In 2024, the average summer temperature in Germany was 18.5 degrees Celsius. This was basically unchanged compared to the year before. While figures fluctuated during the given timeline, there were regular peaks, and in general, temperatures had grown noticeably since the 1960s. Not beating the heat German summers are getting hotter, and as desired as warm weather may be after months of winter (which, incidentally, also warms up year after year), this is another confirmation of global warming. Higher summer temperatures have various negative effects on both nature and humans. Recent years in Germany have seen a growing number of hot days with a temperature of at least 30 degrees, with 11.5 recorded in 2023. However, this was a decrease compared to the year before. The number of deaths due to heat and sunlight had peaked in 2015. Rain or shine All the German states saw less sunshine hours in 2023 compared to the previous year. The sunniest states were Baden-Württemberg, Bavaria and Saarland. Meanwhile, summer precipitation in Germany varied greatly during the same timeline as presented in this graph, but 2022 was one of the dryest years yet.

Based on current monthly figures, on average, German climate has gotten a bit warmer. The average temperature for January 2025 was recorded at around 2 degrees Celsius, compared to 1.5 degrees a year before. In the broader context of climate change, average monthly temperatures are indicative of where the national climate is headed and whether attempts to control global warming are successful. Summer and winter Average summer temperature in Germany fluctuated in recent years, generally between 18 to 19 degrees Celsius. The season remains generally warm, and while there may not be as many hot and sunny days as in other parts of Europe, heat waves have occurred. In fact, 2023 saw 11.5 days with a temperature of at least 30 degrees, though this was a decrease compared to the year before. Meanwhile, average winter temperatures also fluctuated, but were higher in recent years, rising over four degrees on average in 2024. Figures remained in the above zero range since 2011. Numbers therefore suggest that German winters are becoming warmer, even if individual regions experiencing colder sub-zero snaps or even more snowfall may disagree. Rain, rain, go away Average monthly precipitation varied depending on the season, though sometimes figures from different times of the year were comparable. In 2024, the average monthly precipitation was highest in May and September, although rainfalls might increase in October and November with the beginning of the cold season. In the past, torrential rains have led to catastrophic flooding in Germany, with one of the most devastating being the flood of July 2021. Germany is not immune to the weather changing between two extremes, e.g. very warm spring months mostly without rain, when rain might be wished for, and then increased precipitation in other months where dry weather might be better, for example during planting and harvest seasons. Climate change remains on the agenda in all its far-reaching ways.

This statistic shows the average temperature in Germany in the summers of 2023 and 2024, by federal state. In summer 2024, the average temperature in Berlin was 19.7 degrees Celsius. This made Berlin the warmest federal state in the period in question, followed by Brandenburg and Baden-Württemberg.

Temperature in Germany increased to 10.88 celsius in 2023 from 10.78 celsius in 2022. This dataset includes a chart with historical data for Germany Average Temperature.

In 2023/2024, the average winter temperature in Germany was 4.1 degrees Celsius. That winter was part of a growing list of warmer winters in the country. Figures had increased noticeably compared to the 1960s.

Warmer in the winter

Everyone has a different perception of what actually makes a cold or warm winter, but the fact is that winter temperatures are, indeed, changing in Germany, and its 16 federal states are feeling it. Also in 2022/2023, Bremen and Hamburg in the north recorded the highest average figures at around 4 degrees each. The least warm states that year, so to speak, were Thuringia, Saxony, and Bavaria. The German National Meteorological Service (Deutscher Wetterdienst or DWD), a federal office, monitors the weather in Germany.

Global warming

Rising temperatures are a global concern, with climate change making itself known. While these developments may be influenced by natural events, human industrial activity has been another significant contributor for centuries now. Greenhouse gas emissions play a leading part in global warming. This leads to warmer seasons year-round and summer heat waves, as greenhouse gas emissions cause solar heat to remain in the Earth’s atmosphere. In fact, as of 2022, Germany recorded 17.3 days with a temperature of at least 30 degrees Celcius, which was more than three times the increase compared to 2021.

In April 2025, the average precipitation amounted to 31 liters per square meter, an increase compared to the previous month. The rainiest state in Germany was Saarland.

Quality controlled and gap-filled continuous air temperature data from the urban weather station at Freiburg-Werthmannstrasse (FRWRTM, 7.8447ºE, 47.9928, 277 m) using a passively ventilated and shielded temperature and humidity probe (Campbell Scientific Inc., CS 215) operated in a Stevenson Screen 2m above ground level in the vegetated backyard of Werthmannstrasse 10, 79098 Freiburg im Breisgau, Germany.

