European Alps (In Austria, Switzerland and France, over the period 1962-2013) show a clear and regionally consistent acceleration of mass loss over recent decades over the entire European Alps (140).
Global glacier mass changes and their contributions to sea level rise have been quantified from 1961 to 2016. This was done by repeated mapping and differencing of glacier surface elevations from in situ, air-borne and spaceborne surveys, in combination with time series of in situ point measurements on glaciers (200).
Glaciers distinct from the Greenland and Antarctic ice sheets cover an area of approximately 706,000 square kilometres globally (201). Their total ice volume equals 0.4 metres of potential sea level rise (202). The quantification of global glacier mass changes reveals that glaciers contributed 27 ± 22 millimetres to global mean sea level rise from 1961 to 2016, or a contribution of 0.5 ± 0.4 mm per year when a linear rate is assumed. The present glacier mass loss is equivalent to the sea level contribution of the Greenland Ice Sheet (203), and clearly exceeds the loss from the Antarctic Ice Sheet (204). It accounts for 25% to 30% of the total observed sea level rise, which ranged between 2.6 and 2.9 ± 0.4 mm per year over the satellite altimetry era (1993 to mid-2014) (205).
When annual rates are averaged over periods of five years, sea level contributions ranged between 0.2 and 0.3 mm per year until the 1980s, and then increased continuously to reach 1.0 mm per year in recent years (2011–2016) (200).
Under present ice-loss rates, most of today’s glacier volume would thus vanish in the Caucasus, Central Europe, the Low Latitudes, Western Canada and the USA, and New Zealand in the second half of this century. However, the heavily glacierized regions of the world would continue to contribute to sea level rise beyond this century, as glaciers in these regions would persist but continue to lose mass (200).
Climate change scenarios
Scenarios for the 5th IPCC report
For the Fifth Assessment Report of the IPCC a different type of climate change scenarios has been defined than for the previous IPCC reports: the so-called Representative Concentration Pathways (RCPs) (73). In contrast to the previous SRES scenarios, RCP scenarios do not specify socioeconomic scenarios, but assume pathways to different target radiative forcing at the end of the twenty first century. For instance, scenario RCP8.5 assumes an increase in radiative forcing of 8.5 W/m2 by the end of the century relative to pre-industrial conditions. The reduction in emissions necessary to limit the forcing to this level is supposed to be reached with air quality legislation but without a strict climate policy.
Scenarios that are often used for climate projections are the relatively low RCP2.6, the moderate RCP4.5 and the high-end RCP8.5 scenario. Climate models project further increases in global average temperature over the 21st century. For the period 2081-2100 (relative to 1986-2005), increases of between 0.3°C and 1.7°C for the lowest emissions scenario (RCP2.6 (Representative Concentration pathway)), and between 2.6°C and 4.8°C for the highest emissions scenario (RCP8.5) are estimated (124).
The EU and UNFCCC target of limiting global average temperature increase to less than 2°C above pre-industrial levels is projected to be exceeded between 2042 and 2050 by the three highest of the four RCPs (128). The lowest, RCP2.6, implies strong reduction in greenhouse gas emissions over this century and negligible or even negative emissions at the end of the century (129).
The moderate SRES A1B scenario leads to a global mean temperature increase in 2100 in the likely range of 2.8-4.2°C, which approximates to RCP6 and lies clearly between RCP4.5 and RCP8.5 (74). The RCP4.5 has a target radiative forcing of 4.5 W/m2 and requires, for example, a decline in overall energy and fossil fuel use and a substantial increase in renewable energy forms (116).
Scenarios for previous IPCC reports
Future climate change projections are based on scenarios of potential socio-economic futures that determine the level of greenhouse gas emissions to the atmosphere. Researchers generally use the scenarios that have been published by the IPCC in 2000, the so-called IPCC SRES scenarios (49).
Each socio-economic scenario provides a description of possible future developments. IPCC distinguishes between so-called A and B families; these are widely applied and cover a broad range of possible population growth and economic development. The A scenarios represent a vision of the future where economic development is the priority, whereas the B scenarios represent a future where environmental sustainability plays a central role. Family ‘1’ scenarios are more globalised than ‘2’ scenarios.
A1 describes a future world of very rapid economic growth and a population that increases from 5.3 billion in 1990 to peak in 2050 at 8.7 billion and then declines to 7.1 billion in 2100. Rapid introduction of new and efficient technologies is assumed, as is convergence among regions, including large reductions in regional differences in GDP. The A1 family has 3 subgroups:
- A1F1: high use of fossil fuels
- A1T: high use of non-fossil energy sources
- A1B: an intermediate case
The B1 family describes a convergent, more equitable world, and has the same population scenario as the A1 family; however, there is introduction of clean and resource-efficient technologies, and global solutions are found for economic, social and environmental sustainability. Additional climate initiatives are not assumed.
The A2 and B2 families have higher, continuously increasing population scenarios (to 15.1 and 10.4 billion in 2100, respectively).
Air temperature changes in Europe in the 21st century
The annual average land temperature across Europe is projected to continue increasing faster than global average temperature (146).
The first results of the high-resolution (horizontal resolution of 12.5 km) future climate simulations (from the EURO-CORDEX program) address changes in annual mean temperature and total precipitation across Europe for three scenarios (A1B, RCP4.5 and RCP8.5) and the time periods 2021–2050 and 2071–2100 (compared with 1971–2000). For these simulations a large number of regional and global climate models have been used (72). According to these results, under a high-end climate change scenario (RCP8.5), large parts of Northern Scandinavia, Eastern Europe and the Alpine ridge might be exposed to a warming of more than 4.5°C in 2071–2100; this could be avoided by the moderate scenario (RCP4.5).
By the end of this century (2071-2100 relative to 1971-2000), annual average land temperature over Europe is projected to increase in the range of 1°C to 4.5°C under the RCP4.5 scenario, and 2.5°C to 5.5°C under the RCP8.5 scenario. This is more than the global average (130, 146, see also (29) for similar results for the period 2080-2100 compared with the 1961-1990 average. The warming is projected to be greatest over Eastern Europe, Scandinavia and the Arctic in winter (December to February), and over Southwestern and Mediterranean Europe in summer (June to August) (30,32,130). The temperature rise in parts of France and the Iberian Peninsula may exceed 6°C, while the Arctic could become on average 6°C and possibly 8°C warmer than the 1961–1990 average (31).
In summer, the warming in large parts of Central, Southern and Eastern Europe is shown to be especially connected to higher temperatures on warm days, more than to a general warming at all temperature intervals. Much of the warming in winter is connected to higher temperatures on cold days (32). The yearly maximum temperature is expected to increase much more in southern and central Europe than in northern Europe (36).
The summer climatic zones shift northward by at least 400-500 km by the end of the twenty first century. The mean number of days/year exceeding 30°C at the model grid point nearest to Paris increases from 9 days under current climate to 50 days under future climatic conditions (38).
It is estimated that countries in Central Europe may experience the same number of hot days as currently occur in Southern Europe and that Mediterranean droughts may start earlier in the year and last longer. Most affected are probably the southern Iberian Peninsula, the Alps, the eastern Adriatic seaboard, and southern Greece (35).
