The questions below are originally written with the perspective of providing user guidance related to EURO-CORDEX results, however many questions are broad and applies in a more general sense. This could mean that they can also be used by users of CORDEX information from other regions in the world and also in general every climate projection data user. Please note that there is a comprehensive user guide document provided by EURO-CORDEX and can be found at https://www.euro-cordex.net/imperia/md/content/csc/cordex/guidance_for_euro-cordex_climate_projections_data_use__2021-02_1_.pdf
The main topics are based on different categories suggested by the climate modellers as “typical” categories for which questions often arise. As an additional input questions asked by attendees at the C3S webinar series (C3S webinars on regional climate projections for Europe | Copernicus) held in early March 2021.
The main topics has primarily been chosen based on experience by the regional climate modellers in dialogue with ECMWF. The main topics fall into eleven different categories as listed below. For each question (underlined) it is attempted to answer in two levels where the first level is a short more general answer (in italics), while the second layer goes slightly more into detail sometimes including references for further reading when appropriate. Depending on the question, sometimes the answer includes a short recommendation (in bold face) on what a user may need to consider when using EURO-CORDEX climate projection data and some specific links (in red), when the information can be already found elsewhere. Extensive literature references can be found below (also at the specific questions).
Questions related to added value of RCMs with respect to GCMs
Regional climate models have frequently been shown to add value over GCMs. Questions are related to what added value is and how it can be assessed.
Click here to expand the questions
What is added value?
Added value of high-resolution regional climate models with respect to the underlying global climate models means that they are better at representing processes in the climate system and thereby better at simulating regional climate including its variability and trends. It could also mean that they are better at accurately projecting future climate change.
In principle, added value in RCMs over coarse-scale GCMs stems from the RCMs better resolving land-sea contrasts, height of mountains and various important processes in the climate system such as mid-latitude cyclones and frontal systems. While this constitutes added value on a basic level it can be more difficult to actually measure it. In particular, as large-scale conditions from a global climate model can be flawed, a regional climate model cannot necessarily change this situation even if some features may be better represented (Sørland et al. 2018).
How can added value be measured?
Added value for the historical period can be found from comparing output from the regional and global climate model to observations. Added value implying better accuracy for projecting future climate change is more difficult to assess. One possibility is to investigate how regional climate models perform in other regions where the climate is warmer, which to some extent could be seen as a proxy for a future warmer climate. Another, to evaluate to what extent historical trends are captured.
In particular, for the EURO-CORDEX regional climate models used in C3S, studies have shown added value for areas of complex terrain (e.g. Torma et al. 2015) and for daily-scale variability, including high-intensity precipitation (e.g. Prein et al. 2015). Potential added value from future climate model projections with so called convection-permitting models in their representation of heavy precipitation on sub-daily time scales is also to be expected (Kendon et al 2014), but, it is noted that such models are currently not part of the regional climate projections provided by C3S.
References
Kendon, EJ, NM Roberts, HJ Fowler, MJ Roberts, SC Chan, and CA Senior (2014) Heavier summer downpours with climate change revealed by weather forecast resolution model, Nature Climate Change, 4, 570–576, doi:10.1038/nclimate2258
Prein AF, Gobiet A, Suklitsch M, Truhetz H, Awan NK, Keuler K and Georgievski G (2013) Added value of convection permitting seasonal simulations. Clim. Dyn., 41, 2655–2677, doi:10.1007/s00382-013-1744-6.
Sørland S, Lüthi D, Schär C and Kjellström E, (2018) Bias patterns and climate change signals in GCM-RCM model chains. Environ. Res. Lett., 13, 074017, DOI: 10.1088/1748-9326/aacc77.
Torma Cs, Giorgi F, and Coppola E (2015) Added value of regional climate modeling over areas characterized by complex terrain—Precipitation over the Alps, J. Geophys. Res. Atmos., 120, 3957– 3972. doi: 10.1002/2014JD022781.
Questions related to climate model output and data
Questions related to data outputs from the EURO-CORDEX RCMs. Which kind of data exist, which variables are stored, for which levels and at what time intervals
Click here to expand the questions
What is the climate model output representative of?
Climate models calculates the evolution of the state of the climate system over time in relatively short time steps (minutes to hours) on a spatial grid with grid spacing of tens to hundreds of kilometers.
In some more detail, climate models produce output data for a number of variables describing the state of the atmosphere, including temperature, humidity, air pressure and wind speed. These variables are calculated prognostically at a large number of vertical levels. In addition, a number of fluxes are calculated including fluxes of momentum, energy, humidity in the atmosphere and between the atmosphere and the surface that can be either ground or ocean. Also, a number of sub-grid scale features, including cloud properties, radiation processes, precipitation processes are also calculated. such as cloud droplet concentration and cloud water and cloud ice content. The amount of output data that is stored in the simulations may differ between different RCM groups. Most often, a minimum set of variables is stored in order to compare single RCM projections to others.
What RCM-data is available at the CDS?
Output data from the RCMs are usually stored at intervals of once per day or once per three hours.
