The objective of this diploma thesis is to assess the coupling of atmospheric models for macro and meso/microscale applications. Therefore the simulation results for 3 different coupling approaches for the Weather Research and Forecasting model (WRF) with the Grazer Mesoscale Model (GRAMM) are analyzed and a reference run. For regional models it is in general a very challenging task to reproduce local characteristics, e.g. locally developed wind flows, even with nesting techniques. Due to this analysis a better understanding of the approaches and their behavior compared to the measurements should be allowed. Also some system immanent problems are pointed out and become evident when the local model is driven with the data of the regional one. Within this work, besides well-known problem of too high wind speeds provided by WRF, also other parameters are revealed with a significant trigger potential. Many of the problematic periods with strongly overestimated wind speeds go along with an overestimated WRF temperature, significantly higher than the observed one, and/or Planet Boundary Layer Height (PBLH) much higher in elevation. Another important occurring point is a lack of representing the right wind flow characteristics. While the observations show strongly changing wind directions the simulation appears to go different ways. All this cases are shown within the analysis by the use of time series investigations and various other visualizations.