National Climate Report Climate ‒ Yesterday, today and in future

Table of Contents

Foreword...... 2

Conference of the Parties (COP23)...... 3

Constantly changing – Weather and climate in Germany...... 4

Climate, climate variability and climate extremes...... 6

Climate models...... 8

Climate change and climate projections...... 10

Regional diversity – The climate in Germany...... 12

Climate variables and their changes

Temperature...... 14

Precipitation...... 20

Sunshine...... 26

Sea Level...... 28

Phenology...... 30

Extreme events...... 32

Current climate system research...... 38

A climate lexicon...... 40

Publishing details...... 42

1 Foreword

Dear readers,

Earth’s 2016 surface temperatures were the warmest since modern recordkeeping. 2016 is remarkably the third record year in a row in this series. In Germany mean temperatures in 2016 were 9.5 °C which is 1.3 °C warmer than the reference period 1961–1990. At 10.5 °C, 2014 was the warmest year on record in Germany. These temperatures very probably only represent peaks in an ongoing process. The Fifth Assessment Report by the Intergovernmental Panel on Climate Change (IPCC) predicts further signifi cant climate warming throughout the following years of this century.

Climate change represents a huge challenge for almost all of us. Its impact may, for example, be felt in far more days of high heat stress or many more extreme weather events. The international framework for responding to climate change was adopted at the World Climate Conference COP21 in Paris and the COP 22 in Marrakech. These conferences defi ned the goals which must now be met. These can only be achieved with a detailed understanding of the current state of aff airs.

The National Climate Report presents in brief and concise form the knowledge available about yesterday‘s, today‘s and tomorrow‘s climate. This report off ers its readers an in-depth introduction to climate change. In this respect, the National Climate Report delivers the knowledge base for successful adaptation to climate change.

Dr Paul Becker Vice president of Deutscher Wetterdienst

2 International co-operation to protect and explore the climate

In November 2017, the 23rd Conference of the Parties (COP23) to the United Nations Framework Convention on Climate Change (UNFCCC) will be held in Bonn, Germany, under the presidency of Fiji. Like the preceding COP session, the Climate Change Conference in Marrakech in 2016, this year‘s conference will again be focused on the implementation of the efforts for climate protection agreed during the Paris Climate Change Conference (COP21) in December 2015. COP21 brought a decisive breakthrough when 195 countries came, for the first time ever, to a general and legally binding global agreement on climate change. The agreement includes a global plan for action aimed at keeping the global temperature rise well below 2 °C, if possible even below 1.5 °C in order to reduce the impacts of climate change. Now that the USA have announced their intention to withdraw from the agreement, COP23 and its outcomes are awaited eagerly.

The scientific basis for the Paris Agreement was the Fifth Assessment Report (AR5) of the Intergovernmental Panel on Climate Change (IPCC) published in 2013/2014. The IPCC was founded by the United Nations Environment Programme (UNEP) and the World Meteorological Organization (WMO) and acts as both an international coordination body and a scientific organisation. The IPCC itself does not conduct any research, but accumulates the scientific publications of many experts. The IPCC‘s report describes the possible development of the earth‘s climate and the resulting impacts up to the end of the 21st century, revealing evidence that, without any reduction in greenhouse gas emissions, it is very probable that the earth‘s climate and its manifestations will have changed considerably compared to today.

Climate research plays an important role for the society all over the world, in Europe and in Germany. It is not only focused on exploring the natural scientific foundations of climate change or on developing climate models and running global, regional and localised climate simulations, it also examines the expected impacts and helps to identify possible measures for reducing the global warming and adapting to the climate change.

The preparations for the IPCC‘s Sixth Assessment Report have started; its publication is planned before 2022. In it, account will be taken of the latest scientific findings that result from climate simulations.

3 Constantly changing – Weather and climate in Germany

The weather in all its myriad forms shapes our life. It influences our daily choice of what to wear as well as the infrastructure which our economy and society rely on. The human-induced increase in greenhouse gas concentrations, as well as changes in land use are causing our weather and climate to change. The following pages survey climatic conditions in the past and consider future developments in Germany.

From short-term variation to long-term change: has a strong influence on the highly varied structure weather and climate in Germany of the climate. Altitude of the terrain and distance Germany is in the warm-temperate mid-latitude from the coast are the dominant influences on climate zone at the point of transition between temperature. The oceanic influence, which diminishes the maritime climate of western Europe and the from the north-west to the south-east, is responsible continental climate of eastern Europe. The central for Germany‘s relatively mild winters and moderately European climate shows the influence of moist, hot summers. mild Atlantic air masses and dry continental air, which is hot in the summer and cold in the winter. The Deutscher Wetterdienst monitors the weather The prevailing air mass depends on the large-scale in Germany from numerous locations and has been circulation pattern. This means that the seasons doing so in some places for more than 100 years. A can vary quite considerably from year to year and multitude of parameters are registered, including the climate in Germany is marked by a high level of temperature, precipitation and sunshine. The variability as a result. observed values vary from day to day and from year to year. Alongside this variability, the records The topography of Germany, its low mountain ranges made by the measurement systems of the Deutscher and the different types of landscape they encompass, Wetterdienst also help to detect long-term changes.

4 For instance, the measurements show that it has become around 1.4 degrees warmer in Germany over the last 136 years. This change has gone hand in hand with a fall in the number of cold and very cold days and a rise in the number of warm and very warm days.

The amount of precipitation has increased over the last 136 years. This is especially the case in winter and spring. Mean annual precipitation has risen by 9 per cent. There has been little noticeable change in the number of days on which there is at least 10 litres per square metre (l/m2) of precipitation.

Sea level is also measured. This has also been reduced substantially. Further economic growth and observed to change: over the last 100 years, the sea continuing high levels of greenhouse gas emissions level has risen by around 20 cm in the German Bight are likely to result in a change of between 3 °C and and by around 14 cm on the Baltic coast of Germany. 4 °C as well as a further fall in the number of cold and very cold days and a substantial rise in the number of Do human activities influence the climate? warm and very warm days. Emissions of greenhouse gases and large-scale changes in land use are two ways in which human The increase in temperature will very probably activities interfere with the Earth‘s natural climate be accompanied by a further increase in annual system. One focus of research worldwide is therefore precipitation and thus the number of days on which at on analysing the consequences of this interference. least 10 litres of precipitation per square metre (l/m2) will increase. Based on climate models, scientists have computed the impacts of the global and regional climate. A rise in air temperature will be accompanied by Depending on the scenario, these calculations show an increase in sea temperature. This will cause sea that the mean annual temperature in Germany will water to expand as it heats up and the level of the rise by between a minimum of 1 °C and over 4 °C in sea will rise. New studies of ocean warming and of the next 100 years. According to the climate models, the Antarctic and Greenland ice sheets suggest that it will only be possible to limit the rise in temperature the rate at which the sea level is rising will probably to just 1 °C if emissions of greenhouse gases are accelerate.

◂ Mean concentrations of atmospheric CO2, measured at the Mauna Loa observatory in Hawaii. The data form the longest series of direct carbon dioxide measurements worldwide. The figures shown are the monthly values (red curve) and annual mean (black curve). The variations within an individual year are the result of different periods of vegetative growth. (Source: NOAA)

5 Climate, climate variability and climate extremes

Weather, weather conditions and climate: in meteorology and climatology, these three concepts stand for processes which unfold over varying periods of time, with ‘weather’ describing the condition of the atmosphere over a short period of time, ‘weather conditions’ a phase of weeks or several months and ‘climate’ longer periods of time ranging from decades through to geological ages.

What is climate? variations in meteorological conditions. The Earth‘s The World Meteorological Organization (WMO) climate has always varied over time, with ice ages offers a scientifically precise definition of climate as alternating with warm periods. „the synthesis of weather over the whole of a period, essentially long enough to establish its statistical Climate is always a local phenomenon. The climate of ensemble properties“. Etymologically, „climate“ comes Helgoland, for example, is not the same as the climate from the Ancient Greek klĩma, or „I incline“, and of Munich. In line with the WMO definition, climates alludes to the constellation of the Earth in the solar are described by analysing conditions over a period of system, to the Earth‘s axial tilt, the varying distance at least 30 years. of our planet to the sun and the related marked

6 ▴ Deutscher Wetterdienst measuring fi eld in St. Peter-Ording.

Climate variability Mean monthly temperatures in Germany Climate is the sum of weather and weather conditions and thus varies. Describing the climate solely based on average values is not enough in itself. There is 20 °C a high level of variability even at the level of daily weather observations. This variability is also apparent in weather conditions. The same applies on longer 15 °C timescales. Mean temperatures, for instance, are lower in the winter than in the summer. The seasons themselves also diff er from year to year. Winters can 10 °C be milder or colder, summers drier or wetter.

This variability not only shows up in temperatures. 5 °C It is an aspect of all meteorological elements (e.g. Mean 1961–1990 precipitation and sunshine duration). Even climate 2014 change-induced warming features variability in that 0 °C Too warm compared to average not every year is invariably hotter than the year before. Some years may be hotter while others may be Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec colder than the mean. ▴ 2014 was the warmest year in Germany since records began. With the exception of August, average temperatures were in part Climate trends signifi cantly higher in every month than during the international We refer to a climate trend whenever, within a few standard reference period 1961–1990. decades, a change is observed to happen such that, for example, positive deviations in temperature occur. The change may well be interrupted or attenuated for short periods; the decisive factor, however, is that a long-term event, such as a prolonged drought period, the observed direction of change remains consistent or an event which is highly unusual for the time of over a long period of time. Long-term changes of this year. A temperature of 25 °C would not be unusual in kind can, of course, be the result of natural processes, July, for example. On the other hand, 25 °C on New such as variations in Earth‘s orbit or solar activity. But Year‘s Day would be unusual and would qualify as an human activities also impact the climate system. extreme event.

Extreme events Extremes are part and parcel of weather and climate, Extreme events are very rare occurrences which they were part of the climate in the past and will deviate substantially from mean conditions. There are remain part of the climate in the future. Analysis of many diff erent reasons why an event may become an the intensity and the frequency with which such extreme event. An extreme event may just occur on a events occur is a key aspect of current climate single day, such as a hurricane-force gust, or may be research.

7 ▴ Climate models use approximation formulas to describe fundamental natural processes and interactions. Some of these are shown here. (Source: Max Planck Institute for Meteorology)

Climate models

With climate change happening it is not reasonable to extrapolate the climate variations and trends observed in the past one-to-one into the future. Climate models, i.e. computer-aided tools for producing simplifi ed descriptions of natural phenomena, are therefore used to evaluate the future development of the climate.

