Stratospheric circulation : a narrowband filter spectrophotometer for total ozone determination

This thesis describes the use of ultraviolet narrowband interference filters in the spectrophotometric determination of local atmospheric total ozone burdens. An investigation of a prototype interference filter spectrophotometer shows that its total ozone measurements are in error owing to the breadth of the filters' passband skirts and to the transmittance sideband leakages of the filters. Accurate spectral transmittance measurements, extending down to transmittances of 10-7, are described and show that the leakage sidebands generally occur close to the passbands and in the near infrared spectral region. An outline of the design and construction of a new, more accurate filter spectrophotometer is given. The use of stacks of two filters at each wavelength greatly improves the narrowness of the passband skirts and significantly decreases the transmittances of the leakage sidebands. In addition a very effective sideband blocking filter is used. The considerably improved performance of this new spectrophotometer is confirmed by the results of a direct comparison of the instrument with a standard Dobson prism spectrophotometer. A study is presented of the effects of filter non-uniformity, orientation, aging and temperature. This enables the establishment of criteria in these regards for accurate total ozone-measurement. A similar study is made of the instrument's photomultiplier detector. The effects of the filters' finite bandwidths are analysed by a numerical simulation of the spectral characteristics of the whole measurement process. The theory involved is given. Calculations show that double stacked filters perform considerably better than single filters. Over the usual operating ranges of airmass and total ozone, the error in the total ozone measurement due to the filter bandwidth effect, is found to be about ±1.0% for the doubled-filter instrument. Corrections for the error are tabulated