COnsortium For Small-scale MOdeling Model Research Articles

Prediction of tropical cyclone trajectories over the Northern Indian Ocean using COSMO

The present study investigates the performance of a regional numerical weather prediction model; namely, the Consortium for Small-scale Modelling (COSMO) in the prediction of the tropical cyclone (TC) trajectories for varying intensities of the storm. A total of 8 TCs formed over the Northern Indian Ocean from 2017 to 2019 are chosen for the evaluation of the COSMO model. The central pressure ( $$P_\mathrm{Central}$$ ), pressure drop ( $$\varDelta P$$ ), and maximum sustained surface wind speed (MSW) simulated by the COSMO model are validated against the concurrent observations from India Meteorological Department. The forecasted mean track errors are 95 km for a lead time of 24 h, whereas it was about 140 km for a lead time of 48 h. The mean initial positional error in identification of the storm was about 50 km. The intensity of a storm is underestimated in terms of $$\varDelta P$$ and MSW, especially for a lead time of 0–24 h, whereas the model shows a consistent overestimation for a lead time of more than 24 h. During the initial stage of a storm, when its intensity is categorized as a Deep Depression, we notice a maximum amount of uncertainty in the prediction of cyclone track. The COSMO model yields improved predictability of the tracks for storms categorized as Very Severe Cyclonic Storms. As the intensity of a storm increases from a Deep Depression to a Very Severe Cyclonic Storm, the track errors associated with model simulations tend to decrease. Results of the present study illustrate the predictability of TCs from COSMO in terms of the trajectory and intensity of the storm.

High-Resolution Simulation of Polar Lows over Norwegian and Barents Seas Using the COSMO-CLM and ICON Models for the 2019–2020 Cold Season

The lack of meteorological observations at high latitudes and the small size and relatively short lifetime of polar lows (PLs) constitute a problem in the simulation and prediction of these phenomena by numerical models. On the other hand, PLs, which are rapidly developing, can lead to such extreme weather events as stormy waves, strong winds, the icing of ships, and snowfalls with low visibility, which can influence communication along the Arctic seas. This article is devoted to studying the possibility of the numerical simulation and prediction of polar lows by different model configurations and resolutions. The results of the numerical experiments for the Norwegian and Barents seas with grid spacings of 6.5 and 2 km using the ICON-Ru configurations of the ICON (ICOsahedral Nonhydrostatic) model and with a grid spacing of 6.5 km using the COSMO-CLM (Climate Limited-area Modeling) configuration of the COSMO (COnsortium for Small-scale MOdelling) model are presented for the cold season of 2019–2020. All the used model configurations demonstrated the possibility of the realistic simulation of polar lows. The ICON model showed slightly more accurate results for the analyzed cases. The best results showed runs with lead times of less than a day.

Performance Evaluation of High-Resolution Simulations with COSMO over South Italy

This study aims to assess the capabilities of a weather forecasting system based on simulations performed with the COSMO (COnsortium for Small-scale Modeling) model over a domain located in southern Italy, employing a spatial resolution of about 1 km, driven by ECMWF-IFS global data. The model is run daily at the Italian Aerospace Research Center (CIRA), and the evaluation was performed from January to May 2018 using a combination of observational data, specifically data provided by the CIRA meteorological station, wind profiler and ceilometer. Moreover, data provided by radio sounding located at Pratica di Mare and ground stations at two other locations were also used. A model configuration optimized through a tuning procedure over the domain considered was employed, while the evaluation was performed by comparing daily values of several variables and using standard monitoring indices. The results highlight that the model has good capability in reproducing daily values of temperature, while precipitation intensity is generally underestimated, even if rain patterns are well captured (alternating rainy and dry days). Good agreement is also reported for wind speed, especially at 100 and 500 m altitude. Regarding radio sounding data, the COSMO model configuration selected can reproduce the vertical profile of temperature and dew point, with the exception of inversion points. Evaluation against ceilometer data is achieved in terms of cloud height and planetary boundary layer height.

Regional Climate Modelling with COSMO-CLM: History and Perspectives

The history of the COSMO (COnsortium for Small-scale Modelling) model goes back to the early 1990s, when the Deutscher Wetterdienst (German Meteorological Service, DWD) decided to develop a non-hydrostatic model, the Lokalmodell (LM) [...]

