Universität Mainz
Institut für Physik der Atmosphäre (IPA)

Project team: Heini Wernli, Jörg Trentmann

Activities within

Compilation of a climatology of the synoptic-scale characteristics of convective events in Central Europe.

Convective events are categorized into air mass, prefrontal, frontal and postfrontal. It is conceivable that some categories lead to larger damage and/or are more difficult for operational forecasting. Furthermore, they are associated with different moisture sources (local evapotranspiration, advection) and therefore differ in their forecast sensitivity to parameterized surface processes and upstream boundary conditions. To quantify these aspects, a climatology will be compiled based upon gridded rain gauge data and synoptic-scale fields from ERA40. For identified events of convection a set of parameters is determined that leads to a physically meaningful meteorological classification (e.g. thermodynamic stability indices, ambient pressure gradient, regional baroclinicity and moisture origin).

Assessment of the forecasting capabilities of  NWP models for convection in Central Europe

Different forecasting systems can provide different information with varying lead times about convective events. Global models, in particular ensemble prediction systems, can be valuable for an early (2-5 days) indication of meteorological conditions favorable for convection. High-resolution regional models are then used to predict the location, timing and intensity of the event. The quality of these NWP systems depends on the geographical area, the synoptic setting and the model resolution. Operational NWP forecasts of convection in Central Europe will be systematically verified for several summer seasons. An interesting aspect will be the validation of the limited-area high-resolution ensemble prediction system COSMO-LEPS that is based on the LM and run daily at ECMWF.

Process studies and sensitivity experiments with the LM in a very-high resolution mode

It is intended to investigate the physical mechanisms that trigger convection with the aid of case study simulations with the LM. Sensitivity experiments with reduced soil moisture, reduced moisture advection and reduced topography can provide hints on the model sensitivity to these parameters and thereby provide guidance for future model improvements and for the setup of an optimized observational setting. The LM experiments will be done with a horizontal resolution of 1-3 km and turned off parameterization for convection. Recent experiments indicated the potential of treating moist convection explicitly.

Figure 1 displays the horizontal wind at 10 m above the surface (arrows) as simulated with the Lokal Modell (LM) at 1200 UTC on 19 June 2002 in the northern part of the black forest. The LM was employed at 2.8 km resolution which enables to resolve most of the relevant structure of the mountainous region, e.g., the Murgtal and the Kinzigtal. The surface wind field is significantly modified by the topography leading to up valley winds towards the crest of the black forest. At the mountain ridge a convergence line exists leading to cloud formation (black contours). This process is considered as being important for the initiation of convection in this area. Figure 2 shows the temperature at 2 m above the surface (colours) at 2000 UTC on 19 June 2002 as simulated with the LM in southwest-Germany. The white (black) contour lines depict updraft (downdraft) regions at 850 hPa. The arrows represent the horizontal surface wind field. Also shown (thin black contours) are the contours of the topography. The Rhine valley shows the highest temperatures reaching over 25 deg C. A significant cold pool associated with a downdraft region south of Stuttgart (S) is simulated by the model with temperatures of about 15 deg C. Also, significant horizontal outflow of cold air can be seen from this downdraft region.