Meteorological models are widely used to provide dynamical and thermodynamical inputs for air quality models, but they also are known to be one of the largest sources of error in air quality simulations. One of the most critical shortcomings of meteorological models is their poor performance in capturing highly variable flows close to the surface in stable boundary layers (SBLs). The WRF-ARW model is designed with a sixth-order numerical differencing scheme and minimal numerical diffusion that might allow at least the statistical structure of the dominant scales of wind variance in the SBL to be simulated. In this study an instrumented network has been deployed in the ridge and valley topographical regime of central PA. Data from the observing network and output from ARW simulations at 440-m resolution are analyzed to compare spectra in the SBL. The chief sources responsible for near-surface wind variance in the modeled SBL appear to be due to small-scale terrain irregularities and transient internal gravity waves in the flow aloft, much as hypothesized for the producing meandering flow in stable near-surface layers. Numerical investigations along this line may lead to better understanding and possibly even to better forecasts for many aspects of observed plume behavior in nocturnal SBLs.