Spatio-temporal Structure of the Boundary Layer under the Impact of Mountain Waves

During the Hydrological cycle in the Mediterranean Experiment (HyMeX) in autumn 2012 intensive measurements were conducted in the Tavignano Valley, which extends from the centre to the coast of the island of Corsica. On the investigated day, the atmospheric boundary layer (ABL) in the valley showed a distinctive spatio-temporal variability, which resulted from the interaction and superposition of mesoscale dynamically- and thermally-driven processes and dry convection. Based on the observations, not all of the observed ABL characteristics could be explained and hypotheses on the involved processes were formulated in a previous study. To close the observational gaps and to test the hypotheses, high-resolution simulations with the COSMO (Consortium for Small-scale Modeling) model were now performed. The model was able to reproduce the main ABL characteristics and could hence be used to address the processes affecting the ABL. The main features were: in the upper part of the valley, the stable nocturnal ABL was eroded from top and bottom alike by shear-generated turbulent mixing in the vicinity of a mountain wave and buoyancy- and shear-driven surface-based turbulent mixing, leading to a very abrupt increase of the daytime ABL depth. In the lower part of the valley, the ABL remained rather shallow and was dominated by a superimposed thermally-driven sea breeze and upvalley wind. In the afternoon, the formerly deep ABL in the upper part of the valley rapidly decreased when the combined sea breeze and upvalley wind moved up the valley. While the ABL depth was rather horizontally homogeneous in the lower part of the valley and near the coast, it showed a considerable variability in the valley's upper part on scales of a few kilometres due to the varying dominance of the different processes. The local ABL depth also varied considerably in time depending on which influence dominated, i.e. of surface heating, mountain wave or sea breeze and upvalley wind. As the simulated sea breeze strongly depended on the sea-surface temperature, the results were sensitive to the chosen value in the model.

Attribution of precipitation to cyclones and fronts over Europe in a kilometer-scale regional climate simulation

Abstract. This study presents a detailed analysis of the climatological distribution of precipitation in relation to cyclones and fronts over Europe for the 9-year period 2000–2008. The analysis uses hourly output of a COSMO (Consortium for Small-scale Modeling) model simulation with 2.2 km grid spacing and resolved deep convection. Cyclones and fronts are identified as two-dimensional features in 850 hPa geopotential, equivalent potential temperature, and wind fields and subsequently tracked over time based on feature overlap and size. Thermal heat lows and local thermal fronts are removed based on track properties. This dataset then serves to define seven mutually exclusive precipitation components: cyclonic (near cyclone center), cold-frontal, warm-frontal, collocated (e.g., occlusion area), far-frontal, high-pressure (e.g., summer convection), and residual. The approach is illustrated with two case studies with contrasting precipitation characteristics. The climatological analysis for the 9-year period shows that frontal precipitation peaks in winter and fall over the eastern North Atlantic and the Alps (> 70 % in winter), where cold frontal precipitation is also crucial year-round; cyclonic precipitation is largest over the North Atlantic (especially in summer with > 40 %) and in the northern Mediterranean (widespread > 40 %); high-pressure precipitation occurs almost exclusively over land and primarily in summer (widespread 30 %–60 %, locally >80 %); and the residual contributions uniformly amount to about 20 % in all seasons. Considering heavy precipitation events (defined based on the local 99.9th all-hour percentile) reveals that high-pressure precipitation dominates in summer over the continent (50 %–70 %, locally >80 %); cold fronts produce much more heavy precipitation than warm fronts; and cyclones contribute substantially (50 %–70 %), especially in the Mediterranean in fall through spring and in northern Europe in summer.

An assessment of a very severe cyclonic storm in the Arabian sea using the COSMO model

Accurate and reliable representation of convective processes is one of the major sources of uncertainty in numerical weather prediction (NWP) models, especially for those operating in the grey zone resolutions. The performance of NWP models become more sensitive to their grid resolutions, when they are used for simulation of severe weather events, such as a cyclonic storm. In this paper, we present a detailed assessment of an intense convective episode with heavy precipitation associated with the passage of a very severe cyclonic storm “OCKHI” using the Consortium for Small-scale Modelling (COSMO). A set of distinct numerical simulations are carried out using COSMO to address the impact of grid resolution and the treatment of explicit and implicit convection. Results obtained from the present investigation indicate that explicit treatment of convection in the COSMO model led to improved prediction of the cyclonic event in terms of sea level pressure, maximum sustained surface wind speeds and the accumulated rainfall, but reduction of the spatial grid resolution from 7 to 3 km did not show appreciable differences in the forecast fields. Conclusively, the current study recommends switching off the convection parameterization scheme at a grid resolution of 7 km for improved predictability of tropical cyclones.

