Swiss Contribution to the Annual Joint Wmo

Federal Department of Home Affairs FDHA Federal Office of Meteorology and Climatology MeteoSwiss

SWISS CONTRIBUTION TO THE ANNUAL JOINT WMO TECHNICAL PROGRESS REPORT ON THE GLOBAL DATA- PROCESSING AND FORECASTING SYSTEM (GDPFS) INCLUDING NUMERICAL WEATHER PREDICTION (NWP) RESEARCH ACTIVITIES FOR 2012

1 Summary of highlights

MeteoSwiss deploys a vast palette of tools for fulfilling their forecasting, monitoring, and warning duties as the necessary information needs to cover the very broad temporal range from the recent past to two years into future. MeteoSwiss-run observation-based systems monitor (past few hours) and nowcast (next 0-6 hours) the current weather situation, while nowcasting and NWP systems cover the very short range (0-12 hours) and the short range (out to 72 hours) forecast. The NWP model used at MeteoSwiss is the COSMO model, jointly developed in the COnsortium of Small scale MOdelling. For the medium range (out to 10 days), the extended range (out to 30 days) as well as for the long range (out to 2 years) MeteoSwiss bases their products mainly on the ECMWF, but performs significant post-processing of the model output. Probabilistic forecast products are derived for the extended short range (out to 5 days) from the COSMO ensemble prediction system (EPS) run at ECMWF by a COSMO partner, while for the medium range the ECMWF EPS is mainly used. The following highlights reflect recent progress at MeteoSwiss: • Considerable attention has been given to precipitation estimation through the MeteoSwiss high-resolution rain gauge network applying advanced geostatistical methods to the rain gauge network alone and in combination with the radar precipitation estimates. • Nowcasting capabilities at MeteoSwiss are mainly based on observations, but are complemented by heuristic or numerical models of various kinds. Hereby radars play a central role, contributing to object-based thunderstorm tracking, quantitative areal precipitation estimation (deterministic and probabilistic), nowcasting of precipitation by Lagrangian persistence including a treatment of orographic rain. MeteoSwiss made a major investment in renewing their weather radar network and extending it to better cover the two inner-alpine valleys. Resulting from a recently terminated EUMETSAT fellowship at MeteoSwiss, probabilistic information on convection initiation and evolution to allow early detection of severe storms is assessed from satellite imagery. • The state-of-the-art high-resolution NWP model COSMO is exploited for nowcasting, very short range and short range forecasting in various configurations for a number of years. A project is underway to further increase the spatial resolution of both the deterministic model (mesh size 1 km) and the probabilistic model (mesh size 2-3 km). The model output statistics (MOS) is currently being adapted for the COSMO model. The powerful fieldextra toolbox is a MeteoSwiss developed software, which has been adopted as official COSMO software.

04b0238ef04452bfc46d850ec7ebf574.doc • Recently nowcasting has additionally been complemented by the INCA, as system developed by the Austrian weather service ZAMG and run at MeteoSwiss. • Data assimilation is key in improving the very high-resolution NWP skills. The relatively novel assimilation of temperature, humidity and wind retrievals from atmospheric profiler (e.g. microwave radiometers, LIDARS, ceilometers) is, therefore, one focus of the data assimilation research at MeteoSwiss. • Recently, a COSMO extension for dispersion modelling, COSMO-ART, is used for detailed pollen forecasts and as an emergency tool in case of nuclear power plant accidents. The latter relies on the frequent assimilation of wind profiles into the high-resolution model in order to calculate the dispersion of radioactive substances released. Also, a model is deployed which relates solar UV radiation at the surface to human UV exposure and accounts of the indirect radiation components. • Finally, in order to leverage future supercomputers, current weather prediction codes have to be adapted. To this end a specific project is underway with the aim to reengineer the COSMO code to run efficiently on both massively parallel scalar machines as well as heterogeneous systems with GPUs (Graphical Processor Units). • The main forecast products undergo an objective quality control procedure since the 1980ies. Requiring significant manual input the method has recently been automatized. Also, an index- based verification measure has been chosen following the examples implemented in the UK and Germany.

2 Equipment in use

AUTHORS: MARTIN SCHÄFER MeteoSwiss decided for a server strategy which prefers Linux for application servers, and Windows for both office based server applications and Clients. Windows 7 SP 1 / Office 2010 based HP desktops and Dell laptops are therefore used as PC clients, with a an increasing trend towards laptops. A few remaining Solaris based workstations are being replaced and virtualized in data centers, and, where feasible, migrated to Linux. Access to Solaris and Linux machines from client PC’s is accomplished via X-Windows, using Xmanager and/or X2go. The server Hardware consists of SPARC Enterprise M-Series servers, and HP Blade servers for both Linux and Windows based servers. Virtualization of servers continues, using VMware for Linux and Windows, and a Hitachi based SAN/NAS for storage. We focus on Ubuntu 12.04 LTS as preferred Linux distribution, and RedHat Enterprise Linux 6 to complement the Linux application server environment. Sun Solaris 10 is still in use for legacy applications, like e. g. the Data Warehouse, which is based on Oracle Database. Windows Servers are on 2008 R2 operating system level. Application middleware is mainly based on Oracle Weblogic, and we use Informatica PowerCenter as ETL tool. We just recently implemented Icinga as open source monitoring tool, and upgraded BMC ARS Remedy workflow tool to V. 8.0. Network equipment is based on mainly Cisco products.

04b0238ef04452bfc46d850ec7ebf574.doc 3 Data and Products from GTS in use

AUTHOR: ESTELLE GRÜTTER In 2012 the migration to Table Driven Code Forms could be finalized for all data types provided by MeteoSwiss. At present nearly all observational data from GTS are used. Further in use are GRIB data from Bracknell, Washington and Offenbach as well as T4-charts from Bracknell and Washington. Additionally most of MOTNE and OPMET data are used as well. The number of incoming messages of the majority of the different types has again increased in the last year. An enormous increase can be reported for DRIFTER and AIREP/AMDAR messages, while the number of METAR, GRIB, T4 has decreased slightly, BATHY even to a remarkable amount (~ 30 % less). Typical figures on message input for 24 hours are:

SYNOP, SYNOP Ship 34972 TEMP Part A + B 4677 PILOT Part A + B 1508 METAR 157377 TAF short/long 50422 AIREP/AMDAR 31635 GRIB 36382 T4 (BUFR, FAXG3) 27529 BATHY/TESAC 6648 DRIFTER 17373

4 Forecasting system

4.1 System run schedule and forecast ranges AUTHORS: PHILIPPE STEINER / FRANCIS SCHUBIGER / ROLAND MÜHLEBACH

Very short range INCA: 24 runs per day (hourly), out to 6hrs forecast range; operated by MeteoSwiss

Short range Medium and extended range forecasting are based on external NWP sources, but MeteoSwiss runs their own short-range forecasting system. The core of this system is the non-hydrostatic model COSMO (of the Consortium for Small-Scale Modelling, see section 7). At MeteoSwiss, the model is running operationally at two spatial scales: The regional model COSMO- 7 with a horizontal resolution of about 6.6.km is driven by the ECMWF global model IFS. The local

04b0238ef04452bfc46d850ec7ebf574.doc model COSMO-2, having a horizontal grid spacing of about 2.2 km, is nested in COSMO-7. The nesting of NWP models is illustrated in Figure 1.

