We investigate the impact of horizontal resolution, convection parameterization and lateral boundary conditions on the ability of a regional climate model - RegCM4 to simulate the observed temperature and precipitation climatology over Croatia and surrounding regions. An ensemble of RegCM4 simulations over Europe has been generated including two resolutions (50 and 12.5 km), two different deep convection schemes and different boundary conditions (a reanalysis and four global climate models). Using two observational data sets (with and without the undercatch correction) shows that the magnitude of RegCM4 errors in winter precipitation also depends on observational uncertainty. While the mean precipitation amounts in high-resolution simulations are generally overestimated in the cold part of the year, low-resolution RegCM4 simulations are comparable to most of other EURO-CORDEX models. When forced by ERA-Interim, RegCM4 performs relatively poorly in reproducing temporal correlations against observations, but performs well in terms of spatial correlation. Whereas precipitation amounts from high-resolution simulations need further improvements, the RegCM4 skill in simulating near-surface temperature is high and well suited for impact studies.

The lower Neretva river basin includes a fertile valley at the estuary into the Adriatic Sea, where intense agricultural production occurs, and the higher terrain where drinking water resources exist. To provide input for the further assessment of crop-yield production and hydrological risks, climate and climate change were analysed using the Opuzen station air temperature and total precipitation data for the 1961-2015 period. Both historical and future climates (2021-2050) were assessed based on simulations of three regional climate models (RCMs). The RCMs were forced by the observed concentrations of greenhouse gases (GHGs) from 1951 to 2000, and the IPCC A1B scenario of the GHG emissions was applied from 2001 onwards. The models were compared with the observations, and two bias adjustment methods were evaluated. The results generally showed a significant increase in the mean annual and seasonal temperature and a weak decreasing trend in annual and seasonal precipitation. Projections revealed a predominant increase in the mean temperature by the mid-21st century for all three RCMs (between 0.5 and 3.5 degrees C). The precipitation changed by between -60 and +60% throughout the year for the different models, although the changes generally were not statistically significant.

The Dinaric mountains in Croatia present one of the hot spots of European biodiversity, possessing a very large number of species and hosting most endemics. The Dinaric mountains in Croatia strongly affect the climate of the Adriatic region, making a distinct boundary between the maritime and a continental climate. In this chapter, an overview of climatic conditions of the Dinaric area is provided, including observed climatic changes of temperatures and precipitation in the last century. Existing soil types were assessed and described, in particular with respect to soil-forming processes. The role of climate as a dominant factor of soil formation was evaluated in relationship to other factors such as lithology and topography. Also, the occurrence of a specific broad range of forest associations in Dinarides was presented. We examined the correspondence between forest vegetation, soil, and climatic properties in the Dinaric area. As a finale, some future, very possible scenarios of regional climatic development are presented as a serious hazard to the sustainability of natural forest resources. We determined the variety of soil types, ranging from soils that are characteristic for Mediterranean (on limestone) such as Terra rossa with intensive red color, calcomelanosols and calcicambisols in high karst, to soils characteristic for continental climate (on flint or silica) such as dystric cambisols and luvisols. The forest vegetation of the Dinaric mountains constitutes 54 diverse forest ecosystem types, encompassing specific combinations of soil and phytocoenoses. Existing ecosystems form nine broader groups, that is, bioclimates, which are typical for Dinarides.

The hydrostatic regional climate model RCA, version 3 (RCA3), of the Swedish Meteorological and Hydrological Institute was used to dynamically downscale ERA-40 and the ECMWF operational analysis over a 22-yr period. Downscaling was performed at four horizontal resolutions-50, 25, 12.5, and 6.25 km-over an identical European domain. The model-simulated precipitation is evaluated against high-resolution gridded observational precipitation datasets over Switzerland and southern Norway, regions that are characterized by complex orography and distinct climate regimes.RCA3 generally overestimates precipitation over high mountains: during winter and summer over Switzerland and during summer over central-southern Norway. In the summer, this is linked with a substantial contribution of convective precipitation to the total precipitation errors, especially at the coarser resolutions (50 and 25 km). A general improvement in spatial correlation coefficients between simulated and observed precipitation is observed when the horizontal resolution is increased from 50 to 6 km. The 95th percentile spatial correlation coefficients during winter are much higher for southern Norway than for Switzerland, indicating that RCA3 is more successful at reproducing a relatively simple west-to-east precipitation gradient over southern Norway than a much more complex and variable precipitation distribution over Switzerland. The 6-km simulation is not always superior to the other simulations, possibly indicating that the model dynamical and physical configuration at this resolution may not have been optimal. However, a general improvement in simulated precipitation with increasing resolution supports further use and application of high spatial resolutions in RCA3.