The world in grids The climate is modelled by superimposing a three- Climate models consolidate a number of different dimensional grid on the Earth‘s atmosphere and (sub)models into one supermodel. Submodels are oceans. The resolution (grid-point spacing) of global capable of describing all the main processes in the climate models has to be very coarse to enable planet Earth‘s atmosphere, hydrosphere, cryosphere models to be calculated over many years within and biosphere. However, representations in climate an acceptable computing time. Although these models cannot be made to correspond one-to-one models provide adequate descriptions of large- to the entire range of real processes. Firstly, not all scale climate variability their resolution is too low natural processes are sufficiently well understood. to provide detailed information about the regional Secondly, this would require an extremely high level characteristics of climate change in a specifi c area of of computing time. the Earth (e.g. Germany). Regional climate models, which have a much closer grid spacing than global climate models, are used for this purpose. They are based on the results of global model runs.

8 Currently, simulations are available for Germany using grid sizes of 50 km and 12.5 km. This means, for example, that the simulated temperature can only take on a diff erent value every 12.5 km.

Reliable information cannot be provided for single grid cells. Several grid cells must always be grouped together, typically in a matrix of three by three grid cells. A model resolution of, for example, 12.5 km can only produce information for a region of 37.5 km x 37.5 km.

Many models, many results Climate models are being developed by a large number of more or less independently working research groups all around the world. Some components of these models are consequently described in diff erent terms, which in turn can lead to diff erent results. This is due to the simplifi ed basic assumptions regarding natural processes which are needed in order to develop a model.

The existing ensemble range (=group of climate ▴ Example of model grid cells with horizontal projections) is an important indicator of the state of spacing of the atmosphere and the vertical our understanding of natural processes. The greater layers. (Source: Max Planck Institute for the range, the more cautiously statements about, for Meteorology) example, change signals need to be worded.

The fi ner the resolution, the more accurate the model – in this case using the example of the orography of Germany in various model grid resolutions. The impacts of describing a region based on a substantially fi ner grid are clearly visible.▾

Global climate model (very coarse) Regional climate model (coarse) Regional climate model (fi ne) 1,875° (approx. 200 km) 0,44° (approx. 50 km) 0,11° (approx. 12.5 km)

9 Climate change and climate projections

The concept of climate change describes changes in climatic conditions at a single location or across the entire planet. In relation to the parameter of temperature, this change may represent warming or cooling. The climate change which is the subject of much discussion is not the outcome of natural influences (variations in the Earth‘s orbit or solar radiation). Human activity has a significant influence on the global and regional climate.

Climate change – the human factor The scenarios are referred to as RCP2.6, RCP4.5, Human activities impact climate in many different RCP6.0 and RCP8.5. The numbers in each case (e.g. 8.5) ways. In particular: stand for the “positive” global energy imbalance of 1. The combustion of fossil fuels releases large 8.5 W/m2 in the year 2100 compared to irradiance in the amounts of carbon dioxide and other pollutants years 1861–1880. This period represents the state of the into the atmosphere. climate before humans began to have a substantial 2. Deforestation, afforestation and surface sealing influence on the concentration of greenhouse gases in changes the way land is used on a regional and the atmosphere (pre-industrial level). The scenarios do global scale. not model the development of socio-economic factors, The changes observed in the global climate can e.g. population, energy use or emissions of greenhouse only be explained by taking account of both natural gases. However, these can be assigned to the RCPs influences and the impact of human activity. indirectly.

It is not possible to describe precisely how humans What will our emissions future look like? will affect the climate in the years and decades to This report presents the results of simulations based come. However, it is possible to make assumptions on a climate mitigation scenario (RCP2.6) and the about the probable impact of human activity over business-as-usual scenario (RCP8.5). time. In the scientific community, these assumptions are referred to as scenarios. In recent years, a number of conceivable scenarios have been developed which describe the more or less strong influence of human activity on the climate. Four representative scenarios or so-called representative concentration pathways, RCPs, were selected in the run-up to the Fifth Assessment Report by the Intergovernmental Panel on Climate Change (IPCC). These scenarios describe the radiative forcing arising from changes in concentrations of greenhouse gases and the influence of aerosols (small particles, such as soot, in the atmosphere). Put simply, the concept of radiative forcing refers to the “additional/higher” amount of energy radiating down on the planet.

10 The business-as-usual scenario (RCP8.5) describes a world in which energy systems are largely based on fossil fuel combustion. Emissions of greenhouse gases will continue to increase with a permanent increase in radiative forcing up to the year 2100.

For comparative purposes, the text also refers to a further scenario, the SRES scenario A1B. The A1B scenario describes a world of very rapid economic growth and a global population that peaks in mid- century and declines thereafter. This is the scenario on which the projections of the Earth‘s future climate in the IPCC‘s Fourth Assessment Report are based. Most of the climate change predictions communicated in ▴ Development of atmospheric carbon dioxide and the global mean recent years are based on this scenario. temperature through to the year 2300 for the diff erent emissions scenarios. (Source: http://www.climatechange2013.org/images/ What would happen if? – Climate projections fi g u r e s/ WGI_AR5_Fig12-42.jpg, changed) When a global climate model is used to compute possible climate change on the basis of one of the scenarios this is done in the framework of a climate The climate mitigation scenario (RCP2.6) is based projection. A climate projection must not be confused on 2 °C target. The aim is to keep global warming in with a forecast. A climate projection is a “what would the year 2100 below 2 °C above 1860 temperatures. happen if” calculation based on a chosen scenario. The RCP2.6 assumes very substantial and very rapid climate projections for the diff erent scenarios help to reductions in global greenhouse gas emissions down put anticipated changes in the climate into a broader from current levels. Radiative forcing will peak context, such as the range of maximum and minimum before the year 2050 (3.0 W/m2). Forcing will then changes anticipated. The actual changes which take continuously decline down to 2.6 W/m2 in the year place will most probably lie within this range. 2100. This will only be possible in a world in which energy systems are no longer based on fossil fuel This report uses the fi ndings of 54 climate projections combustion. Global greenhouse gas emissions will which cover the period 1971 to 2100. Two diff erent need to have peaked before the year 2020 to achieve 30-year time periods are used to calculate the this and to have declined to a level of insignifi cance diff erence between the baseline period and the state by 2080 (zero emissions). of the climate in the future. An average state is calculated for each period. The baseline period for the observed climate is from 1971 to 2000 as taken from the models. Two periods will be analysed in the future; these are referred to in the following as the short- term and long-term planning horizons. The short-term planning horizon describes the average state in the years 2021 to 2050. The long-term planning horizon is based on the years 2071 through to 2100. Future changes are stated as an average value and as a range. The range is described by means of the lowest and highest variation in the existing data sets.

11 Regional diversity – West GermanWestdeutsche Basin Tieflandsbucht 20 °C 0 hrs 19,710 km² 250 mm 100 hrs The climate in Germany 81 m 15 °C 200 mm

(10–463 m) 200 hrs 10 °C 150 mm It is not always appropriate to use area averages 9.6 °C 5 °C 300 hrs for the entire territory of the Federal Republic of 100 mm 771 mm 400 hrs Germany to describe the climate in Germany. The 0 °C 50 mm relevant situation can often be better described on 500 hrs 1,451 hrs -5 °C 0 mm a small or regional scale. Depending on the event or Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec issue, the necessary regional classifi cation may diff er markedly and in some cases may be limited to a very Central UplandsZentrale and Mittelgebirge Harz und Harz small geographical area. While a very fi ne division 20 °C 0 hrs 29,971 km² 250 mm would be required for certain weather phenomena, 100 hrs it is possible to defi ne larger regions on the climatic 274 m 15 °C 200 mm (40–1,141 m) 200 hrs timescale. 10 °C 150 mm 8.0 °C 5 °C 300 hrs The classifi cation applied here combines the existing 100 mm 790 mm 400 hrs natural regions and landscapes into twelve diff erent 0 °C 50 mm regions. These regions are intended to distinguish the 500 hrs 1,417 hrs -5 °C 0 mm gradient from marine to continental infl uences on the Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec one hand and relief-related structures on the other. Uplands on the rightRechtsrheinische bank of the Mittelgebirge Rhine

The resulting regions are shown in the map. The 20 °C 0 hrs 15,555 km² 250 mm associated climate diagrams show mean temperature, 100 hrs precipitation and sunshine duration during the 340 m 15 °C 200 mm (35–879 m) 200 hrs reference period 1961–1990. Corresponding data 10 °C 150 mm is also shown for the entire territory of the Federal 8.0 °C 5 °C 300 hrs Republic of Germany. 100 mm 1,006 mm 400 hrs 0 °C 50 mm

Changes in the climate in the past and possible -5 °C 500 hrs 1,434 hrs Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 mm developments in the future in Germany are explained on the following pages. If one or several regions diff er markedly, this is stated explicitly. Uplands on the left bankLinksrheinische of the MittelgebirgeRhine

20 °C 0 hrs 16,859 km² 250 mm 100 hrs 368 m 15 °C 200 mm

(80–816 m) 200 hrs 10 °C 150 mm 8.3 °C 5 °C 300 hrs Region 100 mm 863 mm 400 hrs 0 °C 50 mm Area 500 hrs 1,518 hrs -5 °C 0 mm Average altitude Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec (minimum–maximum altitude)

Mean annual temperature Upper Rhine LowlandsOberrheinisches Tiefland

20 °C 0 hrs Total annual precipitation 10,619 km² 250 mm 100 hrs 151 m 15 °C 200 mm Annual sunshine duration (80–645 m) 200 hrs 10 °C 150 mm 9.8 °C Diagrams 5 °C 300 hrs 100 mm Mean monthly temperature 722 mm 400 hrs 0 °C 50 mm Total monthly precipitation -5 °C 500 hrs Monthly sunshine duration 1,569 hrs Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 mm

12 North West GermanNordwestdeutsches Lowlands Tiefland North East GermanNordostdeutsches Lowlands Tiefland

20 °C 0 hrs 20 °C 0 hrs 60,298 km² 250 mm 46,496 km² 250 mm 100 hrs 100 hrs 32 m 15 °C 200 mm 43 m 15 °C 200 mm

(−4–210 m) 200 hrs (0–179 m) 200 hrs 10 °C 150 mm 10 °C 150 mm 8.6 °C 8.4 °C 5 °C 300 hrs 5 °C 300 hrs 100 mm 100 mm 745 mm 400 hrs 577 mm 400 hrs 0 °C 50 mm 0 °C 50 mm