ECMWF IFS

COSMO-7

COSMO-2

Figure 1 NWP system of MeteoSwiss

The primary aim of COSMO-2 is to provide forecasts from nowcasting to very short-range time scales, whereas COSMO-7 is used for the short-range time scale. Both COSMO-7 and COSMO-2 have their own assimilation cycle, which is updated in intervals of 3 hours. Three daily 72 hours COSMO-7 forecasts are calculated, based on the 00, 06 and the 12 UTC IFS (main or boundary conditions) runs. One COSMO-2 forecast is computed every 3 hours in parallel to the computation of the necessary COSMO-7 boundary conditions. The lead time of the COSMO-2 forecast starting at 03 UTC is 45h, and 33h otherwise. The cut-off time for all forecasts is 45 minutes. An on-demand mode can be activated, e.g. in case of incident in nuclear power plants. COSMO-2 is then computed hourly with at least 3 hours assimilation and 6 hours forecast. A sophisticated set of scripts controls the whole operational suite, and allows for a very high reliability of the system, with less than 2% of the forecasts requiring manual intervention. This same environment is also used to run parallel suites, to validate proposed modifications to the system, and to facilitate experimentation by the modelling group. The computing resources and expertise are provided by the Swiss National Supercomputing Centre (CSCS, see www.cscs.ch). COSMO-7 and COSMO-2 are calculated on a Cray XE6 equipped with AMD Opteron 12-core processors, and achieve a sustained performance of 270 GFlops on 1079 computational cores for COSMO-2. Pre- and post-processing run on the service nodes of the machine. An additional machine same architecture and with 4032 computational cores is available for as fail-over and for R&D. A large multi-terabytes long term storage is used for archiving purposes and a 1 GBit/s link connects the MeteoSwiss main building with the CSCS (on the other side of the Alps!).

04b0238ef04452bfc46d850ec7ebf574.doc Medium Range IFS: 2 runs per day (00, 12 UTC), up to 240 hrs; operated by ECMWF

Specialized numerical predictions MOS: 2 runs per day (00, 12 UTC), up to 240 hrs; based on IFS and GME; operated by DWD Kalman filtering: 2 m temperature based on IFS, COSMO-7 and COSMO-2; operated by MeteoSwiss 2m-dewpoint temperature based on COSMO-7 and COSMO-2; operated by MeteoSwiss

4.2 Medium range forecasting system (4-10 days)

4.2.1 Data assimilation, objective analysis and initialization

4.2.1.1 In operation None

4.2.1.2 Research performed in this field None

4.2.2 Model

4.2.3 Operationally available Numerical Weather Prediction Products Even if GME and GFS are available in the medium range, ECMWF is the principal forecasting system at this range.

4.2.4 Operational techniques for application of NWP products (MOS, PPM, KF, Expert Systems, etc..)

4.2.4.1 In operation See section 4.3.4.1 for Fieldextra.

04b0238ef04452bfc46d850ec7ebf574.doc 4.2.5 Ensemble Prediction System (EPS) AUTHORS: ANDRÉ WALSER

4.2.5.1 In operation MeteoSwiss does not run a medium range forecasting system, but contributes to the improvement of the limited-area ensemble prediction system COSMO-LEPS based on global ECMWF Ensemble forecasts (EPS) and on the COSMO Model. COSMO-LEPS has been developed at ARPA-SIMC, Bologna, and runs operationally at ECMWF (see section 7.1.1). It makes probabilistic high-resolution short to early-medium range (5.5 days) forecasts available at MeteoSwiss.

4.2.5.2 Research performed in this field See section 7.1.2.

4.2.5.3 Operationally available EPS Products COSMO-LEPS products are visualized in the form of probability maps, stamp maps and meteograms for various parameters. The maps complement the deterministic COSMO products, while the meteograms combine the output from both systems for a single point. In addition, the COSMO-LEPS forecasts are calibrated with a reforecast dataset including 20 years of forecasts (1 member) for the years 1989-2008. Calibrated products are available for precipitation, temperature and wind gusts in the form of probability maps for certain thresholds and return periods, meteograms and so-called warngrams which show for a given location the probabilities for a set of return periods for the 5-day forecast range with 24h sliding windows.

4.3 Short-range forecasting system (0-72 hrs) AUTHORS: PHILIPPE STEINER / FRANCIS SCHUBIGER

4.3.1 Data assimilation, objective analysis and initialization

4.3.1.1 In operation Data assimilation of COSMO is based on the nudging or Newtonian relaxation method, where the atmospheric fields are forced towards direct observations at the observation time. Balance terms are also included: (1) hydrostatic temperature increments balancing near-surface pressure analysis increments, (2) geostrophic wind increments balancing near-surface pressure analysis increments, (3) upper-air pressure increments balancing total analysis increments hydrostatically. A simple quality control using observation increments thresholds is in action. Currently, the following conventional observations are assimilated both for COSMO-7 and COSMO-2: synop/ship/buoys (surface pressure, 2m humidity, 10m wind for stations below 100 m above msl), temp/pilot (wind, temperature and humidity profiles), airep/amdar (wind, temperature) and wind profiler data. COSMO-2 additionally assimilates radar data, using the 2-dimension latent heat nudging scheme. An empirical quality function for radar quantitative precipitation estimates is in operation, which is based on the frequency of signal occurrence of a particular radar pixel (D. Leuenberger et al, 2010, and references therein). MeteoSwiss uses its own snow analysis which is derived from MSG satellites combined with dense observations. A multi-layer soil model with 8 layers for energy and 6 for moisture is used. Finally, the vegetation and ozone fields are based on climatic values.

04b0238ef04452bfc46d850ec7ebf574.doc The MeteoSwiss Data Warehouse (DWH) is being used as the operational data base for conventional observations. Data from DWH is retrieved at CSCS in BUFR format, and converted to the NetCDF format with the bufrx2netcdf software of DWD. The number of assimilated conventional observations is monitored.

4.3.2 Model

4.3.2.1 In operation A thorough description of the COSMO Model itself can be found on the COSMO web site (see section 7.1). It is a primitive equation model, non-hydrostatic, fully compressible, with no scale approximations. The prognostic variables both for COSMO-7 and COSMO-2 are the pressure perturbation, the Cartesian wind components, the temperature, the specific humidity, the liquid water content, cloud ice, rain, snow and turbulent kinetic energy. COSMO-2 furthermore uses a prognostic graupel hydrometeor class in the microphysical parameterization. COSMO-7 uses the Tiedtke scheme to parameterize convection, whereas in COSMO-2 convection is parameterized by a shallow convection scheme, and the deep convection is explicitly computed. The model equations are formulated on a rotated latitude/longitude Arakawa C-grid, with generalized terrain-following height coordinate and Lorenz vertical staggering. Finite difference second order spatial discretization is applied, and time integration is based on a third order Runge-Kutta split-explicit scheme. Fourth order linear horizontal diffusion with an orographic limiter is active for wind in COSMO-7 only. Rayleigh-damping is applied in the upper layers. For the advection of the humidity constituents a symmetric Strang-splitting in all 3 directions is used at each time step. COSMO-7 is calculated on a 393 x 338 mesh with a 3/50° mesh size (about 6.6 km), on a domain covering most of Western Europe. In the vertical a 60 layers configuration is used; the vertical resolution in the lowest 2 km of the atmosphere increases from about 10 m up to 250 m. The main time step is 60 seconds. COSMO-2 is calculated on a 520x350 mesh, with a 1/50° mesh size (about 2.2 km), on a domain which is centred on the Alps. The COSMO-7 mesh is chosen in such a way that on the integration domain of COSMO-2, each COSMO-7 grid point coincides with a grid point of COSMO-2. COSMO-2 uses the same vertical configuration as COSMO-7. The main time step is 20 seconds. Table 1 summarizes the specifications of the new COSMO system.

04b0238ef04452bfc46d850ec7ebf574.doc Table 1 Specification of COSMO-7 and COSMO-2

COSMO-7 COSMO-2 Number of grid points and levels 393 x 338, 60L 520x350, 60L Horizontal mesh size 3/50° ~ 6.6km 1/50° ~ 2.2km Time step 60s 20s Conv. Observations + Data Assimilation Conv. Observations Radar Since March 2011, MeteoSwiss also provides daily forecasts of birch pollen, which are based on the numerical pollen dispersion model COSMO-ART of the Karlsruhe Institute of Technol-ogy (KIT) (B. Vogel et al, 2009, and H. Vogel et al, 2008). Since spring 2012, grass pollen forecasts are also produced. COSMO-ART provides spatially and temporally highly resolved pollen forecasts hitherto not available.