• Quality controlled in-canopy air temperature data are available and aggregated at 10min, 30min, hourly, daily, monthly and yearly resolution for the year 2024.
• Average, minimum and maximum in-canopy air temperatures are provided on hourly, daily, monthly and annual scales.
• Characteristic hours and days are reported on daily, monthly and annual scales (e.g. summer days with T_max > 25ºC, hot days with T_max > 30º, desert days with T_max > 35ºC, tropical nights with T_min > 20°, frost days with T_min < 0ºC and ice days with T_max < 0ºC, all based on 00:00 - 24:00 UTC).
• Detailed information on gap-filled data is provided.
• Note: All times are provided in UTC, not local time.

For more details read `FRWRTM_2024_AirTemperature_MetaData.txt`.

Influences of climate change on urban climate are often investigated in the context of increased values for high temperatures or precipitation extremes. As a consequence, many studies concentrate on the climate change impacts that happen during summer. When looking at projected temperature and precipitation changes, mean changes are larger in winter than in summer, at least for northern Germany. At the same time, the distribution of temperature is broadened, which implies that winters with temperatures below freezing point or snowfall will still happen in the future.

The aim of this project is to quantify climate change related impacts on the winter climate of Hamburg in detail. For this purpose, an existing canopy resolving model (MITRAS) will be expanded to allow a detailed analysis of precipitation including snow and of frost distribution at the local scale. With this tool the impact on the local winter climate, as well as of adaption measures developed for a changed summer climate are investigated. To ensure model results occur in time a code optimisation and parallelisation is part of this project.

This project was financed within the framework of the Helmholtz Institute for Climate Service Science (HICSS), a cooperation between Climate Service Center Germany (GERICS) and University Hamburg, Germany (https://www.hicss-hamburg.de/projects/urban_winter_hamburg/).

This statistic shows the average precipitation amount in summer in Germany from 1960 to 2024. In 2024, summer precipitation in Germany amounted to 240 liters per square meter. This was a decrease compared to the previous year. Figures have fluctuated over the timeline under review.

Average Summer (Jul-Sep) Temperature (2015-2030) simulated by RegCM3 with ECHAM5 projections as boundary conditions.

Units are degrees Celsius.

These data were generated by the regional climate model RegCM3 with boundary conditions from a GCM future climate projections. The data were downscaled statistically by calculating differences (anomalies) between the RegCM3 results with GCM-driven boundary conditions for 1968-99 and those for a future period, in this case 2015-2030. The anomalies were added (temperatures) or multiplied (precipitation) to a climate baseline from PRISM (Parameter-elevation Regressions on Indepenent Slopes Model - prism.oregonstate.edu) data based on historical observations. The PRISM baseline was calculated as average monthly climate conditions for 1968-1999 reprojected the results to the BLM Albers 4km grid. PRISM data are provided in a 2.5 arc-minute lat-lon grid.

RegCM3 is the third generation of the Regional Climate Model originally developed at the National Center for Atmospheric Research during the late 1980s and early 1990s. Details on current model components and applications of the model can be found in numerous publications (e.g., Giorgi et al, 2004a,b, Pal et al, 2007), the ICTP RegCNET web site (http:users.ictp.itRegCNETmodel.html), and the ICTP RegCM publications web site (http:users.ictp.it~pubregcmRegCM3pubs.htm). The Western North America domain has a horizontal grid spacing of 15 km and 18 vertical levels.

RegCM3 requires time-dependent lateral (wind, temperature, and humidity) and surface [surface pressure and sea surface temperature (SST)] boundary conditions that are updated every 6 hours of simulation. Lateral boundary conditions are derived from General Circulation Model (GCM) output or observations (e.g. NCEP). Additional information can be found at: http:regclim.coas.oregonstate.edu.

Global simulations from the Max Planck Institute (Germany) climate model ECHAM5 were part of a suite of model results used in the 4th Climate Model Inter-comparison Project (CMIP4) and the Intergovernmental Panel for Climate Change 4th Assessment Report. Details and documentation of the model can be found on the CMIP website: http:wwwpcmdi.llnl.govipccmodel_documentationipcc_model_documentation.php.