For the EURO-CORDEX data provided through the CDS see: https://confluence.ecmwf.int/display/CKB/CORDEX%3A+Regional+climate+projections
For a list of available variables and time steps see: https://cds.climate.copernicus.eu/cdsapp#!/dataset/projections-cordex-domains-single-levels)
In the process of running climate models it can be decided which variables to store, and at what temporal resolution or vertical levels. The result is often a compromise between user requirements and the availability of storage space. In some cases, it is possible that also other variables, or data at higher temporal resolution or other vertical levels, are stored locally by the group that produced a particular RCM run. Such data may be available from the producers upon request.
Questions related to climate models
Questions relate to how different processes of the climate system are included in climate models and how these processes may be formulated. In this category there are also questions related to model resolution and need for computer power.
Click here to expand the questions
What is a climate model and how does it work?
A numerical climate model is a computer program building on mathematical descriptions of the relevant processes of the climate system. The model progressively calculates the evolution of the state of the climate system over long periods of time (e.g. decades to centuries) in short time steps (minutes to hours).
The models are formulated considering the governing physical processes for momentum, mass, energy and water conservation etc. These are represented by a set of differential equations that are solved for their time tendencies. These tendencies are subsequently added to the state of the system thereby generating a future state. From the future state new tendencies are calculated that, in turn, are used to derive yet another new state etc.
Regional climate models have been developed as a tool to improve horizontal resolution, and thereby representation of detailed regional and local processes. The EURO-CORDEX regional climate models operate on a computational grid covering parts of the North Atlantic and Europe. To run these regional models, information from the larger scale global climate system is taken as input from a global climate model at the lateral boundaries, typically 4-8 times every day. Also, sea-surface temperatures and sea-ice conditions to be used in the regional model are most often taken from the global climate model unless regional ocean models are included in the regional climate model system. In the case of coupled models, also the regional ocean model needs to take input from the global ocean model at its boundaries. For Europe, such coupled models exist for the Baltic Sea and the North Sea in northern Europe and for the Mediterranean in the south. Such coupled models are, however, not currently part of EURO-CORDEX, however some of them are available for the Med-CORDEX domain.
Which processes are included in a climate model?
Earth System Models (ESMs) include a large number of processes describing the atmosphere, ocean, cryosphere and biosphere and the interaction between them. Most CORDEX regional climate models are relatively simple in comparison focusing mainly on standard processes describing the atmosphere and its interaction with the land surface.
Development of fully coupled ESMs involving a wide range of components of the Earth system has been ongoing for several decades. Components of such models involve: atmospheric dynamics and physical processes, physical ocean models, glacier and land ice models, dynamic vegetation models, models of the biogeochemistry of the oceans, models of atmospheric aerosols and chemistry. ESMs can be run in different configurations depending on the questions to be addressed. Also, some regional climate models have been set up as regional ESMs, RESMs. Such RESMs are generally rare in a wider CORDEX perspective.
Users of regional climate change information should consider whether the RCMs include relevant processes for their purpose. General features for all CORDEX domains include: the relatively crude treatment of atmospheric aerosols in most RCMs and the inability to realistically simulate convective clouds/precipitation in these relatively coarse-scale models. Specifically, for different regions limitations in representing regional features like ocean areas may need attention. For instance, in Europe, the EURO-CORDEX RCMs generally do not include detailed treatment of the Baltic Sea and the Mediterranean, and as those are poorly resolved in GCMs this may have negative consequences on the results of the models.
Questions related to climate projections
Questions related to climate projections concern initialization and what are the data needed in the models. Transient climate simulations and time-slice experiments are also discussed as well as differences between numerical weather prediction and climate projections.
Click here to expand the questions
What is the difference between climate projections and a weather forecast?
A weather forecast starts from a known initial state that is as close to the actual state of the system as given by available observations. The weather forecast can provide relevant information for a few days to at most a few weeks. A climate projection starts from an arbitrary model state and can be run for decades to centuries.
After a few days to some week(s) the predictability of the weather vanishes and continuing a simulation with a weather prediction model will not give an accurate weather forecasts at these time horizons. At longer time scales of weeks to months, processes working on slower time scales can have an impact on the weather and even if single events cannot be forecasted it may be possible to say that it will be warmer or colder than normal, or wetter or drier than normal. Such forecasts are relying on slowly varying processes in the climate system. For instance, if a large-scale warm anomaly in sea surface temperatures is found in an area this will impact the region for a long time which can be simulated by a forecast model. A climate projection, on the other hand, is not useful for predicting the actual state of the system (weather) for the first weeks of a simulation as the initial state is not close to the observed state of the climate system. However, the climate model can be integrated for long time periods and will generate possible weather situations that resemble those in the real climate system. Assessed over long-enough times this would represent the climate and its variability even if the exact evolution over time can’t be compared to the actual evolution of the climate system in a meaningful way.
Any users of forecast data (seasonal to decadal) are not served by the current EURO-CORDEX data.
Users of climate projections must be aware that direct comparisons of time series of weather events between model and the real atmosphere are not meaningful because climate model can provide information on weather information only in statistical sense over a long period (typically 30-years) of time.
How do you start a climate change projection and why is this important?
A simulation with a climate model requires an initial state for each individual grid cells in the model (atmosphere, ocean, land surface). Given this state the exact evolution over time will differ between different simulations. This implies that two simulations starting from slightly different initial conditions in the 19th century will differ in their details also in scenarios for the 21st century.
