Spectrum reconstruction based on the constrained optimal linear inverse methods.

The dispersion effect of birefringent material results in spectrally varying Nyquist frequency for the Fourier transform spectrometer based on birefringent prism. Correct spectral information cannot be retrieved from the observed interferogram if the dispersion effect is not appropriately compensated. Some methods, such as nonuniform fast Fourier transforms and compensation method, were proposed to reconstruct the spectrum. In this Letter, an alternative constrained spectrum reconstruction method is suggested for the stationary polarization interference imaging spectrometer (SPIIS) based on the Savart polariscope. In the theoretical model of the interferogram, the noise and the total measurement error are included, and the spectrum reconstruction is performed by using the constrained optimal linear inverse methods. From numerical simulation, it is found that the proposed method is much more effective and robust than the nonconstrained spectrum reconstruction method proposed by Jian, and provides a useful spectrum reconstruction approach for the SPIIS.

Effect and suppression of secondary fringes in FATWindII.

By considering the instrument as a complex operator on the incident electric field, a model to calculate secondary fringes of the Field-widened, Achromatic, Temperature-compensated Wind Imaging Interferometer (FATWindII) has been built. The distribution of secondary fringes on a charge coupled device detector has been plotted. The effects of secondary fringes on inversion errors of temperature and wind velocity have been presented. The results show that antireflection coating on the air/glass interface cannot meet the accuracy requirement of FATWindII. A theoretical method for calculating the optimal wedge angles of compensating glasses is derived to suppress the secondary fringes while preserving the primary ones. By adopting both methods, coating with antireflection film and shaping wedge compensating glasses, the relative intensity of secondary fringes is reduced to below 2.5% and the inversion errors of temperature and wind velocity introduced by the effects of secondary fringes can be minimized to about 0.05 K and 0.045 ms(-1), respectively.