Future vulnerability of European urban areas to climatological extreme events has been assessed for the period 2021-2050 compared to the reference period 1971-2000 (139). This was done for highly populated urban areas in Central Europe (the east of France, Germany, Austria, Czech Republic, Slovenia, north of Italy). The assessment is based on high-resolution regional climate model simulations and a scenario of moderate climate change (the so-called IPCC A1B scenario). On average for Central Europe the results show an increase of the number of heat waves (17%), as well as the number of single hot days (12%) and tropical nights (30%), and a decrease of the number of frost days (30%). For both hot and frost days the absolute decrease is largest in the northeast whereas the relative changes are largest in the northwest (139).
The results show that the number of hot days (daily maximum temperatures > 30°C) will increase particularly in the western part of central Europe, affecting cities like Paris, Stuttgart and Hamburg. At some locations, the number of hot days is to increase by over 70%. The number of tropical nights (daily minimum temperature > 20°C) shows a similar pattern since ‘hot days’ regions cannot significantly cool down and high temperatures remain throughout the night. The relative increase of the number of tropical nights shows a clear trend from west to east, ranging from roughly 70% for Hamburg and 40% for Paris, down to less than 10% in the vicinity of Vienna. Also the increase in the number of heat waves (daily maximum temperatures > 30°C on at least 5 days in a row) is higher in the western part (with cities like Paris and Zurich) than in the eastern part. In general, a decrease of the number of frost days is projected, by up to 45% at the coastline of the Atlantic Ocean. Remarkably there is no clear projected change in the number of frost days in the alpine area. The increase of the number of days that buildings have to be air-conditioned in the summer dominates over the decrease that buildings need to be heated in the winter (139).
The IPCC concluded in 2019 that until 2050, regardless of the climate scenario, surface air temperature is projected to continue increasing (very high confidence) at an average rate of 0.3°C per decade, with a likely range of ± 0.2°C per decade, locally even more in some regions, generally outpacing global warming rates (0.2 ± 0.1 °C per decade (high confidence) (214).
Averaged over northern Europe as a whole, the annual median warming is between 2.0°C (RCP4.5) and 2.9°C (RCP8.5) by the middle of the 21st century and between 2.7°C (RCP4.5) and 5.0°C (RCP8.5) by the end of the 21st century. Projected warming over northern Europe is stronger during winter than during summer (see table below) (106).
Table. Projected change in median surface air temperature (°C) by the mid- and end of the 21st century relative to present day (1986–2005) for northern Europe for RCP2.6 (32 models), RCP4.5 (42 models) and RCP8.5 (39 models) as well as the 25th and 75th percentiles of the model results (numbers between brackets) (106).
|2046 - 2065||Winter (December - February)||2.7 (1.8 - 3.5)||3.4 (2.9 - 4.7)|
|Summer (June - August)||1.8 (1.2 - 2.5)||2.5 (1.9 - 3.2)|
|Annual mean||2.0 (1.6 - 2.8)||2.9 (2.4 - 3.5)|
|2081 - 2100||Winter (December - February)||3.4 (2.6 - 4.4)||6.1 (5.3 - 7.5)|
|Summer (June - August)||2.2 (1.6 - 3.0)||4.5 (3.5 - 5.8)|
|Annual mean||2.7 (2.1 - 3.5)||5.0 (4.3 - 6.3)|
According to a large number of models (GCM’s), surface mean air temperatures of the larger Mediterranean basin will increase mostly in summer and least in winter in 2070 - 2100 with respect to reference period of 1970 - 2000 under the A2, A1B and B1 emissions scenarios. Especially in summer, it is very likely that extreme climatic conditions will be more severe in the Mediterranean Basin in 2100 (89).
It is expected that summer warm-up will be twice as fast in Southern Europe than in Northern Europe (39).
An analysys based on the outcomes of a large number of models (GCM’s) shows that the Mediterranean region is likely to warm at a rate about 20% larger than the global annual mean surface temperature in the second half of this century, with values particularly large in summer and in the continental areas north of the basin (where warming will be in general 50% larger than at global scale and locally even twice as large). Day temperatures are likely to increase more than night temperatures and summers more than winters leading to an increase of amplitude of both daily and annual temperature range (179).
Heat wave changes in Europe in the 21st century
Since 2003, Europe has experienced several extreme summer heat waves (2003, 2006, 2007, 2010, 2014 and 2015). Such heat waves are projected to occur as often as every two years in the second half of the 21st century under a high emissions scenario (RCP8.5). The impacts will be particularly strong in southern Europe (146). The average number of heat waves increases from one every 3-5 summers (1961-1990) to about 2-3 heat waves per season at the end of this century (2071-2100), heat waves being defined as a spell of at least six consecutive days with maximum temperatures exceeding the local 90th percentile of the control period (1961-1990) (34).
Areas already experiencing strong heat waves (e.g. the Mediterranean region) could experience even more intense heat waves in the future. But other areas (e.g. France, Germany, and the Balkans) could see increases of heat wave intensity that could have more serious impacts because these areas are not currently as well adapted to heat waves (37).
The mean duration of heat waves increases by a factor of between one and eight over most of Europe. Much higher increases of at least a factor of seven are predicted for the mean intensity, the mean number of heat waves and the frequency of heat-wave days, with greatest changes (more than tenfold increases) in the south of France and Spain (38).
The 2003 heat wave can be used within certain limits as an analog to what may occur with more regularity in the future. The physical processes that characterized the 2003 heat wave, such as soil moisture depletion and the positive feedback on summer temperatures, and the lack of convective rainfall in many parts of the continent that generally occur from June–September, are projected to occur with greater frequency in the future (50). Droughts and high temperatures are linked: low soil moisture conditions can induce higher heat wave temperatures in Europe (176).
European capitals ranked by exposure to heat waves
75% of the population of the European Union lives in urban areas, and this percentage is expected to grow to 82% by 2050 (195). Needless to stress the importance of assessing climate change impacts on the urban environment, so cities can adapt in time to climate extremes. Future heat-related mortality in particular is a major concern. Heat waves are one of the deadliest weather-related hazards (196). The death toll of the 2003 heat wave in Europe, for instance, was 40,000 people. More extreme heat waves, with high death tolls, have struck Europe since then, like the one of 2010 impacting Moscow, or the recent 2017 heat wave in western and central Europe.
In order to raise awareness that it is time to act and adapt cities to temperature extremes, future impacts of temperature extremes in European capital cities have been assessed, with emphasis on heat waves. 8 model simulations (combinations of Global Circulation Models downscaled by Regional Climate Models) were used for this, based on a high-end scenario of climate change (RCP8.5). Selected time slices are near future (2021 - 2050) and distant future (2071 - 2100), and near past (1981 - 2010) for comparison. Focus was on 31 European capitals (EU28 plus Moscow, Oslo and Zurich) (194).
Definition heat and cold waves: Heat waves are defined in many different ways. In this study (194), a heat wave is defined as a period of at least 3 consecutive days with maximum temperature above a certain threshold. No fixed value for a threshold was chosen since this would highlight heat waves in southern Europe far too much. A couple of days of 27 °C in Northern Europe, for instance, may be experienced as a heat wave by the local population, but would not be noticed if the heat wave threshold were set at 30 °C. A definition for the heat wave threshold was chosen, therefore, that allows for this threshold to vary in time and space: the threshold is the 10% highest maximum daily temperatures in a certain time of the year, as observed in a certain city over the reference period 1981-2010. This way, a lower threshold is calculated for the North than for the South of Europe.
