Eva Holtanová

Performance of ENSEMBLES regional climate models over Central Europe using various metrics

We show the evaluation of ENSEMBLES regional climate models (RCMs) driven by reanalysis ERA40 over a region centered at the Czech Republic. Attention is paid especially to the model ALADIN-CLIMATE/CZ, being used as the basis of the new climate change scenarios simulation for the Czech Republic. The validation criteria used here are based on monthly or seasonal mean air temperature and precipitation. We concentrate not only on spatiotemporal mean values but also on temporal standard deviation, inter-annual variability, the mean annual cycle, and the skill of the models to represent the observed spatial patterns of these quantities. Model ALADIN-CLIMATE/CZ performs quite well in comparison to the other RCMs; we find its performance satisfactory for further use for impact studies. However, it is also shown that the results of evaluation of the RCMs’ skill in simulating observed climate strongly depend on the criteria incorporated for the evaluation.

Performance of ALADIN-Climate/CZ over the area of the Czech Republic in comparison with ENSEMBLES regional climate models

Nowadays Regional Climate Models (RCMs) are increasingly used for downscaling of information from the coarse resolution of global climate models (GCMs) and they represent a more and more popular tool for assessment of future climate changes and their impacts at regional scales. In spite of continual progress of RCMs, their outputs still suffer from many uncertainties and biases. Therefore, it is necessary to assess their ability to simulate observed climate characteristics and uncertainties in their outputs before they are applied in subsequent studies. In the present study, the assessment of RCM performance in simulating climate in the reference period of 1961–1990 over the area of Czech Republic is presented. Furthermore, we focused on the intercomparison of the models’ results, mainly on the comparison of the Czech model ALADIN-Climate/CZ with outputs of other RCMs. Simulation of ALADIN-Climate/CZ in 25-km horizontal resolution, and thirteen RCM simulations from the ENSEMBLES project were assessed. Attention was paid especially to comparison of simulated and observed spatial and temporal variability of several climatic variables. The monthly and seasonal values of surface air temperature, precipitation totals and relative humidity were examined for evaluation of temporal variability and 30-year seasonal and monthly values with respect to spatial variability. Climate model performance was evaluated in several ways, namely by boxplots, maps of variability characteristics, skill scores based on mean square error and Taylor diagrams. Model errors detected by model evaluation depend on many factors (e.g. considered variables and their characteristics, area of analysis, time scale of interest and the method of assessment). On the basis of incorporated performance criteria model ALADIN-Climate/CZ belonged to a better group of RCMs in most cases. However, it was definitely the worst in simulating spring monthly means of air temperature and relative humidity in all seasons.

Heat wave of August 2012 in the Czech Republic: comparison of two approaches to assess high temperature event

We present an analysis of a period of high air temperature that occurred in the second half of August 2012 in the Czech Republic (CZ). We use and compare the results of two different approaches for the evaluation of high air temperature events. The Weather Extremity Index (WEI) evaluates the extremity and spatial extent of the meteorological event of interest. The second method is based on the duration of daily maximum air temperature above specific thresholds. In 2012, the high air temperature in the CZ lasted from 18 August to 24 August (18/8 to 24/8). It was connected with the inflow of hot air from northern Africa between the low pressure trough over the eastern Atlantic and the region of high pressure in central Europe. The high air temperature culminated on 20/8 when its maximum was greater than 30°C across the whole of the CZ. The highest daily maximum air temperature on record in the CZ with a value of 40.4°C was observed at the Dobřichovice station. Our results demonstrate that the studied period was quite extraordinary, occurring so late in the summer with a relatively large areal extent and extremity of detected maximum air temperature. Furthermore, the WEI was found useful for identification of very extreme high air temperature events and facilitated intercomparison in terms of extremity and spatial extent. However, WEI cannot be used for detection of periods with a persistent relatively high air temperature that could have severe impacts on both human activities and natural ecosystems but during which the extremity of observed air temperature values is not very high.

Changes in air temperature means and interannual variability over Europe in simulations by ALADIN-Climate/CZ: dependence on the size of the integration domain

This paper presents an evaluation of two simulations by regional climate model (RCM) ALADIN-Climate/CZ with different sizes of integration domain and their driving simulation of global climate model (GCM) ARPÉGE-Climat over central Europe during the period 1961–2010. After a brief evaluation of seasonal means of air temperature characteristics (daily mean, maximum and minimum temperature) and their variability, we focus on the ability of the simulations to represent observed changes in seasonal mean air temperature and its variability. We distinguish between intrinsic and trend-induced variability. Moreover, the dependence on the size of the integration domain and the improvement brought about by the nested RCM compared to the driving GCM are analysed. Both the evaluation of temperature means and variability and the evaluation of their changes confirm that RCM simulation on a larger domain often differs more considerably from driving GCM than simulation on a smaller domain. The larger domain simulation usually produces smaller biases for mean temperature but has worse agreement with observations in terms of variability. Further, the areas with the greatest discrepancies are often in the south-eastern region. In addition, noticeable differences between RCM and GCM simulations mostly emerge for the Alpine region and eastern regions. A possible explanation is the complex terrain and larger distance from domain boundaries. Moreover, it was analysed whether removing of the linear trend enhances the representation of changes in variability. This was not generally confirmed. Only in summer the simulations underestimated the observed increase in intrinsic variability less considerably than for total variability. However, biases were still considerable in many cases.

Similarities within a multi-model ensemble: functional data analysisframework

Despite the abundance of available global and regi onal climate model outputs, their use for evaluatio n of past and future climate changes is often complicated by subs tantial differences between individual simulations, and the resulting uncertainties. In this study, we present a methodol ogy framework for the analysis of multi-model ensem bles based on functional data analysis approach. A set of two met rics that generalize the concept of similarity base d on the behaviour of 15 entire simulated climatic time series, encompassing both past and future periods, is introduced. As fa r as our knowledge, our method is the first to quantitatively assess simila rities between model simulations based on the tempo ral evolution of simulated values. To evaluate mutual distances of t he time series we used two semimetrics based on Euc lidean distances between the simulated trajectories and on differenc es in their first derivatives. Further, we introduc e an innovative way of visualizing climate model similarities based on a n etwork spatialization algorithm. Using the layout g raphs the data are 20 ordered on a 2-dimensional plane which enables an u n mbiguous interpretation of the results. The metho d is demonstrated using two illustrative cases of air temperature ove r th British Isles and precipitation in central Eu rope, simulated by an ensemble of EURO-CORDEX regional climate models and their driving global climate models over the 1971– 2098 period. In addition to the sample results, interpretational aspects of the applied methodology and its possibl e extensions are also discussed. 25