The aim of this study is to evaluate the improvement in simulated precipitation of the third version of the Rossby Centre Climate model (RCA3), when run at horizontal resolutions of 50, 25, 12 and 6 km. The analysis is focused on the countries of Switzerland and Norway, where complex orography can play a dominant role in precipitation formation. Observation data sets used for model validation were for Switzerland, the RhiresD dataset, an approximately 2-km resolution product of the Swiss National Meteorological Service (MeteoSwiss), and over Norway, the KLIMAGRID dataset (Mohr, 2007), a 1-km resolution product of the Norwegian Meteorological Institute (METNO), is used. The range of horizontal resolutions in experiments facilitates a systematic analysis of the impact of increasing the horizontal resolution on precipitation, in two different climate regimes.

Several temperature indices are calculated for summer 1989-2008 in Croatia from the two different experiments with regional climate model RegCM4. Experiments were done for two different horizontal resolutions (50 and 12km) and compared with observed gridded EOBS data set and observations from 20 meteorological stations in Croatia. The higher horizontal distribution resolves better topography and reduces biases in daily minimum and maximum temperature. For all indices, except HWDI, the increase from 50 to 12 km gives a better spatial distribution, although the simulated values are overestimated. Of all analysed indices, the indices at the 12-km resolution based on percentile are the closest to the observed ones.

Five indices of extreme precipitation are analysed over the Croatian Adriatic region on a seasonal and annual basis from 19 meteorological stations and a 3-member ensemble of the Regional Climate Model (RegCM3) simulations in the reference (1961−1990) and the near-future (2011−2040) climate. Future climate integrations are performed under the IPCC A2 emission scenario. Uncertainty of projected changes is assessed by comparing RegCM3 results with those from a subset of the ENSEMBLES regional climate models. Observed wet extremes exhibit large values in the areas close to the coastal mountains. Interannual variability in the number of very wet days (R95) and the fraction of precipitation associated with very wet days (R95T) is relatively large in all seasons, indicating a variable synoptic activity over the Adriatic region from one year to the next, as well as variable intense showers in the summer. The most prominent feature is a statistically significant decrease in the observed maximum number of consecutive dry days (CDD) during the autumn. When compared against the observations, RegCM3 overestimates observed precipitation and maximum 5 d precipitation amounts (Rx5d) in all seasons except the summer with statistically significant differences. Simulated number of dry days (DD) is generally underestimated and, consequently, R95 is mostly overestimated. Of all indices considered, R95T is best simulated by the model. The interannual variability of precipitation and indices is generally well reproduced. The projected changes in the mean and indices of extreme precipitation in the near future are weak overall, except in the autumn, in both the RegCM3 ensemble and selected ENSEMBLES models. Although the Mediterranean region is characterised as one of the regions most responsive to climate change, our results indicate that over the eastern Adriatic region, significant changes may not occur in the near future.

Promjene indeksa oborinskih ekstrema u bližoj buducnosti (2011-2040.) u odnosu na referentno razdoblje (1961-1990.) analizirane su prema rezultatima simulacija tri clana ansambla regionalnog klimatskog modela RegCM3. Simulacije buduce klime uvažavaju A2 scenarij Međuvladinog panela o klimatskim promjenama (IPCC). U bližoj buducnosti model predviđa smanjenje oborine u proljece i jesen te na godisnjoj razini, a povecanje oborine se može ocekivati zimi i ljeti. Promjene intenziteta dnevnih oborina zimi, u proljece i u ljeto u skladu su s promjenama sezonske kolicine oborine, dok se u jesen i za godinu intenzitet može povecati zbog porasta susnih dana odnosno smanjenja oborinskih dana. Udio u ukupnoj oborini koji dolazi od dana s oborinom iznad 95. percentila dnevnih kolicina oborine (vrlo vlažni dani) može se povecati u nekim dijelovima Jadrana i to ne samo u sezonama u kojima se ocekuje povecanje oborine, vec i u sezonama u kojima rezultati modela upucuju na smanjenje oborine.