In this study, the intensity of a heat wave day is defined as the number of degrees maximum temperature exceeds a certain threshold, calculated from observations in the past. The sum of these intensities for all days of the heat wave presents an estimate of heat wave intensity. This sum is called the Heat Wave Magnitude Index in this study (197).
In a similar way, the intensity of projected future cold waves was quantified, focused on sequences of days colder than the 10% lowest minimum daily temperatures in a certain time of the year over the reference period 1981-2010. For both cold and heat waves three labels were used to characterize these extreme evens: “severe”, “extreme”, and “very severe” events.
Changing heat waves in Europe’s capital cities: In the near future, extreme heat waves are more likely to occur in Madrid and Valletta, and (with decreasing likelihood), Lefkosia, Ljubljana, Rome, Sofia, Athens, Bucharest, Budapest, and Zurich. By then, very severe heat waves are most likely in Valletta and Madrid, and to a slightly lesser degree in Rome and Sofia. In the distant future, the probability of occurrence of very severe heat waves will be high in a large number of cities: Athens, Bratislava, Bucharest, Budapest, Lefkosia, Ljubljana, Madrid, Rome, Wien, Zagreb, Zurich, Paris, and Lisbon. By then, the probability of an extreme heat wave to occur is relatively low only in Amsterdam. Still, also for Amsterdam, extreme heat events are a serious future threat, despite the city being ranked with lowest heat wave risk in 2100 in comparison to other European capitals (194).
The Maltese capital Valletta particularly emerges from this study, being ranked Europe’s capital city with the most severe heat waves by the end of this century. These impacts will likely be even more serious due to Valletta's geographic location in the middle of Mediterranean Sea where high humidity may add to these high impacts for the local population (198). Another important finding is the projected dramatic increase in heat wave occurrence in the continental European cities of Bucharest, Budapest, Ljubljana, Prague, Paris and Zagreb.
Vanishing cold waves in Europe’s capital cities: Cold waves will still represent some threat in 2050 but they will not be the major threat in Europe’s capitals. By the end of this century, cold waves are projected to completely vanish. The increasing negative impacts of extreme heat will by far outweigh the positive consequences of vanishing cold waves (194). Specifically in urban areas, the crucial negative impacts of extreme heat include health risks, human discomfort, associated higher concentrations of pollutants, lower water quality, increase in the energy demand for cooling, and decrease in labour productivity (199).
Record-breaking summer temperatures
The probability that summers in the future (2061-2080) will be warmer than any experienced in the past roughly 100 years (2020-2014) was quantified for two scenarios: a high-end (business-as-usual) scenario of climate change (RCP 8.5) and an intermediate scenario that reflects a moderate mitigation pathway of limiting fossil fuel use (RCP 4.5). Large numbers of simulations were used for this, based on one climate model. The warmest summer that was simulated for the period 1920-2014 was defined as the historical record (164). In this study, summer characteristics for the past 100 years were modeled. This period is too short to capture the extreme heat wave of 2003. This summer was an outlier. The modeled summers, therefore, do not include extreme summers like the one observed in 2003.
The results show that, on a global basis, the probability of exceeding the historical record temperature increases from less than 10% in a present-day summer to over 80% in 2070 with projected warming under the business-as-usual scenario (RCP 8.5). This increase can be reduced to nearly half (to 41%) by moderate climate mitigation (RCP 4.5). Europe is one of the world’s regions that benefit most from mitigation by seeing risk reductions of over 50% (164).
Future return periods of the heat waves of 2003 and 2010
According to regional multi-model experiments (11 RCMs, driven by different GCMs and forced with the A1B emission scenario), the probability of a summer experiencing mega heat waves will increase by a factor of 5 to 10 within the next 40 years. However, the magnitude of the 2010 event was so extreme that despite this increase, the likelihood of an analog over the same region remains fairly low until the second half of the 21st century (54). Under a high-end scenario of climate change (the RCP8.5 scenario), very extreme heat waves (which are much stronger than either the 2003 or the 2010 heatwave) are projected to occur as often as every two years in the second half of the 21st century (132). The projected frequency of heatwaves is strongest in southern and south-eastern Europe (132).
Model projections indicate that weekly heat spells of the magnitude of the second week of August 2003 will probably occur in 2020–2049 with a best-guess return period of ~10 years in Eastern Europe and ~15 years in Western Europe. However, a weekly 2010-like event remains very rare in the same period (best guess return periods >30-year over both regions). By the end of the 21st century, such extreme weekly heat spells are expected every ~8 years in Eastern Europe and ~4 years in Western Europe, whereas some models show regular 2003-like anomalies (about every second summer). The estimated return period involve major model uncertainties and should be carefully interpreted, given the high natural variability of such extreme events (54). Events similar to the European record hot summer of 2003 would be very likely at least 24% less frequent in a world at 1.5°C global warming compared to 2°C global warming (158).
Future intensity of heat waves similar to the one of 2010
The 2010 heat wave in eastern Europe and Russia ranks among the hottest events ever recorded in the region. It was likely the hottest summer in the last 500 years in this part of the world. There were over 50,000 extra heat-related deaths, and economic losses were more than US$15 billion (54).
In addition to the widespread, anticyclonic conditions that led to this heat wave, depletion of soil moisture was a crucial driver behind the extreme heat. Evaporation from the soil and transpiration by plants was very low due to the very dry soil. As a result, little of the sun’s radiation was used for evaporation and transpiration, and thus turned into a latent heat flux that cools the surface. Instead, most of the sun’s radiation was turned into surface warming (90).
The low soil moisture conditions were a necessary basis for this extreme heat wave. Still, under present climate conditions, soil moisture levels are generally high enough, and surface evaporation is strong enough to cap maximum summer temperatures (191). These constraints may weaken, however, under future warming (192), scientists warn in an article in Nature Climate Change (190). In future summers, soil moisture levels quickly decline after the start of summer, and the soil will be largely depleted of moisture long before the end of summer. Even an above-average wet spring will no longer be able to constrain summer temperatures under atmospheric conditions similar to those of the 2010 heat wave.
They simulated the atmospheric heat wave conditions of 2010, in a warmer world. They focused on 2075 under a high-end scenario of global warming (RCP 8.5). ‘Future mid-latitude heat waves analogous to the 2010 event will become even more extreme than previously thought’, they conclude, ‘with temperature extremes increasing by 8.4 °C over western Russia.’ Because of the disappearance of the evaporative cooling constraint, atmospheric conditions similar to those of 2010 will lead to much higher temperatures than the mean climate change effect.
Heat waves in Central Europe
In the last three decades of the previous century Central Europe has experienced 22 heat waves. Future changes in the frequency of occurrence of these heat waves were studied for this part of Europe, that includes most of Germany, the Czech Republic, Slovakia, the Southwest of Poland, Northern Austria and Hungary (160). In this study a heat wave is defined by at least three consecutive hot days. A hot day occurs when on average maximum daily temperature over Central Europe exceeds the 90th percentile of the distribution of daily maximum summer temperatures for the period 1970 - 1999. This value of the 90th percentile for this historical period is also kept as the reference for quantifying future hot days and heat waves.