-5 °C 500 hrs -5 °C 500 hrs 1,495 hrs Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 mm 1,627 hrs Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 mm

East German BasinsSüdostdt. and Becken+Hügel Hills

20 °C 0 hrs 46,831 km² 250 mm 100 hrs 149 m 15 °C 200 mm

(20–604 m) 200 hrs 10 °C 150 mm 8.7 °C 5 °C 300 hrs 100 mm 577 mm 400 hrs 0 °C 50 mm

-5 °C 500 hrs 1,564 hrs Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 mm

Nordwestdeutsches Tiefland

Nordostdeutsches Tiefland Eastern UplandsOestliche Mittelgebirge Westdeutsche Tieflandsbucht 20 °C 0 hrs Rechtsrheinische22,811 Mittelgebirge km² 250 mm

Zentrale Mittelgebirge und Harz 100 hrs 515 m 15 °C 200 mm Südostdeutsche(115–1,465 Becken und Hügel m) 10 °C 200 hrs Linksrheinische Mittelgebirge 150 mm 6.9 °C Oberrheinisches Tiefland 5 °C 300 hrs 100 mm Südwestdeutsche Mittelgebirge 873 mm 0 °C 400 hrs Östliche Mittelgebirge 50 mm

Alpenvorland -5 °C 500 hrs 1,553 hrs Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 mm Alpen

Deutschland Alpine foothillsAlpenvorland Germany 20 °C 0 hrs 20 °C 0 hrs 35,7375 km² 250 mm 33,244 km² 250 mm 100 hrs 100 hrs 250 m 15 °C 200 mm 530 m 15 °C 200 mm

(−4–2,962 m) 200 hrs (305–1,516 m) 200 hrs 10 °C 150 mm 10 °C 150 mm 8.2 °C 7.8 °C 5 °C 300 hrs 5 °C 300 hrs 100 mm 100 mm 789 mm 400 hrs 979 mm 400 hrs 0 °C 50 mm 0 °C 50 mm

-5 °C 500 hrs -5 °C 500 hrs 1,544 hrs Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 mm 1,637 hrs Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 mm

South West GermanSüdwestdt. Uplands Mittelgebirge Alps Alpen

20 °C 0 hrs 20 °C 0 hrs 56,827 km² 250 mm 4,170 km² 250 mm 100 hrs 100 hrs 446 m 15 °C 200 mm 1,117 m 15 °C 200 mm

(110–1,493 m) 200 hrs (420–2,962 m) 200 hrs 10 °C 150 mm 10 °C 150 mm 7.9 °C 5.1 °C 5 °C 300 hrs 5 °C 300 hrs 100 mm 100 mm 879 mm 400 hrs 1,935 mm 400 hrs 0 °C 50 mm 0 °C 50 mm

-5 °C 500 hrs -5 °C 500 hrs 1,577 hrs Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 mm 1,621 hrs Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 mm

13 Temperature

The mean annual temperature in Germany in the reference period 1961–1990 was 8.2 °C. While mean temperatures are somewhat lower in Germany‘s low mountain ranges and along the foothills of the Alps (between 6.9 °C and 8.0 °C), and are just 5.1 °C in the Alps themselves, mean temperatures are significantly higher in the Upper Rhine Lowlands (9.8 °C) and the West German Basin (9.6 °C) in particular.

Changes in air temperature since 1881 The mean annual air temperature for Germany rose by 1.4 °C between 1881 and 2016. The long-term mean for the standard reference period 1961–1990 of 8.2 °C rose to 8.9 °C in the current 30-year period 1981– 2010.

The observable trend of global atmospheric warming is overlaid by the natural variability of the climate system in which there are recurrent periods during which the increase in temperatures stagnates or by phases in which the temperature even falls temporarily. Falls in temperature are usually the result of periodic variations which are closely coupled with ocean circulation. These periodic variations overlap the influence of external climate forcing mechanisms which, in addition to naturally occurring phenomena such as insolation and volcanic activity, include human-induced changes in the concentration of greenhouse gases in the atmosphere as a result, for example, of air pollution and changes in land use.

The period from 1910–1950 and the time following 1980s in particular were marked by rising temperatures, while temperatures during the periods in between remained largely at the same level. Temperatures remained relatively constant at the end of the 19th century as well. ▴ Map of grid values (1 km x 1 km) showing mean annual temperatures in Germany in the standard reference period 1961–1990.

14 2014 10.5 °C 10.3 °C

10.0 Annual averages Mean for the 1961–1990 reference period Linear trend 9.5

9.0

8.5

8.0

7.5

7.0

6.5 1940 6.0 6.6 °C 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010

▴ It has become warmer in Germany: mean annual temperature (area average from station measurements taken at 2 m) from 1881–2016.

The biggest changes (1.5 °C) since 1881 have occurred temperatures just below freezing in the Alps, the in the West German Basin, the Uplands on the left Alpine foothills and in the low mountain range areas. bank of the Rhine and in the Upper Rhine Lowlands; The highest mean temperatures of approximately temperatures have risen least (1.0 °C) in the North 1.7 °C to 2.4 °C are measured in the Upper Rhine East German Lowlands. Lowlands and the West German Basin. It is also relatively mild in the North West German Lowlands 2014 was the warmest year in Germany since 1881. at 1.2 °C. Ten of the seventeen warmest years have been in the 21st centur y. In the summer months, in contrast, the regional differences in mean temperatures are, with the Seasonal variations exception of the Alps, less pronounced: at a mean The long-term mean air temperature for the winter temperature of 16.3 °C for Germany as a whole the months (December, January, February) is 0.3 °C highest mean temperatures for the months of June, for Germany as a whole, in other words just above July and August are 18.0 °C in the Upper Rhine Valley freezing point. Air temperatures are marked by high and 17.1 °C in the East German Basin; in contrast, spatial differentiation with negative temperatures or temperature in the low mountain ranges reach 16 °C.

15 The ten warmest years in Germany since 1881

2014 10.3 °C 2000 9.9 °C 2007 9.9 °C during the reference period 1961–1990. In the period 2015 9.9 °C 1981–2010, these values had risen to an average of eight to nine days, and as much as 13 days in the 1994 9.7 °C Upper Rhine Lowlands. There are still fewer than 2002 9.6 °C seven hot days per year along the Alpine foothills and 2011 9.6 °C three in the Alps themselves. Nonetheless, these are 1934 9.5 °C twice as many days in the Alpine foothills and three 1989 9.5 °C times as many in the Alps compared to the 1961–1990 1990 9.5 °C reference period.

The highest number of ice days would be expected in the Alps and in the Eastern Uplands. During the Change in indicator days reference period 1961–1990, average temperatures The number of hot days (daily maximum temperature here did not rise above freezing on 42 days per year ≥ 30 °C) on average for the whole of Germany has (56 days in the Alps). It is relatively mild in the Upper increased since the 1950s from around three days Rhine Lowlands and in the West German Basin with, per year to the current average of nine days per on average, fewer than 16 ice days per year. There year. The mean number of ice days (daily maximum may be as many as 20 to 26 ice days per year in temperature < 0 °C) has decreased during the same the North West and North East German Lowlands. period from 28 days to 19 days. Most heat situations Compared to the mean values for the 1961–1990 and are in the warmest regions of Germany (the Upper 1981–2010 periods, the least changes are in the Alps. Rhine Lowlands, in the West German Basin and the The biggest changes have been observed in the West East German Basins and Hills). On average, there German Basin, where the number of ice days has were between five and nine hot days per year in fallen by a quarter.

More hot days and fewer ice ▸ days in Germany. The graph 60 Days shows the annual values and Number of ice days per year Linear trend for ice days corresponding linear trend for 50 Number of hot days per year Linear trend for hot days Germany from 1951–2016. 40

30

20

10

0

10

0 1950 1960 1970 1980 1990 2000 2010

16 The future A further increase in temperatures in Germany is expected (very high agreement). For the short-term planning horizon (2021–2050), this increase is equal to around 1.0 °C to 1.3 °C (medium agreement). There is only a minor diff erence in the changes in the climate projections (climate mitigation scenario and business- as-usual scenario). The results vary between 0.7 °C and 2.1 °C. Warming is somewhat more pronounced in southern Germany.

Temperature developments for the long-term planning horizon are largely determined by the scenario selected. Based on the climate mitigation scenario, temperatures are expected to rise by 1.2 °C (medium agreement). Stabilisation at the level of the short-term planning horizon will be achieved by a very large reduction in greenhouse gas emissions in the scenario defi nition. Temperatures have risen by 2.5 °C since pre- ▴ Range of existing climate projections for mean annual industrial times. Regional diff erences are practically temperature in Germany. The diagram shows the available change non-existent. Under the conditions of the business-as- signals for the short-term (2021–2050) and long-term (2071–2100) usual scenario, temperatures will rise by around 3.7 °C planning horizon, in each case compared to the 1971–2000 baseline period. The results for the climate mitigation scenario (RCP2.6, (medium agreement). The results vary between 2.7 °C green) are juxtaposed in each case with the results for the business- and 5.3 °C. Warming is more pronounced in southern as-usual scenario (RCP8.5, blue). The diagram forms symbolise regions. the area between the smallest and biggest change signal within the scenario being considered. The width of the form signalises The results available for the business-as-usual the probability of occurrence (the wider the form, the higher the probability). The diagram also shows the mean (black dot) and scenario roughly correspond with the results of the percentile (25 %, 50 % and 75 %) as white lines. The results existing climate projections based on the SRES produced by each of the models are shown next to the forms as scenario A1B. black lines.