Most climate model projections in global climate models start in pre-industrial time, which is often taken to be sometimes in the second half of the 19th century. For this time period (or any other historical time period) there are not enough observations to start a model from a purely observational state. Instead, an arbitrary state is taken from a long-term global climate model integration with the same climate model representing pre-industrial conditions.
Important in a climate model integration is that the model starts from a spun-up state to avoid spurious drift in the system that would result from an integration starting from a state far away from equilibrium for the model. For global climate models, notably the slow time scales of the ocean imply that spin-up times of several centuries are needed. This is achieved in the pre-industrial control runs from which initial states for climate projections are taken. For regional models not involving the global ocean circulation, such as the EURO-CORDEX models, climate processes are acting at shorter time scales. Depending on area, slower processes involving soil moisture may, still, call for longer spin-up times of several years.
What determines the evolution of a climate projection?
The exact evolution of the climate in a climate projection is governed by the initial state of the system, of the imposed forcing conditions and of how the climate model responds to the forcing.
As the initial state in a global climate model differs from the real state of the climate system there will necessarily be a mismatch between various modes of variability in the real system and in the climate integration. This means, for instance, that El Nino episodes or phases with stronger or weaker westerlies over the North Atlantic (strong or weak North Atlantic Oscillation conditions) may be out of phase. Such discrepancies can influence climate at scales covering years to decades and is one important source of uncertainty in climate change projections (e.g. Hawkins and Sutton, 2009).
Forcing conditions are imposed using, first, historical forcing conditions as reconstructed until a certain point in time, when scenarios for future climate forcing comes into place. For the RCP-scenarios used in CORDEX, the historical forcing is applied until 2005. From 2006 and onwards, future scenarios are used for providing forcing conditions. For the newer SSP-RCP-scenarios that are used in CMIP6, the transition takes place at the shift from 2015 to 2016.
Summarizing for a CORDEX regional climate model projection, the evolution is determined by:
- The initial conditions in the underlying global climate model, typically starting in 1850.
- The initial conditions in the regional model starting around 1960.
- The forcing conditions from the global climate model imposed on the boundaries.
- The local forcing conditions, including changes in greenhouse gases, aerosols and land-use, imposed in the RCM.
- How the RCM responds to these changes in forcing.
Note that this list is not in order of importance but rather in a sequential way starting from the GCM simulation ending up at the regional and local scale as simulated by the RCM.
References
Hawkins E and Sutton R (2009) The potential to narrow uncertainty in regional climate predictions Bulletin of the American Meteorological Society, 90, pp. 1095-1107, 10.1175/2009BAMS2607.1
Questions related to CMIP and CORDEX
Questions related to the existing networks of climate model cooperations and which climate model projections to use.
Click here to expand the questions
What is CMIP?
CMIP stands for the Coupled Model Intercomparison Project and is a joint global collaboration between climate modelers organized under WCRP (the World Climate Research Program) of WMO (World Meteorological Organization).
CMIP is now in its sixth phase (Eyring et al., 2016) and results from CMIP6 have been assessed in the IPCC AR6 reports. Similarly, CMIP5 (Taylor et al., 2012) was assessed in the IPCC AR5 reports.
Results from CMIP5/CMIP6 can be accessed in the C3S Climate Data Store (CDS) at https://cds.climate.copernicus.eu/#!/search?text=cmip5 and https://cds.climate.copernicus.eu/cdsapp#!/dataset/projections-cmip6
What is CORDEX?
CORDEX stands for the Coordinated Regional climate Downscaling EXperiment and is a joint global collaboration under the WCRP of WMO.
The CORDEX vision is to advance and coordinate the science and application of regional climate downscaling through global partnerships.
For Europe, in particular, CORDEX is organized in EURO-CORDEX (https://euro-cordex.net/) and Med-CORDEX (https://www.medcordex.eu/).
Results from CORDEX (all 14 CORDEX domains) can be accessed in the CDS at Copernicus Climate Data Store | Copernicus Climate Data Store
Which data should I use, CMIP-data or CORDEX-data?
The answer strongly depends on the question that you want to answer. The CORDEX-data adds value to the CMIP-data as the regional climate models are run at higher horizontal resolution. This can be of particular importance for assessing changes in extremes. CMIP-data, on the other hand, represent a more comprehensive data set in the number of projections for the future. This means that CMIP-data may be better suited for assessing what is robust and what is more uncertain related to future climate change. Ideally, both data sets should be used simultaneously to address both questions.
References
Eyring V, Bony S, Meehl GA, Senior CA, Stevens B, Stouffer RJ and Taylor KE (2016) Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization Geosci. Model Dev. 9, 1937–58.
Taylor KE, Stouffer RJ and Meehl GA (2012) An overview of CMIP5 and the experiment design Bull. Am. Meteor. Soc. 93 485–98
Questions related to downscaling techniques
Questions related to dynamical and statistical downscaling techniques.
Click here to expand the questions
What is downscaling?
To increase resolution in climate model output dynamical or statistical downscaling can be applied to include information representative of finer scales than those represented by the model. This is in contrast to simple interpolation that, despite providing more high-resolution data, does not add such additional fine-scale information.