Compared to this historic period, the frequency of heat waves is projected to increase by a factor 2 in the near future (2020 - 2049). For the late twenty-first century (2070 - 2099), the projected frequency increase of heat waves depends on the rate of climate change: under a high-end scenario of climate change (RCP 8.5), 3-4 heat waves per summer are projected, compared to about two heat waves under a moderate scenario of climate change (RCP 4.5). These projections are based on a large number of combinations of global and regional climate models (160).
The 1994 heat wave is found to be the most distinctive during the 1970 - 1999 period. It lasted 16 days and was associated with large excess mortality in the Czech Republic (161), Poland (162) and other Central European countries. This heat wave has been ranked as the most severe in Central Europe over the whole 1950 - 2012 period (163). Such extraordinary heat waves will probably still be rather rare in the near future. At the end of this century, however, at least one event per decade similar to the 1994 heat wave is projected for Central Europe (160).
Heat wave impacts on electricity supply
Extreme heat waves have an impact on western European electricity supply due to the increased electricity demand for cooling (83) and the power limitation of thermoelectric plants due to regulation constraints concerning downstream water temperature (40). At least the two most severe heat waves during 1979–2008 (those of 2003 and 2006) had a negative impact on western European electricity supply, with shut downs of thermoelectric plants (84).
Therefore, changes of western European heat wave characteristics have been studied for the end of the twenty first century (2070–2099), compared with 1979–2008 (82). Heat waves were defined as periods of at least 3 consecutive days of extremely high daily maximum temperature affecting at least 30 % of Western Europe. For the high temperature threshold the 98th percentile of the daily maximum temperature during March-October was chosen; for western Europe, this more or less agrees with a threshold of 30°C, commonly used to define a “hot day”.
The study is based on 19 climate change models and three scenarios of future greenhouse gasses concentrations (the so-called RCP2.6, RCP4.5 and RCP8.5 scenarios) that cover a range of moderate to high-end climate change projections. The results indicate a strong increase of the number of heat waves. Besides, the average future heat wave lasts longer, is more extended and more intense. Heat waves with similar or higher severity than observed for the 2003 heat wave remain rather rare, except for the most extreme scenario of climate change (RCP8.5) (82).
For the more moderate scenarios of climate change (RCP2.6 and RCP4.5), the median values of highest simulated severities for all model results are comparable to the 2003 heat wave; for the high-end scenario (RCP8.5), heat waves (median of all model results) with 5 times higher severity than the 2003 heat wave are simulated. For this scenario, planners would need to adapt to yet unprecedented heat wave severities. The spread in results between different models is large for all heat wave characteristics, however. Far more extreme values for heat wave characteristics have been calculated than the median values of all model results: the variability in results due to different models is larger than the variability due to different climate change scenarios (82).
Dry spell changes in Europe in the 21st century
The first results of the high-resolution (horizontal resolution of 12.5 km) future climate simulations (from the EURO-CORDEX program) address changes in annual mean temperature and total precipitation across Europe for three scenarios (A1B, RCP4.5 and RCP8.5) and the time periods 2021–2050 and 2071–2100 (compared with 1971–2000). For these simulations a large number of regional and global climate models have been used (72). According to these results, a small increase in the length of extended dry spells is projected for Central Europe in 2071–2100 (compared with 1971–2000) for the moderate scenarios RCP4.5 and A1B. A decrease in the length of extended dry spells is calculated for this period in A1B for parts of Scandinavia. Under RCP8.5 more but shorter dry spells are projected in the alpine region (72).
Changes in temperature variability in Europe in the 21st century
Numerous studies conclude that a change in variance of the temperature Probability Density Function (PDF) has a greater impact on the frequency and duration of extremes than just a shift in the mean (8). Recent studies show that by 2071 to 2100, summer temperature variability is expected to have increased by up to 100% in central Europe (33). Winter temperature variability and the number of cold and frost extremes are projected to decrease further (1,62). In agreement with these results it was concluded a robust signal across a large number of global and regional climate model projections that in summer and south-central Europe hot extremes warm stronger than the mean, and in winter and northern Europe cold extremes warm stronger than mean temperatures (56).
Precipitation changes in Europe in the 21st century
Annual and seasonal precipitation
The first results of the high-resolution (horizontal resolution of 12.5 km) future climate simulations (from the EURO-CORDEX program) address changes in annual mean temperature and total precipitation across Europe for three scenarios (A1B, RCP4.5 and RCP8.5) and the time periods 2021–2050 and 2071–2100 (compared with 1971–2000). For these simulations a large number of regional and global climate models have been used (72). According to these results, in large parts of Central Europe and Northern Europe annual precipitation may increase in 2071–2100 (compared with 1971–2000) up to about 25 % and decrease in Southern Europe. For winter in this period, RCP8.5 projects strongest increases in heavy precipitation (up to 35 %) in Central and Eastern Europe. The moderate RCP4.5 scenario shows a similar pattern of changes, though less pronounced. The spatial pattern for the A1B precipitation changes qualitatively agrees with the described changes for RCP4.5 and RCP8.5, and the magnitude of the changes mostly lies in-between the two RCPs (72).
According to previous studies it is expected that precipitation will increase between 1 and 2% per decade in Northern Europe while it will decrease by less than 1% per decade in Southern Europe. At the same time seasonal differences will increase throughout Europe. Most of the continent will get wetter in winter, while during summer drying is expected in the South (-5% per decade) (38). Summer precipitation may decrease substantially (in some areas up to 70%, depending on the scenario) in Southern and Central Europe. For Northern Europe both a slight decrease (38,44) and increase (39) of summer precipitation have been reported. Relatively small precipitation changes were found for spring and autumn (44).
Models project an increase in winter precipitation in Northern Europe, whereas many parts of Europe may experience dryer summers. But there are uncertainties in the magnitude and geographical details of the changes. Geographically, projections indicate a general precipitation increase in Northern Europe and a decrease in Southern Europe. The change in annual mean between 1980–1999 and 2080–2099 (for the intermediate IPCC SRES A1B projections) varies from 5 to 20% in Northern Europe and from – 5 to – 30% in Southern Europe and the Mediterranean (1).
This century, storms will get wetter and more erratic as the world warms, and in most of Europe (except for the most southern part) the heaviest rains will occur more often (183).
Very high daily precipitation, defined as the 95th percentile of precipitation on wet days, may increase strongly over western, central and northern Europe at the end of the 21st century, with the smallest increase (about 20%) for a low scenario (RCP2.6) and the largest (40-70 %) for a high-end (RCP8.5) scenario of climate change (108). Extreme daily precipitation is projected to increase in southern Europe. For a high-end scenario of climate change, this increase may be more than 20% locally in the Po and Veneto basins in Italy, the Rhône in France, Northern Greece, and the basins covering Slovenia and Croatia in the Adriatic (178).