Improving the Processes in the Land Surface Scheme TERRA: Bare Soil Evaporation and Skin Temperature

Newly improved formulations of the bare soil evaporation and the surface temperature are presented, using the multilayer land surface scheme TERRA of the Consortium for Small-scale Modeling (COSMO) atmospheric model. The simulations were carried out in offline mode with atmospheric forcing data from the Meteorological Observatory Lindenberg–Richard-Aßmann-Observatory of the German Meteorological Service. The results show that the bare soil evaporation simulated by the reference version of TERRA is substantially overestimated under wet conditions, and underestimated under dry conditions. Furthermore, the amplitude of the diurnal cycle of the surface temperature is systematically underestimated. In contrast, the diurnal cycles of the temperatures in the soil are overestimated instead. The new description of the bare soil evaporation in TERRA is based on a resistance formulation analogue to Ohm’s law, while the surface temperature is now based on the skin temperature formulation by Viterbo and Beljaars. The new formulation improves the simulated bare soil evaporation substantially. In particular, the overestimation under wet conditions is reduced, also acting against an extensive drying of the soil during the annual cycle. Additionally, the underestimation under dry conditions is reduced as well. Furthermore, the simulated amplitude of the diurnal cycle of the surface temperature is substantially increased. In particular, a nocturnal warm bias is systematically reduced. In addition to this, the new formulations were also applied in coupled mode in the COSMO model, resulting in improved diurnal cycles of near-surface temperature and dew point.

Large impact of tiny model domain shifts for the Pentecost 2014 mesoscale convective system over Germany

Abstract. The mesoscale convective system (MCS) that affected Germany at Pentecost 2014 (9 June 2014) was one of the most severe for decades. However, the predictability of this system was very low as the operational deterministic and ensemble prediction systems completely failed to predict the event with more than a 12 h lead time. We present hindcasts of the event using the COnsortium for Small-scale MOdeling (COSMO) model at a convection-permitting (2.8 km) resolution on a large (1668 km×1807 km) domain. Using this large domain allowed us to successfully simulate the whole life cycle of the system originating from the French Atlantic coast. However, even with the large domain, the predictability of the MCS is low. Tiny changes to the model domain produced large changes in the MCS, removing it completely from some simulations. To demonstrate this we systematically shifted the model domain by just one grid point in eight different directions, from which three did not simulate any convection over Germany. Our analysis shows that there were no important differences in domain-averaged initial conditions or in the preconvective environment ahead of the convective system. The main reason that one-third of these seemingly identical initial conditions fail to produce any convection over Germany seems to be the proximity of the track of the initial convective system to the coast and colder sea surface. The COSMO model simulates small horizontal displacements of the precursors of the MCS which then determine if the cells dissipate close to the sea or reach a favorable area for convective development over land and further evolve into an MCS. This study demonstrates the potentially huge impact of tiny model domain shifts on forecasting convective processes in this case, which suggests that the sensitivity to similarly small initial-condition perturbations could be a helpful indicator of days with low predictability and should be evaluated across other cases, models, and weather regimes.