Average Summer (Jul-Sep) Temperature (2045-2060) simulated by RegCM3 with ECHAM5 projections as boundary conditions. Units are degrees Celsius. These data were generated by the regional climate model RegCM3 with boundary conditions from a GCM future climate projections. The data were downscaled statistically by calculating differences (anomalies) between the RegCM3 results with GCM-driven boundary conditions for 1968-99 and those for a future period, in this case 2015-2030. The anomalies were added (temperatures) or multiplied (precipitation) to a climate baseline from PRISM (Parameter-elevation Regressions on Indepenent Slopes Model - prism.oregonstate.edu) data based on historical observations. The PRISM baseline was calculated as average monthly climate conditions for 1968-1999 reprojected the results to the BLM Albers 4km grid. PRISM data are provided in a 2.5 arc-minute lat-lon grid. RegCM3 is the third generation of the Regional Climate Model originally developed at the National Center for Atmospheric Research during the late 1980s and early 1990s. Details on current model components and applications of the model can be found in numerous publications (e.g., Giorgi et al, 2004a,b, Pal et al, 2007), the ICTP RegCNET web site (http://users.ictp.it/RegCNET/model.html), and the ICTP RegCM publications web site (http://users.ictp.it/~pubregcm/RegCM3/pubs.htm). The Western North America domain has a horizontal grid spacing of 15 km and 18 vertical levels. RegCM3 requires time-dependent lateral (wind, temperature, and humidity) and surface [surface pressure and sea surface temperature (SST)] boundary conditions that are updated every 6 hours of simulation. Lateral boundary conditions are derived from General Circulation Model (GCM) output or observations (e.g. NCEP). Additional information can be found at: http://regclim.coas.oregonstate.edu/. Global simulations from the Max Planck Institute (Germany) climate model ECHAM5 were part of a suite of model results used in the 4th Climate Model Inter-comparison Project (CMIP4) and the Intergovernmental Panel for Climate Change 4th Assessment Report. Details and documentation of the model can be found on the CMIP website: http://wwwpcmdi.llnl.gov/ipcc/model_documentation/ipcc_model_documentation.php.

The effective water balance results from the difference between water supply and potential evapotranspiration during the main vegetation period (May-August). The water supply is made up of the precipitation during this period, the amount of water present and extractable in the soil (described by the usable field capacity in the effective root area) and, if necessary, a capillary rise. The capillary rise is the result of the rise rate per day and the culture-dependent duration of the rise. The ascent rate is largely dependent on the soil type and the distance between the lower limit of the effective root zone and the ground or backwater body. The use-differentiated soil overview map 1:1,000,000 (BÜK1000N) served as the pedological basis. The data from CORINE Land Cover (CLC2006) were used for land cover and land use. The climate scenario data (https://www.dwd.de/ref-ensemble) were made available by the German Weather Service (DWD) in a resolution of 5 x 5 km. This is an ensemble of 16 bias-corrected area data sets (combination of global and regional climate models) that describe the scenario RCP8.5 (RCPs Representative Concentration Pathways) and assume an additional radiative forcing of 8.5 W/m². The nine raster data sets with a resolution of 5 x 5 km each represent the mean, 15th and 85th percentile of the effective water balance in the main vegetation period in Germany for the climatic periods 1971-2000, 2031-2060 and 2071-2099.

Quality controlled and gap-filled continuous air temperature data from the urban rooftop weather station at Freiburg-Chemiehochhaus (FRCHEM, 7.8486ºE, 48.0011ºN, 323.5 m) using an actively ventillated and shielded psychrometer operated 2m above roof level.

• Quality controlled air temperature data are available and aggregated at hourly, daily, monthly and yearly resolution for the year 2022.
• Average, minimum and maximum air temperatures are provided for the aggregation intervals.
• Characteristic hours and days are reported on daily, monthly and annual scales (e.g. summer days with T_max > 25ºC, hot days with T_max > 30º, desert days with T_max > 35ºC, tropical nights with T_min > 20°, frost days with T_min < 0ºC and ice days with T_max < 0ºC, all based on 00:00 - 24:00).
• Detailed information on gap-filled data is provided.
• Note: All times are provided in UTC, not local time.

For more details read `FRCHEM_2022_AirTemperature_MetaData.txt`.