Extreme precipitation events are projected to increase by 17% in Northern and 13% in Central Europe during the 21st century, with no changes projected in Southern Europe (40). From a comparison of scenarios of European precipitation extremes based on six different regional climate models (RCMs) it was concluded that the frequency of precipitation extremes increases from the period 1961–1990 to 2071–2100 by a factor of 2 to 5 over Northern Europe in winter. The corresponding numbers for Central Europe are from no change to a frequency increase by a factor of about 2. The simulated response in winter qualitatively conforms to the observed trends in heavy precipitation over Europe, which shows an increase in winter primarily north of 45°N (41). It is premature, however, to infer the detection of an anthropogenic influence on heavy precipitation in Europe (42).
According to regional climate model calculations for the period 2071–2100, and compared with the period 1961–1990, the mean precipitation increases in Northern Europe partly due to an increased number of wet days and partly due to increased precipitation intensity. In southern Europe, both the number of wet days and total precipitation is generally projected to decrease. At the same time there are extensive areas in southern Europe that are likely to experience relatively more intense precipitation (32).
For summer, there is a much larger variation in the change of precipitation extremes between the RCMs than in winter. All models show a tendency toward larger increases or smaller decreases (depending on the model) with more and more extreme events. The larger-scale pattern shows a gradient from increases in northern Scandinavia to decreases in the Mediterranean region (41). A future simulation of summer precipitation for the future period 2031 – 2036 and a high-end scenario of climate change (RCP8.5) with a detailed weather model suggests that, compared with the period 1990 - 1995, the summers in Europe will possibly be drier, with longer dry spells, shorter wet spells and heavier precipitation (177).
In Southern Europe, the longest dry period within a year may be prolonged by one month at the end of 21st century. In Central Europe, prolongation of longest dry period is by one week, and no prolongation is projected for Northern Europe. Thus regions in Europe that are now dry are projected to become even more vulnerable (1,38).The longest yearly dry spell could increase by as much as 50%, especially over France and Central Europe (43). Studies on the impact of stringent CO2 mitigation indicate that some significant increases in drought may be inevitable even under strongest mitigation scenarios (58).
Future vulnerability of European urban areas to climatological extreme events has been assessed for the period 2021-2050 compared to the reference period 1971-2000 (139). This was done for highly populated urban areas in Central Europe (the east of France, Germany, Austria, Czech Republic, Slovenia, north of Italy). The assessment is based on high-resolution regional climate model simulations and a scenario of moderate climate change (the so-called IPCC A1B scenario). On average for Central Europe the results show an increase of the number of heavy precipitation events by40 %. In Europe, heavy precipitation events are most likely in the summer (June, July and August). For Central Europe the relative increase in the number of days with heavy precipitation in these months (daily accumulated precipitation > 20 mm) is largest in northern Germany, Belgium and the Netherlands (> 30%) (139).
In their 5th Assessment Report the IPCC presented a precipitation increase averaged over northern Europe at the end of the 21st century for April through September ranging from 3 to 4% for the mid-21st century and 5 to 8% at the end of the century, under an intermediate (RCP4.5) to high-end (RCP8.5) scenario of climate change (107). For October through March the projected increase in precipitation averaged over northern Europe ranges from 8% (RCP4.5) to 11% (RCP8.5) for the mid-21st century and from 11% (RCP4.5) to 20% (RCP8.5) at the end of the 21st century (107).
According to a large number of models (GCM’s), precipitation amounts of the larger Mediterranean basin will decrease in all seasons at almost all parts of the basin in 2070 - 2100 with respect to reference period of 1970 - 2000 under the A2, A1B and B1 emissions scenarios. Increased drought conditions and heavy precipitation events will very likely occur more frequently in 2100 with respect to the reference period (89). Heavy rainfall events are likely to intensify by 10-20%, in all seasons except summer (189). Similar results were found for more recent model projections (based on the RCP8.5 scenario of climate change): widespread reduction of precipitation will be particularly serious for the Iberian, Balkan, and Anatolian peninsulas (179).
A global atmospheric temperature increase of 2°C will probably be accompanied by a reduction in summer precipitation of around 10-15% in Southern France, Northwestern Spain and the Balkans, and up to 30% in Turkey and Portugal (187). Scenarios with 2-4 °C temperature increases in the 2080s for Southern Europe would imply widespread decreases in precipitation of up to 30% (especially in spring and summer months) (188).
Hail events changes in Europe in the 21st century
Future projections of hail events are subject to large uncertainties, because small-scale hail events cannot be directly represented in global and regional climate models. However, model-based studies for central Europe show some agreement that hailstorm frequency will increase in this region (146).
Snow cover changes in Europe in the 21st century
The IPCC concluded in 2019 that at lower elevation in many regions such as the European Alps, the snow depth or mass is projected to decline by 25% (likely range between 10 and 40%), between the recent past period (1986-2005) and the near future (2031-2050), regardless of the greenhouse gas emission scenario. This corresponds to a continuation of the ongoing decrease in annual snow cover duration (on average 5 days per decade, with a likely range from 0 to 10). By the end of the century (2081-2100), reductions of up to 80% (likely range from 50 to 90%) are expected under RCP8.5, 50% (likely range from 30 to 70 %) under RCP4.5 and 30% (likely range from 10 to 40 %) under RCP2.6 (216).
Model simulations project widespread reductions in snow cover over the 21st century (45,146). Decreases of between 9 and 17% in the annual mean northern hemisphere snow cover by the end of 21st century are projected by individual models (45). Although winter precipitation is projected to increase in northern and central Europe, less frequent frost occurrences associated with higher temperatures are projected to reduce the number of days with snow cover. Decreases of more than 60 snow-cover days are projected to occur (for the period 2071–2100 compared with 1961–1990) around the northern Baltic Sea, on the west slopes of the Scandinavian mountains and in the Alps (46). The beginning of the snow accumulation season is projected to be later and the end earlier, and snow coverage during the snow season is projected to decrease (47).
Climate model projections for the period 1950–2100, assuming the SRES A1b emissions scenario, show that precipitation on future days, where average daily temperature is below freezing, decreases in large parts of Europe in a future warmer climate. This change occurs in addition to the number of such days decreasing strongly in the future. Observations also show that this decrease has already set in in the twentieth century. The result is in marked contrast to the climatological winter precipitation which increases in most global-warming scenarios. It has been argued that this is because these days in the future will occur for more extreme circulation types associated with, on average, drier weather conditions (64). These results were confirmed for western and central Europe for the period 2071-2100 compared with 1981-2010 (model projections based on RCP4.5 and RCP8.5 scenarios) (67).
Ice cover changes in Europe in the 21st century
Ice provides a range of ecosystem services, including fish harvest, cultural traditions, transportation, recreationand regulation of the hydrological cycle, to more than half of the world’s 117 million lakes. An extensive loss of lake ice will occur within the next generation. From a comprehensive large-scale assessment of lake ice loss in the Northern Hemisphere it was estimated that 14,800 lakes in the Northern Hemisphere currently experience intermittent winter ice cover, increasing to 35,300 and 230,400 at 2°C and 8°C global warming, respectively, impacting up to 394 and 656 million people (193).
Wind climate changes in Europe in the 21st century
A belt stretching from the United Kingdom to Poland will experience an increase in extreme wind speed, while Southern Europe and the Mediterranean will rather see a decrease in strong winds (69). An extension of the Atlantic storm track to the northeast is projected in combination with a narrowing of the storm track over western Europe (109). Projections of changes in storm frequency and intensity are still very uncertain, however, and model-dependent. From different studies different, and sometimes contradictory, results have been reported.