Projections of Alpine Snow-Cover in a High-Resolution Climate Simulation

The recent development of high-resolution climate models offers a promising approach in improving the simulation of precipitation, clouds and temperature. However, higher grid spacing is also a promising feature to improve the simulation of snow cover. In particular, it provides a refined representation of topography and allows for an explicit simulation of convective precipitation processes. In this study we analyze the snow cover in a set of decade-long high-resolution climate simulation with horizontal grid spacing of 2.2 km over the greater Alpine region. Results are compared against observations and lower resolution models (12 and 50 km), which use parameterized convection. The simulations are integrated using the COSMO (Consortium for Small-Scale Modeling) model. The evaluation of snow water equivalent (SWE) in the simulation of present-day climate, driven by the ERA-Interim reanalysis, against an observational dataset, reveals that the high-resolution simulation clearly outperforms simulations with grid spacing of 12 and 50 km. The latter simulations underestimate the cumulative amount of SWE over Switzerland over the whole annual cycle by 33% (12 km simulation) and 56% (50 km simulation) while the high-resolution simulation shows a spatially and temporally averaged difference of less than 1%. Scenario simulations driven by GCM MPI-ESM-LR (2081–2090 RCP8.5 vs. 1991–2000) reveal a strong decrease of SWE over the Alps, consistent with previous studies. Previous studies had found that the relative decrease becomes gradually smaller with elevation, but this finding was limited to low and intermediate altitudes (as a 12 km simulation resolves the topography up to 2500 m). In the current study we find that the height gradient reverses sign, and relative reductions in snow cover increases above 3000 m asl, where important parts of the cryosphere are present. In addition, the simulations project a transition from permanent to seasonal snow cover at high altitudes, with potentially important impacts to Alpine permafrost. This transition and the more pronounced decline of SWE emphasize the value of the higher grid spacing. Overall, we show that high-resolution climate models offer a promising approach in improving the simulation of snow cover in Alpine terrain.

The on-line coupled atmospheric chemistry model system MECO(n) – Part 5: Expanding the Multi-Model-Driver (MMD v2.0) for 2-way data exchange including data interpolation via GRID (v1.0)

Abstract. As part of the Modular Earth Submodel System (MESSy), the Multi-Model-Driver (MMD v1.0) was developed to couple online the regional Consortium for Small-scale Modeling (COSMO) model into a driving model, which can be either the regional COSMO model or the global European Centre Hamburg general circulation model (ECHAM) (see Part 2 of the model documentation). The coupled system is called MECO(n), i.e., MESSy-fied ECHAM and COSMO models nested n times. In this article, which is part of the model documentation of the MECO(n) system, the second generation of MMD is introduced. MMD comprises the message-passing infrastructure required for the parallel execution (multiple programme multiple data, MPMD) of different models and the communication of the individual model instances, i.e. between the driving and the driven models. Initially, the MMD library was developed for a one-way coupling between the global chemistry–climate ECHAM/MESSy atmospheric chemistry (EMAC) model and an arbitrary number of (optionally cascaded) instances of the regional chemistry–climate model COSMO/MESSy. Thus, MMD (v1.0) provided only functions for unidirectional data transfer, i.e. from the larger-scale to the smaller-scale models.Soon, extended applications requiring data transfer from the small-scale model back to the larger-scale model became of interest. For instance, the original fields of the larger-scale model can directly be compared to the upscaled small-scale fields to analyse the improvements gained through the small-scale calculations, after the results are upscaled. Moreover, the fields originating from the two different models might be fed into the same diagnostic tool, e.g. the online calculation of the radiative forcing calculated consistently with the same radiation scheme. Last but not least, enabling the two-way data transfer between two models is the first important step on the way to a fully dynamical and chemical two-way coupling of the various model instances.In MMD (v1.0), interpolation between the base model grids is performed via the COSMO preprocessing tool INT2LM, which was implemented into the MMD submodel for online interpolation, specifically for mapping onto the rotated COSMO grid. A more flexible algorithm is required for the backward mapping. Thus, MMD (v2.0) uses the new MESSy submodel GRID for the generalised definition of arbitrary grids and for the transformation of data between them.In this article, we explain the basics of the MMD expansion and the newly developed generic MESSy submodel GRID (v1.0) and show some examples of the abovementioned applications.

Spatial verification approaches as a tool to evaluate the performance of high resolution precipitation forecasts

