Climate Model Ensemble

For the regionalisation and derivation of climate impacts from future climatic development paths (corresponding to the three RCP scenarios), it is necessary to refine the global/general climate model simulations (GCMs) for large regions such as Europe. This is usually achieved by using these as a drive for higher resolution regional models (RCMs).These are able to reflect the regional characteristics of the temperature and precipitation distribution more realistically, as effects caused by topography and land use are better represented.

The EU project www.impact2c.eu, completed in 2015, provides such a regional climate model ensemble of 2 to 4 single model simulations for all three RCPs (Gobiet et al., 2015). In addition, the simulation values were adjusted to the observation values of the EOBS data set (bias correction) for the historical simulation period (1971-2000).

KlimafolgenOnline is based on a lattice observation dataset from station data, which comprises more stations and measured variables (than the EOBS dataset Europe). Since the observation data for the bias correction of the model data come from Impact2c (Europe) on the one hand and the observation data in Climate Impactonline (Germany) on the other hand from different data products, there is a significant deviation between historical observation and the historically simulated climate for Germany.

The logic of Climate ImpactsOnline includes a direct synopsis of observation and future simulation data to reduce the complexity of the above model and data specific inhomogeneities for laypersons. Therefore, in this new version, in order to make the transition between the observed climate (1901-2010) and the simulated climate (2011-2100) seamless, the Impact2c ensemble difference of the future simulated climate to the historically simulated climate (1971-2000) is added to the actual observed values of the period 1971-2000. This results in a homogeneous data set that maintains the accuracy of the observation data and connects the climate-induced signal in an easily understandable way.

Climate Scenarios

The portal shows three different climate scenarios, so-called Representative Concentration Paths (RCPs). These were developed by the scientific community for the fifth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC).

The RCP scenarios 2.6, 4.5 and 8.5 used in the portal take into account different levels of greenhouse gas-induced radiation drives for the period 2011-2100. The intensity of the future climate protection from 2011 to 2100, which influences the radiation propulsion as well as the greenhouse gas concentration, is decisive for the different temperature increases in each scenario.

The scenario RCP 2.6 (Two degree scenario) shows a possible development that is likely to comply with the global two-degree limit. A requirement for this is strong climate protection.

The setting of the two-degree limit is a political objective, since science assumes that the climate consequences would still be manageable if the temperature were to fall below this level. Further information can be found in the German synthesis report of the Fifth Assessment Report of the IPCC. This document is available as PDF in the menu under climate knowledge.

At the UN climate summit in Paris in December 2015, 195 states adopted a new agreement against global warming. According to this, global warming is to be limited to "well below two degrees" compared with pre-industrial conditions. In addition, efforts are to be made to stabilise the global climate at 1.5 °C. The aim is to reduce global warming to "well below two degrees" compared to pre-industrial conditions.

The RCP 4.5 scenario (Moderate scenario) follows a significantly reduced emission trend and by 2100 achieves a global mean temperature increase of approx. 3° degrees compared with pre-industrial levels.

The RCP 8.5 scenario (Business as usual scenario) corresponds most closely to current emission trends and is therefore known as the business as usual scenario.
It considers a possible climate development with further increasing greenhouse gas emissions without additional climate protection measures. This results in an average global temperature increase of between 3.6 °C and 4.1 °C by 2100 compared to pre-industrial Nineau, if the currently available scenarios of the global climate models are taken as a basis.

Interpolation

For the presentation of the climate indicators derived from the daily climate data, a method was applied in which the available information at the locations of the measuring stations was interpolated to grid points. Their spatial resolution is approx. 12 x 12 km. The searched value at the corresponding point is determined by a weighted summation of the station data within a given radius of influence by subsequently sectioniding the sum by the number of stations (Shepard, 1988). The method is similar to the Cresman scheme used by Schrodin (1995). The calculation algorithm was adopted by the German Weather Service (DWD).

