Noise assessment of a Fourier transform infrared spectroradiometer subject to the stability of a conventional laboratory blackbody source.

Abstract

A method is described for the measurement of the noise-equivalent spectral radiance (NESR) of Fourier transform infrared (FTIR) spectroradiometers at all wave numbers of a selected range. The method requires minimal detailed knowledge of the sensor and no support equipment beyond a blackbody source. The NESRs of the FTIR spectroradiometer are determined at every wave-number increment in the 700-1300 cm(-1) range, for six resolutions, with a conventional blackbody source and ensembles of differential spectra. The NESRs are well behaved and consistent with the expected dependence on resolution; however, they depend on source temperature at the highest (1 cm(-1)) and lowest (32 cm(-1)) resolutions, with little or no statistical dependence at intermediate resolutions. Residual source drift is shown to be the likely cause of the dependence at 1 cm(-1); the dependence on the source at 32 cm(-1) resolution is shown to be most probably due to photon noise. At intermediate resolutions the sensor noise is dominant.