The spatial resolution of Numerical Weather Prediction (NWP) models has increased significantly in recent years. While high-resolution models are able to produce more detailed precipitation structures, their true benefit lies in more realistic statistics rather than the information provided for a specific grid point. Unfortunately, NWP model verification using traditional grid-point-by-grid-point methods has not managed to keep pace due to the limited amount of point observations available in comparable resolution. Spatial verification methods represent a possible solution since they reward closeness or resemblance by relaxing the requirement for exact matches between the forecast and observations.An intense convective event in the Mediterranean region is used as a test case to analyze the forecasting performance of the Consortium for Small-scale Modeling model (COSMO) at two different resolutions. Satellite estimates of precipitation are used as ground truth. The precipitation forecasts are assessed using various spatial methods and averaging techniques, each of which provides distinctly useful information on model performance.The results from this particular test case indicate that a neighbourhood verification framework as well as the use of the Structure, Amplitude and Location (SAL) index, which belongs to the object-based methods, can identify the scales and precipitation thresholds at which the fine resolution configuration (COSMO-GR3) provides more accurate forecasts than the coarser resolution configuration (COSMO-GR7) and is therefore worth the additional computational burden. Additional insight is gained by comparing the results of various spatial methods with traditional verification metrics based on point observations. Finally, it is demonstrated that the interpolation method used in the adaptation of data at various scales can, in some cases, influence verification results as much as model resolution.

Numerical modeling analysis of the mesoscale environment conducive to two tornado events using the COSMO.Gr model over Greece

The COnsortium for Small-scale MOdeling (COSMO) was formed in October 1998, and its general goal is to develop, improve and maintain a non-hydrostatic limited-area atmospheric model. The COSMO model has been designed both for operational numerical weather prediction (NWP) as well as various scientific applications on the meso-β and meso-γ scale. Two tornado case studies were selected to investigate the ability of COSMO model to depict the characteristics of severe convective weather, which favoured the development of the associated storms.The first tornado (TR01) occurred, close to Ag. Ilias village, 8 Km north-western of Aitoliko city over western Greece on February 7, 2013, while the second tornado (TR02) was developed close to Palio Katramio village, 8 Km southern from Xanthi city over northern Greece on November 25, 2015. Although both tornadoes had a short lifetime, they caused significant damages. The COSMO.GR atmospheric model was initialized with analysis from the European Centre for Medium-Range Weather Forecasts (ECMWF). The resulting numerical products with spatial resolution of 0.02° (∼2km) over the geographical domain of Greece depicted very well the severe convective conditions close to tornadoes formation. The Energy Helicity Index (EHI) diagnostic variable in both numerical simulations showed a gradual increase of values closing to the location and time of the tornadogenesis. Similar to EHI, the storm relative helicity (SRH) spatio-temporal analysis followed a gradual increase prior to the tornadogenesis events and was reduced after them.

Towards European-scale convection-resolving climate simulations with GPUs: a study with COSMO 4.19

Abstract. The representation of moist convection in climate models represents a major challenge, due to the small scales involved. Using horizontal grid spacings of O(1km), convection-resolving weather and climate models allows one to explicitly resolve deep convection. However, due to their extremely demanding computational requirements, they have so far been limited to short simulations and/or small computational domains. Innovations in supercomputing have led to new hybrid node designs, mixing conventional multi-core hardware and accelerators such as graphics processing units (GPUs). One of the first atmospheric models that has been fully ported to these architectures is the COSMO (Consortium for Small-scale Modeling) model.Here we present the convection-resolving COSMO model on continental scales using a version of the model capable of using GPU accelerators. The verification of a week-long simulation containing winter storm Kyrill shows that, for this case, convection-parameterizing simulations and convection-resolving simulations agree well. Furthermore, we demonstrate the applicability of the approach to longer simulations by conducting a 3-month-long simulation of the summer season 2006. Its results corroborate the findings found on smaller domains such as more credible representation of the diurnal cycle of precipitation in convection-resolving models and a tendency to produce more intensive hourly precipitation events. Both simulations also show how the approach allows for the representation of interactions between synoptic-scale and meso-scale atmospheric circulations at scales ranging from 1000 to 10 km. This includes the formation of sharp cold frontal structures, convection embedded in fronts and small eddies, or the formation and organization of propagating cold pools. Finally, we assess the performance gain from using heterogeneous hardware equipped with GPUs relative to multi-core hardware. With the COSMO model, we now use a weather and climate model that has all the necessary modules required for real-case convection-resolving regional climate simulations on GPUs.