4.3.2.2 Research performed in this field

Development of a deterministic 1km implementation of COSMO Many of the key physical processes of Alpine meteorology (valley winds, orographically influenced/triggered precipitation, convection, fog) are still at least partly unresolved in COSMO-2, which employs 2.2km horizontal grid spacing. Apart from the canonical improvement of the resolution of topography and land-use, evidence from research (e.g. Langhans et al. 2012, Bryan et al. 2007) also suggests improvements in near-surface winds as well as convection and entailing precipitation. Thus, since beginning of 2012 and within the framework of the COSMO-NExT project, MeteoSwiss is developing a deterministic 1km implementation of COSMO named COSMO-1. The model domain will span the greater Alpine region and similar to COSMO-2, the model will be implemented using a rapid update cycle (RUC) with a new forecast every 3h. Research currently focusses on the dynamical core (new fast-waves solver), turbulence parameterization (representation of turbulence in the “grey-zone”), external parameters (utilization of new datasets for topography and soil-type). Redesign of the COSMO model code for future HPC architectures The available computer power is the most important constraint limiting the horizontal resolution, the complexity of the model system, and the number of ensemble members of numerical weather prediction and climate models. Emerging and future supercomputing architectures are expected to bring several changes: an increasing number of compute cores competing for resources such as memory bandwidth and communication bandwidth, only slow increases of the throughput of I/O subsystems, heterogeneous computes nodes with accelerators such as GPUs. In order to leverage future supercomputers current weather prediction codes have to be adapted. To this end, the HP2C COSMO and HP2C OPCODE projects carried out in the framework of the Swiss HP2C (High Performance High Productivity Computing) initiative as well as the COSMO priority project POMPA (Performance on Massively Parallel Architectures) aim at reengineering the numerical weather prediction and regional climate model COSMO (Consortium of Small-Scale Modelling) with the goal to run efficiently on both massively parallel scalar machines as well as heterogeneous systems with GPUs (Graphical Processor Units). The projects encompass a redesign of the model code in order to use memory bandwidth more efficiently, have the capability to run the same code on CPUs and GPUs, and improve the I/O strategy. Currently, the GPU version of COSMO (using compiler directives in some parts and a domain specific embedded language (DSEL) in others) is being tested on a prototype hardware (fat node with 2 CPUs and 8 NVIDIA Kepler GPUs).

04b0238ef04452bfc46d850ec7ebf574.doc 4.3.3 Operationally available NWP products A suite of post processing modules is available: • Kalman filtering of model output for 2 meter temperature, dew point temperature and windspeed on different levels (also on hub heights of wind turbines) • thunderstorm prediction derived by a learning system (with the boosting method), • probabilistic precipitation forecasts using the neighbourhood method, • visualization software based on the NCL package (NCAR command language) and on in- house developments at MeteoSwiss; static maps, 2- and 3-dimensional loops with texture based flow visualization are created; • a trajectory model providing guidance on transport route (hot air balloons, pollutants); • a Lagrangian particle dispersion model to calculate dispersion and deposition of radioactive materials, having also a backward mode to calculate the origin of pollutants or of pollen. Based on these modules, a standard set of products is provided to the MeteoSwiss bench forecasters and used as guidance for short-range forecasts. In case of necessity the two last modules can be run by the on-duty forecasters at any time (on-demand mode). Besides that a large quantity of tailor made products, based on direct model output, are disseminated to internal and external clients.

4.3.4 Operational techniques for application of NWP products

4.3.4.1 In operation Fieldextra Software: MeteoSwiss has developed a flexible tool called "fieldextra" which supports the manipulation of NWP model data, especially COSMO model data, and gridded observations. This tool is used both as pre- and post-processing component of the MeteoSwiss NWP production suite, is an official COSMO software. Fieldextra is designed as a toolbox, with the production as primary target. A set of primitive operations are provided, which can be combined in any meaningful way. During execution the input data is read once only, but as many output as desired are produced. The program is controlled by a collection of Fortran namelists, stored in a control data file. Comprehensive checks of user defined parameters are performed, and a rich set of diagnostic is produced. A lot of effort has been invested to optimize both the memory footprint and the execution time, and to offer a robust well written and well tested code. Simple data processing and more complex data operations are supported, for example: selecting data satisfying some complex condition, comparing or merging multiple fields, horizontal and vertical re- gridding, computation of regional conditions, computation of stability indices, computation of EPS derived quantities. Both point values and gridded fields can be generated in output, and GRIB 1, GRIB 2, NetCDF and a rich set of ASCII formats are offered.

4.3.4.2 Research performed in this field Within the MeteoSwiss project “COSMO-MOS”, a regression based statistical post-processing system for improved local forecasts has been designed that is particularly tailored to limited area NWP models with frequent model version releases. So far, this system supports two statistical approaches for modelling the expected forecast error. First, multiple linear regression with automatic model selection is particularly suitable for target variables that can be easily transformed into approximated normal distributions (e.g. temperature, wind speed etc). Second, an extended logistic regression approach (as being suggested by Wilks, Meteoro-logical Applications, 2009) is provided for target variables that are related to hazard assess-ments, such as wind gusts, that turns a deterministic forecast into a full

04b0238ef04452bfc46d850ec7ebf574.doc calibrated and sharp probability distribution for each forecast. The best suited set-up of this system including sampling strategies, length of training period, selection of potential predictors and update cycle of the estimation procedure is identified with a number of sensitivity studies.

4.3.5 Ensemble Prediction System

4.3.5.1 In operation See section 4.2.5.1

4.3.5.2 Research performed in this field Within the MeteoSwiss project “COSMO-NExT” an ensemble system with a convection-permitting mesh-size of about 2 km is in development. It is planned to run the ensemble with about 20 members with perturbed initial conditions by an ensemble data assimilation cycle using a Local Ensemble Transform Kalman Filter (LETKF), perturbed lateral boundary conditions by a global EPS and perturbed model physics by stochastically perturbed parameterized tendency and parameter perturbations. The system will run twice a day with a forecast range of 120 hours.

4.3.5.3 Operationally available EPS Products See section 4.2.5.3

4.4 Nowcasting and Very Short-range Forecasting Systems (0-6 hrs) AUTHORS: URS GERMANN / ALESSANDRO HERING / PAOLO AMBROSETTI

4.4.1 Nowcasting system

4.4.1.1 In operation Tracking and characterization of convective cells by radar (system TRT) MeteoSwiss runs operationally the real-time object-oriented nowcasting tool TRT (Thunderstorms Radar Tracking), as a part of its severe thunderstorms nowcasting, warning and in-formation system. For a detailed description see “WMO_GDPS-Report_2006”.

Quantitative precipitation estimation by radar (product RAIN) The quantitative precipitation estimate (QPE) nowcasting radar product RAIN is the best radar estimation of precipitation amount on the ground in Switzerland. For a detailed description see “WMO_GDPS-Report_2006”.

Automatic heavy precipitation alert system (system NASS) The multiple-radar-based nowcasting application NASS is specifically designed for situations with heavy precipitation. NASS was implemented to generate automatic alerts whenever accumulated radar rainfall exceeds a predefined threshold for periods of 3, 6, 12 and 24 hours. For a detailed description see “WMO_GDPS-Report_2006”.

04b0238ef04452bfc46d850ec7ebf574.doc 4.4.1.2 Research performed in this field Ensemble technique for radar precipitation fields (technique REAL) As part of the WMO-WWRP forecast demonstration project MAP D-PHASE and the European concerted research action COST-731 MeteoSwiss developed an ensemble technique to characterize the residual errors in radar precipitation fields. Each member of the radar ensemble is a possible realization of the unknown true precipitation field given the observed radar field and knowledge of the space-time error structure of radar precipitation estimates. Feeding the alternative realizations into a hydrological model yields a distribution of response values, the spread of which represents the sensitivity of runoff to uncertainties in the input radar precipitation field. The presented ensemble generator is based on singular value decomposition of the error covariance matrix, stochastic simulation using the LU decomposition algorithm, and autoregressive filtering. The real-time implementation of the radar ensemble generator coupled with a semi-distributed hydrological model in the framework of MAP DPHASE is one of the first experiments of this type worldwide. For a detailed description see: Germann et al, Q. J. R. Meteorol. Soc., 135, 445-456, 2009.