17 Seasonal mean temperatures and anticipated changes

1961–1990 1971–2000 2021–2050 2021–2050 2071–2100 2071–2100 (RCP2.6) (RCP8.5) (RCP2.6) (RCP8.5) Spring 7.7 °C 8.1 °C +0.9 °C +1.1 °C +1.0 °C +2.9 °C Summer 16.3 °C 16.6 °C +1.1 °C +1.3 °C +1.0 °C +3.5 °C Autumn 8.8 °C 8.7 °C +1.2 °C +1.5 °C +1.2 °C +3.9 °C Winter 0.3 °C 0.8 °C +1.1 °C +1.6 °C +1.2 °C +4.0 °C Year 8.2 °C 8.6 °C +1.0 °C +1.3 °C +1.2 °C +3.7 °C

Regional diff erences In all seasons, warming is more pronounced in the Alps In the Alps in particular, the projected rates of and Alpine foothills than in Germany as a whole. warming, both for the climate mitigation scenario and Warming is especially noticeable for the long-term for the business-as-usual scenario, are still higher planning horizon in winter with a mean temperature of than the changes projected for Germany as a whole. 4.5 °C (business-as-usual scenario, medium agreement) Here the change for the short-term planning horizon compared to projected rates of warming for Germany is between +1.3 °C (climate mitigation scenario) and as a whole of a mean 4 °C (business-as-usual scenario). +1.5 °C (business-as-usual scenario) compared to the 1971–2000 baseline period. Mean rates of warming of between 1.3 °C (climate mitigation scenario) and 4.4 °C (business-as-usual scenario, medium agreement) are projected for the long-term planning IN BRIEF horizon. Observed facts In the coastal region of Germany‘s North West and • Warming trend in Germany continues North East Lowlands, the projected changes for the long-term planning horizon are slightly below the • Mean annual temperatures have risen by 1.4 °C mean values. Rates of warming of between 1.2 °C in 136 years (climate mitigation scenario, medium agreement) and • Changing extremes: more hot days, fewer ice 3.4 °C (business-as-usual scenario, medium days agreement) are projected here. Short-term planning horizon Seasonal variations • Mean temperatures in Germany as a whole will Similar warming can be observed in each of the rise by an average of 1.0 °C to 1.3 °C seasons, with the exception of spring where warming Long-term planning horizon is less pronounced. Increasing temperatures are accompanied by a marked increase in temperature • In the climate mitigation scenario, warming will extremes. Extremes associated with low temperatures stabilise at 1.2 °C are becoming less common while extremes associated • In the business-as-usual scenario, mean with higher temperatures are increasing (very high temperatures in Germany as a whole will rise by agreement). As a result, heatwaves will occur with an average of 3.7 °C greater frequency. Greater warming in the Alps and the Alpine foothills.

18 Changing global temperatures

2016 was the warmest 2016 year since records began. 1.0 °C 0.89°C

The year was thus in line 0.8 with the long-term global Annual deviation 5-year moving average 0.6 warming trend. 16 of the 17 warmest years since 0.4 records began have been 0.2 measured since 2001. 0.0 The global mean -0.2 temperature for the year 2016 was 1.1 °C above -0.4 pre-industrial mean -0.6 1909 temperatures. Warming -0.8 –0.57°C observed in recent 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 decades is greater over ▴ Global mean annual temperature from 1880 to 2016. The graph shows the diff erence from the landmasses than over the reference period 1961–1990. (Source: NASA‘s Goddard Institute for Space Studies (GISS)) oceans.

The results of climate projections show a continuous rise The COP21 goals (see page 3) can only be seen in context in the global mean temperature. In the climate mitigation by aggregating previous observations of global warming scenario, the global climate will be 1 °C warmer at the with the warming projected on the basis of the climate end of the 21st century than in the period 1986–2005. The scenarios. The fi rst term is the warming (0.6 °C) between business-as-usual scenario would result in mean warming the periods 1850–1900 and 1986–2005. Warming in line of 3.7 °C. The amount of warming will diff er from region with the climate mitigation scenario (0.6 °C + 1.0 °C, total to region. The highest rates of warming will occur on the 1.6 °C) would enable compliance with the COP21 2 °C target. continents and on both polar caps. If temperatures rise as outlined in the business-as-usual scenario (0.6 °C + 3.7 °C, total 4.3 °C), the COP21 targets would be missed by a wide margin.

▴ Mean temperature change for the 2081–2100 period based on the climate mitigation scenario (RCP2.6, left) and the business- as-usual scenario (RCP8.5, right) relative to the 1986–2005 period. Hatching indicates regions where changes in temperature are less than natural climate variability. Stippling indicates regions where changes in temperature exceed natural climate variability. (Source: 5th IPCC Assessment Report 2013, Working Group I, fi gure SPM.8)

19 Precipitation

Precipitation in Germany at the point of transition between the maritime Atlantic climate and the continentally influenced mild climate of central Europe is determined by decreasing humidity with increasing distance from the North Sea and a propensity for precipitation that increases with height above sea level. An increase in mean precipitation has been observed over the last hundred years and total precipitation amounts are expected to continue to rise in the future.

Precipitation in Germany On average, 789 mm of precipitation falls each year in Germany. This corresponds to a total of 789 litres per square metre. Many parts of north eastern and central Germany receive mean annual precipitation of less than 600 mm; in the higher regions of the Alps and the Black Forest, over 1,500 mm of precipitation is normal. Precipitation in lowland areas is highest near to the North Sea. Towards the south-east, the amount of precipitation falls as the continental influence increases. In low mountain ranges, average precipitation rises with increasing height above sea level. The orientation of mountain ranges and other orographic features also have a modifying impact on precipitation patterns. On average (reference period 1961–1990), annual precipitation is lowest in the North East German Lowlands and in the East German Basins and Hills at 577 mm and highest in the Alps at 1,935 mm.

In some years and in particular areas, there may be considerably less or substantially more precipitation. Since records began in Germany, for example, the lowest precipitation of 209 mm fell in the year 1911 (Aseleben, Saxony-Anhalt) and the highest of 3,503 mm in 1970 (Balderschwang, Bavaria).

▴ Map of grid values (1 km x 1 km) showing mean annual precipitation in Germany in the reference period 1961–1990.

20 1,200 mm Annual totals 2002 Mean for the 1961–1990 reference period 1,018 mm 1,000 linear trend

800

600

1959 400 551 mm

200

0 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010

▴ It has become wetter in Germany: annual precipitation (area average from station measurements) from 1881 to 2016.

Changes in annual precipitation since 1881 Seasonal variations The magnitude of precipitation varies greatly in time On average in Germany, precipitation was somewhat and space. Since 1881 the wettest year on average in higher in the hydrological summer half years (May the whole of Germany was the year 2002, in which to October) in all the years of observation. Around 1,018 mm of precipitation were measured; the driest 57 % of the annual precipitation falls in the summer year was the year 1959 (551 mm). With at times very half year, and around 43 % in the winter half year. strong variations from year to year or from decade to In some regions the precipitation in the summer decade, annual precipitation in Germany as a whole half-year is even greater, for example in the Alpine increased in the 136 years from 1881 by 74 mm or 9 % Foothills (63 %). In other regions the dominance of (relative to the 1961–1990 reference period). This the summer half-year is much less pronounced, for increase was uneven: from the 1880s through to the example in the Uplands on the right bank of the Rhine 1920s, there was initially a strong increase in (51 %). Looking at just the three summer months of precipitation; subsequently up to the present, June, July and August, the driest summer was in the precipitation has increased at a slower rate. This latter year 1911 (124 mm), and the wettest summer in the increase was accompanied by short term variations, year 1882 (358 mm). The normal total is 239 mm such that it was somewhat drier in the 1940s and the (average 1961–1990). From 1881 up to the present, 1970s, while the 1960s, 1980s and the phase around summer precipitation has hardly changed at all. During the millennium were comparatively wet. the same period, precipitation has increased in the transitional seasons of spring and autumn, with this trend being significantly more pronounced in spring than in autumn.

The most striking development has been in the three winter months (December, January, February), however. Winter precipitation has increased since the winter of 1881/82 up to the present by 48 mm or 26 % relative to the reference period 1961–1990. This means that the increase in average annual precipitation can largely be explained by the increase in precipitation during the winter. Despite this overall trend for the winter, there are nonetheless significant differences in this season from year to year. The winter with the least

21 precipitation (69 mm) in the year 1890/91 contrasts Germany (more than 27 days). There has been no with the winter with the most precipitation (304 mm) in observable change since the 1950s in the number the year 1947/48. The normal total is 181 mm (average of days with more than 20 mm of precipitation per 1961–1990). day. The variability in the annual number of heavy precipitation events is very high and, at a total of The Alps are the wettest region in every season. In 5 days per year on average, the frequency is relatively the driest regions, in contrast, there are differences rare in Germany as a whole. Regional differences, between the seasons: precipitation is lowest in the on the other hand, are very large. There are three or North East German Lowlands in spring and summer, fewer events each year in north-east Germany and with 132 mm and 182 mm respectively (average on the coasts; in southern Germany and in all the 1961–1990), while the East German Basins and Hills mountain regions, these events occur on more than region is drier, with 128 mm in autumn and 123 mm 7 days per year. in winter. There were no significant changes in the amount of precipitation in either region between In addition to discussing changes in the frequency of 1881 and 2016. The strongest annual trend relative to extreme precipitation the question to which extent 1961–1990 is in the North West German Lowlands at warming is accompanied by drying out is of great +14 %, where the greatest increase, as is the case for importance, especially in the summer. Extreme Germany as a whole, is for the winter at +33 %. drying-out can have massive economic impacts, e.g. on inland waterway shipping and agriculture. Change in indicator days Dry periods are determined based on the frequency As regards special precipitation events, there are of episodes in which no precipitation falls for at least two contrary aspects to consider: too much and too 10 days in a row. However, as is the case with extreme little. Looking at the days on which at least 10 mm of precipitation, these events are also rare (on average precipitation falls, the average is 21 days throughout 1.3 cases per summer in Germany) and there are the whole of Germany, irrespective of the large inter- therefore no reliable statistics available which clearly annual variations. This figure has hardly changed demonstrate observable changes since the 1950s. in the last 66 years. There is, however, a clear and There are also pronounced natural variations with observable north-south and west-east gradient of alternating phases of greater or less dryness similar frequency, with the fewest events in the north-east in form to the varying phases of extreme precipitation (average below 13 days) and most events in southern events.

35 Days

Number of days with ≧ 10 mm precipitation per year linearer Trend 30

25

20

15

10

5

Mean number of days with 10 mm and ▶ 0 more precipitation (area average from 1950 1960 1970 1980 1990 2000 2010 station measurements) from 1951 to 2016.

22 The future A marked change in the mean annual precipitation totals over the short-term planning horizon (2021–2050) is not expected in Germany (very high agreement). An increase in mean annual precipitation of 5 % is calculated (medium agreement). There is little difference between the scenarios. The results vary between −2 % and +14 % and are almost identical throughout all the territory of the Federal Republic of Germany. One fundamental issue of note is that a modelled change of less than 10 % cannot be distinguished from natural climate variability. This threshold applies to all following values. The attributes medium to very high agreement above and in the text below relate to the scientific plausibility and uniform tendency of the model results.

