The development of Germany’s climate since the end of the 19th century
A climate can be described in terms of the average condition of the atmosphere, characteristic extreme values and in terms of the frequency curve of meteorological phenomena such as air temperature, precipitation and wind in a specific location. A climate is the outcome of complex interactions among all components that make up the system of land, atmosphere and oceans. Also part of this system are the biosphere with seasonal changes in vegetation, the hydrosphere, the soil and the cryosphere (ice). The fact that climate changes in the course of time is known at least from our knowledge about the last ice age which covered major parts of present-day Germany with a thick blanket of ice. The evaluation of observational data since the mid-19th century indicates progressive global warming which cannot be attributed to natural causes, and today it is considered an established scientific fact that further temperature increases are to be expected. The mean value of average temperatures on the surface of land and water has steadily increased in the course of recent years. Since the 1960s, each decade was warmer than the previous one, and the data available so far for the current decade indicate that the decade from 2011 to 2020 will be marked by a new maximum level. According to the analyses carried out by the American research organisations NASA and NOAA, the global average temperature is currently around 1°C above the level prevailing in the mid-18th century (Fig. 2). It should be noted that most of the warming occurred in the course of the past 35 years: 15 of the 16 warmest years in global records were recorded in the years since 2001, with 2016 being regarded globally as the warmest year so far, and the past few years from 2015 until 2018 were, in global terms, the four warmest years since systematic records began.
For Germany sufficient data exist from 1881 onwards enabling us to identify detailed climate changes nationwide. However, this can be said only for variables such as monthly observations of temperature and precipitation. The relevant daily data as well as other measured variables such as sunshine duration are generally not available nationwide before 1951. It is possible, however, on the basis of available data to retrace at least the average conditions of the two most important meteorological variables up to the end of the 19th century and thus essentially also to the beginning of human impacts on the climate. While the impact of additional greenhouse gases on the development of temperatures in the course of the past 139 years is patently obvious, the correlation with changes in precipitation is less obvious. This is partly due to a change triggered by general warming in large-scale atmospheric conditions. Nevertheless, precipitation is a crucial factor in the availability of water and is of equally great interest as temperature itself. Other meteorological factors were not illustrated in this context because they were considered to be of somewhat minor importance. Besides, these time series, as they are shorter by 50%, allow only a limited comparison with temperature and precipitation processes. The latter is basically also true in respect of the examination of extreme events, as daily measuring values are required for such examinations. Nevertheless, owing to their high damage potential, it is precisely these events which represent the greatest threat to our society. It was therefore decided to carry out an analysis of changes observed so far, despite the limited availability of data.
Average climate changes
The assessment of mean values for atmospheric conditions was based on summarising the monthly data available for seasonal and annual mean values. In addition, the data collected at selective points in meteorological stations were applied scientifically to Germany nationwide.
The annual air temperature as an aggregated mean for Germany between 1881 and 2018 was determined statistically to have risen by 1.5°CI (see figure 2). This value is by 0.5°C higher than the global temperature throughout the same period of time. Apart from such long-term evaluations, it is customary, in line with recommendations by the World Meteorological Organisation (WMO), to collect mean values for a period of 30 years in order to ascertain the nature of the climate and its changes. This makes it possible to eliminate the influence of short-term atmospheric fluctuations from a statistical appraisal of the climate, at the same time making it possible to trace the ups and downs of the climate as a whole. To this end, the WMO has suggested to focus on the period of 1961-1990. A comparison of the climate reference period (1961–1990) with the actual reference period (1981–2010) confirms that the air temperature mean in Germany rose from 8.2°C to 8.9°C.