Shepard, D. (1988): A two-dimensional interpolation function for irregular spaced data. ACM National Conference Proceedings, Havard College, Cambridge Massachusetts, 517-523.
Schrodin, R. (1995): Dokumentation des EM/DM-Systems. Dokumentation des DWD, Abteilung Forschung, Offenbach.
https://swift.dkrz.de/v1/dkrz_a88e3fa5289d4987b4d3b1530c9feb13/ReKliEs-De/Supplement/Info/Interpolationsverfahren_PIK.pdf

Models

A model chain developed at PIK is used to estimate climate impacts on sectors in Germany. The models used here have been established at PIK, coordinated and applied together in studies and projects in this linkage. Examples of this can be found in the form of sectionerse regional studies for Germany.

On the basis of the regional climate model simulations (impact2C) used here, the impacts were estimated for various sectors.

Climate impact models:
SWIM - Soil and Water Intergrated Model (Sector water)
IRMA - Integrated Regional Model Assessment (Agriculture sector)
4C - FORESEE - Forest Ecosystems in a Changing Environment (Forest sector)

Detailed information on the models can be found in the sector-specific information.

Three examples of regional studies in which these models were applied to climate impacts in Germany are listed here:

PIK-Report No.121 - Klimawandel in der Region Havelland-Fläming
A. Lüttger, F.-W. Gerstengarbe, M. Gutsch, F. Hattermann, P. Lasch, A. Murawski, J. Petraschek, F. Suckow, P. C. Werner (Januar 2011)

PIK-Report No.112: Die Ertragsfähigkeit ostdeutscher Ackerflächen unter Klimawandel
F. Wechsung, F.-W. Gerstengarbe, P. Lasch, A. Lüttger (eds.) (Dezember 2008)

PIK-Report No.99: KLARA - Klimawandel - Auswirkungen, Risiken, Anpassung
M. Stock (Hrsg.) (Juli 2005)

Bandwidth

The regional climate model ensemble available from Impact2c is relatively small. That's why model uncertainty and internal (year-to-year) variability were brought together, sorted and the value at which 10%, 50% and 90% of all values are below, are described to estimate the bandwidths (50s percentile on the map, 10s, 50s and 90s percentile in the time series diagram).

For a 30-year period and a model ensemble of 4 simulations, this means a available pool of 4 x 30 = 120 annual values for the respective climate indicators.

Extreme Event Data

The number of extreme events per year/per season in Germany was estimated using data from the ClimXtreme Catalogue of Extreme Events. ClimXtreme is a network of different research institutions in Germany with the common goal to advance research on extreme events in the context of climate change.

This catalogue is a work in progress and is therefore subject to change. This portal aims to visualize the existing data and does not claim complete depiction of all events that occurred. It is possible that less data is available in early years, especially in the pre-satellite era (before 1979).

The data was processed by counting the number of listed events per time period (year/season), where events that spanned over the turn of a year were assigned to the first year in order to not count them twice. The percentiles were calculated using Python, see Fig. 1.

(Currently, this data is only visible in expert mode. To access it, please go to Settings -> Expert Mode -> On. The page will reload and you will then be able to choose the new parameters from the parameter list.)

Update of the data

Due to the continuous development in the field of climate and climate impact research, an update of the future climate simulations was carried out in summer 2019. The statistical downscaling of the earlier version was replaced by. dynamic climate scenarios, which were developed as part of the Impact2C.

This is one of the most comprehensive ensembles of regional climate simulations. This allows us to better and better understand the consequences of global warming with its impact on Germany.

For insectionidual derived climate indicators, this means a reassessment in terms of their properties and climate sensitivity. Specifically, this means that extreme summer temperature conditions will be somewhat less pronounced in the updated data in the future. In contrast to the old version, average annual precipitation increases slightly and heavy precipitation increases significantly in the summer half-year. This has corresponding effects on all sectors, such as forest fire risk, wet days and agricultural yields.

Observation data

Based on direct and long measurement data of the German Weather Service at 180 synoptic stations and 1038 precipitation stations, a homogenized station data set was generated and then interpolated to a grid. Thus a complete data set of all meteorological measured variables is available on a daily value basis from 1901-2010 at 1218 stations or all approx. 4000 grid points.