The surface energy budget components from two simulations of the regional climate model RegCM4.2 over the European/North African domain during the period 1989–2005 are analysed. The simulations differ in specified boundary forcings which were obtained from ERA-Interim reanalysis and the HadGEM2-ES Earth system model. Surface radiative and turbulent fluxes are compared against ERA-Interim. Errors in surface radiative fluxes are derived with respect to the Global Energy and Water Cycle Experiment/Surface Radiation Budget satellite-based products. In both space and time, we find a high degree of realism in the RegCM surface energy budget components, but some substantial errors and differences between the two simulations are also present. The most prominent error is an overestimation of the net surface shortwave radiation flux of more than 50 W/m 2 over central and southeastern Europe during summer months. This error strongly correlates with errors in the representation of total cloud cover, and less strongly with errors in surface albedo. During other seasons, the amplitude of the surface energy budget components is more in line with reference datasets. The errors may limit the usefulness of RegCM simulations in applications (e.g. high-quality simulation-driven impact studies). However, by using a simple diagnostic model for error interpretation, we suggest potential sensitivity studies aiming to reduce the underestimation of cloud cover and overestimation of shortwave radiation flux.

The 2m temperature (T2m) and precipitation from five regional climate models (RCMs), which participated in the ENSEMBLES project and were integrated at a 25-km horizontal resolution, are compared with observed climatological data from 13 stations located in the Croatian coastal zone. The twentieth century climate was simulated by forcing RCMs with identical boundary conditions from the ERA-40 reanalysis and the ECHAM5/MPI-OM global climate model (GCM); climate change in the twenty-first century is based on the A1B scenario and assessed from the GCM-forced RCMs’ integrations. When forced by ERA-40, most RCMs exhibit cold bias in winter which contributes to an overestimation of the T2m annual cycle amplitude and the errors in interannual variability are in all RCMs smaller than those in the climatological mean. All models underestimate observed warming trends in the period 1951–2010. The largest precipitation biases coincide with locations/seasons with small observed amounts but large precipitation amounts near high orography are relatively well reproduced. When forced by the same GCM all RCMs exhibit a warming in the cold half-year and a cooling (or weak warming) in the warm period, implying a strong impact of GCM boundary forcing. The future eastern Adriatic climate is characterised by a warming, up to +5 °C towards the end of the twenty-first century; for precipitation, no clear signal is evident in the first half of the twenty-first century, but a reduction in precipitation during summer prevails in the second half. It is argued that land-sea contrast and complex coastal configuration of the Croatian coast, i.e. multitude of island and well indented coastline, have a major impact on small-scale variability. Orography plays important role only at small number of coastal locations. We hypothesise that the parameterisations related to land surface processes and soil hydrology have relatively stronger impact on variability than orography at those locations that include a relatively large fraction of land (most coastal stations), but affecting less strongly locations at the Adriatic islands.

This study investigates the performance of two planetary boundary layer (PBL) parameterisations in the regional climate model RegCM4.2 with specific focus on the recently implemented prognostic turbulent kinetic energy parameterisation scheme: the University of Washington (UW) scheme. When compared with the default Holtslag scheme, the UW scheme, in the 10-year experiments over the European domain, shows a substantial cooling. It reduces winter warm bias over the north-eastern Europe by 2 °C and reduces summer warm bias over central Europe by 3 °C. A part of the detected cooling is ascribed to a general reduction in lower tropospheric eddy heat diffusivity with the UW scheme. While differences in temperature tendency due to PBL schemes are mostly localized to the lower troposphere, the schemes show a much higher diversity in how vertical turbulent mixing of the water vapour mixing ratio is governed. Differences in the water vapour mixing ratio tendency due to the PBL scheme are present almost throughout the troposphere. However, they alone cannot explain the overall water vapour mixing ratio profiles, suggesting strong interaction between the PBL and other model parameterisations. An additional 18-member ensemble with the UW scheme is made, where two formulations of the master turbulent length scale in unstable conditions are tested and unconstrained parameters associated with (a) the evaporative enhancement of the cloud-top entrainment and (b) the formulation of the master turbulent length scale in stable conditions are systematically perturbed. These experiments suggest that the master turbulent length scale in the UW scheme could be further refined in the current implementation in the RegCM model. It was also found that the UW scheme is less sensitive to the variations of the other two selected unconstrained parameters, supporting the choice of these parameters in the default formulation of the UW scheme.