Nowcasting heavy orographic precipitation using Doppler radar and radiosounding (project COST-731) MeteoSwiss developed as part of COST-731 a novel heuristic system for nowcasting heavy precipitation in the Alps. The system uses as input estimates of the mesoscale wind field as derived from real-time Doppler radar measurements and information on air mass stability from radio-soundings and ground stations. Both mesoscale flow and upstream air mass stability are predictors of the amounts and geographic distribution of heavy orographic precipitation, and can therefore be exploited for nowcasting. Since 2012 the system runs at MeteoSwiss in real-time, in a pre-operational mode. For a detailed description see: Panziera L, Germann U. 2010. The relation between airflow and orographic precipitation on the southern side of the Alps as revealed by weather radar. Q. J. R. Meteorol. Soc. 136: 222–238. DOI:10.1002/qj.544

Context and Scale Oriented Thunderstorm Satellite Predictors Development (project COALITION) Through a 3-year fellowship funded by EUMETSAT MeteoSwiss is developping nowcasting applications into an entity-oriented model, which merges severe convection predictors retrieved from different sources (MSG, Weather Radars, NWP, lightning climatology and orographic gradients) with evolving thunderstorm properties. The heuristic model will calculate probabilistic information about time, space and intensity evolution of severe convection for use by decision makers. Focus is given to early detection of severe storms over the European Alpine region. The project was terminated in 2012 and at MeteoSwiss the system runs now in real-time, in a pre-operational mode.

Improving Preparedness and Risk Management for Flash Floods and Debris Flow Events (project IMPRINTS) Over complex terrain such as the Alps current nowcasting systems based on Lagrangian persistence of radar precipitation fields fail to produce useful forecasts, because the orography interferes with the evolution of precipitation, in particular by means of blocking and enhancement. As part of the FP7 research project IMPRINTS (2009-2012), MeteoSwiss investigates orographic forcing of precipitation and incorporate the findings into current Lagrangian persistence nowcasting systems. If successful, the resulting radar nowcasting system will be implemented in the Swiss radar data processing chain and will be extended by ensemble techniques and an algorithm for blending radar nowcasts with NWP model output

04b0238ef04452bfc46d850ec7ebf574.doc For a detailed description see: Panziera, L., U. Germann, M. Gabella and P. V. Mandapaka, 2011. NORA–Nowcasting of Orographic Rainfall by means of Analogues. Q. J. R. Meteorol. Soc. 137: 2106–2123 Mandapaka, P.V., U. Germann, L. Panziera and A. Hering, 2011. Can Lagrangian Extrapolation of Radar Fields Be Used for Precipitation Nowcasting over Complex Alpine Orography?, Weather and Forecasting, 27: 28-49 Mandapaka, P.V., U. Germann, L. Panziera, 2013. Diurnal cycle of precipitation over complex Alpine orography: inferences from high resolution radar observations. Quarterly Journal Royal Met. Soc., published online. DOI: 10.1002/qj.2013

Real-time radar-raingauge merging (project CombiPrecip) CombiPrecip is an ongoing project which aims at producing accurate precipitation estimation maps by combining raingauges and radar data in real-time. The underlying technology is geostatistical in nature, where both spatial and temporal information has been taken into account in a so called co- kriging with external drift modelling scheme. The technique is coupled with innovative engineering to mitigate artifacts in the extrapolation regime and in the presence of strong convective cells where lack of sufficient representativeness of raingauge data typically causes problems. CombiPrecip is in a state of continuous improvement, but current results show a significant improvement over radar-only rainfall maps especially in terms of bias. For a detailed description see: Sideris I.V., M. Gabella, R. Erdin and U. Germann, 2013. Real-time radar-raingauge merging using spatiotemporal co-kriging with external drift in the alpine terrain of Switzerland, Q. J. Roy. Meteor. Soc., accepted.

4.4.2 Models for Very Short-range Forecasting Systems

4.4.2.1 In operation Integrated Nowcasting through Comprehensive Analysis (INCA) The nowcasting analysis and forecasting system INCA, developed by the Austrian NWS ZAMG is run operationally at MeteoSwiss. This novel approach produces meteorological fields, with high resolution in time and space (gridded values) for several parameters, incorporating available information like numerical models and diverse kinds of observation (both in-situ and remote sensed), as well as high resolution orography. Several customer oriented products have been developed und made operational, particularly for the rain and snow forecast of in the Nowcasting range (both internally and externally)

4.4.2.2 Research performed in this field Integrated Nowcasting through Comprehensive Analysis (INCA) Some optimization in the model data interpolation in INCA have been introduced.

04b0238ef04452bfc46d850ec7ebf574.doc 4.5 Specialized numerical predictions

4.5.1 Assimilation of specific data, analysis and initialization (where applicable)

4.5.2 Specific Models (as appropriate related to 4.5) AUTHORS: PHILIPPE STEINER / FRANCIS SCHUBIGER / ALEXANDER HAEFELE / LAURENT VUILLEUMIER A) COSMO-ART provides spatially and temporally highly resolved pollen forecasts. B) Model relating solar UV ground radiation measurements to related human UV exposure. C) CN-MET is an integrated analysis and forecasting system consisting of a high resolution numerical weather prediction model and a dense observation network. CN-MET provides meteorological information for dispersion calculations in the case of an accident in a nuclear power plant.

4.5.2.1 In operation A) The pollen module of the numerical dispersion model COSMO-ART (Vogel et al. 2008) was developed by the Karlsruhe Institute of Technology (KIT) in collaboration with MeteoSwiss. Since 2011 MeteoSwiss performs daily runs with COSMO-ART to provide high-resolution birch pollen forecasts. Recently these were complemented by grass pollen forecasts. An example is given in Figure 2.

Figure 2 Forecast of birch pollen concentrations for Switzerland as generated by the numerical dispersion model COSMO-ART.

04b0238ef04452bfc46d850ec7ebf574.doc B) No C) Yes

4.5.2.2 Research performed in this field A) MeteoSwiss is involved in a number of research activities within the continuous development of COSMO-ART. Key areas include the modeling of pollen emissions namely parameterizations of pollen emission and descriptions of the pollen season. Further, precise plant distribution maps are a prerequisite for successful application of COSMO-ART. Pauling et al. (2011) developed a set of methods designed to provide this important input to COSMO-ART. Furthermore, the reduced pollen production with increasing altitude is taken into account. B) Solar ultraviolet (UV) radiation is one of few environmental exposures that can both cause and protect against diseases. While UV exposure can prevent diseases of vitamin D insufficiency, it can cause eye diseases and is responsible for 50–90% of all skin cancers (World Health Organization (WHO), 1992). For a given individual, the anatomical distribution of UV exposure is highly heterogeneous, poorly correlated to ground irradiance, and depends on the time of exposure and orientation to the sun (Parisi, Kimlin, Wong, & Fleming, 1996). Variations in UV doses received across individuals are even greater as they are strongly influenced by behavioral and host factors such as posture, orientation to the sun, skin complexion, clothing and other sun-protective behaviors (Autier, Boniol, & Doré, 2007) (Parisi, Kimlin, Lester, & Turnbull, 2003). UV protection messages often focus on direct radiation and short-term, acute exposure (avoidance of erythema), implicitly assuming that direct UV radiation is the key contributor to the overall UV exposure. However, little is known regarding the relative contribution of the direct, diffuse and reflected UV radiation to the anatomical exposure as individual dosimetric measurement cannot separate the three radiation components. A model, Simulating UV Exposure (SimUVEx) (Vernez, Milon, Francioli, Bulliard, Vuilleumier, & Moccozet, 2011), has been developed that uses continuous datasets of erythemally-weighted UV ground irradiance to estimate the dose and anatomical distribution of UV received by exposed individuals (Fig. 1). The data are fed into a ray-tracing algorithm in which direct I(t), diffuse D(t) and reflected R(t) (reflection from ground) components are separately taken into account. Exposure to various anatomical locations is obtained by exposing a 3D virtual manikin to a local radiation sphere, discretized into n sub-surfaces, for a given duration. The exposure levels and doses computed during the simulation can be visualized as comprehensive 3D images using expressive rendering techniques.

04b0238ef04452bfc46d850ec7ebf574.doc Figure 3 Schematic view of the SimUVEx model.

