The Statistics of Radio Astronomical Polarimetry: Bright Sources and High Time Resolution

TL;DR

This paper develops a four-dimensional statistical framework for analyzing radio pulsar polarization, explaining observed phenomena through source-intrinsic noise and small-number statistics, and clarifying the nature of polarization in giant pulses.

Contribution

It introduces a new formalism for pulsar polarization analysis that accounts for source noise and small-number effects, challenging previous interpretations of polarization measurements.

Findings

Self noise explains excess polarization dispersion in bright pulsars.

Degree of polarization of unresolved pulses is fundamentally undefined.

Mode-separated pulse profiles can be derived without assumptions about mode polarization.

Abstract

A four-dimensional statistical description of electromagnetic radiation is developed and applied to the analysis of radio pulsar polarization. The new formalism provides an elementary statistical explanation of the modal broadening phenomenon in single pulse observations. It is also used to argue that the degree of polarization of giant pulses has been poorly defined in past studies. Single and giant pulse polarimetry typically involves sources with large flux densities and observations with high time resolution, factors that necessitate consideration of source-intrinsic noise and small-number statistics. Self noise is shown to fully explain the excess polarization dispersion previously noted in single pulse observations of bright pulsars, obviating the need for additional randomly polarized radiation. Rather, these observations are more simply interpreted as an incoherent sum of covariant, orthogonal, partially polarized modes. Based on this premise, the four-dimensional covariance matrix of the Stokes parameters may be used to derive mode-separated pulse profiles without any assumptions about the intrinsic degrees of mode polarization. Finally, utilizing the small-number statistics of the Stokes parameters, it is established that the degree of polarization of an unresolved pulse is fundamentally undefined; therefore, previous claims of highly polarized giant pulses are unsubstantiated. Unpublished supplementary material is appended after the bibliography.