Neutral surface layer turbulence over complex terrain

Accurate turbulence estimates are important input to atmospheric dispersion models since they characterize downwind dispersion and hence, potential pollutant concentrations. When only basic wind information is available, an atmospheric modeler must first estimate roughness length (z{sub 0},) at the location of interest, ({mu}*) from similarity theory using average wind speed ({mu}) and z{sub 0}, and finally apply experimentally derived relationships to determine the turbulence intensities. Even when turbulence coefficients are measured, the turbulence profile must be estimated in the surface layer, using, for example, the power law recommended in a US Environmental Protection Agency guidance document. In this study, turbulent intensities and wind profiles are analyzed in eight direction sectors during near neutral stability. ``Local`` and ``regional`` roughness lengths are calculated from wind speed profiles and from longitudinal turbulence intensities ({sigma}{sub {mu}}) at both sites. With ``regional`` roughness length, complex terrain features are in effect the roughness elements. Profiles of median, 15-minute averaged turbulence intensities {sigma}{sub {mu}}, {sigma}{sub {nu}} and {sigma}{sub w} are calculated at both sites. Profiles of median {sigma}{sub {theta}} and {sigma}{sub {phi}} are also calculated using four mean values of regional z{sub 0} at both sites. Finally, differences between widely-used turbulence relationships and the relationships determined in this study, and their possible effect on model results, are discussed