Three ambient irradiance time series are required as input parameters to the model for the direct, the diffuse and the ground reflected irradiance (W/m2), as well as the sun position, defined by its azimuth p(t) and zenith d(t) angles. The model derives radiances distributed on a local sphere from these inputs using simplifying hypotheses. Such input data can be obtained from meteorological stations equipped with multiple broadband radiometers (e.g. one for direct, one for diffuse radiation with a shadowing disc and one turned upside down for reflected radiation). Alternatively these data can be obtained from atmospheric radiation transfer models (RTM). In clear-sky situations, RTM give accurate estimates of the irradiance components, but their use is more questionable in cloudy situations. For complex cloudy situations, efforts are currently devoted to reconstructing UV irradiance based on proxies. Deriving UV exposure from irradiance is based on simplifying the reflected R(t) and diffuse D(t) components as hemispherical quasi-isotropic sources (azimuthal isotropy, and limited zenith dependence) with time-dependent intensities (Figure 3, center). The direct component I(t) is described as a parallel source of radiation varying in intensity with time and in direction with the sun position. Traditional 3D human modeling and animation approaches, based on articulated skeleton and 3D surface skinning (Di Giacomo, Kim, Moccozet, & Magnenat-Thalmann, 2007) are used to produce a 3D virtual manikin with variable morphologies in standard working positions. Postures are defined as a set of angles values defined at joint articulations of a simplified skeleton. Its surface is depicted as a single 3D mesh of connected triangles, whose density depends on the desired resolution (here optimizing computing time and level of body detail results in ~4000 triangles). Manikin models are produced with the open source software MakeHuman (MakeHuman). Each triangle of the manikin receives a specific amount of radiation according to its exposure to the different virtual sources of radiation. This is the average of the exposures computed at the vertices, which are functions of radiation intensity, orientations of triangles and light sources as well as shading from body parts (Figure 4).

Figure 4 Computation of energy received at a vertex

C) The meteorological surveillance of the four nuclear power plants in Switzerland is of first importance in a densely populated area such as the Swiss Plateau. The project “Centrales Nucléaires et Météorologie“ CN-MET aimed at providing a new security tool based on one hand on the development of a high resolution numerical weather prediction (NWP) model. The latter is providing essential nowcasting information in case of a radioactive release from a nuclear power plant in Switzerland. On the other hand, the model input over the Swiss Plateau is generated by a dedicated network of surface and upper air observations including remote sensing instruments (wind profilers

04b0238ef04452bfc46d850ec7ebf574.doc and temperature/humidity passive microwave radiometers). This network is built upon three main sites ideally located for measuring the inflow/outflow and central conditions of the main wind field in the planetary boundary layer over the Swiss Plateau, as well as a number of surface automatic weather stations (AWS). The network data are assimilated in real-time into the COSMO-2 high-resolution NWP model using a rapid update cycle described in section 4.1. This set-up has replaced the former security tool based on in situ observations (in particular one meteorological mast at each of the power plants) and a local dispersion model. It is used to forecast the dynamics of the atmosphere in the planetary boundary layer (typically the first 4 km above ground layer) and over a time scale of 24 h. This tool provides at any time (e.g. starting at the initial time of a nuclear power plant release) the best picture of the 24-h evolution of the air mass over the Swiss Plateau and furthermore generates the input data (in the form of simulated values substituting in situ observations) required for the local dispersion model used at each of the nuclear power plant locations.

4.5.3 Specific products operationally available None

4.5.4 Operational techniques for application of specialized numerical prediction products (MOS, PPM, KF, Expert Systems, etc..) (as appropriate related to 4.5)

4.5.5 Probabilistic predictions (where applicable)

4.5.5.3 Operationally available probabilistic prediction products None

4.6 Extended range forecasts (ERF) (10 days to 30 days) AUTHOR: CHRISTOPH SPIRIG

04b0238ef04452bfc46d850ec7ebf574.doc 4.6.1 Models

4.6.1.1 In operation Within the framework of the national NCCR climate research programme, MeteoSwiss established a quasi-operational monthly forecasting system over the past years. This is based on forecast data from the ECMWF monthly prediction system VarEPS. The system was deployed in autumn 2008 and has since become fully operational. Calibration and visualization techniques are applied in a similar way as in the long range seasonal forecasts described below (Section 4.7). Since December 2011 MeteoSwiss has additionally implemented the updated second weekly run of the ECMWF monthly forecast system. This runs on an operational basis every Monday and essentially provides an update to the main forecast on Thursday.

4.6.1.2 Research performed in this field In a recent research project at MeteoSwiss, the ECMWF monthly forecasts were coupled to a statistical weather generator and an impact model in order to forecast agricultural pest on a lead-time up to 32 days. Results showed a clear skill improvement over the full forecast range when incorporating monthly forecasts as compared with deterministic benchmark forecasts using climatological information for predicting the timing of insect life phases.

Figure 5 : Improvement of root mean square error (RMSE) in predicting the timing of insect life phases when using monthly forecasts “MOFC-WG” instead of climatology “benchmark” (Hirschi et al., 2012)

4.6.2 Operationally available NWP model and EPS ERF products Operational products based on ECMWF VarEPS include maps of weekly categorical probability forecasts of surface temperature, precipitation and geopotential height over various regions as well as tercile data as tables for selected station locations and regional averages. These products are provided to customers of MeteoSwiss. Additionally, since 2010, probabilistic forecasts of average temperature and precipitation tendencies of forecast days 12-18 are provided weekly (on Fridays) to the public in the form of a “weekly climate outlook” on the internet.

04b0238ef04452bfc46d850ec7ebf574.doc Figure 6 Example of a weekly climate outlook for Southern Switzerland.

4.7 Long range forecasts (LRF) (30 days up to two years) AUTHOR: CHRISTOPH SPIRIG

4.7.1 In operation MeteoSwiss issues long range forecasts (up to 7 months) on the basis of the ECMWF seasonal forecast model system (System 4). The model data are post-processed, evaluated and disseminated by MeteoSwiss. The post-processing technique of climate model output includes a climate-conserving recalibration technique (CCR), which has been developed by MeteoSwiss within the framework of the national NCCR climate research programme. ECMWF releases the forecasts on the 8th of each month, the MeteoSwiss products based on these forecasts are issued the following day.

4.7.2 Research performed in this field Building upon our experiences in evaluating seasonal forecasts and tailoring products for both commercial customers and the general public, we participate in the EU FP7 framework project EUPORIAS. EUPORIAS is a four-year project coordinated by the UK Met Office, aiming at developing reliable climate predictions for a number of key sectors (including water, energy, health, transport, agriculture, and tourism) on timescales from seasons to years ahead. The EUPORIAS project aims to maximize the usefulness of seasonal to decadal (S2D) forecasts by assessing users’ needs and using this information to develop tools to forecast impacts. MeteoSwiss is involved in several work packages, namely in downscaling/post-processing of long term forecasts, development of relevant climate indicators and investigation of user’s needs.

4.7.3 Operationally available EPS LRF products The operational products of seasonal forecasts (up to 7 months) include climagrams, probability charts and tercile data for surface temperature, precipitation and geopotential height, which are available for customers. The skill of seasonal temperature forecasts is also monitored and is provided in the form of skill maps. For the wider public, probabilistic seasonal forecast information is issued in form of a quarterly climate outlook bulletin for Switzerland. The recalibrated seasonal forecast products (using the method of CCR, Weigel et al., 2008) are available for surface temperature and issued as climagrams, probability charts and tercile data.

04b0238ef04452bfc46d850ec7ebf574.doc Figure 7 Examples of seasonal forecast products, top: Climagrams of temperature (recalibrated version on the right), bottom: Map of generalized discrimination skill score (Weigel and Ma-son, 2011)

5 Verification of prognostic products

AUTHOR: LUDWIG Z’GRAGGEN

5.1 Removal of OPKO and KOMIFRI by the new verification scheme MOVI Although OPKO is an understandable quote, this verification scheme has some disadvantages. First, it takes 1.5 hours to verify solely one day. Secondly, the objectivity is not always guaranteed, because a text must be changed into a number. KOMIFRI evaluates medium range forecasts. For the region East, West and South only two stations per region are taken into account. For this reason, the representativeness for a larger region is not given. To eliminate these weaknesses, it has been decided to introduce a new verification scheme. Attention was paid that the verification scheme is internationally accepted. For this reason, the scheme, which is used at the British and German weather service, was considered to be suitable and, therefore, essentially adopted. The new verification scheme is named “MOVI”, this means “MeteoSwiss Objective Verification index”. “MOVI” describes an index, which is based on the so called GlobalScore. GlobalScore puts together the reduction of the variance for the parameter temperature and the Equitable Threat Score for the

04b0238ef04452bfc46d850ec7ebf574.doc parameters sunshine duration, precipitation and wind. The index is set to 100 for the reference period from October 2009 to September 2011. All other evaluated time periods relate to the reference period.