Global climate models typically provide output at resolution of about 100-200 km. This may not be adequate for various impact studies at local scales. Consequently, there may be a need for further refining this information to local scales. Simple interpolation can provide information at any resolution but does not add additional information on finer scales. As an example, a small-scale (with respect to the climate model) mountain, will in the real climate system show different climate characteristics compared to its low-lying surroundings. In a coarse-scale model where this mountain is not present, interpolation between two surrounding low-level grid points will not give information that is adequate for the high-altitude mountain. In statistical and dynamical downscaling, such information can be added to provide more realistic information also at the local site.
Why is there a need for downscaling?
The coarse resolution of global climate models implies that they have a relatively poor representation of local features such as land-sea distribution and height of mountains. Also, relevant processes of the climate system, such as low-pressure systems and associated areas of precipitation implies that such models are not always fit for purpose for representing regional or local features. This is particularly problematic for some high-impact events such as high-intensity precipitation and wind storms.
GCMs provide good information about climate characteristics on large regional scales. This includes differences between continents and oceans and between high and low latitudes in general. But, also, they can provide information about climate details in parts of a continent like southern and northern Europe etc. However, at a typical GCM-resolution of 100-200 km grid spacing, complex features of the European geography such as the Mediterranean and Baltic Sea, are not well resolved. Also, mountain chains are far too low implying that their role as barriers for precipitation are often underestimated. Moreover, high-intensity events on daily and sub-daily time scales are poorly represented by such coarse models.
Techniques for downscaling global model information, either dynamically through regional limited area models, or through empirical statistical methods, have been developed for this purpose.
What is dynamical downscaling? What is required to run a regional climate model?
Dynamical downscaling with limited-area models is principally very similar to the global climate models as they are numerical models of the climate system. The key difference is that they are applied at higher horizontal resolution for a limited area taking lateral boundary conditions from the global model as input.
In addition to lateral boundary conditions for the atmosphere from the global model, regional climate models mostly need sea-surface temperatures and sea-ice conditions. Exceptions are if they are applied in continental areas or if the regional models also include modules for treating the ocean and sea ice. Such regional earth system models have been set up for the Mediterranean Sea and for the Baltic Sea and the North Sea albeit the actual number of existing simulations is still low.
The EURO-CORDEX projections provided by C3S RCMs have not been run in coupled mode, implying that sea-surface conditions are taken directly from the underlying global model. This also applies for most other CORDEX domains. For the Med-CORDEX domain, however, there exists also simulations in which the Mediterranean Sea has been explicitly simulated in a regional ocean model coupled to the atmosphere.
Depending on what the data should be used for the relevance of sea surface temperature and sea-ice conditions should be carefully considered.
What is empirical or statistical downscaling?
Empirical downscaling implies that large-scale climate information, for instance from a climate model, is downscaled with help from observations. Statistical relationships between the local scale observation and the large-scale model fields are identified for the historical time period. For future climate change the same empirical relationships are assumed to be valid which may not always be the case.
As empirical downscaling relies on observations it is required that such observations exist and that they are representative for the scales that should be addressed. A complication with the empirical downscaling methods is that downscaling is most often done one variable at the time. An implication is that time series of several variables may not be completely consistent.
It is also noted that the assumption that the empirical relationships are the same also in the future climate may not always be true. Examples when this assumption may be violated include areas where non-linear changes are seen. This could for instance involve areas with retreating snow cover where temperature distributions changes differently in their different parts (e.g. Kjellström, 2004). Another example would be areas where strong drying is seen that would exacerbate warming and thereby further increase the response.
It is recommended that users of empirically downscaled information carefully consider if such potential problems may have any implication on their applications.
What are advantages/disadvantages with dynamical/statistical downscaling?
A major advantage with empirical downscaling over dynamical downscaling it can be relatively easily used for downscaling large ensembles of climate model data requiring only limited computational capacity. The key drawback is that the empirical methods assumes that future relationship between local and large scales remain the same as in the historical climate. Dynamical downscaling, on the other hand, allows for such changes over time. Furthermore, regional models provide internally consistent climate states implying that several variables from the model can be used simultaneously.
References
Kjellström, E., 2004. Recent and future signatures of climate change in Europe. Ambio, 33(4-5), 193-198.
Questions related to ensemble design and model selection
Questions on how the GCM-RCM-RCP multi-model ensemble was designed in EURO-CORDEX and how the addition of simulations contributed by the C3S was chosen. Moreover questions related to how single or several models (or simulations) can be chosen from the large ensemble for various purposes are answered.
Click here to expand the questions
How are GCM-RCM ensembles designed?
Traditionally, GCM-RCM ensembles have been produced on an ad hoc basis where RCMs have been downscaling a number of GCMs that have supplied boundary conditions. This implies that GCM-RCM ensembles most often do not fully represent the full spread in larger GCM ensembles such as the full CMIP5 and CMIP6 ensembles.
Future climate change is uncertain as future forcing conditions are not fully known. This uncertainty can be handled by producing simulations for different scenarios. Another source of uncertainty relates to the sensitivity of the climate system, i.e. how strong it reacts to changing forcing conditions. This response can be different in different regions. In addition, internal natural variability also contributes to the uncertainty as it may be that single years, decades or even longer periods can be warmer/colder/wetter/drier than average just by chance.
Preferably, GCM-RCM ensembles could be designed in such a way as to sample as much of the spread as possible. Notably, this requires that lateral boundary conditions are provided by all GCMs. Users of RCM projections should be aware that these usually do not represent the full spread of the available GCM ensembles and that assessments of uncertainty and robustness of results may differ.