In a recent review on storminess over the North Atlantic and north-western Europe, projected changes were summarized in both storm frequency and storm intensity on the basis of numerous recent studies that assessed potential future change in these two aspects of storms on the basis of climate scenario simulations with different kinds of models. For the North Sea region, most of these studies showed a future increase in storm frequency and storm intensity (110).
From statistical analyses (ranking and extreme value statistics) for a large part of Europe a general and consistent tendency towards an increased frequency of windstorm-related losses over most of Western, Central and Eastern Europe was concluded for IPCC B1 and A2 scenarios, and slightly inconsistent findings for A1B scenario. From these analyses it was concluded that losses may reach unseen magnitudes at the end of the 21st century, which for some countries (e.g. Germany) may exceed 200% of the strongest event in present day climate simulations. In these analyses, it was assumed that storm damages occur only at 2% of all days; the minimum wind speed that is expected to produce any loss, therefore, is defined as the regional 98th percentile of the daily maximum wind speed (59).
These statistical analyses show 3 different tendencies for the period 2060-2100 compared with 1960-2000 (59):
- Countries with shorter return periods of storms and higher losses for all 3 climate scenarios: Germany, Belgium, the Netherlands, Poland, Estonia, Austria, Croatia, Bosnia and Hungary;
- Norway with longer return periods and lower losses for all 3 climate scenarios;
- All other countries in the studied part of Europe (Czech Republic, Finland, Great Britain, Ireland, Italy, Latvia, Lithuania, Portugal, Slovakia, Slovenia, Spain, Switzerland) have typically higher losses under future climate conditions and in some cases shorter return periods. Some countries, e.g. Italy and Sweden, actually show a tendency to longer RPs (A1B scenario).
In addition to these statistical analyses, simulations by a global climate model for the period 2060-2100 show that maximum storm losses for countries of Western Europe could increase by ~65% by the end of the 21st century, according to the IPCC A1B and A2 scenarios (59). Similar results were found in earlier studies for Central Europe (60), and some European countries (61). The significance of changes in storm magnitude strongly depends on country and scenario. For many countries, findings point towards higher loss events, significant for at least one of the tree studied IPCC climate change scenarios (B1, A1B, A2). An exception is Norway, for which weaker losses are found (59). For the Mediterranean, model simulations show a reduction in the total number of cyclones and windstorms crossing the Mediterranean region under climate change conditions (SRES A1B scenario) for the first half of the twenty-first century compared with the end of the previous century (68).
Cyclones in Europe in the 21st century
Several recent models suggest that the frequency of Atlantic tropical cyclones could decrease as the climate warms. However, these models are unable to reproduce storms of category 3 or higher intensity (48). An operational hurricane-prediction model has been applied to calculate a realistic distribution of intense hurricane activity for present-day conditions. The model projects nearly a doubling of the frequency of category 4 and 5 storms by the end of the 21st century, despite a decrease in the overall frequency of tropical cyclones. The largest increase is projected to occur in the Western Atlantic, north of 20°N (48). Changes in the storm track with increased cyclone density over western Europe in winter and reduced cyclone density on the southern flank of the storm track over western Europe in summer are projected to occur towards the end of the 21st century (109).
Medicanes: cyclones in the Mediterranean Sea
Hurricanes that form in the Mediterranean Sea are called Medicanes: a combination of both words. Medicane winds reach strength over 12 Beaufort. They are a rare phenomenon. They can develop when storms pass the warm water of the Mediterranean Sea, just like tropical cyclones develop in the Atlantic Ocean (207). In previous decades, Medicanes in the Mediterranean were mostly observed south of the Gulf of Lion‐Genoa (western Mediterranean), and (to a lesser degree) southeast of Sicily (Ionian Sea). They form predominantly from autumn to spring (206).
Medicanes resemble tropical cyclones (208). Sea surface temperature is a key factor that drives the formation of these hurricanes. Global warming will increase sea surface temperature of the Mediterranean Sea. This will affect the occurrence and strength of Medicanes. Changes in the characteristics of Medicanes have been studied for the near future (2016 - 2035) and the far future (2081 - 2100), compared with the situation in the recent past (1986 - 2005). A global climate model was used for this; scenarios were calculated under an intermediate scenario of climate change (RCP4.5)). Dramatic shifts were found in Medicane‐related hazards by the end of this century (206).
Medicanes will occur less frequent (see also (209)) but they are likely to become more vigorous. They will become more tropical: they will be associated more frequently with stronger winds and will lead to more intense precipitation. Flood risk posed by these storms will increase. Also, they may change the path they follow: they may become more common in the Ionian Sea and less common in the western Mediterranean Sea. The model results show that Medicanes will probably not change substantially in the next few decades. They will change significantly by the end of the century, however. In fact, the study was carried out under an intermediate scenario of climate change. When climate change turns out more strongly, the Mediterranean Sea will warm up more strongly and this will amplify the Medicanes' potential destructiveness even more (206).
Impacts on glaciers in the 21st century
Changes in glacier mass for all glaciers in the world have been modelled for climate change projections based from 14 global climate models and two emission scenarios (midrange scenario RCP4.5 and high end scenario RCP8.5; these are the scenarios for the Fifth Assessment Report of the Intergovernmental Panel on Climate Change). These results suggest a mean reduction of the current global glacier volume by 29% (RCP4.5) or 41% (RCP 8.5) over the period 2006–2100. This equals a contribution of these glaciers to sea level rise of 155 ± 41 mm (RCP4.5) and 216 ± 44 mm (RCP8.5); these numbers are multi-model means ± one standard deviation (65).
The regions that are projected to lose more than 75% of their current volume on average are, for RCP4.5: Western Canada and US (85 ± 11%), Scandinavia (76 ± 25%), North Asia (88 ± 7%), Central Europe (84 ± 10%), Caucasus (89 ± 7%) and Low Latitudes (88 ± 6%). For RCP8.5 all of these regions lose on average more than 90% of their current volume by 2100 (65).
Many glaciers are projected to disappear regardless of future emissions (213).
Arctic sea-ice cover in the 21st century
The Arctic summer sea-ice cover normally reaches its annual lowest extent in September. The lowest sea ice extent so far was reached in September 2012. The recent Paris Agreement has inspired many scientists to study the differences in impacts between 1.5 °C and 2.0 °C global warming (above pre-industrial conditions). One of these studies focused on the impact on summer Arctic sea ice extent (185).
This model study showed a huge impact of limiting warming to 1.5 °C compared with 2.0 °C warming, or more. If warming is limited to 1.5°C, sea ice cover in September will be less than the record 2012 minimum only 55% of the time in the late twenty-first century. At 2.0 °C this will be the case in almost all September months. The likelihood of any September to be completely ice-free under a 2.0 °C scenario is 35% by 2100, according to this study. If warming is limited to 1.5 °C, this likelihood is only a few percent. Besides, under scenarios that reach or exceed 2.0°C, ice-free conditions for multiple summer months and for several years in a row also become possible by the late twenty-first century (185).
Still, even when warming is limited to 1.5 °C, the Arctic summer sea-ice cover experiences significant reductions compared to today's cover. Nevertheless, to limit warming to 1.5 °C rather than 2.0°C or more has the potential to avoid frequent ice-free conditions as well as to prevent a loss of multiyear sea ice in the Arctic (185).
