Summary and Results from the Meeting of the WP-Coordinators, Offenbach, DWD, June 15 2000

1.)  Some of the Work Packages are already finished and some are going to begin according to the plan. All work packages with top priority are in good progress. Some WPs have a slight delay and those with a significant delay, which is mostly due to technical problems or changes in staff, have to be put on the WP list for the next period. Especially, the EUCOS impact study by MCH (MeteoSwiss, i.e. SMI) in WP 1.4 will be done with LM at ECMWF boundary data. This requires the development of the corresponding interpolation program IFS2LM within a very short time period.
 

2.) For time-critical WPs as well as technical problems, a short-term plan of action was formulated:
 

3. The following WPs have to be continued/started in the next Cosmo period due to significant delays

• Tuning of the nudging scheme using statistical evaluation (WP 1.3)
• Latent heat nudging (WP 1.6)
• Horizontal diffusion (WP 2.2)
• Implementation and testing of 2-way nesting (WP2.4)
• SVAT soil model LSPM (WP (discussion at the Cosmo meeting)
• Model tests and studies at 2.5 km resolution (WP 2.8 and 3.7)
• PBL verification of LM (masts, WP 5.5 and 5.8)
• Reference version of LM and update procedure (WP 6.3.1)
• Testing of LM2LM (WP 6.3)

4.) From verification results presented so far, some basic model deficiencies can be summarized:
 

• Within alpine valleys, the atmosphere is too dry (almost no precipitation) this might result from unresolved 2dx topography as well as too steep mountain slopes (too step for the present numerics)
• smaller amount of low level cloudiness compared to GME, which might result from problems in the surface evaporation processes
• below frontal clouds, the 2m-temperature appears to be larger than observed: the cloud seem to be too transparent for radiation
• during nighttime the 2m-temperature has a quite large cold bias, especially during winter (this could also result from a too small amount of low clouds)
• the evaluation of monthly mean forecasts show a trend to decrease cloudiness with the forecast time. This can come from problems in the surface scheme, in the turbulence parameterization (PBL entrainment), or from numerical problems. The reason for this behaviour has to be investigated quite urgently.
• southeasterly flow over the alps is sometimes unrealistically simulated (either phase errors or wrong synoptic developments in the northern parts of the alps)
• unrealistical large amounts of precipitation over mountain tops
• 10m winds appear to be too high over Switzerland
• there is a different spin-up behaviour of precipitation in northern Italy for winter and summer seasons which is not understood.

 

•  horizontal averaging of the forcing functions for the convection scheme. This reduces the maxima of convective precipitation over    mountains (DWD, Doms)
• filtering of 2 and 3 dx components of the topography (DWD, Gassmann)
• positive definite Leapfrog advection for humidity (DWD, Gassmann)
• Check of surface evaporation (DWD, Heise)
• Reformulation of cloud-radiation interaction (DWD, Ritter and Doms)
• Check of local water budget (NN)

6.) The Cosmo Web/Ftp Server will be installed at HNMS  (Sun Ultra 10)
 

7.) The basic goals for the next Cosmo period have been discussed. The following points have been put on the list:

• consolidation and upgrade of the LM-system
• not too many basic research projects but to try to increase the quality of the system in both scientific and technical aspects
• better communication with the NWP community (i.e. publications, presentations, conferences etc)
• experimental work on the lower mesoscale (i.e. 2.5km resolution)
• development of a Cosmo verification system