This paper summarizes the results of meteorological and dispersion measurements during Savannah River Laboratory's Project STABLE (Stable Boundary Layer Experiment) field program. The field program took place at the Savannah River site on three nights during 12-17 April 1988. Meteorological data were collected from a 304-m tower, an array of eight 60-m towers, two sodars, a tethersonde, and a sonic anemometer. Based on the classification scheme of Kurzeja et al. (1991) the first and third nights were classified as unsteady type IV nights because of the passage of microfronts on each night. The second night exhibited a continuous level of high turbulence with a weak surface-based inversion and was classified as a steady type III night. The third night was especially interesting because of the considerable directional wind shear with height and the occurrence of two turbulent episodes. The directional shear may have been related to the passage of a high pressure center during the night. The turbulent episodes lasted from 5 to 30 min and had a horizontal extent of at least 30 km. They were preceded by cooling near the surface and a shift to lower frequencies of the Brunt-Vaisala frequency N(BV). It is suggested that a decrease with time of N(BV) or an increase in the standard deviation of the vertical component of velocity (sigma-w) or sigma-w/N(BV) might be used to forecast the onset of turbulent episodes reaching the surface. Sulfur hexafluoride (SF6) tracer was released continuously during the experiments and measured 10-20 km downwind by a mobile continuous analyzer. The tracer transport was shown to be consistent with low-level winds on the 304-m tower, except on the third night. It was hypothesized that on the third night the spatial variability of the wind field was much more extreme than on the other nights and, hence, the winds from the 304-m tower did not reflect the true plume movement. A spatially averaged mean wind was more successful in explaining the observed horizontal plume movement. The wider-than-expected across-arc concentration distribution on the third night was attributed to continuously varying shear; whereas the occasional secondary maxima in the tracer pattern were due to discontinuities in the vertical wind shear.