A closer scrutiny of the temporal development shows that the rise in temperature did not take place evenly. In fact, there were phases of warming as well as periods of stagnation, interspersed from time to time with shorter periods during which temperatures tended to decrease slightly. One reason for this is the wide range of variations in atmospheric conditions from year to year with regard to, globally speaking, a relatively small region like Germany. Figure 2 demonstrates in fact that the variability of temperatures in Germany (bar) is much greater than the global temperatures (plane). However, in the course of periods extending over several decades, the so-called decadal climate variability also plays a crucial role. These are periodic variations extending over several years or even a few decades, which are closely linked with ocean currents. Dependent on temperatures on the surface of the oceans, there will be phases in which the atmosphere either warms up or cools down. These phases overlie the influence on the climate from external driving factors. However, apart from natural elements such as solar irradiation and volcanic activity, human influences have to be taken into account, such as changes in land use, air pollution owing to sulphur output from industrial plant as well as emissions of greenhouse gases such as carbon dioxide. Periods of a greater cooling effect exerted by ocean circulation on the atmosphere can therefore lead to a total concealment of the long-term trend, even at times when the total of extreme climate drivers alone would lead to warming. As soon as the oceans’ influence is reversed, temperatures can be observed to increase.
In Germany the temperature rise observed hitherto seems to be homogenous throughout the country. In principle this applies also to the various meteorological seasons. Just in summer (June to August) the value for a surface area mean of 1.4°C deviates slightly from the annual mean. For the other seasons the same temperature increase of 1.5°C is the same as for the year as a whole. Roughly the same can be said for spatial differences. In this case, the annual mean temperature rise ranges from 1.3°C to 1.6°C, with the warming tendency in the western and southern Länder a little higher so far whereas in the northern Länder such as Brandenburg and Berlin, this tendency is a little lower than the Länder mean. Greater deviations from this general spatial distribution can be observed exclusively in the winter months. During the winter season temperature rise in the north-eastern Länder by values between 1.2°C to 1.3°C has so far been generally lowest whereas other Federal states such as Bavaria recorded a temperature rise up to 1.7 °C.
I All statements made in the text regarding changes in temperature and precipitation, as well as the indices for extremes based on those variables, were calculated by means of (least square) linear trend; they are considered statistically sound provided they achieve a significance level of at least 99%.
Contrary to temperature, there are distinct differences in changes to precipitation in Germany, especially by season but also in spatial terms. In summer the rainfall mean has remained largely unchanged whereas in winter especially, conditions have become significantly more humid. Likewise, the amounts of precipitation have increased at times of seasonal change, although this increase is distinctly lower and statistically unproven. Overall, the surface area mean for Germany since 1881 shows an increase in the annual mean precipitation of 8.7%. However, there are major differences from a spatial point of view. Especially the states in the north-west of Germany show distinct increases in wet conditions of up to 16% in Schleswig-Holstein, whereas precipitation figures from Mecklenburg-Vorpommern to Saxony-Anhalt and Thuringia show only a slight increase in the annual mean (less than 10 %). In Saxony conditions have actually become slightly drier during the same period. Spatially the picture is basically similar for the transitional seasons of spring and autumn.
The most distinct changes have so far been observed for the winter season. As demonstrated in Fig. 3, the surface area mean for average precipitation levels has increased by 25% since winter 1881 / 1882. The spatial distribution of changes is obviously similar to that of temperature at this time of year. In other words, the least increases i.e. values of less than 25% have so far been recorded in the north-eastern states of Germany. In the other German Länder the rainfall has increased more than can be said for the nationwide average. In the light of these spatially differing variables for warming and increases in precipitation, it can be said that the differences in the continentality of regions, i.e. in relation to the influence of land and sea on the climate at a specific location, show a slight rising tendency in the course of the 20th century. With regard to the summer months, there has been hardly any change so far. While it is true that the precipitation mean at that time of year has decreased by 3.8% since 1881, it must also be said that the overall minimal decrease, which is within the range of natural variability, does not allow any conclusions even regarding tendency (see Fig. 4).
As the term implies, extreme situations are rare in that they deviate strongly from usual situations. Consequently, statistical analyses are less resilient than evaluations of average situations. So-called once-in-a-century events (i.e. extreme events which statistically occur once in 100 years) have to be determined e.g. on the basis of series of measurements which typically extend to a little more than a hundred years. A relatively easy and very descriptive method of determining changes in extreme events are so-called climatological key days on which threshold values are recorded, i.e. so-called threshold value events. This is, in fact, an evaluation of days on which e.g. the maximum temperature exceeds a specific threshold value, as e.g. the number of hot days with a maximum temperature of at least 30°C. Apart from key days, it is possible to utilise other indices which can also be used for recording extreme climate events such as heat or drought periods. Listed below are various indices for the analysis of changes in extreme events regarding temperature and precipitation levels.