What considerations are behind the design of the C3S contribution to EURO-CORDEX?
A decision was made to build on already existing and published EURO-CORDEX simulations at 0.11° horizontal resolution with the rationale of producing a GCM-RCM-RCP matrix as large as possible. The strategy developed was such that a number of different “dimensions” can be explored with the resulting data including sensitivity to forcing scenario, choice of GCM and choice of RCM. In addition, a number of simulations were run to make it possible to address natural variability.
Due to resource limitations it was early on decided that C3S-funded regional simulations would not do a homogeneous non-complete filling of the entire GCM-RCM-RCP matrix (several hundreds), but would rather fill up selected matrix slices (e.g. one GCM-all RCMs-one RCP, many GCMs-a few RCMs-one RCP, a few GCMs-a few RCMs-all RCPs). The plan for the collective simulations produced within C3S has therefore been to distribute the allocated effort to approximately 65 simulations (the new simulations therefore roughly speaking double the available EURO-CORDEX simulations produced within other efforts) between three main goals:
- To enable studies of variability by performing several ensemble member downscaling simulations with the same GCM-RCM-RCP combination.
- To fill as large sub-matrix as possible for the RCP8.5 emission scenario.
- To add a significant amount of simulations to the weak RCP2.6 scenario, where noise is relatively large compared to the climate change signal.
How can subsets of simulations be chosen in an optimal way?
As future climate change is uncertain it is important to address this uncertainty in any study of future conditions. A strategy for model selection can be to select single models in a way that preserves as much of the initial spread in the ensemble as possible. Another strategy may be to select only a few that are representative for different ways of evolution in a story-telling sense.
Depending on which climate change impact is studied careful consideration should be given to how any subset is chosen.
Questions related to evaluation of climate models
Questions on observations and how these may be used to evaluate climate models. If direct observations are lacking, other types of data may be used – such as reanalysis products.
Click here to expand the questions
How can climate models and downscaling techniques be evaluated?
Climate models are evaluated in their capacity to simulate the recent past climate and the evolution of the historical climate. Output from the climate models are compared to observations. Large discrepancies need to be explained and ideally lead to the development of improved model versions. As climate models are very similar to numerical weather prediction models, evaluation can benefit from such models that are routinely applied and evaluated on a daily basis.
In particular for regional climate models, evaluation is performed in two steps. In the first step, the RCM is evaluated in experiments when it is downscaling reanalysis data. In those, so called, hindcast experiments, the large-scale evolution of the climate system, including interannual variability, is as close to the observed state as it is possible. Consequently, model output can be more directly compared with observations, also in a time-series mode (i.e. the sequence of weather events should be realistic).
In the second step, each individual climate change projection can be evaluated for how well the historical climate is captured. An example is given for the EURO-CORDEX area in Vautard et al. (2020). In this case, it is essential that the underlying global climate model is evaluated as its representation of the large-scale conditions is key to the performance of the regional climate model. Also empirical downscaling methods need to be evaluated. A common approach is to perform cross-validation where data is split in two sets. One set is then used to find the empirical relationship, and the other set is used to evaluate it. Different ways to evaluate RCMs and statistical downscaling techniques are discussed in Maraun et al. (2015). For these types of evaluations, when downscaling of GCM output is considered, it is not meaningful to compare the actual time evolution of the climate with observations due to the internal variability in the climate system (i.e the sequence of weather events is not realistic and they can be reliable only in statistical sense for a longer time – typically 30 years – period).
It is recommended that users of downscaled climate information are aware of to what degree the downscaling methods can provide data that are of good enough quality for their respective questions.
What data can be used for evaluating climate models?
A wide range of observational data are used for evaluating climate models. This includes both in-situ measurements of near-surface variables as well as data from the atmosphere and oceans and remote sensing data from satellites and weather radars. It also includes so called reanalysis data that is a composite of observations as analysed by a weather prediction model.
Various data sets are used for model evaluation in different regions. This includes both global datasets and more detailed regional datasets. Many such datasets consist of information from near-surface observations that have been gridded but it could also be data from satellites. In addition to observations and remote sensing data, also reanalysis data are extensively used. These data are a blend of observations and weather forecast models. Particularly, for the evaluation of the EURO-CORDEX regional climate models a strong emphasize has been on the gridded data from E-OBS with daily data for temperature and precipitation for Europe extending back to 1950 (Cornes et al. 2018).
Users of EURO-CORDEX regional climate models should be aware that the observational material used for evaluation of regional climate models differ between different areas in Europe. For some areas, notably areas of complex topography, there are limitations in these data and, consequently, the skill of the climate models may not be fully known.
