Can statistics of turbulent tracer dispersion be inferred from camera observations of SO 2 in the ultraviolet

Abstract. Turbulence is one of the unsolved problems of physics. Atmospheric turbulence and in particular its effect on tracer dispersion may be measured by cameras sensitive to the absorption of ultraviolet (UV) sun-light by sulfur dioxide (SO2), a gas that can be considered a passive tracer over short transport distances. We present a method to simulate UV camera measurements of SO2 with a 3D Monte Carlo radiative transfer model which takes input from a large eddy simulation (LES) of a SO2 plume released from a point source. From the simulated images the apparent absorbance and various plume density statistics (centerline position, meandering, absolute and relative dispersion, skewness, and fractal dimension) were calculated. These were compared with corresponding quantities obtained directly from the LES. Mean differences of centerline position, absolute and relative dispersion, and skewness between the simulated images and the LES were found to be smaller than a quarter of one camera pixel, with standard deviations between 1/2 and 3/2 camera pixel. Furthermore sensitivity studies were made to quantify how changes in solar azimuth and zenith angles, aerosol loading (background and in plume), and surface albedo impact the UV camera image plume statistics. Changing the values of these parameters within realistic limits have negligible effect on the centerline position, meandering, absolute and relative dispersions, and skewness of the SO2 plume. Thus, we demonstrate that UV camera images of SO2 plumes may be used to derive plume statistics of relevance for the study of atmospheric turbulent dispersion.