Statistically backed statements on changes in the frequency of cases where threshold values have been exceeded are already available: The frequency of hot days has increased in Germany nationwide, whereas ice days (days with maximum temperatures of < 0°C) have become more and more infrequent during the past 60 years. At the same time, the frequency of intensive hot periods has increased, and the heat intensity has increased nationwide in Germany since 1951.
It is more difficult, however, to make reliable statements regarding trends of heavy precipitation events.. On one hand, such events display great variability both spatially and temporally. On the other hand, especially during summer months, convective events (the development of showers and thunderstorms) are considered relevant in cases where they occur either within the space of an hour or less. Although it is possible to observe tendencies towards a greater frequency of heavy precipitation events in the course of the past 65 years, it has so far not been possible, owing to the lack of available data, to make any statistically backed climatological statements on changes in heavy precipitation events.
For the analysis of temperature extremes the amount of hot days and ice days was taken into account. Furthermore, the most intensive annual 14-day heat period with a daily maximum of at least 30 °C air temperature for the period 1951–2018 was evaluated for eight German cities.
Since 1951 there has been an increase in the number of hot days in terms of the surface area mean for Germany from a mean of approximately three days per annum to a current mean of approx. ten days per annum (see Fig. 5, left). More than ten hot days have never been recorded in Germany before 1994. The years with the most hot days were 2018, 2003 and 2015. This increase is backed up by statistics, notwithstanding great variability of this index from year to year. In contrast, the decrease in the mean of ice days by approximately 27 days per annum to currently approx. 28 days per annum is much less distinctive and statistically not proven (see Fig. 5, right).
Figure 6 shows the most intensive annual 14-days heat period for several cities, with a daily maximum air temperature mean of at least 30°C for the period 1950–2018. Regarding the cities examined, it is clear to see that the frequency and intensity of the intensive heat periods examined in this context show a rising tendency from north to south. Generally speaking, the highest daily maximum temperature mean in heat periods in the more northerly cities lies below 33°C, although this value is often exceeded in southern cities. There are fewer recordings for Munich than is typical for the south, because the recording station is located at a relatively high altitude (515 m). Furthermore, it can be seen that such extreme heatwaves have occurred more frequently since the 1990s; e.g. Hamburg never experienced such events between 1950 and 1993, whereas five extreme heatwaves have occurred there since 1994.
The temporal development of surface area mean values contained in temperature indices is clearly reflected in their spatial development. Likewise, the major spatial differences between individual German regions are clearly visible. Between 1959 and 1968 the mean of hot days was predominantly between zero and four days per annum. Just along the Rhine trough and in north-east Germany south of Berlin between four and eight such days occurred, while in parts of the southern Rhine trough up to ten such days occurred (see Fig. 7). Until the decade of 1999–2008, the number of hot days increased on average by up to 18 days per annum. The extreme north of Schleswig-Holstein was the only area where that decade again showed fewer than two hot days per annum. In the course of the past ten years the number of hot days, especially in eastern Germany and in the RhineMain area, has again increased markedly. As a result, the multi-annual mean in large areas of the south and east shows more than ten such days per annum.
Relatively warm air is able to absorb more water vapour than relatively cold air. This is why principally, consistent relative air humidity is expected to coincide with greater precipitation. Besides, it can be assumed that, especially on the so-called convective scale, i.e. the development of showers and thunderstorms, an intensification of processes leading to the development of clouds and precipitation can be expected as a result of changes in meteorological conditions. The heavy precipitation occurring under such conditions would then even increase disproportionately compared to the increased content of water vapour in the air. The term heavy rain is used for major precipitation amounts per time unit. It typically results from convective clouds (e.g. cumulonimbus clouds). Heavy rain can lead to a fast rise in water levels and flooding which are often accompanied by soil erosion. The three warning stages operated by the DWD for different durations are illustrated in Table 1.