References
Cornes R, van der Schrier G, van den Besselaar EJM and Jones PD (2018) An Ensemble Version of the E-OBS Temperature and Precipitation Datasets, J. Geophys. Res. Atmos., 123. doi:10.1029/2017JD028200
Maraun D, Widmann M, Gutiérrez JM, Kotlarski S, Chandler RE, Hertig E, Wibig J, Huth R and Wilcke RAI (2015), VALUE: A framework to validate downscaling approaches for climate change studies. Earth’s Future, 3: 1–14., https://doi.org/10.1002/2014EF000259
Vautard R, Kadygrov N, Iles C, Boberg F, Buonomo E, Bülow K, Coppola E, Corre L, van Meijgaard E, Nogherotto R, Sandstad M, Schwingshackl C, Somot S, Aalbers E, Christensen OB, Ciarlo JM, Demory M-E, Giorgi F, Jacob D, Jones RG, Keuler K, Kjellström E, Lenderink G, Levavasseur G, Nikulin G, Sillmann J, Solidoro C, Sørland SL, Steger C, Teichmann C, Warrach-Sagi K and Wulfmeyer V (2020) Evaluation of the large EURO-CORDEX regional climate model ensemble. J. Geophys. Res. DOI: 10.1029/2019JD032344
Questions related to forcing scenarios
Questions related to components and mechanisms influencing the climate such as greenhouse gas content of the atmosphere, aerosol particle concentrations, land-use changes and how these are incorporated into the climate models. Different generations of scenarios and how comparable these are also part of this topic.
Click here to expand the questions
Are future forcing conditions considered similarly in regional climate models as in global climate models?
Consistency facilitates interpretation of the results if forcing is the same in global and regional climate models and if it is treated in a similar way. In practice, however, details in the models differ and thereby forcing is often treated differently. This applies also for CORDEX regional climate model projections provided by the C3S.
Examples of differences include different radiation schemes that result in different treatment of both greenhouse gases and aerosols. For instance, some regional climate models have simplified radiation schemes where several greenhouse gases are lumped together and treated as CO2-equivalents instead of separately calculating the impacts of the different greenhouse gases (e.g. CO2, CH4, N2O, etc.). Further, many CORDEX regional climate models have a more simplified treatment of aerosols compared to their driving global models and thereby changing aerosol concentrations are not always included in the RCMs. Consequently, this may lead to different results as shown for Europe by e.g. Boé et al (2020). Land-use is another feature that may differ between GCMs and RCMs and in most of the EURO-CORDEX RCM projections land-use is not evolving with time. Further information about the details of some of the RCMs can be found at https://search.es-doc.org/?project=cordexp.
It is suggested that users of regional climate information carefully evaluate whether such differences in forcing are of relevance for their specific questions.
Are the scenarios from SRES similar to the RCP scenarios? And what about RCP scenarios, how do they differ from the SSP-RCPs?
Scenarios from different generations differ and cannot always easily be compared. The RCP scenarios used in CORDEX and CMIP5 differ both from the older SRES emission scenarios and from the newer SSP-RCP scenarios used in CMIP6.
Emission scenarios, such as those provided by SRES (Special Report on Emission Scenarios), or forcing scenarios such as the RCP (Representation Concentration Pathways) scenarios provide forcing data for the climate models. Such data includes long-term evolution of concentrations of a number of greenhouse gases (e.g. CO2, CH4, N2O, CFCs, SF6), different aerosol components (e.g. sulfate, nitrate, soot, dust, sea salt) and changes in land-use. For a more elaborate description of the SRES and RCP scenarios see the Appendix of the EURO-CORDEX User Guidance document https://www.euro-cordex.net/imperia/md/content/csc/cordex/guidance_for_euro-cordex_climate_projections_data_use_2021-02_1.pdf.
More details about the three generation of scenarios are available in the following reports and scientific papers
- SRES: https://www.ipcc.ch/site/assets/uploads/2018/03/sres-en.pdf
- RCP: The representative concentration pathways: an overview | Climatic Change,
- SSP: https://www.sciencedirect.com/science/article/pii/S0959378016300681
As a way of facilitating comparison of climate model projections with forcing scenarios from different generations, comparisons can be done at certain warming levels such as +2°C warming relative to pre-industrial conditions (Kjellström et al., 2018). By such an approach the specifics of the individual scenarios are of less importance. A caveat though is that regional forcing and response to changes may be different at different time periods despite a similar global mean response (Bärring and Strandberg, 2018).
It is suggested that users of regional climate information carefully evaluate to what extent the forcing scenarios are of relevance for their specific questions.
Which scenarios have been used in EURO-CORDEX?
EURO-CORDEX has focused on the RCP scenarios. RCP2.6, RCP4.5 and RCP8.5, where the numbers relate to the forcing in W/m2 at the end of the 21st century relative to that of pre-industrial conditions.
For a detailed picture of which forcing scenarios for which GCM-RCM data exist in the CDS, see CDS documentation and catalogue entries.
Documentations:
CMIP5: https://confluence.ecmwf.int/display/CKB/CMIP%3A+Global+climate+projections
CMIP6: https://confluence.ecmwf.int/display/CKB/CMIP6%3A+Global+climate+projections
CORDEX: https://confluence.ecmwf.int/display/CKB/CORDEX%3A+Regional+climate+projections
CDS catalogue entries:
CMIP5:
https://cds.climate.copernicus.eu/cdsapp#!/dataset/projections-cmip5-monthly-single-levels
https://cds.climate.copernicus.eu/cdsapp#!/dataset/projections-cmip5-monthly-pressure-levels
https://cds.climate.copernicus.eu/cdsapp#!/dataset/projections-cmip5-daily-single-levels
https://cds.climate.copernicus.eu/cdsapp#!/dataset/projections-cmip5-daily-pressure-levels
CMIP6: https://cds.climate.copernicus.eu/cdsapp#!/dataset/projections-cmip6
CORDEX: https://cds.climate.copernicus.eu/cdsapp#!/dataset/projections-cordex-domains-single-levels
For future application when EURO-CORDEX may have generated CMIP6-based scenarios it is noted that the new SSP-RCP scenarios differ from the RCPs even if the nominal forcing levels at the end of the century (indicated by the numbers, e.g. 8.5 (W/m-2) in SSP5-8.5 vs RCP8.5) are the same. This results from the fact that the specific pathways to 2100 differ in terms of how the different greenhouse gases evolve over time. For more information on differences between RCPs and SSP-RCPs see for instance Tebaldi et al. (2021).
