Determination of ozone temperature by neural network

One of the factors that governs the rate of depletion of atmospheric ozone is its temperature. Although direct measurement of ozone temperature is not feasible, it is possible to infer the temperature from the broadening of spectral absorption bands; instruments with this capability are now being developed. This paper investigates the use of a neural network to provide a quick and robust method of temperature estimation from measured ozone spectra.