Stable hydrogen and carbon isotope ratios of wood lignin methoxy groups (δ13CLM and δ2HLM values) have been shown to be reliable proxies of past temperature variations. Previous studies showed that δ2HLM values even work in temperate environments where classical tree-ring width and maximum latewood density measurements are less successful. Here, we analyse annually resolved δ13CLM values from 1916-2015 of four beech trees (Fagus sylvatica) from a temperate site near Hohenpeißenberg in southern Germany and compare these data with regional to continental scale climate observations. Initial δ13CLM values were corrected for the Suess effect (a decrease of δ13C in atmospheric CO2) and physiological tree responses to increasing atmospheric CO2 concentrations considering a range of published discrimination factors. The calibration of δ13CLM chronologies against instrumental data reveals highest correlations with regional summer (r = 0.68) and mean annual temperatures (r = 0.66), as well as previous-year September to current-year August temperatures (r = 0.61), all calculated from 1916-2015 and reaching p < 0.001. Additional calibration trials using detrended δ13CLM values and climate data, to constrain effects of autocorrelation on significance levels, returned rsummer = 0.46 (p < 0.001), rannual = 0.25 (p < 0.05) and rprev.Sep-Aug = 0.18 (p > 0.05). The new δ13CLM chronologies were finally compared with previously produced δ2HLM values of the same trees to evaluate the additional gain of assessing past climate variability using a dual-isotope approach. Compared to δ13CLM, δ2HLM values correlates substantially stronger with large-scale temperatures averaged over western Europe (rprev.Sep-Aug = 0.69), whereas only weak and mainly insignificant correlations are obtained between precipitation and both isotope chronologies (δ13CLM and δ2HLM values). Our results indicate great potential of using δ13CLM values from temperate environments as a proxy for local temperatures, and in combination with δ2HLM values, to assess regional to sub-continental scale temperature patterns.

Business Climate Index: sa: Business Climate: West Germany data was reported at 97.400 1991=100 in Jan 2004. This records an increase from the previous number of 96.900 1991=100 for Dec 2003. Business Climate Index: sa: Business Climate: West Germany data is updated monthly, averaging 93.900 1991=100 from Jan 1969 (Median) to Jan 2004, with 421 observations. The data reached an all-time high of 113.200 1991=100 in Jun 1969 and a record low of 76.700 1991=100 in Aug 1982. Business Climate Index: sa: Business Climate: West Germany data remains active status in CEIC and is reported by Ifo Institute - Leibniz Institute for Economic Research at the University of Munich. The data is categorized under Global Database’s Germany – Table DE.S006: Business Climate Index: 1991=100: Seasonally Adjusted: IFO Institute.

description: Difference of Average Summer (Jul-Sep) Temperature (2045-2060 vs 1968-1999) simulated by RegCM3 with ECHAM5 projections as boundary conditions. Units are degrees Celsius. These data were generated by the regional climate model RegCM3 with boundary conditions from a GCM future climate projections. The data were downscaled statistically by calculating differences (anomalies) between the RegCM3 results with GCM-driven boundary conditions for 1968-99 and those for a future period, in this case 2015-2030. The anomalies were added (temperatures) or multiplied (precipitation) to a climate baseline from PRISM (Parameter-elevation Regressions on Indepenent Slopes Model - prism.oregonstate.edu) data based on historical observations. The PRISM baseline was calculated as average monthly climate conditions for 1968-1999 reprojected the results to the BLM Albers 4km grid. PRISM data are provided in a 2.5 arc-minute lat-lon grid. This dataset depicts differences between the downscaled data for 2015-2030 and 1968-1999. RegCM3 is the third generation of the Regional Climate Model originally developed at the National Center for Atmospheric Research during the late 1980s and early 1990s. Details on current model components and applications of the model can be found in numerous publications (e.g., Giorgi et al, 2004a,b, Pal et al, 2007), the ICTP RegCNET web site (http://users.ictp.it/RegCNET/model.html), and the ICTP RegCM publications web site (http://users.ictp.it/~pubregcm/RegCM3/pubs.htm). The Western North America domain has a horizontal grid spacing of 15 km and 18 vertical levels. RegCM3 requires time-dependent lateral (wind, temperature, and humidity) and surface [surface pressure and sea surface temperature (SST)] boundary conditions that are updated every 6 hours of simulation. Lateral boundary conditions are derived from General Circulation Model (GCM) output or observations (e.g. NCEP). Additional information can be found at: http://regclim.coas.oregonstate.edu/. Global simulations from the Max Planck Institute (Germany) climate model ECHAM5 were part of a suite of model results used in the 4th Climate Model Inter-comparison Project (CMIP4) and the Intergovernmental Panel for Climate Change 4th Assessment Report. Details and documentation of the model can be found on the CMIP website: http://wwwpcmdi.llnl.gov/ipcc/model_documentation/ipcc_model_documentation.php.; abstract: Difference of Average Summer (Jul-Sep) Temperature (2045-2060 vs 1968-1999) simulated by RegCM3 with ECHAM5 projections as boundary conditions. Units are degrees Celsius. These data were generated by the regional climate model RegCM3 with boundary conditions from a GCM future climate projections. The data were downscaled statistically by calculating differences (anomalies) between the RegCM3 results with GCM-driven boundary conditions for 1968-99 and those for a future period, in this case 2015-2030. The anomalies were added (temperatures) or multiplied (precipitation) to a climate baseline from PRISM (Parameter-elevation Regressions on Indepenent Slopes Model - prism.oregonstate.edu) data based on historical observations. The PRISM baseline was calculated as average monthly climate conditions for 1968-1999 reprojected the results to the BLM Albers 4km grid. PRISM data are provided in a 2.5 arc-minute lat-lon grid. This dataset depicts differences between the downscaled data for 2015-2030 and 1968-1999. RegCM3 is the third generation of the Regional Climate Model originally developed at the National Center for Atmospheric Research during the late 1980s and early 1990s. Details on current model components and applications of the model can be found in numerous publications (e.g., Giorgi et al, 2004a,b, Pal et al, 2007), the ICTP RegCNET web site (http://users.ictp.it/RegCNET/model.html), and the ICTP RegCM publications web site (http://users.ictp.it/~pubregcm/RegCM3/pubs.htm). The Western North America domain has a horizontal grid spacing of 15 km and 18 vertical levels. RegCM3 requires time-dependent lateral (wind, temperature, and humidity) and surface [surface pressure and sea surface temperature (SST)] boundary conditions that are updated every 6 hours of simulation. Lateral boundary conditions are derived from General Circulation Model (GCM) output or observations (e.g. NCEP). Additional information can be found at: http://regclim.coas.oregonstate.edu/. Global simulations from the Max Planck Institute (Germany) climate model ECHAM5 were part of a suite of model results used in the 4th Climate Model Inter-comparison Project (CMIP4) and the Intergovernmental Panel for Climate Change 4th Assessment Report. Details and documentation of the model can be found on the CMIP website: http://wwwpcmdi.llnl.gov/ipcc/model_documentation/ipcc_model_documentation.php.

