Here we present the first multi-model ensemble of regional climate simulations at kilometer-scale horizontal grid spacing over a decade long period. A total of 23 simulations run with a horizontal grid spacing of $$\sim $$∼ 3 km, driven by ERA-Interim reanalysis, and performed by 22 European research groups are analysed. Six different regional climate models (RCMs) are represented in the ensemble. The simulations are compared against available high-resolution precipitation observations and coarse resolution ( $$\sim $$∼ 12 km) RCMs with parameterized convection. The model simulations and observations are compared with respect to mean precipitation, precipitation intensity and frequency, and heavy precipitation on daily and hourly timescales in different seasons. The results show that kilometer-scale models produce a more realistic representation of precipitation than the coarse resolution RCMs. The most significant improvements are found for heavy precipitation and precipitation frequency on both daily and hourly time scales in the summer season. In general, kilometer-scale models tend to produce more intense precipitation and reduced wet-hour frequency compared to coarse resolution models. On average, the multi-model mean shows a reduction of bias from $$\sim \,$$∼ −40% at 12 km to $$\sim \,$$∼ −3% at 3 km for heavy hourly precipitation in summer. Furthermore, the uncertainty ranges i.e. the variability between the models for wet hour frequency is reduced by half with the use of kilometer-scale models. Although differences between the model simulations at the kilometer-scale and observations still exist, it is evident that these simulations are superior to the coarse-resolution RCM simulations in the representing precipitation in the present-day climate, and thus offer a promising way forward for investigations of climate and climate change at local to regional scales.

The transferability of the regional climate model REMO with a standard setup over different regions of the world has been evaluated. The study is based on the idea that the modeling parameters and parameterizations in a regional climate model should be robust to adequately simulate the major climatic characteristic of different regions around the globe. If a model is not able to do that, there might be a chance of an “overtuning” to the “home-region”, which means that the model physics are tuned in a way that it might cover some more fundamental errors, e.g., in the dynamics. All simulations carried out in this study contribute to the joint effort by the international regional downscaling community called COordinated Regional climate Downscaling EXperiment (CORDEX). REMO has been integrated over six CORDEX domains forced with the so-called perfect boundary conditions obtained from the global reanalysis dataset ERA-Interim for the period 1989 to 2008. These six domains include Africa, Europe, North America, South America, West Asia and the Mediterranean region. Each of the six simulations was conducted with the identical model setup which allows investigating the transferability of a single model to regions with substantially different climate characteristics. For the consistent evaluation over the different domains, a new evaluation framework is presented by combining the Köppen-Trewartha climate classification with temperature-precipitation relationship plots and a probability density function (PDF) skill score method. The evaluation of the spatial and temporal characteristics of simulated precipitation and temperature, in comparison to observational datasets, shows that REMO is able to simulate the mean annual climatic features over all the domains quite reasonably, but still some biases remain. The regions over the Amazon and near the coast of major upwelling regions have a significant warm bias. Wet and dry biases appear over the mountainous regions and East Africa, respectively. The temperature over South America and precipitation over the tundra and highland climate of West Asia are misrepresented. The probable causes leading to these biases are discussed and ideas for improvements are suggested. The annual cycle of precipitation and temperature of major catchments in each domain are also well represented by REMO. The model has performed well in simulating the inter- and intra-seasonal characteristics of different climate types in different regions. Moreover, the model has a high ability in representing the general characteristics of different climate types as measured by the probability density function (PDF) skill score method. Although REMO seems to perform best over its home domain in Europe (domain of development and testing), the model has simulated quite well the climate characteristics of other regions with the same set of parameterization options. Therefore, these results lead us to the conclusion that REMO is well suited for long-term climate change simulations to examine projected future changes in all these regions.

