Development of a predictive haze index to reflect observed visibility degradation in the city of Cape Town

Abstract

Measurement and forecasting of visibility is essential in assessing the extent of visibility degradation (haze) due to air pollution in Cape Town. Rapid impairment in visibility over the years forced the City ofCape Town to initiate the Brown Haze I (BH I) study in September 1997. The main objective of the Brown Haze I Study was to determine the contribution of all major pollutant sources to the brown haze, and to determine other factors that contribute to haze formation. However to date no study has attempted to devise quantitative methods to monitor visibility in Cape Town. This study reports on a new method of measuting visibility using Digital Image Processing. An hourly series of time-stamped images of three scenic vistas was recorded over a period of 8 months (20060318 to 20061031) using three co-located Sanyo E-6311xP-H digital cameras equipped with E-50500D 20 to lOOrnrn Variable-focal Auto Iris Lenses. The cameras were orientated in three directions (approximately due west, north and east). These time- stamped images were archived for subsequent digital image processing. The average image intensities of selected areas of each image were calculated using the histogram function of digital image processing software (PhotoshopTM 7.0), to obtain the inherent and apparent contrast ratios. The inherent and apparent contrast measurements were then used to calculate light extinction coefficient (bex') and the visual range (VR).