We provide the storyline data (in NetCDF format) used in the paper: ���The July 2019 European heatwave in a warmer climate: Storyline scenarios with a coupled model using spectral nudging��� published in Journal of Climate. The data is structured in four .tar.gz files (Preindustrial, Present, 2 and 4 K warmer climates) containing all variables used in this each climate. The data from the five ensemble members (E1 to E5) have been included separately in 3-months files. Atmospheric variables (Files are named as: {variable}_E{ensemble member}_{starting month}{year}.nc: Latent heat flux (ahfl) Sensible heat flux (ahfs) Monthly Global Mean 2m Temperature (GMTT2mMonthly) Maximum 2m Temperature (t2max) Mean 2m Temperature (t2mean) Minimum 2m Temperature (t2max) Soil Wetness (ws) Only for present climate: 850 hPa Temperature (T850) Total Cloud Cover (TCC) 500 hPa Geopotential Height (Z500) Five layers soil moisture (Only for present climate, Files are named as: From20172019in2017Climatessp370{ensemble member}_{year}{starting month}.01_jsbid.nc) Oceanic variables (from FESOM, Files are named as: {variable}_E{ensemble member}_{year}{starting month}01.nc: Sea Ice Concentration (SIC) Sea Surface Temperature (SST) Please, note that FESOM uses an unstructured mesh. This work was supported by the Helmholtz-Climate-Initiative through the HI-CAM project (Drivers cluster). Author Goessling acknowledges the financial support by the Federal Ministry of Education and Research of Germany in the framework of SSIP (Grant 01LN1701A). The simulations were performed at the German Climate Computing Center (DKRZ) using the ESM-Tools (Barbi et al. 2021). We thank Sebastian Rast (MPI-M) for support with the spectral nudging in ECHAM, and the ESM-Tools staff for their assistance during the simulation.

The data herein presents a new chironomid record and associated chironomid-based temperature reconstruction covering the time interval ca. 98–46 ka from the palaeolake Füramoos, Southern Germany. These data also include non-chironomid invertebrate remains, including Ceratopogonidae, Daphnia and Ephemeroptera as well as Characean oospores.

