A Ground-Based Albedo Upper Limit for HD 189733b from Polarimetry

TL;DR

This study uses polarimetry to set an upper limit on the geometric albedo of exoplanet HD 189733b, constraining its reflectivity and scattering properties through extensive observations and modeling.

Contribution

First polarimetric upper limit on HD 189733b's albedo using ground-based observations, incorporating a physically-motivated scattering model to interpret polarization data.

Findings

Set a 3-sigma upper limit of 58 ppm on polarization variability.

Constrained the planet's geometric albedo to be less than 0.37.

Results are consistent with some previous measurements but challenge higher albedo claims.

Abstract

We present 50 nights of polarimetric observations of HD 189733 in $B$ band using the POLISH2 aperture-integrated polarimeter at the Lick Observatory Shane 3-m telescope. This instrument, commissioned in 2011, is designed to search for Rayleigh scattering from short-period exoplanets due to the polarized nature of scattered light. Since these planets are spatially unresolvable from their host stars, the relative contribution of the planet-to-total system polarization is expected to vary with an amplitude of order 10 parts per million (ppm) over the course of the orbit. Non-zero and also variable at the 10 ppm level, the inherent polarization of the Lick 3-m telescope limits the accuracy of our measurements and currently inhibits conclusive detection of scattered light from this exoplanet. However, the amplitude of observed variability conservatively sets a $3 \sigma$ upper limit to the planet-induced polarization of the system of 58 ppm in $B$ band, which is consistent with a previous upper limit from the POLISH instrument at the Palomar Observatory 5-m telescope (Wiktorowicz 2009). A physically-motivated Rayleigh scattering model, which includes the depolarizing effects of multiple scattering, is used to conservatively set a $3 \sigma$ upper limit to the geometric albedo of HD 189733b of $A_g < 0.37$. This value is consistent with the value $A_g = 0.226 \pm 0.091$ derived from occultation observations with HST STIS (Evans et al. 2013), but it is inconsistent with the large $A_g = 0.61 \pm 0.12$ albedo reported by (Berdyugina et al. 2011).