Regional differences In the long-term planning horizon (2071–2100) annual ▴ Range of existing climate projections for total annual precipitation precipitation in Germany is projected to increase by in Germany. The diagram shows the available change signals for +9 % (medium agreement). The degree of change will the short-term (2021–2050) and long-term (2071–2100) planning be almost identical throughout all the territory of the horizon, in each case compared to the 1971–2000 reference period. The results for the climate mitigation scenario (RCP2.6, green) are Federal Republic of Germany. juxtaposed in each case with the results for the business-as-usual scenario (RCP8.5, blue). The diagram forms symbolise the area The number of days with precipitation of at least between the smallest and biggest change signal within the scenario 10 mm per day is expected to increase in all regions, being considered. The width of the form signalises the probability over both the short-term and long-term planning of occurrence (the wider the form, the higher the probability). The diagram also shows the mean (black dot) and the percentile horizons. Only in the Alpine region do some models (25 %, 50 % and 75 %) as white lines. The results produced by each project fewer such days. A less pronounced increase of the models are shown next to the forms as black lines. is projected for days with precipitation of 20 mm and more. However, the range for heavy and extreme precipitation within the ensemble is very large in parts and the results are therefore not particularly robust. The regional differences in changes in mean annual precipitation totals are not very pronounced.

23 Seasonal mean precipitation values and expected changes

1961–1990 1971–2000 2021–2050 2021–2050 2071–2100 2071–2100 (RCP2.6) (RCP8.5) (RCP2.6) (RCP8.5) Spring 186 mm 179 mm +5 % +8 % +3 % +13 % Summer 239 mm 234 mm −2 % ±0 % ±0 % −9 % Autumn 183 mm 191 mm +3 % +4 % ±0 % +7 % Winter 181 mm 183 mm +7 % +7 % +4 % +17 % Year 789 mm 788 mm +3 % +5 % +2 % +9 %

Seasonal variations Based on all the RCP-scenarios, precipitation in winter is projected to increase by +7 % over the short term 2021–2050 planning horizon (medium agreement). It is not possible to predict the direction of change for the summer. The range of results extends from small increases through to a slight fall. In the transitional seasons increases in mean total precipitation of +3 % (autumn) or +8 % (spring) are shown for this planning horizon (medium agreement). IN BRIEF In spring and autumn the change for the long-term Observed facts planning horizon (2071–2100) may be between +1 % and +13 % (medium agreement); in contrast, • 9 % increase in annual precipitation in 136 years the change in winter may be up to +17 % (medium • Increase in precipitation in spring, autumn and agreement). For the summer the range goes from winter, but not in summer no change in the climate mitigation scenario (±0 %) • Indications of an earlier start and later end to the to more pronounced decreases in the business-as- season with convective precipitation and more usual scenario (−9 %) in this planning horizon. pronounced intense rainfall events The business-as-usual scenario ranges between an increase of +30 % (very low agreement) and a Short-term planning horizon decrease of −50 % (very low agreement). In each of • No clear change in mean total annual the regions the range of results for the summer is also precipitation (+5 %) very wide and the corresponding results therefore do not appear very robust. Long-term planning horizon • Annual precipitation is expected to increase in The results available for the business-as-usual Germany by +9 % scenario diff er from the results of earlier climate projections based on the SRES scenario A1B. The For both planning horizons, increases in business-as-usual scenario no longer shows the large precipitation are simulated for the winter months. reductions in summer precipitation over the long- The long-term planning horizon projects a range from no change to slight decreases in precipitation term planning horizon which are shown in the SRES for the summer. scenario A1B.

24 Changes in global precipitation

Global precipitation varies greatly, both spatially and temporally, due to the many relevant natural variations resulting, for example, from typical circulation patterns, such as the El Niño Southern Oscillation (ENSO) and the North Atlantic Oscillation (NAO).

Precipitation over Europe has increased by 6–8 % over the last century. This increase manifests itself in quite diff erent ways in northern and southern Europe: in northern Europe precipitation has increased by 10–40 %; precipitation in the Mediterranean area and in parts of south-east Europe, in contrast, has decreased by up to 20 %. As in Germany, regional trends are also apparent. Precipitation totals have increased in northern and western Europe in particular and most strikingly in the winter months (20–40 %). In contrast, there has been a decrease in precipitation in southern Europe in all seasons. Signifi cant reductions in precipitation are also apparent in central Europe in summer. ▴ Change in annual precipitation between two 30-year periods 1981–2010 and 1951–1980. (Data source: Global The results of the climate projections show that, for the most Precipitation Climatology Centre (GPCC) operated by part, changes in precipitation follow a pattern whereby drier DWD under the auspices of the World Meteorological regions are becoming even drier and wetter regions even Organization (WMO)) wetter. The changes which are expected in connection with the climate mitigation scenario (RCP2.6) will, however, be very moderate towards the end of the 21st century compared between south-west Europe through the Balkans to central to the period 1986–2005, particularly over Europe. The Asia by an annual average of 20–40 % and by as much as signals in the business-as-usual-scenario (RCP8.5) are 50–75 % in the summer. Annual average increases of 10–30 % becoming much clearer, although here too there is a clear are limited to Scandinavia. In the summer, in contrast, Europe dichotomy in Europe: precipitation is projected to decrease as a whole will probably receive less precipitation.

▴ Mean change in precipitation for the 2081–2100 period based on the climate mitigation scenario (RCP 2.6, left) and the business- as-usual scenario (RCP 8.5, right) relative to the 1986–2005 period. Hatching indicates regions where changes in precipitation are smaller than natural climate variability. Stippling indicates regions where changes in precipitation exceed natural climate variability. (Source: 5th IPCC Assessment Report 2013, Working Group I, fi gure SPM.8)

25 Sunshine

On average, the sun shines in Germany 254 minutes per day or 1,554 hours a year. The sun shines most in southern Germany and in north-east Germany for up to 280 minutes per day; the sun may even shine for over 300 minutes per day on the Baltic coast. Daily sunshine duration is lowest in the Central Uplands and Harz region, where the measured long-term mean is just 230 minutes per day.

Past and present Development of sunshine since 1951 The presentation of annual totals up to 2016 shows In the period 1951–2016, mean daily sunshine duration three rough periods of different sunshine durations: increased by nine minutes. This is largely due to more a phase of higher annual values from 1951 to 1976, sunshine during the spring and summer months (in followed by an accumulation of lower annual totals each case +16 minutes per day). There has been a through to roughly the end of the 1980s with values slight decrease in observed sunshine duration in the rising again after this date. Between around 1950 and autumn months (6 minutes per day). The highest rate 1980, there was a worldwide phase of decreasing solar of increase at +23 minutes has been observed in the radiation which was in part ascribed to increasing air West German Basin, while there was no change in the pollution. Successful air pollution control measures Eastern Uplands in the period 1951–2016. subsequently resulted in an increase in surface solar radiation.

2,400 Hrs Annual values Mean for the 1961–1990 reference period 2003 Linear trend 2,014 Std. 2,000

1,600

1,200 1977 1,362 Std. 800

400 Annual sunshine duration ▸ (area average from station 0 measurements) in Germany 1950 1960 1970 1980 1990 2000 2010 from 1951–2016.

26 The future Seasonal mean daily sunshine duration and anticipated changes Sunshine duration is not calculated directly in the climate models, but is indirectly derived from short- wave radiation. Radiation is associated with cloud cover conditions, one of the major challenges facing climate modelling. The range of 1961–1990 1971–2000 2021–2050 2021–2050 2071–2100 2071–2100 (RCP2.6) (RCP8.5) (RCP2.6) (RCP8.5) modelled values therefore varies Spring 304 min. 312 min. +12 min. −12 min. −12 min. −24 min. greatly between the models. The results are consequently Summer 400 min. 403 min. −6 min. ±0 min. −12 min. 6 min. much less informative than, Autumn 205 min. 199 min. ±0 min. ±0 min. ±0 min. ±0 min. for example, the results for Winter 102 min. 105 min. −12 min. −12 min. −12 min. −24 min. temperature changes. Year 254 min. 256 min. −6 min. −6 min. −6 min. −12 min.

Over the short-term planning horizon 2021–2050 daily sunshine duration is projected to fall by 6 minutes (medium agreement). This fall is particularly noticeable in the business-as-usual scenario in the winter and spring (high agreement), with no changes expected to IN BRIEF occur in the summer and autumn (high agreement). These changes are projected uniformly throughout Observed facts Germany. • Average of 254 minutes sunshine per day • Small increase in sunshine duration in Germany The changes are expected to increase over the since 1951 long-term 2071–2100 planning horizon. Mean daily sunshine duration is expected to drop by between 6 • A lot of sun in the south and extreme north-east, more frequently overcast in the centre minutes and 12 minutes (medium agreement). This fall will be particularly noticeable in spring and winter Short-term planning horizon with up to 24 minutes less sunshine per day (very low • Minimum reduction in sunshine duration agreement). There will probably be no changes in throughout Germany sunshine duration in autumn. In the business-as-usual scenario some projections forecast increases in mean Long-term planning horizon daily sunshine duration of up to one hour in summer • Strengthening of these tendencies (very low agreement). Pronounced reduction over both planning horizons in winter and spring (only RCP8.5).

27 Sea Level

Written by the Federal Maritime and Hydrographic Agency

In the period since records began, the mean sea level has risen by around 2 to 4 mm per year along the entire North Sea coast. Climate models predict a further rise in the future. New studies of ocean warming and of the Antarctic and Greenland ice sheets suggest that the rate at which temperatures are rising will probably accelerate.

