Statistical Properties of a Polarization Vector’s Ellipticity Angle
M. M. McKinnon · 2025
Read Statistical Properties of a Polarization Vector’s Ellipticity Angle by M. M. McKinnon (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-17560.txt.
Introduction
McKinnon works out the statistics of a radio wave's ellipticity angle in a 2025 paper. Pulsars and kindred sources send elliptically bent waves, and reading that bend guides models of how the waves form.
For a steady wave strength, the paper derives the angle's spread, mean, scatter, and trust ranges. Scatter shrinks as the signal strengthens, as with the better known position angle. Two twists set this angle apart: its spread is lopsided, and noise biases the reading, worst at weak signals and steep true angles. Joint formulas cover the case where wave strength itself wobbles.
Catalog record
| CREATOR | M. M. McKinnon |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 59 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-17560
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Statistical Properties of a Polarization Vector’s Ellipticity Angle M. M. McKinnon Abstract The orientation of a polarization vector on the Poincaré sphere is defined by its position angle (PA) and ellipticity angle (EA). The radio emission from pulsars, magnetars, and fast radio bursts can be elliptically polarized, and measurements of the EA have become increasingly important in interpretations and models of their polarization. An in-depth understanding of the statistical properties of the measured polarization angles is a prerequisite to their detailed interpretation. While the statistics of the PA have been understood for some time, the statistics of the EA do not appear to be as well developed as those of the PA. The statistical properties of the EA are derived when the amplitude of the polarization vector is constant, to include its probability density, mean, standard deviation, and confidence limits. Similar to the PA, the standard deviation and confidence limits of the EA vary inversely with the polarization signal-to-noise ratio. However, unlike the PA, the probability density of the EA is generally asymmetric, its standard deviation and confidence limits are dependent upon the intrinsic value of the EA, and the measured EA is biased by the instrumental noise, particularly at low signal-to-noise ratios and large values of the intrinsic EA. General expressions for the joint probability density of the polarization angles and the probability density of the EA are also derived when the amplitude of the polarization vector fluctuates due to the superposition of incoherent modes of orthogonal polarization. I INTRODUCTION Pulsar average profiles have typically shown how the total intensity, circular polarization, linear polarization, and polarization position angle (PA) vary across the pulsar’s pulse. In retrospect, measurements of the polarization vector’s ellipticity angle (EA) have been absent from most of these profiles. Both the PA and the EA are needed to completely define the orientation of the vector on the Poincaré sphere. The observations of PSR B0329+54 by Edwards & Stappers (2004) and PSR B0809+74 by Edwards (2004) were some of the first to investigate the behavior of the EA in detail. At some pulse longitudes in both pulsars, individual samples of the EA are concentrated near the equator of the Poincaré sphere, but the average EA makes a large excursion toward one of the poles of the sphere. These excursions tend to be accompanied by a large discontinuity in the PA of \Delta\psi\simeq\pi/2 . The PA discontinuities are indicative of orthogonal modes of polarization (OPMs), which are generally attributed to the natural modes of wave propagation in the pulsar’s magnetosphere (Melrose & Stoneham 1977; Melrose 1979; Allen & Melrose 1982; Barnard & Arons 1986). The PA and EA in parts of PSR B0031-07 (Ilie et al. 2020) behave similarly to those of PSRs B0329+54 and B0809+74. The PA and EA in the central component of PSR B0329+54 combine to form an intriguing partial annulus in one hemisphere of the sphere. Edwards & Stappers (2004) suggested the annulus may be the result of generalized Faraday rotation in the pulsar magnetosphere (Kennett & Melrose 1998). They attributed the large EA excursions observed elsewhere in the pulse to either a transition between polarization modes that were not precisely orthogonal or an OPM transition combined with a circularly polarized emission component. More recent observations of pulsars, magnetars, and fast radio bursts (FRBs) have revealed large variations in the EA as functions of pulse longitude, frequency, or time. In some cases, the trajectories of their polarization vectors trace arcs or great circles on the Poincaré sphere. The magnitude of the EA can be large as the vector approaches a pole of the sphere. Dyks et al. (2021) attributed the longitude-dependent variations in the EA of PSR B1451-68 to changes in the relative intensity of coherent OPMs. They recognized that the variations in PA and EA caused by such an OPM transition follow a great circle on the Poincaré sphere, whereas variations due to changes in the phase offset between the modes generally trace a small circle on the sphere. Oswald et al. (2023) proposed that the OPMs in a pulsar’s magnetosphere are partially coherent and demonstrated that the frequency-dependent variations in the EA of PSR J0820-1350 and the longitude-dependent EA variations in PSR J1157-6224 are consistent with changes in the phase offset between the modes. They also demonstrated that the longitude-dependent variations in the EA of PSR J0134-2937 are consistent with a change in the relative intensity of the modes. Lower et al. (2024) found that frequency-dependent variations in the PA and EA of the magnetar XTE J1810-197 follow a circle on the Poincaré sphere. The possible explanations for this behavior include Faraday conversion in the relativistic magnetized plasma that surrounds the magnetar or mode coupling within its magnetosphere. Bera et al. (2025) showed that temporal variations in the PA and EA of FRBs 20210912A and 20230708A follow great circles on the sphere. The variations are consistent with propagation effects in a birefringent medium located in the outer magnetosphere of the FRB progenitor or in the magnetized plasma surrounding it. Cao et al. (2025) observed large variations in the PA and EA across single pulses of PSR B1919+21. They attributed the variations to changes in the phase offset of partially coherent polarization modes. The aforementioned examples illustrate a variety of EA behaviors and emphasize the importance of the EA in the interpretation of the observations, particularly within the context of different polarization models. An in-depth understanding of the statistical properties of the polarization angles is a prerequisite to their detailed interpretation. The statistical properties of the PA have been understood for quite some time (e.g., Davenport & Root 1958; Papoulis 1965; Naghizadeh-Khouei & Clarke 1993). The probability density function (pdf) of the PA, f_{\psi}(\psi) , is unimodal and symmetric about its intrinsic value, \psi_{o} . When the signal-to-noise ratio (SNR) in linear polarization is large ( s\gg 1) , f_{\psi}(\psi) is Gaussian, with a mean of \psi_{o} and a standard deviation of \sigma_{\psi}=1/(2s) . The PA pdf is uniform when s=0 . It is straightforward to numerically derive the confidence limits of f_{\psi}(\psi) , because it is symmetric about \psi_{o} (Wardle & Kronberg 1974; Naghizadeh-Khouei & Clarke 1993; Everett & Weisberg 2001). And since the pdf is Gaussian at high SNR, the confidence limits can be approximated by its standard deviation. McKinnon & Stinebring (1998) generalized f_{\psi}(\psi) to include OPMs. This pdf is generally bimodal. The angular separation between the two peaks of the pdf is \Delta\psi=\pi/2 , the difference between the heights of the peaks is proportional to the difference in the mean intensities of the modes, and the widths of the peaks vary inversely with the polarization fluctuations (McKinnon 2022). The statistical properties of the EA do not appear to be as well developed as those for the PA. McKinnon (2003, 2006) derived the pdf of the polarization vector’s colatitude, \theta , which is complementary to twice the EA ( \theta=\pi/2-2\chi ). The derivation was limited to circularly polarized radiation. For this specific case, the pdf of \theta evolves from a sine distribution, f_{\theta}(\theta)=\sin(\theta)/2 , when the SNR in total polarization is s=0 to a Rayleigh distribution when s\gg 1 . The EA pdf is also bimodal when the OPMs are