Linear Polarization Variations and Circular Polarization are Common Among Airless Bodies

Abstract

Using the POLISH2 polarimeter at the Lick Observatory Shane 3-m and Nickel 1-m telescopes, we discover rotation phase-locked variations in linear polarization to be common among asteroids and a NEO in a clear, 383 to 720 nm bandpass. Essentially all bodies in our eight-year study harbor statistically significant, repeatable linear polarization variations at the $0.01\% = 100$ ppm level or above (1 Ceres, 2 Pallas, 3 Juno, 4 Vesta, 6 Hebe, 7 Iris, 12 Victoria, 15 Eunomia, 16 Psyche, 132 Aethra, 216 Kleopatra, and 65803 Didymos). Since polarimetry is a differential technique, such variations cannot be due to shape changes and must be caused by heterogeneity in surface albedo and/or composition. While (4) Vesta has long been known to exhibit large, repeatable polarization variations across its surface, we discover the variations on (6) Hebe, (12) Victoria, and (65803) Didymos to be 1.5 to 3.5 times as large. We hypothesize that the \textit{DART} impact with Dimorphos blanketed Didymos with depolarizing ejecta, which suggests pristine variations across Didymos to have been even larger. As the only NEO in this study with data quality sufficient to investigate polarization variations, Didymos' huge variations suggest they may be common among NEOs. We also discover optical circular polarization to be enhanced for low-albedo, M type asteroids, which is correlated with their large radar albedos. Thus, we present optical circular polarimetry as an alternative method for the identification of metalliferous bodies.

Figures (4)

• Figure 1: Lick 3-m POLISH2 polarization phase curves of Main Belt Asteroids (MBAs), NEOs (1998 OR2 and 65803 Didymos), and Uranus (control target). Lick 1-m POLISH2 observations of (1) Ceres are also shown. Note the axis breaks to accommodate high phase angle observations of (65803) Didymos. Top: Fractional linear polarization $p = P/I$ is plotted against phase angle $\alpha$, and linear polarization vanishes at the inversion angle $\alpha_0 \sim 20^\circ$. This angle is unique to each airless body and is tied to its mean surface index of refraction Muinonen2002aMasiero2009GilHutton2017. Fractional polarization is defined to be negative for $\alpha < \alpha_0$ ("negative branch") and positive for $\alpha > \alpha_0$ ("positive branch"). Middle: Polarization orientation $\Theta'$ with respect to the plane perpendicular to the Sun-body-observer scattering plane. Here, the inversion angle $\alpha_0 \sim 20^\circ$ marks the discontinuity where polarization rapidly rotates by $\Delta \Theta' = \pm 90^\circ$ from parallel to the scattering plane (for $\alpha < \alpha_0$) to perpendicular to the scattering plane (Rayleigh-like, for $\alpha > \alpha_0$). The control target Uranus displays Rayleigh-like polarization orientation, perpendicular to the scattering plane ($\Theta' = 0.39^\circ \pm 0.72^\circ$), even down to $\alpha = 1.6^\circ$ due to its gaseous atmosphere. Thus, airless and gaseous bodies may be immediately distinguished from each other at low $\alpha$ simply from their polarization orientation $\Theta'$. Bottom: Absolute value of fractional circular polarization $|v| = |V|/I$ vs. $\alpha$, where the metalliferous (16) Psyche, (69) Hesperia, and (97) Klotho harbor elevated circular polarization (see section \ref{['sec_circ']}). Linear fits to $|v|$ vs. $\alpha$ are shown for both the metalliferous sample (solid line) and the non-metalliferous sample (dashed line). The surface of the non-metalliferous Didymos may be contaminated by material excavated from Dimorphos during the DART impact or may simply be a consequence of high phase angle observation.
• Figure 2: Left: Fractional linear polarization $p$ versus $\alpha$ for (12) Victoria obtained on UT 14 to 17 Sep 2021 with Lick 3-m POLISH2. The polarization phase curve varies both from nightly changes in $\alpha$ (blue points) and from intrinsic rotational variations across the asteroid's surface (red points). Right: (12) Victoria observations phased to the 8.66 h rotation period of the asteroid. Observations made at overlapping rotational phases are used to subtract polarization variations due to phase angle changes and uncover variations intrinsic to the asteroid (red points).
• Figure 3: Rotation phase-locked measurements of airless bodies obtained with POLISH2 in an unfiltered, 383 to 720 nm bandpass at the Lick 1-m (for 1 Ceres) and Lick 3-m (all other bodies). Top panels: Nightly fractional linear polarization $p = P/I$ phased to known rotation period from the JPL Small-Body Database, where rotation phase = 0 is arbitrary. Absolute linear polarization $|p|$ increases toward the top of each panel regardless of whether the object was observed on the positive or negative branch (Figure \ref{['phase']}). Center panels: Phase-locked linear polarimetry binned in rotation phase. Best, third-order Fourier fits are shown as red curves. Bottom panels: Linear polarization orientation $\Theta'$ relative to the plane perpendicular to the Sun-body-observer scattering plane. For positive branch (Rayleigh-like) scattering, $\Theta' = 0^\circ$, while $\Theta' = \pm 90^\circ$ for negative branch linear polarization. (3) Juno harbors rotational variations in $\Theta'$ with $> 8 \sigma$ confidence (binned data in black with sinusoidal fit curve in red).
• Figure 4: Airless bodies with POLISH2 detections of rotational polarization variations sorted by $\Phi \times \Delta p / p$ (normalized peak-to-peak value of rotation phase-locked linear polarization variations) into bins for low, medium, high, and extreme surface heterogeneity. Values of $\Delta p / p$ are multiplied by the phase function $\Phi (\alpha)$ at the time of observation to correct for the shadowed disk present at high phase angles. This reduces $\Delta p / p$ and is only significant for 65803 Didymos at $\alpha \sim 76^\circ$. Relative visual geometric albedo is shown by the greyscale shading of each data point, and taxonomic type is indicated at the top. Both quantities are obtained from the JPL Small-Body Database. No clear correlation between surface heterogeneity $\Delta p / p$ and visual albedo exists. The extreme surface heterogeneity of (12) Victoria is unexpected. Our measurements of (65803) Didymos suggest that surface heterogeneity was higher pre-DART impact, which implies heterogeneity may be higher for NEOs than for Main Belt Asteroids (see section \ref{['sec:didy']}). Values and upper limits from the literature are labeled in red Degewij1979Broglia1989Broglia1990Broglia1992Broglia1994Nakayama2000Castro2022, where all studies other than Degewij1979 report detections with significantly larger variations than we detect in this work.