Sea level ‒ a volatile measure account vertical land movements and are referred to Mean sea level and its future changes are of great as relative changes in sea level. importance for the long-term planning of coastal Neither global nor regional climate models are defence works. Several components contribute to currently capable of adequately simulating freshwater changes in sea level: fluxes from glacier and ice sheet meltwater and a) steric changes relating to changing temperature corresponding values must currently be estimated and salinity and added to steric and dynamic values. b) dynamic changes based on changing ocean currents Observed changes in sea level c) enhanced freshwater fluxes to global oceans North Sea: Sea level records for the German Bight go resulting from increased glacial melting back to 1843 (Cuxhaven), although most date from d) enhanced freshwater fluxes due to increased the 1930s. There are large differences in the rate of melting of the Greenland and/or Antarctic ice relative sea level rise of between 1.7 mm/year and sheets 4.1 mm/year depending on geographical location. All e) vertical land movements sea level records contain a large amount of decadal variability. In some decades, the sea level has risen by The contributions made by components a) to d) over 4 mm/year while in others it has fallen slightly. produce absolute changes in sea level while actual Around 0.5–1.5 mm/year of this rise must be deducted changes measured at tide gauges also take into again, however, to account for subsidence on the

530 cm

520 Jahresmittelwerte 510 linearer Trend

500

490

480 Mittlerer Meeresspiegel am ▸ 470 Pegel Cuxhaven 1843–2015. (Quelle: Universität Siegen, 460 Bundesamt für Seeschifffahrt 1840 1860 1880 1900 1920 1940 1960 1980 2000 und Hydrographie)

28 German North Sea coast. Similar absolute sea level and 98 cm. However, the report also pointed out that rises have been observed on the English and Scottish insufficient attention had been given to the potential east coasts, on the Dutch coast and along the north- contribution from the Antarctic and Greenland ice eastern Atlantic in general (by 1.7 mm/year). Tide- sheets as their physical processes are unknown or driven changes in sea level in the North Sea do not have not yet been described mathematically. change in parallel to the mean rise in sea level. Mean high water levels measured at the Cuxhaven tide Much more is now known about these factors. gauge since 1950 have been rising more strongly and Oceanographic observations and bathymetric mean low water levels more weakly than the mean measurements at the edges of the ice sheets are also water level. This may be due to waterway construction increasingly showing that warmer ocean water is measures along the Elbe and altered morphological destroying the ice shelves and that more and more conditions around the Elbe-Weser triangle. melting is taking place at the point of contact between glaciers and the underlying bedrock. As a result, Baltic Sea: The absolute sea level on the Baltic Sea glaciers are draining into the sea much faster. The coast has risen by around 1.4–2.0 mm/year. Apart consequence of this is that sea level will rise at a much from the south-west Baltic Sea, relative sea level is faster rate than the values for 2013. This faster rise falling in all other coastal regions as a result of the can already be observed. At present, the figures for ongoing post-glacial rebound. the continued rise in sea level on the German coasts in the business-as-usual scenario are estimated to rise Future changes in sea level by well over one metre by the end of the 21st centur y. The Fifth Assessment Report by the However, this does not take account of the emerging Intergovernmental Panel on Climate Change (IPCC) possibility that both ice sheets could collapse. in 2013 outlined sea level rises in relation to different greenhouse gas scenarios through to the end of the Given the high capacity of the oceans to store heat 21st century. In the climate mitigation scenario sea sea level will continue to rise well after the end of levels would rise by between 26 cm and 55 cm and the 21st century regardless of the continuing pace of in the business-as-usual scenario by between 52 cm global warming.

29 Phenology

Weather conditions and climate both have an infl uence on plant growth and development. The science which studies these phenomena is known as phenology (Greek: “the study of appearances”). Data on phenological observations are among the most valuable indicators of changes in environmental conditions and have been collected around the world for centuries.

Phenological development in Germany The early summer begins with the fl owering of black The phenological year begins in pre-spring. In the elder (sambucus nigra). In relation to the reference reference period 1961–1990, this phenological season period, this occurs on 7 June and lasts three weeks. began on average in Germany on 3 March. The In the last 26 years, the actual duration of fl owering season begins with the fl owering of the common hazel has hardly changed; however, the umbels of the (corylus avellana). Early spring begins on average elderberry fl ower as early as 27 May. The fi rst 33 days later on 5 April with the fl owering of forsythia blossoming large-leaved lime trees (tilia platyphyllos) (forsythia × intermedia); full spring begins another signal the transition to high summer. On average 32 days later on 7 May, when apple trees (malus) over the 30-year period from 1961 to 1990, high come into blossom, and lasts on average for 31 days. summer began on 28 June. In recent years, the onset Comparison of the international standard reference of high summer has moved forward by 10 days. High period 1961–1990 with the following 1991–2016 period summer is the longest of the phenological seasons in shows that pre-spring now begins on 16 February the vegetation period and has risen over Germany and lasts six days longer than in the reference period. as a whole from 42 days to 45 days. Late summer Early spring (27 March) and full spring (27 April) also has arrived when the fi rst early ripening apples can begin earlier. be picked from the trees. In the reference period

The various plant development ▸ phases are assigned to phenological seasons. The “phenological clock” shows the seasons and their WINTER 1961–1990 English oak (leaf fall) phenological indicator phases Duration in days: 121 (mean for Germany). Comparison of the periods 1961–1990 and LATE AUTUMN 1991–2016 English oak Duration in days: 104 1991–2016 reveals the shift in (leaf discolouration) 18 phenological seasons. Dec Jan PRE-SPRING 18 Nov Feb 39 33 Hazel (flowering) HIGH AUTUMN 19 English oak (fruits) 27 Oct Mar Sep Apr 31 Aug May EARLY AUTUMN 22 26 Black elder JunJul 32 30 EARLY SPRING (fruits) 23 Forsythia (flowering) 22 26 45 31

42 LATE SUMMER 21 Apple, early ripening FULL SPRING (fruits) Apple (flowering)

HIGH SUMMER Large-leaved lime (flowering) EARLY SUMMER Black elder (flowering)

30 1961–1990, this was on 9 August. More recently, the “Yellow Transparent”, “James Grieve” or “Retina” varieties of apple are ripe and can be enjoyed a week earlier. Over the last 26 years, the late summer has become about 3 days shorter.

While during the reference period early autumn began with the fi rst ripe fruits of the black elder on 4 September, its onset is also now much earlier, being observed around 25 August. Recently, the duration of early autumn has increased by 4 days. In the reference period, high autumn followed with the fi rst ripe fruits from the English oak (quercus robur) on 26 September. It now occurs around six days earlier and, on average, lasts eight days longer. This means ▴ Apple, indicator plant of full spring: beginning of fl owering in 2016. that there is only a shift of two days in the date at which the leaves of the English oak begin to discolour, which is the fi rst sign of the beginning oflate autumn. and longitude as well as height above sea level. The In the period 1961–1990, this took place on average on earliest phenological dates are usually recorded in 15 October. Now it occurs on 17 October. The duration the Upper Rhine Lowlands and the latest either in the hasn’t changed. upland regions or far up in the north on the coast of the Baltic Sea. On average, spring moves from south The phenological winter begins when the English to north and from west to east. Spring‘s progress oak begins dropping its leaves. During the reference slows if it has to cross a mountain range. The growth period, this occurred on 2 November. There has been period begins coming to an end soonest in those no striking change since. Between 1991 and 2016, on places where vegetation started latest. average, the winter only began one day later. While it took around 121 days for winter to pass and the new These changes are most strongly aff ected by changes phenological year to begin in the reference period in air temperature, which are most pronounced in 1961–1990, the period between the leaf fall from the winter and spring. Changes in solar radiation have English oak and the renewed fl owering of the hazel most eff ect in spring. has been 17 days shorter on average during the last 26 years. The future Phenological models can be used in conjunction with Regional diff erences climate projections to produce predictions of further The dates and fi gures in the previous section are changes in phenological development. Studies show, mean values for Germany as a whole. As phenological for example, that phenological events may be expected phases depend directly on the temperature, however, to occur ever earlier towards the end of the present there are some signifi cant regional diff erences. Onset century especially in spring. In full spring, apple is times are strongly infl uenced by geographical latitude projected to fl ower around another 15 days earlier.

31 Extreme events

Everyone remembers one extreme event – a disastrous storm, extreme heat or catastrophic flooding. Extreme events frequently cause human suffering and great destruction. How has the frequency of extreme events changed in the past and what changes are to be expected in the future?

Extreme = rare Temperature Extreme events are by their nature rare events. The mean temperature has risen substantially in They are characterised by highly unusual weather recent decades. This means that there have been conditions. Extreme events occurred in the past, more days and periods of very high temperatures. and they will occur again in the future. Well-known A good example is the summer of 2015 during which examples in the distant past include St. Mary temperature records for Germany were broken twice: Magdalene‘s flood in the year 1342, which affected 40.3 °C in Kitzingen. numerous rivers in central Europe, or the “Year Without a Summer” following the eruption of Mount In order to explain extreme events such as the hot Tambora in 1816. summer of 2015 correctly from a climatological perspective it is helpful if the time series used to We have also observed extreme events in the recent capture them statistically are as long as possible. past. Examples include the floods of 2002 and 2013, For this purpose, climatological parameters are used both caused by very heavy rains, the heatwave of which are capable of describing the duration, intensity August 2003 or the Lothar (1999) and Kyrill (2007) and frequency of extreme events. An analysis has storms. been performed for five German cities, which shows a frequency of at least one 14-day period of hot weather It is therefore perfectly reasonable to ask what else (heatwave) per year with a mean daily maximum air will climate change bring with it? As extremes are by temperature of 30 °C or above for the period 1950–2016. definition very rare events, statistical analyses are The mean maximum temperature over such a period is less robust. The recurrence interval is often once used as a measure of the heatwave‘s intensity. every hundred years (once-in-a-century event). The available time series are not much longer than this. The analysis reveals a north-south gradient in the It is thus no easy matter to capture an event on this frequency and intensity of heatwaves; in northern scale statistically. Germany (Hamburg), there was no single longer

32 ◂ Mean values of the warmest °C 14-day periods per year based 37 Hamburg on daily maximum temperatures. 1994 A bar is drawn every time the 31.8 °C value reaches at least 30 °C. The 28 height of each bar represents the 37 Dresden 1994 calculated 14-day mean value. 32.7 °C 28 37 Frankfurt/Main 2003 35.8 °C

28 37 Mannheim 2003 36.5 °C

28 37 Munich 2003 33.1 °C 28 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015

period of hot weather which complied with the above Precipitation definition in the year 2015. In general, the highest Several meteorological causes must converge for mean daily maxima during the heatwaves in the cities precipitation to fall in extreme quantities, such further to the north are below 33 °C; temperatures as 292 mm of precipitation within 7 hours on in major cities further south are often higher. Munich 29 July 2014 in Münster and 353 mm of precipitation experiences fewer heat events than typical for the within 24 hours on 12/13 August 2002 in Zinnwald. south as the station is located at a relatively high Strong local lifting mechanisms are needed to force altitude (515 m). moisture out of the air. The maximum possible volume of water in the air is directly linked to the air It is also apparent that such extreme heatwaves have temperature. The warmer the air mass is, the more been occurring more frequently since the 1990s; for water it can absorb. This means around 7 % more example, such events did not occur at all in Hamburg in water for each 1° increase in temperature (Clausius- the period 1950–1993, whereas since 1994 there have Clapeyron relation). This relationship is the reason been four extreme heat waves. why the volume of extreme precipitation in a defined time period is much higher in the summer than in the Given current and projected ongoing warming, it is winter. The large-scale weather situation must ensure very probable that such high temperatures, and higher continuous incoming flows of more warm and moist extremes, will occur more frequently. They will often air masses. be associated with lengthy spells of hot weather. There are clear indications for this in the results of regional Extreme, thundery short-duration and high intensity climate projections. However, as yet there are no precipitation is usually geographically limited. reliable estimates of the maximum temperatures which Areas of widespread extreme precipitation mainly may be reached in the future. arise when low pressure complexes (e.g. such as the

It almost seems as if cold winters may be a thing of the past. However, the cold winters of 2009/10, 2010/11 Extreme precipitation quantities in Germany and 2012/13 suggest otherwise. These recent winters Quantity Location Duration Date were influenced by long and very cold periods in some 126 mm Füssen (Ostallgäu) 8 min. 25/05/1920 regions due to continuous incoming flows of Arctic air. 245 mm Münster (LANUV) 2 hrs. 28/07/2014 Given that the Barents Sea is now increasingly ice-free, research is currently underway on how the probability 312 mm Zinnwald-Georgenfeld 1 day 12/08/2002 of such weather situations will develop in the future. 779 mm Aschau-Stein 1 month July 1954 In general, however, global warming is reducing the ▴ Observed extreme precipitation quantities in Germany. The return intensity of such weather conditions. period for this events is once in 100 years.

