During MAP phase I numerical experimentation aims to evaluate the predictive skill of current weather prediction models, to improve data assimilation and initial conditions of the models, and to help design the field phase. More than 30 scientists from 19 different institutions and from 11 countries are active in the WG-NM. A dozen models (including 5 non-hydrostatic ones) are currently used for MAP preliminary numerical studies.
The INM/WMO Symposium on Cyclones and Hazardous Weather in the Mediterranean offered the WG-NM the opportunity to hold its first official meeting. A session of the symposium was devoted to MAP-related numerical experiments and many other MAP topics have been addressed during this symposium. Further results were also presented during the Belgirate meeting.
WG-NM activity focuses on the following tasks:
This task is included in the HERA EU-project. Four meteorological services contribute: CNRM, DWD, MeteoSwiss, SMR/ER. The work is coordinated at CNRM by Joel Hoffman. The forecast comparison will be made using the 25km grid used by C. Frei for the precipitation climatology reduced to the common domain of the 4 models. Each group was asked to provide its precipitation data on this grid. The required parameters are 6h-integrated large scale and convective precipitation. In addition orography, surface pressure, 2m temperature, and 10m winds on the model grid were requested. The HERA commitment concerns only the precipitation evaluation, the other parameters will be used mainly for interpretation. The parameters will be provided in 6 hour intervals from T+6 up to T+48 for the 00UTC and 12UTC forecast runs. The period covers the August to November season of the years 1994 to 1997. Results should be available before summer 1998.
Climate-mode simulations are run by DLR and MeteoSwiss/ETH. Meteo-France is currently completing a climate-mode simulation with ALADIN (15 Sept - 15 Oct 1996) that could be used as well for MAP purposes.
Idealized studies are being carried out at ISAO (orographic influence on baroclinic disturbances), and at NCAR and IMGI (idealized flow over the Alps).
From the results presented in Palma and Belgirate it appears that a large effort is devoted to the study of the selected MAP episodes and preferentially of the flash flood cases. The Brig, Piedmont and South Ticino floods have been studied now by at least 3 groups. A common set of figures from each group will be available at the MDC for these 3 cases. The figures will include surface fields (orography, if available 24 hour integrated total precipitation from 6 UTC to 6 UTC, 3 hour integrated precipitation (large scale, convective and total), surface wind, mean sea level pressure, latent and sensible heat fluxes) and some upper level fields (wind , potential temperature, vertical velocity, relative humidity, and geopotential at 850, 500 and 300 hPa; potential vorticity at 850 hPa, and on the 310 and 315 K surfaces).
These simulations aim to aid the field phase design (target areas, location of wave breaking). The participants are therefore asked to identify the areas sensitive to a change in the initial conditions or physical parameterizations. Different mobile instruments will be put in the field during the experiment (wind profilers, ground-based radar and lidar, additional soundings,...). It is strongly encouraged to check how these instruments are located with respect to the sensitive areas.
A nudging scheme for the Swiss Model has been developed at MeteoSwiss capable of assimilating surface and upper air data. The scheme is currently being refined and is a prerequisite for the OSSE project (see below). Nudging towards surface data is developed at SMR/ER.
MeteoSwiss has a one-year project to assess the potential benefit of a wind-profiler network in Switzerland. This study will also be used to define the optimal locations of the profilers that will be installed during MAP. Simulations concerning the observation of mesoscale precipitating systems with ground-based and airborne Doppler radars during MAP/SOP will be attempted in France using outputs from numerical models for MAP-related cases (e.g. South Ticino case with the Canadian MC2 model, Brig case with the French MESO-NH model). This study will help to precisely define the scanning characteristics with these radars and to determine the accuracy these measurements could give.