A Probabilistic Calibration Procedure for the CORSAIR Polarimeter

We present a novel Bayesian model and a corresponding robust, probabilistic calibration procedure for the CORSAIR polarimeter that can be applied to other polarimeters. Our calibration procedure combines existing Mueller matrix representations of polarimeters with Bayesian methods, and computes the posterior distribution of the parameters by collecting data from the polarimeter at different states. We show that the algorithm is able to converge and recover a well-constrained posterior of the free parameters with a credible interval that is consistent with the ground truth values. Posterior predictive checks indicate that our generative model with inferred parameters can reproduce the calibration data within the predictive uncertainty, and captures the dominant systematic effects of the calibration procedure. We further show that we can propagate calibration uncertainties in the distributions to downstream reconstructions of Stokes measurements and magnetic field estimates. We find that the contribution of calibration uncertainty towards the reconstructed results is minimal relative to that of the photon noise uncertainty, indicating that estimates using our Bayesian calibration algorithm can achieve photon noise-limited measurements in the magnetic field parameters. Finally, we test the Bayesian calibration algorithm on a lab prototype of the CORSAIR polarimeter, and show that it converges and closely recovers theoretical estimates of the free parameters from real-world measurements.