33 general „central European low pressure system“ weather system) weaken only very gradually, when slow-moving weather fronts or quasi-stationary air mass boundaries reside over the area or when low pressure areas move in a north-easterly ‘Vb’ direction from northern Italy with precipitation areas which subsequently move westwards to impact eastern Germany.

Another important aspect is the sequence of events. It is possible, for example, for a stable circulation pattern ‒ as was the case in May/June 2016 ‒ to last for a period of several days before turning into a regional trigger of extreme precipitation. This can result in a series of several extreme precipitation events following one after the other. However, there are numerous unanswered questions in connection with heavy, locally limited short-term For the period 1951–2010, the DWD has developed precipitation events. This is why the Deutscher special maps showing spatial distribution of heavy Wetterdienst recently analysed high-resolution precipitation amounts across Germany on a grid precipitation data in more detail. of 5405 boxes (each covering 67 km2). These maps allow the derivation of site-specific statements on The probable maximum precipitation (PMP) heavy precipitation. It is possible to calculate heavy represents the theoretically greatest depth of precipitation amounts as a function of the return precipitation that is physically possible for an area period (from T = 1 a to T = 100 a) for any period of during a given period at a certain time of year under duration (selected time segment with precipitation, defined climate conditions. The DWD estimates interruptions included, irrespective of the start and the PMP using a physically based evaluation of end of the observed precipitation event) between meteorological data. The PMP is then expressed in D = 5 minutes and D = 72 hours. The return period is the form of maximised area precipitation. defined as the average time period (in years) in which the precipitation total reaches or exceeds a value Maximised area precipitation in Germany depends on once. the relevant duration period, location and size of the area under study. The maximised area precipitation in Germany for a duration D = 24 hours in most areas of Return periods of precipitation amounts 25 km2 is 400 mm. There is some evidence that, if the climate changes significantly in Germany, the current D 5 15 30 1 3 6 24 72 min. min. min. hrs. hrs. hrs. hrs. hrs. maximised area precipitation could exceed current estimates in the future. T = 1 a 8 15 19 24 40 60 120 180 T = 10 a 18 32 40 45 80 110 200 320 Possible long-term climate trends in observed T = 100 a 30 45 60 80 100 140 280 450 extreme precipitation can only be studied on the basis ▴ Possible heavy precipitation amounts (in mm) per period of of precipitation measurements made over several duration (D) and return period (T) 1, 10 and 100 years in Germany. decades. Only then is it possible to differentiate between short-term and long-term variations and real long-term trends. The fact that the equipment available for measuring precipitation is not always able to capture small-scale extreme precipitation events makes it more difficult to analyse trends. It is also necessary to distinguish between the frequency and intensity of precipitation.

34 Analyses of daily precipitation in the winter show As described above, there is a direct connection that the number of days with large amounts of between air temperature and probable maximum precipitation increased by around 25% in the period precipitation. The warming observed to date, as well 1951‒2006. A large amount of rainfall is defined as as the further warming projected for the future on the an event which occurs once every 100 days during basis of climate modelling, hold the potential for higher a reference period . These increases occur in all the volumes of precipitation. regions of Germany. This trend has increased slightly in the spring and autumn seasons.

With the available observational data and known methods it is not possible at present to identify trends in the number of days with large amounts of precipitation in the summer. Short and long-term cyclical variability is dominant. Fundamentally, however, the amount of extreme precipitation is significantly lower in the winter than in the summer.

The data used to analyse precipitation over periods of less than 24 hours is much poorer. Analyses of the large-area radar data available for the last 15 years point to a regional increase in short-lived bursts of extreme precipitation. Owing to the short length of the time series used, these findings are not significant from a climatological perspective and may be the result of short- and medium-range variations. The station data obtained from a time series of over 50 years produce specific trend patterns which vary geographically and for different durations. The relative changes for most regions in Germany do not, however, exceed 5 %.

Water management users are currently enhancing their recommendations in order to deal with urban torrential flooding. Nonetheless, more work must be done to close Climate change can also affect the precipitation the diagnosed research and development gap between volumes in Germany in other ways. These could be short-timescale risk management solutions, such changes in the large-scale circulation patterns and as urban drainage systems (15 minutes as the most the tendency for the resulting weather conditions relevant period of duration), and procedures for river to remain unchanged. Evaluations of observations flooding that are based on longer timescales (12 hours and climate model simulations show an increase in and longer). global precipitation amounts of ~2 % per 1 degree of increase in temperature. This value is lower than With rising air temperatures climate change is the ~7 % increase described for water volumes. increasing the potential for extreme precipitation A comparable increase in precipitation amounts events. This process is strengthened even more by the would be based on consistent relative air humidity. exponential, rather than linear, relationship between However, observations and model calculations for the temperature and water content. The current generation past show that there has been a slight fall in relative of regional climate models shows that there is a trend humidity in Germany. Other factors influencing towards more precipitation extremes. At the same precipitation are the changed concentrations of time, the resolution of these models is not fine enough greenhouse gases and aerosols. These are currently to provide detailed local data for these processes. the object of research.

35 Analyses of maximum daily precipitation totals Wind per year show minimum increases in extreme Significant storm events, such as “Christian” or precipitation totals at many stations. These trends are “Xaver” in 2013, regularly enliven discussion around only significant at a few stations (< 10 %). The finest possible changes in the frequency with which storms regional resolution (grid box size) is currently 150 to occur or about general long-term trends in wind 200 km for global climate models and 12 to 25 km speeds. The answers are elusive. This is partly due for regional climate models. This means that both to the non-trivial nature of wind speed measurement. model systems are not capable of directly simulating In order to reduce the influence of the land surface processes which can trigger thunderstorms. Processes as far as possible, wind – in contrast to all other such as the formation of thunderstorms can be meteorological parameters – is ideally measured at described by simplified parameterisations. a height of 10 m above ground level. Nonetheless, measured wind speeds are sensitive to changes in the The regional climate models of the current generation vicinity of stations (e.g. tree growth) or changes of simulate a tendency for an increase in rainfall the measurement location. For this reason, almost all extremes, but, due to resolution which is too coarse wind time series show inhomogeneities. What is more, for these processes, the models are not able to provide the available time series is usually only a few decades detailed local-scale data. Convection-allowing regional long, which is too short a period to determine long- models are currently the subject of research. term trends over a period of 100 years, for example. Particularly interesting storms or hurricanes are rare events and their statistical evaluation is therefore Hail dependent on the longest possible time series. Hail events are local and rare events which can cause great damage to infrastructure and losses in One way of nonetheless obtaining data on the agriculture. As hail usually falls in small geographical development of wind speed and the occurrence of areas, it has not been possible to record all such storms over the last 100 years or so is to refer to events in the past. This gap is being closed by radar geostrophic wind. This is based on differences in air data collected since 2011. The results from their pressure and is closely related to the „true“ wind. It analysis show that there are more hail events (i.e. a has been possible to make high-quality measurements higher number of days with hail) per year in the south of air pressure since the end of the 18th centur y. than in the north. However, it is not possible, on the Periods of between 10 years to a few decades of basis of the available observational data, to identify higher or lower wind speed (so-called multi-decadal any trends in the changing number of hail events. Alternatively, data can be used which indirectly provide information about hail events, namely those convection parameters which describe the potential for storm and hail formation. Statistical analyses of hail-related convection parameters show a slight increase in this potential over the last 20–30 years.

The spatial resolution of the regional climate models which are currently being used is not sufficient to model hail directly. Hail can only be roughly estimated by means of parameterisations. This means that it is not possible to make statements about future trends. Analyses of the convection potential do not show any uniform trends over the short-term planning horizon.

36 14 m/s

Annual averages 13

12

11

10

9 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010

▴ Annual average geostrophic wind, calculated from ground level air pressure data at the Hamburg, Emden and List stations. The diagram shows the period 1880 to 2016.

variations) can be identified for geostrophic wind Tornadoes for the German Bight using air pressure data from Tornadoes are short-lived and localised violently Hamburg, Emden and List on Sylt. Only a weak falling rotating columns of air at the bottom of a convective trend is apparent over the entire time series which is, cloud and in contact with the ground. Depending however, much smaller than the year-to-year variation on their severity they can cause severe damage. and as such is not statistically significant. More tornadoes are being detected at present than in the past. Weaker tornadoes, which cause much The results of climate model projections reveal a less damage, still often remain undetected. Since similar pattern also showing multi-decadal variations the year 2000 between 20 and 60 cases have been in the past and in the future but no long-term trend. recorded per year. There are probably far more This is also the case for storms for which there is no tornadoes than the number actually detected would discernible significant change for the future. suggest. More violent and destructive tornadoes seldom occur in Germany. On average, meteorologists expect between five and ten of those tornadoes a year. Because of the lack of reliable observations/counts ▾ Damage caused by a tornado in Affing (Bavaria) on 13 May 2015. in the past it is not possible to stata whether there has been an increase in the number of tornadoes in Germany or not.

The available regional climate projections do not allow for the conclusion that tornadoes will occur more frequently in Germany in the future. The projected higher energy potential of the atmosphere in the future could result in a higher frequency of violent tornados and thus a higher risk of potentially very destructive tornadoes.

37 Current climate system research

Changes in the climate are based on complex interdependencies. These extend over long periods of time and may take different regional forms. This is why it is so difficult for people to perceive climate change with the ordinary senses. The national and international research community regards study of the climate system and its impacts on society as a key topic.

