Analysis of channeled spectropolarimetry using singular value decomposition

Channeled spectropolarimetry is a technique for measuring the spectral dependence of the polarization state of light. Passive polarization optics are used to encode the spectral dependence of the four Stokes components s k into a single irradiance spectrum. We treat the technique as a linear operator and compute its singular value decomposition numerically. The resulting singular functions divide into three distinct groups representing s 0 , s 1 and mixtures of s 2 and s 3 . The corresponding singular values indicate that measurements of the latter two groups will have signal-to-noise ratios reduced form that of s 0 by factors of 0.6 and 0.4 respectively. The structure of the singular vectors is in agreement with a separate estimate of the system's resolution.