References
Boé J, Somot S, Corre L and Nabat P (2020) Large discrepancies in summer climate change over Europe as projected by global and regional climate models: causes and consequences. Clim Dyn 54, 2981–3002. Large discrepancies in summer climate change over Europe as projected by global and regional climate models: causes and consequences | Climate Dynamics
Bärring L and Strandberg G (2018) Does the projected pathway to global warming targets matter? Environmental Research Letters, 13 (2018), 10.1088/1748-9326/aa9f72
Kjellström E, Nikulin G, Strandberg G, Christensen OB, Jacob D, Keuler K, Lenderink G, van Meijgaard E, Schär C, Somot S, Sørland SL, Teichmann C, and Vautard R (2018) European climate change at global mean temperature increases of 1.5 and 2 °C above pre-industrial conditions as simulated by the EURO-CORDEX regional climate models, Earth Syst. Dynam., 9, 459-478, DOI: 10.5194/esd-9-459-2018.
Tebaldi C, Debeire K, Eyring V, Fischer E, Fyfe J, Friedlingstein P, Knutti R, Lowe J, O’Neill B, Sanderson B, van Vuuren D, Riahi K, Meinshausen M, Nicholls Z, Tokarska KB, Hurtt G, Kriegler E, Lamarque J-F, Meehl G, Moss R, Bauer SE, Boucher O, Brovkin V, Byun Y-H, Dix M, Gualdi S, Guo H, John JG, Kharin S, Kim Y, Koshiro T, Ma L, Olivié D, Panickal S, Qiao F, Rong X, Rosenbloom N, Schupfner M, Séférian R, Sellar A, Semmler T, Shi X, Song Z, Steger C, Stouffer R, Swart N, Tachiiri K, Tang Q, Tatebe H, Voldoire A, Volodin E, Wyser K, Xin X, Yang S, Yu Y and Ziehn T, (2021) Climate model projections from the Scenario Model Intercomparison Project (ScenarioMIP) of CMIP6, Earth Syst. Dynam., 12, 253–293, https://doi.org/10.5194/esd-12-253-2021.
Questions related to natural internal climate variability
Questions related to natural variability. What it is and how it can be addressed.
Click here to expand the questions
What is natural internal climate variability and why is it important?
A range of processes in the climate system, including interaction between its different parts, give raise to a number of long-term variations that occur naturally without any external forcing. Depending on the phases and future variations in these, and other variations, climate features vary over years and decades. As climate projections are generally not in phase with the observed climate such natural variability can impose relatively large differences between climate projections and the actual observed climate. This is particularly a problem for shorter time periods and in a regional to local perspective. Evaluating long time periods can partially circumvent this problem and is a reason for why climate scientists often refer to 30-year periods or more when discussing climate change.
As the initial state in a global climate model differs from the real state of the climate system there will necessarily be a mismatch between various modes of variability in the real system and in the climate integration. Such modes of variability include, for instance, changes in sea surface conditions in the central Pacific Ocean associated with El Nino and La Nina episodes, or phases with stronger or weaker NAO (North Atlantic Oscillation) indices, of strong relevance for the European climate. Mismatches like this can influence the performance of climate models at scales covering years to decades and are also important sources of uncertainty in climate change projections, especially in a regional to local perspective (e.g. Hawkins and Sutton, 2009).
For European climate change this means that for some variables – for which natural variability is large – long-term trends may not be detected until far into the future. For others, with relatively smaller variability, and larger signal, trends may already be discernable (e.g. Kjellström et al., 2013).
How can internal variability be handled?
The best tool for assessing the internal variability is to use large ensembles of climate model projections. By this approach, the probability of having some, or several, projections close to the actual evolution increases.
Similarly, as in weather prediction, large initial-value single model ensembles have been put forward from climate modelling groups during the last decades. Recently, a number of such large ensembles have been assessed by Deser et al. (2020). As pointed out in that study it can be important to realize that the actual evolution of the climate system is just one realization. Consequently, it may be difficult to interpret climate change aspects as result from changing forcing conditions based only on observations as what is observed may be a feature of such variability.
Most CORDEX RCMs have been used to downscale single member GCM simulations. However, for EURO-CORDEX some models have been downscaling three different ensemble members (Documentation for the role of internal variability over Europe). There are also a few examples of regional climate models that have been used for downscaling large ensembles (Addor and Fischer 2015; Aalbers et al. 2016).
References
Aalbers EE, Lenderink G, van Meijgaard E, and van den Hurk B (2016) To what extent is climate change detection at the local scale ‘clouded’ by internal variability? In EGU General Assembly Conference Abstracts. 18, s. 10121.