Results Short range forecast

Last year, only the 1-year period from December 2010 to November 2011 was taken into account. The reason, that the 2-year period from December 2009 to November 2011 was not taken into account, was the strong similarity of this period with the reference period from October 2009 to September 2011. The two periods differs only by two month. Now and in the future, only two-year period will be compared with the reference period.

Figure 8 Monthly values of GlobalScore from Dec. 2010 to Nov. 2012 for forecast day 1 at the forecast centers Geneva (West), Zürich (East), and Locarno-Monti (South)

The period from December 2010 to November 2012, as Figure 8 shows, has lower GlobalScore values from April 2012 to November 2012 than before, especially in the region of the South. This is due to the instable weather. Situation with anticyclonic weather were almost completely missing in this period. Therefore, the forecast was rather difficult. In the south of Switzerland, the forecast of the temperature is very difficult do to the lakes of cold air in the Engadin. When the forecast of the cloudiness is wrong, what can easily happen in unstable weather situations, a great difference of the minimum temperature up to 10 degrees results. In such cases, the reduction of variance and consequently also the GlobalScore receives a rather low value as we can see on the July 2012. Nevertheless, the MOVI value is with 102.1 still higher than the reference period with a value of 100, see Table 1 und 2.

Table 2 Global Score and MOVI of the reference period from October 2009 to September 2011 for forecast day 1 forecast center West East South Total GlobalScore 0.363 0.362 0.330 0.352 MOVI 100 100 100 100

Table 3 Global Score and MOVI from December 2010 to November 2012 for forecast day 1 forecast center West East South Total

04b0238ef04452bfc46d850ec7ebf574.doc GlobalScore 0.376 0.372 0.331 0.360 MOVI 103.5 102.7 100.3 102.1

04b0238ef04452bfc46d850ec7ebf574.doc Medium range forecast

Figure 9 Monthly values of GlobalScore from Dec. 2010 to Nov. 2012 for forecast days 2-5 at the forecast centers Geneva (West), Zürich (East), and Locarno-Monti (South).

Generally, the GlobalScore values for forecast day 2-5 than of for forecast day 1. This is not unexpected. It’s clear that the quality of forecast decreases when forecast days are further in the future.

Also in the medium range forecast, the values of the GlobalScore were rather low from the spring 2012 to the End of the year 2012, see Figure 9. The influence of the unstable weather in this period is even greater than in the short range forecast. This is not surprising because the forecast becomes increasingly dependent from the model outputs. When the weather changes very much in a short time period, the model outputs are uncertain and are varying strongly from one run to the other. For this reason, the quality of forecast is lower than in anticyclonic situations. In the regions East on South, the period from December 2010 to November 2012 the MOVI is with values of 97.7 and 98.3 even inferior than the reference period, see Table 3 and 4. Only in the region West, the MOVI value is with 101.5 better than the reference period. But, also in this region, the period of Spring 2012 to the End of 2012 is worse, especially the September 2009 with the lowest GlobalScore value of the whole period and of all regions, as we can see in the Figure 9.

5.2 Research performed in this field None

6 Plans for the future (next 4 years)

6.1 Development of the GDPFS AUTHOR: ESTELLE GRÜTER

04b0238ef04452bfc46d850ec7ebf574.doc 6.1.1 Major changes in the operational DPFS which are expected in the next year An intensive data exchange with other governmental and private platforms is further promoted. The data is mainly used for the prevention of natural hazards and warnings. The optimization of the warning infrastructure and the availability of more data is ongoing work in progress. To guarantee more real-time data, there’s further on work being done at the whole end-to-end data chain. Control data processing tools will be furthermore developed and implemented as service oriented Infrastructures and especially in form of web services. Data quality control and enhancement will be completed by spatial interpolation methods, as well as an automating of the manual data treatment process, in order to increase the efficiency. The usage of the analytical grid database within the data ware house (DWH) will still increase and in addition be adapted to upcoming use-cases. In 2013 scanning activities for historical data collections will be completed in order to obtain a higher availability of historical data for research and analyses of past weather phenomenon. A quality assurance procedure for climatological and meteorological point measurement stations was developed and will be introduced this year. In order to evaluate more systematically users requirements the Rolling Review of Requirement (RRR) process of WMO will be implemented and integrated in the existing Change Management workflow of MeteoSwiss.

6.1.2 Major changes in the Operational DPFS which are envisaged within the next 4 years The processing system, having a DWH structure since 2002, is the basis for the national „Meteorological and Climate Data Warehouse“. The integration of different measurement technologies and platforms is further envisaged, as well as the integration of measuring systems of public and private partners. Moreover, the further integration of measured data und metadata will be promoted. An important issue within the next 4 years will be the time optimization of the real-time data availability. In this process the whole data chain will be considered from end-to-end and optimized. For better meeting the needs of climatologic research and homogenization of historical data, the systematic information system for historical metadata of stations will be implemented. The quality acceptance procedure for measurement stations will be implemented and the process for doing so nationally established. Software to guide the procedure will be developed and shall be available for any operator of a measurement station. A better combination of remote sensing measurement methods and point data measurements (e.g. the combination of radar and rain gauges) will be promoted. Another important step in the future development is the use of the DWH for numerical weather predictions in models such as COSMO. This means assimilation of observation data into the model, verification and analysis of predictions with the help of observation data and the storage of selected data in the DWH. Running applications using ECMWF Data shall be migrated to official dissemination (RMDCN) that is made for business relevant cases. A superordinate architecture will be created for production tools using meteorological and climate data. This means running applications will be taken into account and adapted to the future needs for data extraction and product creation within this architecture. There will be an overall evaluation of climatologic and meteorological production tools in terms of operational costs and utility. This will be

04b0238ef04452bfc46d850ec7ebf574.doc combined with a future business strategy of MeteoSwiss. In the long term, some of these tools will be consolidated some removed. Maybe some commercial of the shelf tools (such as scheduling tools and reporting tools) will be implemented, if ever they are more cost efficient. In addition, a solution will be developed to store, archive and administer products.

New implemented tools for forecasting and warning will be consolidated in terms of business relevance. There will be a high effort to increase the cost efficiency and to create more private-public- partnerships. The product delivery to the private sector will be focused especially as an input for decision making in public and industries. A systematic and long-term archiving of data will be prepared and delivered to the Swiss federal archives. Therefore the Swiss federal archives are developing solutions.

6.2 Planned research Activities in NWP, Nowcasting, Long-range Forecasting and Specialized Numerical Predictions

6.2.1 Planned Research Activities in NWP AUTHOR: PHILIPPE STEINER At the beginning of 2012, MeteoSwiss started a project for the renewal of the operation NWP system called COSMO-NExT. The main objectives of COSMO-NExT are the development of a O(1km) mesh- size deterministic COSMO model for the very short range as well as a O(2-3km) mesh-size COSMO- based ensemble system out to +5 days, both for the Alpine domain. The third component of the R&D project is the development of a Local Ensemble Transform Kalman Filter (LETKF) data assimilation system, which will provide the initial data for both forecasting systems. Finally, a fourth work package looks at infrastructural questions and is closely linked to ongoing HPC projects both within Switzerland (i.e., HP2C COSMO and HP2C OPCODE, see above) and the COSMO Consortium (Priority Project POMPA, see above). The plan is to have the new NWP systems ready for operations by 2016. MeteoSwiss started two activities related to improving NWP modeling at local scales (and in complex topography), mainly a • better assimilation of ground-based remote sensing observations (microwave radiometers, ceilometers, wind lidars) through the new COST TOPROF Action (approved in May 2013), and • setting up of a coordinated European network of wind profilers (already in operation) and ceilometers for improving NWP modeling in general and for volcanic ash plume transport now- and forecasting in particular.