The references below are cited in full in a separate map 'References'. Please click here if you are looking for the full references for Europe.
- EEA, JRC and WHO (2008)
- IPCC (2007)
- Böhm et al. (2001), in: EEA, JRC and WHO (2008)
- Klein Tank (2004), in: EEA, JRC and WHO (2008)
- Klein Tank et al. (2002), in: EEA, JRC and WHO (2008)
- Luterbacher et al. (2004), in: EEA, JRC and WHO (2008)
- Schär and Jendritzky (2004), in: EEA, JRC and WHO (2008)
- Della-Marta et al. (2007)
- Klein Tank and Können (2003); Alexander et al. (2006); Moberg et al. (2006), all in: Della-Marta et al. (2007)
- Klein Tank et al. (2005); Moberg and Jones (2005), both in: Fischer and Schär (2009)
- Stott et al. (2004), in: Nicholls and Alexander (2007)
- Lorenz et al. (2019)
- Jones and Moberg (2003), in: Alcamo et al. (2007)
- Eisenreich (2005)
- Norrant and Douguédroit (2006), in: EEA, JRC and WHO (2008)
- Klein Tank et al. (2002), in: EEA, JRC and WHO (2008)
- Nicholls and Alexander (2007)
- Klein Tank and Können (2003); Groisman et al. (2004), both in:Nicholls and Alexander (2007)
- Scherrer et al. (2004), in: EEA, JRC and WHO (2008)
- Vojtek et al.(2003), in: EEA, JRC and WHO (2008)
- Rodriguez et al. (2005), in: EEA, JRC and WHO (2008)
- Petkova et al.(2004), in: EEA, JRC and WHO (2008)
- Falarz (2002), in: EEA, JRC and WHO (2008)
- Kohler et al.(2006), in: EEA, JRC and WHO (2008)
- Hyvärinen (2003), in: EEA, JRC and WHO (2008)
- Magnuson et al. (2000), in: EEA, JRC and WHO (2008)
- von Storch et al. (2002), in: EEA, JRC and WHO (2008)
- Lionello (2005), in: EEA, JRC and WHO (2008)
- Christensen et al. (2007), in: EEA, JRC and WHO (2008)
- Giorgi et al. (2004); IPCC (2007a), both in: EEA, JRC and WHO (2008)
- IPCC (2007a, 2007b); ACIA (2004), all in: EEA, JRC and WHO (2008)
- Kjellström (2004)
- Schär et al. (2004); Weisheimer and Palmer (2005), both in: Della-Marta et al. (2007)
- Fischer and Schär (2009)
- Beniston et al. (2007), in: Alcamo et al. (2007)
- Räisänen et al. (2004); Kjellström et al. (2007), both in: Alcamo et al. (2007)
- Meehl and Tebaldi (2004)
- Beniston et al. (2007)
- Eisenreich (2005)
- Sillmann and Roeckner (2008), in: EEA, JRC and WHO (2008)
- Frei et al. (2006)
- Kiktev et al. (2003, 2004); Hegerl et al.(2004), all in: Frei et al. (2006)
- Good et al. (2006), in: Alcamo et al. (2007)
- Kjellström, 2004; Räisänen et al., 2004, both in: Alcamo et al. (2007)
- ACIA (2004)
- Jylhä et al. (2007), in: EEA, JRC and WHO (2008)
- Hosaka et al. (2005), in: EEA, JRC and WHO (2008)
- Bender et al. (2010)
- IPCC (2000)
- Beniston (2004)
- Min et al. (2011), in: Coumou and Rahmstorf (2012)
- Wang et al. (2006); Chang and Guo (2007); Seidel et al. (2008); Ulbrich et al. (2009); Vilibic and Sepic (2010); Bender et al. (2011), in: Coumou and Rahmstorf (2012)
- Coumou and Rahmstorf (2012)
- Barriopedro et al. (2011)
- Luterbacher (2004), in: Barriopedro et al. (2011)
- Fischer et al. (2012)
- Levitus et al. (2009), in Chust et al. (2011)
- Warren et al. (2012)
- Pinto et al. (2012)
- Schwierz et al. (2010), in: Pinto et al. (2012)
- Leckebusch et al. (2007); Pinto et al. (2007a); Donat et al. (2011), all in: Pinto et al. (2012)
- Elguindi et al. (2012)
- Van den Besselaar et al. (2013)
- De Vries et al. (2013)
- Radić et al. (2014)
- Cowtan and Way (2013), in: Kintisch (2014)
- De Vries et al. (2014)
- Nissen et al. (2014)
- Leckebusch et al. (2006); Donat et al. (2011a), both in: Norwegian Meteorological Institute (2013)
- Norwegian Meteorological Institute (2013)
- Dong et al. (2014)
- Jacob et al. (2014)
- Moss et al. (2010), in: Jacob et al. (2014)
- Rogelj et al. (2012), in: Jacob et al. (2014)
- Christidis et al. (2015)
- Chrysanthou et al. (2015)
- Mountain Research Initiative EDW Working Group (2015)
- Viviroli et al. (2007), in: Mountain Research Initiative EDW Working Group (2015)
- Barnett et al. (2008), in: Mountain Research Initiative EDW Working Group (2015)
- Scherrer et al. (2012), in: Mountain Research Initiative EDW Working Group (2015)
- Fischer and Knutti (2015)
- Schoetter et al. (2015)
- Savić et al. (2014), in: Schoetter et al. (2015)
- Van Vliet et al. (2012), in: Schoetter et al. (2015)
- Cioffi et al. (2015)
- Franzke (2015)
- Diffenbaugh et al. (2005); Miralles et al. (2014), both in: Horton et al. (2015)
- Horton et al. (2015)
- Ozturk et al. (2015)
- Hauser et al. (2016)
- Otto et al. (2012), in: Hauser et al. (2016)
- Ma et al. (2016)
- Penuelas et al. (2009); Richardson et al. (2013), in: Ma et al. (2016)
- Kljun et al. (2006); Hu et al. (2010), both in: Ma et al. (2016)
- Zaitchik et al. (2006); Fischer et al. (2007), both in: Ma et al. (2016)
- Bador et al. (2016)
- Wergen and Krug (2010); Meehl et al. (2009); Elguindi et al. (2012); Wergen et al. (2014); Coumou et al. (2013), all in: Bador et al. (2016)
- King et al. (2016)
- Lewis and Karoly (2013), in: King et al. (2016)
- King et al. (2015a), in: King et al. (2016)
- King et al. (2015b), in: King et al. (2016)
- Wild et al. (2005), in: King et al. (2016)
- Min et al. (2014), in: King et al. (2016)
- Collins et al. (2013)
- Collins et al. (2013), in: May et al. (2016)
- Christensen (2013a and 2013b), in: May et al. (2016)
- IPCC (2013), in: May et al. (2016)
- Sillmann et al. (2013), in: May et al. (2016)
- Zappa et al. (2013), in: May et al. (2016)
- Feser et al. (2015), in: May et al. (2016)
- Feser et al. (2015a), in: Stendel et al. (2016)
- Ulbrich et al. (2009); Feser et al. (2015a), both in: Stendel et al. (2016)
- Stendel et al. (2016)
- Mueller et al. (2016)
- Jones et al (2008); Christidis et al (2015b) , both in: Mueller et al. (2016)