Addor N and Fischer EM (2015) The influence of natural variability and interpolation errors on bias characterization in RCM simulations. J. Geophys. Res. Atmos. 120. doi:10.1002/2014JD022824.
Deser C, Lehner F, Rodgers KB et al. (2020) Insights from Earth system model initial-condition large ensembles and future prospects. Nat. Clim. Chang. 10, 277–286, Insights from Earth system model initial-condition large ensembles and future prospects | Nature Climate Change
Hawkins E and Sutton R (2009) The potential to narrow uncertainty in regional climate predictions Bulletin of the American Meteorological Society, 90, pp. 1095-1107, 10.1175/2009BAMS2607.1
Kjellström E, Thejll P, Rummukainen M, Christensen JH, Boberg F, Christensen OB, Fox Maule C (2013) Emerging regional climate change signals for Europe under varying large-scale circulation conditions, Clim. Res., 56, 103–119, DOI: 10.3354/cr01146.
Questions related to representation of extremes
Questions are related to how well extremes are captured in global and regional climate models.
Click here to expand the questions
How can the representation of extremes by climate models be evaluated?
The nature of extremes, that are rare and rarer the more extreme they are, implies that it is fundamentally difficult to: i) evaluate model performance, and ii) assess future changes in such extreme events.
These issues are more pronounced the longer the recurrence time (i.e. the average time it would take before an event is repeated) of interest is. For example, with recurrence times of several hundred years or more, observations to which the models can be compared are rare. For less extreme events, such as a one in a ten-, or one in a twenty-year event, the situation is better, and EURO-CORDEX models have been extensively evaluated for such aspects. Also, assessing the extent of changes in extremes in a future warmer climate is more challenging the rarer the events are. This can be partly treated with more data to assess, in order to get more robust assessments. A consequence of this is that large ensembles of climate projections are needed.
For extreme precipitation events on sub-daily time scales related to convection we note that standard regional climate models such as the EURO-CORDEX models are not well-suited. For this purpose, very high-resolution convection-permitting models has been suggested as a better option (e.g. Prein et al., 2013; Kendon et al., 2014).
Users of EURO-CORDEX data that have particular interest in extremes should carefully investigate to what extent the models can represent these extremes. A number of scientific papers have addressed extremes in these models, e.g. Vautard et al. (2020) and Coppola et al. (2021).
What are compound events and how can climate models represent them?
Another examples of extreme events that may have strong impacts are so called compound events, when two or more phenomena occur at the same time. This can be a low-pressure system bringing high sea-level fluctuations on a coast where a river at the same time see a strong discharge due to strong rainfall in the preceding week. Even if the compound resulting event is extreme it is not necessary that any of these two events are extreme in itself.
Compound events can also relate to a series of repeated events that may not be extreme one at the time, but, where the long period of persistent conditions may be extreme. The warm summer in northern Europe 2018 was an example of such an extreme when repeated heat waves and high-pressure dominated summer weather without precipitation led to severe drought and the warmest summer on record in Sweden (Wilcke et al., 2021). Such events, where several extremes need to take place more or less in the right time in relation to each other, are rarer compared to the individual extremes one at the time.
Due to the rarity of compound events and limited amount of data it is difficult to evaluate how models can represent such events.
References
Coppola E, Nogherotto R, Ciarlo JM, Giorgi F, Somot S, Nabat P, Corre L, Christensen OB, Boberg F, van Meijgaard E, Aalbers E, Lenderink G, Schwingshackl C, Sandstad M, Sillmann J, Bülow K, Teichmann C, Iles C, Kadygrov N, Vautard R, Levavasseur G, Sørland SL, Demory M-E, Kjellström E and Nikulin G (2021) Assessment of the European climate projections as simulated by the large EURO- CORDEX regional climate model ensemble. J. Geophys. Res.: Atmospheres, 126, e2019JD032356, DOI: 10.1029/2019JD032356
Kendon, EJ, NM Roberts, HJ Fowler, MJ Roberts, SC Chan, and CA Senior (2014) Heavier summer downpours with climate change revealed by weather forecast resolution model, Nature Climate Change, 4, 570–576, doi:10.1038/nclimate2258
Prein AF, Gobiet A, Suklitsch M, Truhetz H, Awan NK, Keuler K and Georgievski G (2013) Added value of convection permitting seasonal simulations. Clim. Dyn., 41, 2655–2677, doi:10.1007/s00382-013-1744-6.
Vautard R, Kadygrov N, Iles C, Boberg F, Buonomo E, Bülow K, Coppola E, Corre L, van Meijgaard E, Nogherotto R, Sandstad M, Schwingshackl C, Somot S, Aalbers E, Christensen OB, Ciarlo JM, Demory M-E, Giorgi F, Jacob D, Jones RG, Keuler K, Kjellström E, Lenderink G, Levavasseur G, Nikulin G, Sillmann J, Solidoro C, Sørland SL, Steger C, Teichmann C, Warrach-Sagi K and Wulfmeyer V (2020) Evaluation of the large EURO-CORDEX regional climate model ensemble. J. Geophys. Res. DOI: 10.1029/2019JD032344
Wilcke R, Kjellström E, Liu C, Matei K and Moberg A (2020) The extreme warm summer 2018 in Sweden - set in a historical context. Earth System Dynamics, 11, 1107-1121. DOI: 10.5194/esd-11-1107-2020.