The foundations of modern climate research were On the basis of these analyses, six topics of particular laid with the recognition that increasing combustion interest have been determined: of fossil carbon stores will alter the composition of 1. Determining and reducing uncertainties in climate the Earth‘s atmosphere. The link between observed forecasts and climate projections warming of the Earth‘s atmosphere and human 2. Extending the weather forecast and linking it to activities has been clearly demonstrated in recent the sub-seasonal climate forecasts years. This reinforces the necessity for further study 3. Abrupt climate changes of the climate system. In this respect, the research 4. Water cycle in a warmer world community has formulated three key objectives for 5. Air quality and climate change the years ahead: 6. Greenhouse gas cycles in the climate system 1. Deeper understanding of the complex interactions within the climate system These issues call for ongoing work and research 2. Assessment of and response to the risks and initiatives. This means, for example, that the basis for opportunities arising from climate change long-term strategies for the further development and 3. The role of climate research in society refi nement of regional and global observation systems needs to be laid in research initiatives and models can Deeper understanding of the climate system be used to test many diff erent hypotheses. Work must The fundamental interactions of Earth‘s climate also be done to ensure that the relevant processes system are understood. However, the system is so are systematically recorded over the long term. complex that an enormous amount of research still This requires reliable monitoring of anthropogenic needs to be done to improve our understanding of changes and natural variability. certain detailed aspects of the climate. This includes gaps in the understanding of specific processes as Evaluating and dealing with risks and opportunities well as of the interactions between climate system One special challenge for all stakeholders is the components. time lag and distributional discrepancies between the causes and eff ects of climate change. Questions Many topics have been systematically analysed and concerning the costs and benefi ts of climate change existing gaps identifi ed by climate scientists cannot be answered by climate scientists alone. This collaborating at the national and international level. issue and the options for action which it engenders

38 must be tackled at the regional and global level and be in particular, their tasks and the boundaries of understood as a common task requiring co-operation their work. What role should climate researchers between many diff erent scientifi c disciplines. play? Should they confi ne themselves to an inward scientifi c view or should they pursue a political The regional impacts of climate change are felt in agenda? A good example is the IPCC mandate, societies with very diff erent social and economic which is meant to be policy relevant but not policy structures and cultural characteristics. In many cases prescriptive. there exist substantial diff erences in the cultural perception of and approach to risks. These diff erent The regular analysis of knowledge generation poses a approaches call for analytical research and the formidable and enduring challenge for every scientifi c development of options for action tailored to specifi c discipline. What are the assumptions upon which our regions. present knowledge is based? Where has consensus been established, and where is there dissent? It is also Research on the interactions between society and important to ask whether the existing institutional climate change structures for climate research are fi t for purpose. An important issue for the future is the role of Are the various topics dealt with appropriately in scientists and research institutions in society and, relation to each other?

CLIMATE FORECASTS FOR THE NEXT MONTHS TO YEARS

What will weather conditions be like over the next weeks, months and years? Climate forecasts are already able to deliver the basis for decision- making in many regions of the world.

Climate forecasts are predictions of the probability of months or years in the future being warmer/colder or drier/wetter than the long-term average. They are based on forecasts for future months (key word: seasonal forecasts) and years (key word: decadal forecasts). Combination with forecasts made in the past enables a comprehensive statistical evaluation of predictions to be made and trends to be identifi ed on the basis of climatology. In this respect, climate forecasts are completely diff erent from weather forecasts, which provide information about detailed weather events over the next hours to days. Forecasts extending over a period of several months up to ten years must also take into account all the components of the climate system – not merely the lower layer of the atmosphere (the troposphere, up to approximately 9–16 km) but also higher atmosphere layers, the soil, ocean and sea ice. A climate forecast is produced using a climate model coupled with all these components. In order to generate a robust statistical assessment of forecast quality and reliability both historical and current forecasts are computed as an ensemble. This means, that each forecasts is performed several times under slightly varying states thus forming a range of possible solutions. This range of solutions produced is also used to evaluate the uncertainties which arise from the nonlinearity of the climate system. Seasonal forecasts are provided each month by using a supercomputing system, e.g. the European Centre for Medium-Range Weather Forecasts in Reading (United Kingdom). At DWD, those forecasts are analysed and processed, enabling for instance El Niño forecasts. Research on decadal forecasts is currently ongoing. Operational use is planned for the next few years.

39 A climate lexicon

Baseline period Information about changes in the future mean climate conditions is always provided in relation to a baseline period. The statements made in this report regarding possible future changes are made with reference to the period 1971–2000. Statements always describe the mean conditions pertaining over a period of 30 years.

Indicator days An indicator day is a day on which a defined threshold of a climatological parameter is reached, not reached or exceeded (e.g. a summer day is defined as a day Planning horizon with a maximum temperature of ≥ 25 °C) or a certain This report distinguishes between a short-term and weather phenomenon occurs (e.g. a thunderstorm long-term planning horizon. The short-term planning day is a day on which a thunderstorm occurs). horizon describes the time window for the years 2021 to 2050, the long-term horizon the time window for Climate projection the years 2071 to 2100. Statements relating to these A climate projection is the description of a possible periods are always made in relation to the 1971–2000 and plausible future state of the climate system baseline period. and the chronological development leading up to it. Climate projections are usually produced using a Reference period scenario-based climate model. Information about changes in observed mean climate conditions is always provided in relation to Climate forecast a benchmark. In this climate report, information Climate forecasts make statements about the future about the past is provided in relation to the reference state of the climate system based on its past and period 1961–1990 which corresponds with the WMO’s present state. Traditionally, weather forecasts look internationally agreed standard reference period for at the ways in which the weather will develop over long-term climate observation. Statements always the next one to ten days. Climate forecast research describe the mean conditions pertaining over a period is currently being undertaken on the assessment of of 30 years. developments over much larger timescales extending from several months to a decade. Spread Future climate developments are analysed by means Percentile of a group of climate projections (ensemble). The Percentiles and quantiles are percentage values. spread describes the range between model outputs They group the number of studied model results into which produce the greatest and smallest changes. measurement classes which can include a certain percentage of the results. The range between the Scenarios 15th and 85th percentile, for example, covers 70 % of A scenario draws on assumptions to describe a model results. The value which a percentile takes, possible future. One option is to build a plausible e.g. 85th percentile = 9.4 °C, means that 85 % of the chain of assumptions concerning political, economic results are lower than this value and only 15 % are and ecological conditions in the future and to derive higher. changes in greenhouse gas emissions on this basis.

40 Climate modelling terminology The results derived from climate modelling must be reported in texts on climate change using uniform and clearly defined terminology. This will help • enhance confidence in the validity of findings based on the type, quantity, quality and consistency of the evidence and levels of agreement. • provide a measure of the imprecision of findings which is calculated from quantitative analyses.

Multi-model ensembles, by their nature, are “ensembles of opportunity”, which means that they are Term Agreement a collection of available climate projections that meet very high agreement In at least 9 out of 10 cases certain minimum requirements, such as availability high agreement In approx. 8 out of 10 cases in a certain resolution for a certain region over a medium high agreement In approx. 5 out of 10 cases certain period of time. Furthermore, many climate low agreement In approx. 2 out of 10 cases models have more or less strong similarities (with each other). The combination of these two important characteristics of multi-model ensembles, i.e. the which also includes substantial redundancies. As a arbitrary composition of the ensemble and the non- result, it is only possible to describe the degree of coincident similarities between the models, has the agreement of the used model runs. effect that the crowd of climate change signal data generated by such an ensemble doesn‘t behave like an The degree of agreement is italicised, e.g. very high independent, identically distributed random sample agreement.

41 Publishing details

The National Climate Report has been written Layout and typesetting at the instigation and in close co-operation with Elke Roßkamp (Deutscher Wetterdienst) the Federation-Länder expert discussion group “Interpretation of regional climate model data”. Acknowledgement DWD: Cr (Reno Schafranek), 2, 7, 36/37 (Johann Authors Siemens) Falk Böttcher, Dr Thomas Deutschländer, Fire brigade Gerolstein: 34 Andreas Friedrich, Karsten Friedrich, Creative Collection: Cc, 1t, 6c, 16, 18, 24l, 27cl, 27cr, Dr Kristina Fröhlich, Dr Barbara Früh, 27r, 31 Dr Anette Ganske*, Dr Hartmut Heinrich*, Panthermedia.net: Cl (Dominik Zwingmann), Cc Dr Frank Kreienkamp, Dr Gabriele Malitz, (Robert Biedermann), 1c (Hans Eder), 4 (Dario Jens Möller*, Dr Monika Rauthe, Wolfgang Riecke, Sabljak), 6l (Clemens Humeniuk), 6r (Wolfgang Filser), Thomas Schmidt, Dr Birger Tinz, Dr Andreas Walter 10b (Orlando Rosu), 11 (Hendrik Fuchs), 15r (Daniel * Federal Maritime and Hydrographic Agency Loretto), 21 (pekada), 23 (bestshot70), 24cl (Oliver C. Bellido), 24cr (Gabi Faltenbacher), 24r, (Tyler Editorial team Olson), 27l (Ingram Vitantonio Cicorella), 29t (Roland Dr Frank Kreienkamp Schmock), 29b (Ines Weiland-Weiser), 32 (Bernd Leitner), 35 (Ingo Gronostay), 38 (James Steidl), 40 Steering committee (Rilo Naumann), 41 (drizzd) PD Dr Achim Daschkeit (Federal Environment Agency), MEV-Verlag: 1b, 15l Dr Kai-Achim Höpker (State Institute for the Fotolia.com: 5 (Gina Sanders), 8 (AndreasG), 10t (Paul Environment, Measurements and Nature Conservation Paladin), 17 (gradt), 27t (Mykola Velychko) Baden-Württemberg), Carsten Linke (Brandenburg State Office for the Environment), (l: left; c: centre; t: top; r: right, b: bottom; C: Cover) Dr Frank Kreienkamp (Deutscher Wetterdienst)

Online edition This publication is available in electronic form on our website at www.dwd.de/nationalclimatereport. The website also provides links to background material and similar Deutscher Wetterdienst products.

The online issue is subject to license: http://creativecommons.org/ licenses/by-nc-nd/4.0/deed.de

References in text DWD (2017): National Climate Report. 3. revised edition, Deutscher Wetterdienst, Offenbach am Main, Deutschland, 46 pages.

ISSN 2568-2261 (Online)

42

Email: klima.off[email protected] Fax: +49(0)69/8062-2993 Tel: +49(0)69/8062-2912 63067 Offenbach Frankfurter Straße135 Central ClimateOffice Climate andEnvironment Consultancy Deutscher Wetterdienst you havealso access toourpageson: Through ourwebsiteatwww.dwd.de

DWD 3rd revised edition 10.17