- Thomson et al. (2011), in: Mueller et al. (2016)
- Herring et al. (2015)
- Photiadou et al. (2015), in: Herring et al. (2015)
- Kam et al. (2015)
- Sippel et al. (2016)
- Dong et al. (2016)
- NOAA (2016), in: Sippel et al. (2016)
- Fisher and Knutti (2016)
- European Environment Agency (2016)
- Van der Schrier et al. (2013), in: European Environment Agency (2016)
- Kam et al. (2015); EURO4M (2015), both in: European Environment Agency (2016)
- Luterbacher et al. (2016), in: European Environment Agency (2016)
- Vautard et al. (2014), in: European Environment Agency (2016)
- Moss et al. (2010), in: European Environment Agency (2016)
- Jacob et al. (2014), in: European Environment Agency (2016)
- Meinshausen et al. (2011); Collins et al. (2013), both in: European Environment Agency (2016)
- Russo et al. (2014), in: European Environment Agency (2016)
- Schemm et al. (2017)
- James and Browning (1979); Volkert et al. (1987); Mills (2005); Catto et al. (2012); Catto and Pfahl (2013); Reeder et al. (2015); Schemm et al. (2016), all in: Schemm et al. (2017)
- IPCC (2013), in: Schemm et al. (2017)
- Zolina et al. (2014); Murawski et al. (2015), both in: Schemm et al. (2017)
- Christensen and Christensen (2003); Huntingford et al. (2003); Giorgi et al. (2016), all in: Schemm et al. (2017)
- Hoy et al. (2016)
- Fallmann et al. (2017)
- Vincent et al. (2017)
- Rahmstorf et al. (2017)
- Medhaug et al. (2017)
- National Academies of Sciences, Engineering, and Medicine (2016)
- Stott et al. (2015), in: National Academies of Sciences, Engineering, and Medicine (2016)
- Łupikasza (2017)
- European Environment Agency (2017)
- Munich RE (2014), in: European Environment Agency (2017)
- Zemp et al. (2008, 2015); Huss (2012), all in: European Environment Agency (2017)
- Nesje et al. (2008); Engelhardt et al. (2013); Hanssen-Bauer et al. (2015), all in: European Environment Agency (2017)
- Radić et al. (2014), in: European Environment Agency (2017)
- Marzeion et al. (2012); Huss and Hock (2015), both in: European Environment Agency (2017)
- Nesje et al. (2008), in: European Environment Agency (2017)
- Takács (2011), in: European Environment Agency (2017)
- SYKE (2011), in: European Environment Agency (2017)
- Naturwissenschaften Schweiz (2016), in: European Environment Agency (2017)
- Beniston et al. (2017)
- Van Oldenborgh et al. (2009), in: Beniston et al. (2017)
- King and Karoly (2017)
- Ummenhofer et al. (2017)
- Lhotka et al. (2018)
- Kyselý and Huth (2004), in: Lhotka et al. (2018)
- Kuchcik (2001), in: Lhotka et al. (2018)
- Lhotka and Kyselý (2015a), in: Lhotka et al. (2018)
- Lehner et al. (2018)
- Dosio et al. (2018)
- Pal and Eltahir (2016), in: Dosio et al. (2018)
- Sánchez-Benítez et al. (2018)
- Meehl and Tebaldi (2004); Russo et al. (2015); Schoetter et al. (2015), all in: Sánchez-Benítez et al. (2018)
- Peña-Ortíz et al. (2015), in: Sánchez-Benítez et al. (2018)
- AEMET (2017), in: Sánchez-Benítez et al. (2018)
- KNMI (2017); Meteofrance (2017); Meteosuisse (2017), all in: Sánchez-Benítez et al. (2018)
- Cassou and Cattiaux (2016), in: Sánchez-Benítez et al. (2018)
- Stott et al. (2004), in: Suarez-Gutierrez et al. (2018)
- Suarez-Gutierrez et al. (2018)
- Schleussner et al. (2016); Perkins-Kirkpatrick and Gibson (2017); King and Karoly (2017); Sanderson et al. (2017), all in: Suarez-Gutierrez et al. (2018)
- Teuling (2018)
- Gadian et al. (2018)
- Tramblay and Somot (2018)
- Lionello and Scarascia (2018)
- Fontrodona Bach et al. (2018)
- Diffenbaugh et al. (2013); Krasting et al. (2013), both in: Fontrodona Bach et al. (2018)
- Kapnick and Delworth (2013); Krasting et al. (2013); O’Gorman (2014); Räisänen (2008, 2016), all in: Fontrodona Bach et al. (2018)
- Witze (2018)
- Cheng et al. (2019)
- Jahn (2018)
- Macias et al. (2013), in: Cramer et al. (2018)
- Vautard et al. (2014), in: Cramer et al. (2018)
- Forzieri et al. (2014), in: Cramer et al. (2018)
- Toreti et al. (2013); Toreti and Naveau (2015), both in: Cramer et al. (2018)
- Rasmijn et al. (2018)
- Fischer et al. (2007); Vautard et al. (2007); Seneviratne et al. (2010), all in: Rasmijn et al. (2018)
- Lenderink et al. (2007); Fischer et al. (2012), both in: Rasmijn et al. (2018)
- Sharma et al. (2019)
- Smid et al. (2019)
- Guerreiro et al. (2018); UN-habitat (2010), both in: Smid et al. (2019)
- Habeeb et al. (2015); Revich and Shaposhnikov (2012); Robine et al. (2008), all in: Smid et al. (2019)
- Russo et al. (2015), in: Smid et al. (2019)
- Russo et al. (2017), in: Smid et al. (2019)
- Dunne et al. (2013); Estrada et al. (2017); Zander et al. (2015), all in: Smid et al. (2019)
- Zemp et al. (2019)
- RGI (2017), in: Zemp et al. (2019)
- Huss and Farinotti (2012), in: Zemp et al. (2019)
- Khan et al. (2015), in: Zemp et al. (2019)
- IMBIE (2018), in: Zemp et al. (2019)
- Watson et al. (2015), in: Zemp et al. (2019)
- González‐Alemán et al. (2019)
- Emanuel (2005a); Moscatello et al. (2008), both in: González‐Alemán et al. (2019)
- Tous and Romero (2013), in: González‐Alemán et al. (2019)
- Cavicchia et al. (2014b); Romera et al. (2017); Tous et al. (2016); Walsh et al. (2014), all in: González‐Alemán et al. (2019)
- Masson-Delmotte et al. (2018), in: Lorenz et al. (2019)
- Cattiaux et al. (2013); Fischer and Schär (2009); Fischer et al. (2014); Orlowsky and Seneviratne (2012); Seneviratne et al. (2006), all in: Lorenz et al. (2019)
- De Vries et al. (2012); Fischer et al. (2012); Holmes et al. (2016), all in: Lorenz et al. (2019)
- IPCC (2019a)
- IPCC (2018), in: IPCC (2019b)
- Auer et al. (2007); Ceppi et al. (2012), both in: IPCC (2019b)
- IPCC (2019b)