Methods Sample processing: Sediment samples between depths 9.6-6.0 m required no chemical treatment. Samples betweend depths of 13-9.6 m were heated in 10% KOH for 15 minutes at 85°C due to the dificulty of processing those samples. Sediment samples were then sieved through 100 and 200 micrometer sieves. Chironomids were picked from those samples using a Bogorov tray under a stereomicroscope (30 – 50 x magnification). Samples were dried and then mounted on microscope slides using Euparal and identified under at 40 – 100 x magnification with a compound microscope. Next to chironomids the remains of Sialidae, Ceratopogonidae, Daphnia, Ephemeroptera, oribatid mites, Trichoptera, Plecoptera, Sciaridae and Tipulidae as well as Characeae oogonia and Plumatella statoblasts were mounted and identified. The chironmoid based temperature reconstruciton was produced using a two-component weighted averaging partial least squares model.

For post identification data processing and temperature reconstruction analysis please see the associated publication in Quaternary Science Reviews where the results are discussed in detail. The manuscript is fully open access: https://doi.org/10.1016/j.quascirev.2021.107008.

Quality controlled and gap-filled continuous air temperature data from the urban rooftop weather station at Freiburg-Chemiehochhaus (FRCHEM, 7.8486ºE, 48.0011ºN, 323.5 m) using an actively ventillated and shielded psychrometer operated 2m above roof level.

• Quality controlled air temperature data are available and aggregated at 10min, 30min, hourly, daily, monthly and yearly resolution for the year 2024.
• Average, minimum and maximum air temperatures are provided on hourly, daily, monthly and annual scales.
• Characteristic hours and days are reported on daily, monthly and annual scales (e.g. summer days with T_max > 25ºC, hot days with T_max > 30º, desert days with T_max > 35ºC, tropical nights with T_min > 20°, frost days with T_min < 0ºC and ice days with T_max < 0ºC, all based on 00:00 - 24:00 UTC).
• Detailed information on gap-filled data is provided.
• Note: All times are provided in UTC, not local time.

For more details read `FRCHEM_2024_AirTemperature_MetaData.txt`.

Description

These data were used in the study "Non-Stationary Large-Scale Statistics of Precipitation Extremes in Central Europe" (Fauer et al., 2022, https://doi.org/10.21203/rs.3.rs-2542862/v1). Rainfall data were collected from stations by the German Meteorological Service (DWD) and Wupperverband (corrected data). Raw time series data from the German Meteorological Service is publicly available under https://opendata.dwd.de/climate_environment/CDC/observations_germany/climate/. Only the annual and monthly precipitation maxima over different durations are published here. For more detailed information on our work and the modeling of extreme rainfall data, see also Fauer et al. (2021, https://doi.org/10.5194/hess-25-6479-2021).

Files

precipMax_and_covariates.csv: This file contains aggregated rainfall data over different durations and for different stations ("xdat"). Also, it contains covariates for the variables "blocking", "NAO", surface air temperature ("tas"), humidty and their polynomials up the the fourth order.

precip_meta.csv: This file contains additional information of the different stations such as longitude, latitude, altitude, temporal resolution (m=minutely, h=hourly, d=daily), group. The same group is assigned to stations which have a distance of less than 250 meters and can be treated as one station. The value in "group" corresponds to the value "station" in precipMax_and_covariates.csv.

Abstract of the according study

Extreme precipitation shows non-stationary behavior over time, but also with respect to other large-scale variables. While this effect is often neglected, we propose a model including the influence of North Atlantic Oscillation, time, surface temperature and a blocking index. The model features flexibility to use annual maxima as well as seasonal maxima to be fitted in a generalized extreme value setting. To further increase the efficiency of data usage maxima from different accumulation durations are aggregated so that information for extremes on different time scales can be provided. Our model is trained to individual station data with temporal resolutions ranging from one minute to one day across Germany. The models are selected with a stepwise BIC model selection and verified with a cross-validated quantile skill index. The verification shows that the new model performs better than a reference model without large scale information. Also, the new model enables insights into the effect of large scale variables on extreme precipitation. Results suggest that the probability of extreme precipitation increases with time since 1950 in all seasons. High probabilities of extremes are positively correlated with blocking situations in summer and with temperature in winter. However, they are negatively correlated with blocking situations in winter and temperature in summer.

Acknowledgements

We would like to thank the German Weather Service (DWD), especially Thomas Junghänel, and the Wupperverband, especially Marc Scheibel, for maintaining the station-based rainfall gauge and providing us with data.

Funding

This study is part of the ClimXtreme project (Grant number 01LP1902H) and is sponsored by the Federal Ministry of Education and Research in Germany.

Changelog

added humidity as large-scale covariate

large-scale covariates are normalized to mean=0 and standard deviation=1