Calculations about the systematic error of visibility-meters measuring scattered light

Abstract

Instruments determining the visibility from scattered light give different results, depending on the differing size distribution and refractive index of the atmospheric aerosol particles. Sixteen instruments are tested in a reproduceable way by calculating the differences, which would occur over a great variety of size distributions and refractive indices. Two hundred and thirty seven size distributions found at different times and places around the world are used. All instruments show a small jump in scale reading at the change-over from haze to cloud or fog, which occurs between about 0.6 and 1 km standard visibility. The measured values of the integrating nephelometers vary by a factor of 1.26 and less, around an average value for haze or cloud and fog. The results from backscatter-instruments that measure at a scattering angle of exactly 180° vary by a factor of 1.44 (Lidar: 1.68) and less but only when the refractive index of aerosol particles stays constant. Variations in the refractive index lead to much larger deviations. Backscatter-instruments that do not measure at exactly 180° give a larger range of values around a different average even when the refractive index remains constant. Values derived experimentally often show a smaller spread than those derived from our calculations. The reasons for this discrepancy can only be guessed at. The ratio of the backscattering coefficient to the scattering coefficient for the monostatic Rb-lidar systems lies between 0.013 and 0.036 for haze and between 0.035 and 0.083 for clouds and fog. The values are similar for Nd-lidar systems. The values from backscatter-instruments that do not measure at exactly 180° are up to an order of magnitude smaller for clouds and fog.