The growth in wind energy is driving a need to understand how wind turbines perform under a range of atmospheric conditions. Under stably stratified conditions, for example, wind turbine wakes can persist for long distances and impact the performance of downwind turbines. Additionally, variations in surface topography, from shallow depressions to steep mountains, can deflect wind turbine wakes upward or downward. These and other factors must be captured accurately by atmospheric models that provide predictions for wind turbine micro-siting and operational wind power forecasting over complex terrain. Here we present examples of mesoscale to microscale simulations at real wind farms with complex terrain, including implementation of detailed models for turbine wake effects. The focus is on understanding how simulation of wind turbine wakes is influenced not only by topography and thermal stratification, but by numerical techniques such as turbulence closure models and grid nesting strategies. The Weather Research and Forecasting (WRF) model is used here in grid nested configurations starting from the mesoscale (~10 km resolution) and ending with fine scale resolutions (~10 m) suitable for large-eddy simulation (LES). Wind turbine wake behavior is examined under stably stratified and convective conditions over complex terrain with comparison to field campaign data.