It is well established that Africa is particularly exposed to climate extremes including heat waves, droughts, and intense rainfall events. How exposed Africa is to the co‐occurrence of these events is however virtually unknown. This study provides the first analysis of projected changes in the co‐occurrence of five such compound climate extremes in Africa, under a low (RCP2.6) and high (RCP8.5) emissions scenario. These changes are combined with population projections for a low (SSP1) and high (SSP3) population growth scenario, in order to provide estimates of the number of people that may be exposed to such events at the end of the 21st century. We make use of an ensemble of regional climate projections from the Coordinated Output for Regional Evaluations (CORE) project embedded in the Coordinated Regional Climate Downscaling Experiment (CORDEX) framework. This ensemble comprises five different Earth System Model/Regional Climate Model (ESM/RCM) combinations with three different ESMs and two RCMs. We show that all five compound climate extremes will increase in frequency, with changes being greater under RCP8.5 than RCP2.6. Moreover, populations exposed to these changes are greater under RCP8.5/SSP3, than RCP2.6/SSP1, increasing by 47‐ and 12‐fold, respectively, compared to the present‐day. Regions of Africa that are particularly exposed are West Africa, Central‐East Africa, and Northeast and Southeast Africa. Increased exposure is mainly driven by the interaction between climate and population growth, and the effect of population alone. This has important policy implications in relation to climate mitigation and adaptation.

Global and regional climate model simulations are frequently used for regional climate change assessments and in climate impact modeling studies. To reflect the inherent and methodological uncertainties in climate modeling, the assessment of regional climate change requires ensemble simulations from different global and regional climate model combinations. To interpret the spread of simulated results, it is useful to understand how the climate change signal is modified in the GCM-RCM modelmodelgeneral circulation model-regional climate model (GCM-RCM) chain. This kind of information can also be useful for impact modelers; for the process of experiment design and when interpreting model results. In this study, we investigate how the simulated historical and future climate of the Max-Planck-Institute earth system model (MPI-ESM) is modified by dynamic downscaling with the regional model REMO in different world regions. The historical climate simulations for 1950–2005 are driven by observed anthropogenic forcing. The climate projections are driven by projected anthropogenic forcing according to different Representative Concentration Pathways (RCPs). The global simulations are downscaled with REMO over the Coordinated Regional Climate Downscaling Experiment (CORDEX) domains Africa, Europe, South America and West Asia from 2006–2100. This unique set of simulations allows for climate type specific analysis across multiple world regions and for multi-scenarios. We used a classification of climate types by Köppen-Trewartha to define evaluation regions with certain climate conditions. A systematic comparison of near-surface temperature and precipitation simulated by the regional and the global model is done. In general, the historical time period is well represented by the GCM and the RCM. Some different biases occur in the RCM compared to the GCM as in the Amazon Basin, northern Africa and the West Asian domain. Both models project similar warming, although somewhat less so by the RCM for certain regions and climate types. A common feature in regions of tropical climate types is that REMO shows dryer climate conditions than forMax Planck Institute for Meteorology-Earth System Model (MPI-ESM) for RCP 4.5 and RCP 8.5, leading to an opposing sign in the climate change signal. With an increase in radiative forcing from RCP 2.6 to RCP 8.5 and towards the end of the 21st century, some of the detected differences between GCM and RCM are more pronounced.

A new ensemble of climate and climate change simulations covering all major inhabited regions with a spatial resolution of about 25 km, from the WCRP CORDEX COmmon Regional Experiment (CORE) Framework, has been established in support of the growing demands for climate services. The main objective of this study is to assess the quality of the simulated climate and its fitness for climate change projections by REMO (REMO2015), a regional climate model of Climate Service Center Germany (GERICS) and one of the RCMs used in the CORDEX-CORE Framework. The CORDEX-CORE REMO2015 simulations were driven by the ECMWF ERA-Interim reanalysis and the simulations were evaluated in terms of biases and skill scores over ten CORDEX Domains against the Climatic Research Unit (CRU) TS version 4.02, from 1981 to 2010, according to the regions defined by the Köppen–Trewartha (K–T) Climate Classification types. The REMO simulations have a relatively low mean annual temperature bias (about $\pm 0.5$ K) with low spatial standard deviation (about $\pm 1.5$ K) in the European, African, North and Central American, and Southeast Asian domains. The relative mean annual precipitation biases of REMO are below $\pm 50$ % in most domains; however, spatial standard deviation varies from $\pm 30$ % to $\pm 200$ %. The REMO results simulated most climate types relatively well with lowest biases and highest skill score found in the boreal, temperate, and subtropical regions. In dry and polar regions, the REMO results simulated a relatively high annual biases of precipitation and temperature and low skill. Biases were traced to: missing or misrepresented processes, observational uncertainty, and uncertainties due to input boundary forcing.