6.2.2 Planned Research Activities in Nowcasting AUTHORS: ALESSANDRO HERING / URS GERMANN / PAOLO AMBROSETTI Plans for future research activities are related to an intensification of usage of satellite information for nowcasting in the Alps, particularly for the early detection of severe convection, a successive increase of the multi-sensors capabilities of TRT, the quasi operational use of the radar ensemble generator in hydrological applications, further development of the COST-731 nowcasting system, the renewal of the MeteoSwiss weather radar network, the increase in spatial and temporal resolution of the radar systems, and the investigation of snow and hail by the radar network. The blending of observational data with NWP will be further increased in the nowcasting range.

04b0238ef04452bfc46d850ec7ebf574.doc 6.2.3 Planned Research Activities in Long-range Forecasting None

6.2.4 Planned Research Activities in Specialized Numerical Predictions None

7 Consortium

AUTHORS: PHILIPPE STEINER / MICHAL ZIEMIANSKI

7.1 System and/or Model The COSMO model (http://cosmo-model.org/content/model/general/default.htm) is a non-hydrostatic limited-area atmospheric prediction model. It has been designed for both operational numerical weather prediction (NWP) and various scientific applications on the meso-β and meso-γ scale. The COSMO model is based on the primitive thermo-hydrodynamical equations describing compressible flow in a moist atmosphere. The model equations are formulated in rotated geographical coordinates and a generalized terrain following height coordinate. A variety of physical processes are taken into account by parameterization schemes. Besides the forecast model itself, a number of additional components such as data assimilation, interpolation of boundary conditions from a driving model, and postprocessing utilities are required to run the model in NWP-mode, climate mode or for case studies.

7.1.1 In operation Regional numerical weather prediction at MeteoSwiss is entirely based on the COSMO-Model. COSMO-7 (see sections 4.1, 4.3.1 and 4.3.2) covers most of western Europe with 393x338 grid points/layer at a grid spacing of 6.6 km and 60 layers, and the convection-resolving model COSMO-2, covers the Alpine region with a grid spacing of 2.2 km, 520x350 grid points/layer and 60 layers. On behalf of COSMO, ARPA-SIMC operates the regional ensemble prediction system COSMO-LEPS (http://www.cosmo-model.org/content/tasks/operational/leps/default.htm) at the European Centre for Medium Range Weather Forecasts (ECMWF) in the “Framework for Member-State time-critical applications”. COSMO-LEPS is the Limited Area Ensemble Prediction System developed within the COSMO consortium in order to improve the short-to-medium range forecast of extreme and localized weather events. It is made up of 16 integrations of the COSMO model, which is nested on selected members of ECMWF EPS. COSMO-LEPS covers Central and Southern Europe with 511x415 grid points/layer at a grid spacing of 7 km and 40 layers. The system runs twice a day, starting at 00 and 12UTC with a forecast range of 132 hours.

7.1.2 Research performed in this field The joint research and development is mainly undertaken in the eight working groups (http://cosmo- model.org/content/consortium/structure.htm) and a number of priority projects and priority tasks. The current priority projects are: “Kilometre-Scale Ensemble-Based Data Assimilation” (KENDA), see section 7.4.1, “COSMO-EULAG Operationalization” (CELO) which aim is to get an operational version of COSMO model employing dynamical core with explicit conservative properties for very-high model resolutions, “Calibration of COSMO Model” (CALMO) which aims at development of automatic,

04b0238ef04452bfc46d850ec7ebf574.doc multivariate and based on objective methods calibration of parameterizations of physical processes for the model, “Verifiction System Unified Survey 2” (VERSUS2) developing an operational verification package for deterministic and ensemble forecasting, “Performance On Massively Parallel Architectures” (POMPA) for preparation of the COSMO model code for running on future high performance computing systems and architectures, and “Consolidation of Operation and Research Results for the Sochi Olimpic Games” (CORSO) for enhancing and demonstrating COSMO-based NWP systems in winter conditions and for mountainous terrain. The priority task “NWP Test Suite’ focuses on preparation of software environment to perform controlled and thorough testing for any released version of the COSMO model, according to the “COSMO Standards for Source Code Development”. Environmental prediction aspects of the model involving chemistry, aerosol effects and transport (COSMO ART) are developed in close cooperation with Karlsruhe Institute for Technology (KIT) in Germany.

7.2 System run schedule and forecast ranges See section 4.3.2.

7.3 List of countries participating in the Consortium COSMO stands for COnsortium for Small-scale MOdelling. The general goal of COSMO is to develop, improve and maintain a non-hydrostatic limited-area atmospheric model, the COSMO-model, which is used both for operational and for research applications by the members of the consortium. The consortium was formed in October 1998 at the regular annual DWD (Germany) and MeteoSwiss (Switzerland) meeting. A Memorandum of Understanding (MoU) on the scientific collaboration in the field of non-hydrostatic modeling was signed by the Directors of DWD (Germany), MeteoSwiss (Switzer-land), USAM (Italy, then named UGM) and HNMS (Greece) in March/April 1999. The MoU has been replaced by an official COSMO Agreement, which was signed by the Directors of these four national meteorological services on 3 October 2001. In 2002, the national weather service of Poland (IMGW) joined the Consortium in effect from 4 July. The National Institute of Meteorology and Hydrology (NMA) of Romania and the Fed-eral Service for Hydrometeorology and Environmental Monitoring of the Russian Federation joined the Consortium in effect from 1 October 2009.

04b0238ef04452bfc46d850ec7ebf574.doc Currently, the following national meteorological services are COSMO members:

Germany DWD Deutscher Wetterdienst

Switzerland MCH MeteoSchweiz

Italy USAM Ufficio Generale Spazio Aereo e Meteorologia

Greece HNMS Hellenic National Meteorological Service

Poland IMGW Institute of Meteorology and Water Management

Romania NMA National Meteorological Administration

Russia RHM Federal Service for Hydrometeorology and Environmental Monitoring

These regional and military services within the member states are also participating:

Germany AGeoBw Amt für GeoInformationswesen der Bundeswehr

Italy CIRA Centro Italiano Ricerche Aerospaziali

Italy ARPA-SIMC ARPA Emilia Romagna Servizio Idro Meteo Clima

Italy ARPA Piemonte Agenzia Regionale per la Protezione Ambientale Piemonte

Four national meteorological services, namely INMET (Brazil), DHN (Brazil), DGMAN (Oman) and NCMS (United Arab Emirates) as well as the regional meteorological service of Catalunya (Spain) use the COSMO model in the framework of an operational licence agreement including a license fee. National meteorological services of developing countries (e.g. Egypt, Kenya, Rwanda) can use the COSMO model free of charge. Lateral boundary conditions based on the global model GME of Deutscher Wetterdienst are provided free of charge to all COSMO users.

7.4 Data assimilation, objective analyses and initialization

7.4.1 In operation The data assimilation system for the COSMO model is based on the observation nudging technique. The variables nudged are the horizontal wind, temperature, and humidity at all model layers, and pressure at the lowest model level. The other model variables are adapted indirectly through the inclusion of the model dynamics and physics in the assimilation process during the relaxation. At present, radiosonde, aircraft, wind profiler, surface synoptic, ship, and buoy data are used operationally. For model configurations at the convection-permitting scale, radar-derived precipitation rates are included additionally via the latent heat nudging method. If nudging is used for data assimilation, an extra initialization is not required. Separate two-dimensional analysis schemes based on the successive correction technique are deployed for the depth of the snow cover and the sea surface temperature.

04b0238ef04452bfc46d850ec7ebf574.doc As for COSMO-LEPS, the following initialization is performed: the upper-level initial conditions of the individual members are interpolated from the ECMWF EPS elements providing the boundaries. On the other hand, the initialization at the lower boundary is performed by taking the surface fields of COSMO-EU (the operational COSMO model of DWD running with a grid mesh of 7 km), including soil temperature and humidity, and blending them with those provided by ECMWF.

7.4.2 Research performed in this field The focus of research efforts lies on the development of a novel data assimilation scheme based on the Local Ensemble Transform Kalman Filter technique in the frame of the KENDA priority project. Its main purpose will be to deliver perturbed initial conditions for convection-permitting ensemble prediction systems. For more information, see http://www.cosmo-model.org/content/tasks/priorityProjects/kenda/default.htm. The current research includes, in between, work on assimilation of high-resolution observations: • assimilation of radial velocity from Doppler radars and development of radar observation operator • assimilation of GPS slant path delay data • assimilation of SEVERI-based cloud top height. The new assimilation system already undergoes extensive testing showing promising characterisitics.