Evaluation of ABL parametrization schemes in the COSMO, a regional non-hydrostatic atmospheric model over an inhomogeneous environment

The boundary-layer processes characterized through turbulent motion are generally very small scale and sub grid and are represented through different parametrization schemes in the atmospheric models. Here, we evaluate the performance of the atmospheric boundary layer (ABL) parametrization schemes in the COnsortium for Small-scale MOdelling (COSMO), a non-hydrostatic atmospheric model by comparing the model simulations with the concurrent in situ observations over Thiruvananthapuram (8.5 $$^{\circ }$$ N, 76.9 $$^{\circ }$$ E, India). With a view to investigating the role of master length scale (l) in simulation of the ABL features, the default parametrization scheme based on the 1-D diagnostic closure, is modified by adopting a new mathematical formulation for l, and the new approach is implemented in the COSMO model. A total of three parametrization approaches, including the two in-built schemes, are designed and the model simulations with these three distinct schemes are carried out for a total of 9 days. Results obtained from the present study reveal the role of l in the estimation of eddy diffusivity coefficients ( $$K_m$$ and $$K_h$$ ) and the associated vertical turbulent mixing. The study also attempts to highlight the significance of ABL parametrization schemes over non-homogeneous environment, hitherto least explored.

Spatial and temporal variability of clouds and precipitation over Germany: multiscale simulations across the "gray zone"

Abstract. This paper assesses the resolution dependance of clouds and precipitation over Germany by numerical simulations with the COnsortium for Small-scale MOdeling (COSMO) model. Six intensive observation periods of the HOPE (HD(CP)2 Observational Prototype Experiment) measurement campaign conducted in spring 2013 and 1 summer day of the same year are simulated. By means of a series of grid-refinement resolution tests (horizontal grid spacing 2.8, 1 km, 500, and 250 m), the applicability of the COSMO model to represent real weather events in the gray zone, i.e., the scale ranging between the mesoscale limit (no turbulence resolved) and the large-eddy simulation limit (energy-containing turbulence resolved), is tested. To the authors' knowledge, this paper presents the first non-idealized COSMO simulations in the peer-reviewed literature at the 250–500 m scale. It is found that the kinetic energy spectra derived from model output show the expected −5/3 slope, as well as a dependency on model resolution, and that the effective resolution lies between 6 and 7 times the nominal resolution. Although the representation of a number of processes is enhanced with resolution (e.g., boundary-layer thermals, low-level convergence zones, gravity waves), their influence on the temporal evolution of precipitation is rather weak. However, rain intensities vary with resolution, leading to differences in the total rain amount of up to +48 %. Furthermore, the location of rain is similar for the springtime cases with moderate and strong synoptic forcing, whereas significant differences are obtained for the summertime case with air mass convection. Domain-averaged liquid water paths and cloud condensate profiles are used to analyze the temporal and spatial variability of the simulated clouds. Finally, probability density functions of convection-related parameters are analyzed to investigate their dependance on model resolution and their impact on cloud formation and subsequent precipitation.

Identifying sensitivities for cirrus modelling using a two-moment two-mode bulk microphysics scheme

Cirrus cloud genesis is an inherently multiscale and non-linear problem. The synoptic scale provides the environment, the mesoscale determines the forcing and the actual nucleation events occur on a microscopic scale. This makes the parameterisation in numerical weather prediction models a challenging task. In order to improve the prediction of cirrus clouds and ice supersaturation formation in the German Weather Service (DWD) model chain, the controlling physical processes are investigated and parameterised in a new cloud ice microphysics scheme. The new scheme is an extended version of the ice-microphysical scheme operational in the numerical weather prediction models of DWD. The developed two-moment two-mode cloud ice scheme includes state-of-the-art parameterisations for the two main processes for ice formation, namely homogeneous and heterogeneous nucleation. Homogeneous freezing of supercooled liquid aerosols is triggered in regions with high atmospheric ice supersaturations (145–160%) and high cooling rates. Atmospheric small-scale fluctuations are accounted for by use of the turbulent kinetic energy. Heterogeneous nucleation depends mostly on the existence of ice nuclei in the atmosphere and occurs primarily at lower ice supersaturations. Thus, heterogeneously nucleated ice crystals deplete ice supersaturation via depositional growth and can therefore inhibit subsequent homogeneous freezing. The new cloud ice scheme accounts for pre-existing ice crystals, contains a prognostic budget variable for activated ice nuclei and includes cloud ice sedimentation. Furthermore, a consistent treatment of the depositional growth of the two-ice particle modes and the larger snowflakes is applied by using a relaxation time scale method which ensures a physical representation for depleting ice supersaturation. The new cloud ice scheme is used to identify the relative roles of heterogeneous and homogeneous nucleation in the formation of cirrus clouds and ice supersaturation. A parcel model is used in order to investigate the differences between the operational and new cloud ice scheme. The time scales for the homogeneous nucleation event and for the depositional growth are emphasised. The importance of the new ice nucleation scheme is demonstrated by conducting idealised simulations of orographic cirrus in the COSMO (Consortium for Small-Scale Modeling) model environment.