It must be borne in mind, however, that several other factors and processes play an essential role in the development of precipitation, thus leading to regional differences. Precipitation will not increase evenly in all areas, and in some areas conditions might become drier.
Depending on prevailing measuring conditions, a differentiation is often made between daily precipitation totals and shorter intervals down to durations of 5 minutes. However, many investigations are limited to a minimum temporal resolution of 60 minutes. The frequency of heavy precipitation at a duration stage of 24 hours (see also table 1) in Germany has already increased by approximately 25% in the winter months of the past 65 years. In contrast, no distinct trend was identified for the summer months. Generally speaking, the intensity of heavy precipitation at this timescale can be described as similar.
In contrast, there are relatively few findings available for heavy precipitation of short duration occurring predominantly in summer in Central Europe. Admittedly, there are some indications for an increase in the intensity of convective events as temperatures rise. However, there is a distinct requirement for further research regarding this timescale. Trend analyses of heavy precipitation are principally hampered by the fact that not all particularly intense precipitation events of limited spatial extent are necessarily captured by meteorological stations. It is true that, in addition, there are radar data for contiguous areas, but such timescales are too short to permit making any robust trend statements.
Nevertheless, radar data have, for the first time, made it possible to capture and enumerate the occurrence of heavy rain for contiguous areas. Figure 8 shows for the first time that the hours of heavy precipitation of particularly high intensity amounting to more than 25 l/ m² in 1 hour or more than 35 l/m² in 6 hours respectively in Germany (see Fig. 8, middle) are more evenly distributed than the total hours of moderately heavy rain (see Fig.8, left) where the spatial distribution is distinctly linked to Germany’s topography. This showed for the first time that spatially extremely small-scale heavy rain of short duration and with high damage potential can occur anywhere and affect anyone in Germany; in other words, these conditions constitute a risk which is not limited to the southern states of Germany. A temporal extension of this kind of heavy rain analysis will in future allow a trend analysis for the related frequency of events where threshold values are exceeded.
Apart from the issue of changes in heavy precipitation events it is crucial, especially in summer, to what extent warming is accompanied by additional soil dehydration. Agriculture is particularly vulnerable to drought. In agriculture the term drought always refers to the condition of plants which owing to lack of water resources, either have to limit their photosynthesising activity or even die. Inadequate availability of water in the soil can be caused either by the absence or lack of precipitation or by high evaporation rates of plants; these rates are higher in dry and warm weather than under cold and humid conditions.
An ideal indicator for the degree of water supply available to plants is the soil humidity which is expressed in percent of usable field capacity (% nFK). The nFK is a relative measurement for the amount of soil water available for absorption by plants. When soil humidity drops to beneath 30% to 40% nFK, the plant’s photosynthetic output diminishes and consequently its growth declines sharply. The longer a plant remains in this condition, the more severely it can be damaged. It was therefore considered essential to examine the number of days on which the critical soil humidity values of 30% nFK for the cultivation of winter wheat were not reached. The examination was focused on the main growth period of winter wheat, which will typically last from March until July or August. The type of soil also has a major influence on soil humidity. Heavy soil (such as sandy clay) is able to store more water for plants than light soil (such as clay-rich sand); that is why the former is able to bridge longer droughts than the latter.
As shown in Figure 9, the mean number of days with soil humidity values of less than 30% nFK has increased significantly in Germany since 1961, both for heavy soil (left) and for light soil (right). Owing to the lower water storage capacity of light soil, the number of days on which the critical threshold value is not reached is generally greater for light soil than for heavy soil. Eastern Germany and the Rhine-Main area are particularly affected by increasingly dry soil (see Figure 10).
The UBA’s motto, For our environment (“Für Mensch und Umwelt”), sums up our mission pretty well, we feel. In this video we give an insight into our work.
Wörlitzer Platz 1 06844 Dessau-Roßlau GermanyPlease contact us exclusively by e-mail: buergerservice [at] uba [dot] de. Unfortunately, we are currently not available by telephone. There may currently be delays in answering enquiries. We ask for your understanding. The visitors' room in Dessau-Roßlau is temporarily closed.