7.5 Operationally available NWP products See section 4.3.3. As for COSMO-LEPS, the available operational products include the following: • “ deterministic products”: different weather scenarios (one per member) for the model variables, at several forecast ranges; • “ probabilistic products”: probability of exceeding user-defined thresholds for the different model variables, at several forecast ranges, • “point wise products”: meteograms over station points in terms of the main model variables. See also 4.2.5.3 for local products derived from COSMO-LEPS.

7.6 Verification of prognostic products

7.7 Plans for the future (next four years)

7.7.1 Major changes in operations

7.7.2 Planned Research Activities A 5-year science plan (http://cosmo-model.org/content/consortium/reports/sciencePlan_2010-2014.pdf) summarizes the current strategy and defines the main goal of the joint development work within COSMO. While the

04b0238ef04452bfc46d850ec7ebf574.doc Science Plan undergoes a thorough revision process, at the moment, its main goal remains stable: to develop a model system for short to very short range forecasts with a convective-scale resolution to be used for operational forecasting of mesoscale weather, especially high impact weather. The research- oriented strategic elements to achieve the goal are: an ensemble prediction system, an ensemble- based data assimilation system and a verification and validation tool for the convective scale, extension of the environmental prediction capabilities of the model, use of massively parallel computer platforms. The actions for achieving the goal are undertaken within the current priority projects and task (see section 7.1.2) which will be complemented by the future projects. In the near future, the planned research activity will include a new priority project on convective-scale ensembles involving, in between: • Application of results of KENDA for definition of the ensemble initial conditions • Methodology of physics perturbations including a new stochastic physics scheme.

04b0238ef04452bfc46d850ec7ebf574.doc 8 References

Autier, P., Boniol, M., & Doré, J.-F. (2007). Sunscreen use and increased duration of intentional sun exposure: still a burning issue. Int J Cancer ; :., 121, 1–5. Bauer, A., Diepgen, T. L., & Schmitt, J. (2011). Is occupational solar ultraviolet irradiation a relevant risk factor for basal cell carcinoma? A systematic review and meta-analysis of the epidemiological literature. Br. J. Dermatol., 165, 612–625. Calpini B., Ruffieux D., Bettems J.-M., Hug C., Huguenin P.,. Isaak H.-P, Kaufmann P., Maier O., and Steiner P., Ground-based remote sensing profiling and numerical weather prediction model to manage nuclear power plants meteorological surveillance in Switzerland. Atmos. Meas. Tech., 4, 1617–1625, 2011, doi:10.5194/amt-4-1617-2011 Di Giacomo, T., Kim, H., Moccozet, L., & Magnenat-Thalmann, N. (2007). Control structure and multi- resolution techniques for virtual human representation. Dans L. De Floriani, & M. Spagunolo, Shape Analysis and Structuring (pp. 245–268). Berlin: Springer Verlag. Diffey, B. L. (1991). Solar ultraviolet radiation effects on biological systems. Phys. Med. Biol., 36, 299–328. Dornelles, S., Goldim, J., & Cestarl, T. (2004). Determination of the minimal erythema dose and colorimetric measurements as indicators of skin sensitivity to UV-B radiation. Photochem. Photobiol., 79, 540–544. Elwood, J. M., & Jopson, J. (1997). Melanoma and sun exposure: an overview of published studies. Int. J. Cancer, 73, 198–203. Hirschi, M., Spirig, C., Weigel, A.P., Calanca, P., Samietz, J., and M.W. Rotach. 2012. Monthly weather forecasts in a pest forecasting context: Downscaling, recalibration and skill improvement, J. Appl. Met. Clim., 51, 1633-1638. Hülsen, G., & Gröbner, J. (2007). Characterization and calibration of ultraviolet broadband radiometers measuring erythemally weighted irradiance. Appl. Opt., 46, 5877–5886. International Commission on Non-ionizing Radiation Protection (ICNIRP). (2004). Guidelines on limits of exposure to ultraviolet radiation of wavelengths between 180 nm and 400 nm (incoherent optical radiation). Health Phys., 87, 171–186. Leuenberger D., Laudanna del Guerra F., and Rossa. 2010. Application of an empirical quality function for radar QPE in an NWP model A.. ERAD2010, available from http://www.erad2010.org/pdf/POSTER/Thursday/03_NWP/10_ERAD2010_0140_extended.pdf Lucas, R. M., McMichael, A. J., Armstrong, B. K., & Smith, W. T. (2008). Estimating the global disease burden due to ultraviolet radiation exposure. Int. J. Epidemiol., 37(3), 654-667. MakeHuman. (s.d.). Open source tool for making 3D characters. Consulté le August 16, 2012, sur http://www.makehuman.org/ McKinlay, A. F., & Diffey, B. L. (1987). A reference action spectrum for ultra-violet induced erythema in human skin. Dans Human Exposure to Ultraviolet Radiation: Risks and Regulations (pp. 83–87). Amsterdam: Elsevier Science Ltd. Parisi, A. V., Kimlin, M. G., Lester, R., & Turnbull, D. (2003). Lower body anatomical distribution of solar ultraviolet radiation on the human form in standing and sitting postures. J. Photochem. Photobiol. B, 69, 1–6. Parisi, A. V., Kimlin, M. G., Wong, J. C., & Fleming, R. A. (1996). The effects of body size and

04b0238ef04452bfc46d850ec7ebf574.doc orientation on ultraviolet radiation exposure. Photodermatol. Photoimmunol. Photomed., 12, 66–72. Pauling, A., M.W. Rotach, R. Gehrig, andB. Clot. 2011. Contributors to the European Aeroallergen Network (EAN) (2011). A method to derive vegetation distribution maps for pollen dispersion models using birch as an example. Int J Biometeorol. DOI 10.1007/s00484-011-0505-7. Siani, A. M., Casale, G. R., Sisto, R., Colosimo, A., Lang, C. A., & Kimlin, M. G. (2011). Occupational exposures to solar ultraviolet radiation of vineyard workers in Tuscany (Italy). Photochem. Photobiol., 87, 925–934. Thieden, E., Philipsen, P. A., Heydenreich, J., & Wulf, H. C. (2004). UV radiation exposure related to age, sex, occupation, and sun behavior based on time-stamped personal dosimeter readings. Arch. Dermatol., 140(2), 197-203. Vernez, D., Milon, A., Francioli, L., Bulliard, J.-L., Vuilleumier, L., & Moccozet, L. (2011). A Numeric Model to Simulate Solar Individual Ultraviolet Exposure. Photochem. Photobiol., 87(3), 721-728. Vernez, D., Milon, A., Vuilleumier, L., & Bulliard, J.-L. (2012). Anatomical exposure patterns of skin to sunlight: relative contributions of direct, diffuse and reflected ultraviolet radiation. Br. J. Dermatol., 167(2), 383-390. Vogel B., Vogel H., Bäumer D., Bangert M., Lundgren K., Rinke R., Stanelle T.. 2009. The comprehensive model system COSMO-ART - Radiative impact of aerosol on the state of the atmosphere on the regional scale. Atmos Chem Phys 9:8661-8680. Vogel H., Pauling A., Vogel B.. 2008. Numerical simulation of birch pollen dispersion with an operational weather forecast system. Int J Biometeorol 52: 805-814. Weigel A.P. and Mason S.J. 2011. The Generalized Discrimination Score for ensemble forecasts. Mon. Wea. Rev. 139, 3069-3074. Weigel, A.P., Liniger, M. A., and C. Appenzeller. 2008. Seasonal Ensemble Forecasts: Are Recalibrated Single Models Better than Multimodels?, Mon. Wea. Rev., 137, 1460-1479. World Health Organization (WHO). (1992). Solar and Ultraviolet Radiation. IARC monographs on the evaluation of carcinogenic risks to humans. London: World Health Organization. Wright, C., Diab, R., & Martincigh, B. (2004). Anatomical distribution of ultraviolet solar radiation. S. Afr. J. Sci., 100, 498–500.

04b0238ef04452bfc46d850ec7ebf574.doc