Gridded, locally calibrated, probabilistic temperature forecasts based on ensemble model output statistics

We propose a further refinement of the the non‐homogeneous Gaussian regression approach for temperature, which transforms the output of an ensemble prediction system into predictive Gaussian distributions at each location of interest. Model fitting is partly done within a regression framework using a penalized version of the least‐squares loss function. This is conceptually simpler than the original approach and at the same time is able to prevent overfitting. While calibration is initially performed at observation locations only, geostatistical methods are used to provide predictive distributions on the entire grid. The incorporation of land‐use information in this interpolation scheme further improves predictive performance, even though a simpler statistical model than in the original approach is used. The assessment of predictive performance and calibration is carried out with dynamical forecasts of 2 m temperatures by the COSMO‐DE‐EPS, an application of the COSMO (Consortium for Small‐scale Modeling) model system which covers Germany and neighbouring countries.

Perturbation of initial and boundary conditions for a limited‐area ensemble: multi‐model versus single‐model approach

AbstractWith the aim of developing of an ensemble forecasting system for the short‐range, the validity of a multi‐analysis multi‐model approach for initial‐ and boundary‐condition perturbations has been explored by the Consortium for Small‐scale Modelling (COSMO). This approach has been adopted in the COSMO Short‐Range Ensemble Prediction System (COSMO–SREPS), in which members are driven by a few operational deterministic runs. In this work, we show the extent to which this approach is beneficial with respect to the downscaling of a single‐model global ensemble, adopted by the operational COSMO Limited‐Area Ensemble Prediction System (COSMO–LEPS). Both COSMO–SREPS and COSMO–LEPS are made up of 16 integrations of the COSMO model with 7 km horizontal mesh size. Results based on the 2010–2011 winter season show that the 16 member COSMO–LEPS outperforms the 16 member COSMO–SREPS, which has boundaries perturbed by three global model runs only, but the reverse is true when the three‐member versions of the two systems are compared. Hence, if only a small‐size Limited‐Area Model (LAM) ensemble is feasible, then the multimodel approach for boundaries gives better results. The dependence of the scores on ensemble size is also studied, showing that all the quality measures tend to saturate before the full size is reached. Finally, the impact of combining the two ensembles is assessed, showing that a positive impact results from adding multi‐model‐driven members to COSMO–LEPS.

Quantifying polynya ice production in the Laptev Sea with the COSMO model

Arctic flaw polynyas are considered to be highly productive areas for the formation of sea-ice throughout the winter season. Most estimates of sea-ice production are based on the surface energy balance equation and use global reanalyses as atmospheric forcing, which are too coarse to take into account the impact of polynyas on the atmosphere. Additional errors in the estimates of polynya ice production may result from the methods of calculating atmospheric energy fluxes and the assumption of a thin-ice distribution within polynyas. The present study uses simulations using the mesoscale weather prediction model of the Consortium for Small-scale Modelling (COSMO), where polynya area is prescribed from satellite data. The polynya area is either assumed to be ice-free or to be covered with thin ice of 10 cm. Simulations have been performed for two winter periods (2007/08 and 2008/09). When using a realistic thin-ice thickness of 10 cm, sea-ice production in Laptev polynyas amount to 30 km3 and 73 km3 for the winters 2007/08 and 2008/09, respectively. The higher turbulent energy fluxes of open-water polynyas result in a 5070% increase in sea-ice production (49 km3 in 2007/08 and 123 km3 in 2008/09). Our results suggest that previous studies have overestimated ice production in the Laptev Sea.Keywords: Mesoscale modelling; Laptev Sea; polynya; ice production.(Published: 17 October 2013)Citation: Polar Research 2013, 32, 20922, http://dx.doi.org/10.3402/polar.v32i0.20922