X-rayX-ray photo-polarimetryphoto-polarimetry ofof HDHD 189733b189733b

FrédéricFrédéric MarinMarin andand NicolasNicolas GrossoGrosso

November 13th, 2017 Strasbourg - France IntroductionIntroduction

What is HD 189733 ?

HD 189733 (also V452 Vulpeculae) is a binary star system situated at a distance of ~ 19.45 pc from the Sun (van Leeuwen 2007), in the constellation of Vulpecula (the Fox)

Primary star: main-sequence star of stellar type K1.5 V mass 0.846 M_sol (de Kok et al. 2013) radius 0.805 R_sol (Boyajian et al. 2015) rotational period 11.953 days

HD 189733A emits ~ 1028 erg/s (0.25 – 2 keV band) → 10-5x its bolometric luminosity (~ 10x larger than the ratio of X-ray to bolometric luminosity observed from the Sun at its maximum, Poppenhaeger et al. 2013)

Secondary star: spectral type M4 V located at 216 AU from HD 189733A orbital period 3200 years (Bakos et al. 2006) ~ two orders of magnitude fainter in X-rays than HD 189733A

IntroductionIntroduction

A planetary system

In 2005, ELODIE (echelle-type spectrograph) discovered a Hot Jupiter exoplanet around HD 189733

exoplanet HD 189733b - M = 1.162 M Jupiter - R = 1.26 M (de Kok et al. 2013) Jupiter - orbits around HD 189733A in 2.219 days - 0.031 AU from its host star (Triaud et al. 2009) - orbital plane is parallel within 4° of our line of sight - orbit nearly circular (Berdyugina et al. 2008)

Photometric transits of HD 189733 observed with the 1.20-m OHP telescope (Bouchy et al. 2005)

→ HD 189733b stands as the perfect object for Credit: X-ray: NASA/CXC/SAO/K. Poppenhaeger et al; observations and numerical modeling Illustration: NASA/CXC/M. Weiss PolarimetryPolarimetry ofof HDHD 189733b189733b

Measuring exoplanet polarization to constrain atmospheric models

Berdyugina et al. (2008,2011) obtained polarimetric measurements of HD 189733 in the B band

→ well distributed over the orbital period → two polarization maxima near planetary elongations with a peak amplitude of ~ 2.10-4

Assuming Rayleigh scattering → effective size of the scattering atmosphere 30% larger than the radius of the opaque

Polarimetric data (Stokes q and u with 1 sigma error body previously inferred from transits bars on the scale of 10-4) for HD 189733 → lower limit of the geometrical albedo ~ 0.14 (Berdyugina et al. 2008)

+ the phase dependence of polarization indicates that the planetary orbit is oriented almost in a north-south direction (orbit inclination ~98° and eccentricity ~0.0)

PolarimetryPolarimetry ofof HDHD 189733b189733b

Measuring exoplanet polarization to constrain their atmospheres

But !

Using better instruments, Wiktorowicz et al. (2015) and Bott et al. (2016) reported an absence of large amplitude polarization variations

→ true scattered light of an exoplanet is difficult to detect !

Polarimetric data (Stokes q and u with 1 sigma error bars on the scale of 10-4) for HD 189733 (Berdyugina et al. 2008)

Phase-binned observations of HD 189733 vs. orbital phase. Lick 3-m/POLISH2B-band data are shown in blue, Palomar 5-m/POLISH unfiltered data are shown in black. Multiple scattering models with albedos 0.231, 0.434, and 0.604 are shown and compared to a single scattering model with albedo 0.61 – B11 (Wiktorowicz et al. 2015) HDHD 189733b189733b inin X-raysX-rays

How does the system look in the X-ray band ?

Spectroscopic detection of exoplanets have been quite successful in all wavebands, except at X-ray energies !

Only one detection: Poppenhaeger et al. (2013) (combination of 5 non-flaring transits)

→ the X-ray data favors a transit depth of 6 - 8% (optical transit depth of 2.41%)

Deep transit due to a thin outer planetary atmosphere which is transparent at optical wavelengths, but dense enough to be opaque to X-rays, implying high temperatures in the outer atmosphere at which hydrogen X-ray transit in comparison with optical transit data from Winn et al. (2007); is mostly ionized vertical bars denote 1σ error bars of the X-ray data, dashed lines show the best fit to a limb-brightened transit model Long observing program with XMM-Newton from Schlawin et al. (2010). achieved but the results are still awaited to confirm or reject the results of Poppenhaeger et al. (2013)

Meanwhile, we decided to analytically check those observations using the MC code STOKES

ModellingModelling

STOKES Monte Carlo radiative transfer code for modeling multi-wavelength polarization

Originally made to explore active galactic nuclei Adapted for exoplanets

Quiescent coronal model

STOKES code Goosmann & Gaskell (2007) Marin et al. (2012,2015)

The moderately active corona of HD 189733A emits X-ray photons from an optically thin plasma in collisional ionization-equilibrium

The X-ray spectrum is a bremsstrahlung continuum emission plus line emission from metals

Instrumental XMM-Newton/pn spectrum of HD 189733A simulated from X-ray observations, grouped with a minimum of 25 counts. The red, blue, and black lines are the cool, warm, and total plasma components, respectively. ModellingModelling

STOKES Monte Carlo radiative transfer code for modeling multi-wavelength polarization

Originally made to explore active galactic nuclei Adapted for exoplanets

Quiescent coronal model

STOKES code Goosmann & Gaskell (2007) Marin et al. (2012,2015)

The moderately active corona of HD 189733A emits X-ray photons from an optically thin plasma in collisional ionization-equilibrium

The X-ray spectrum is a bremsstrahlung continuum emission plus line emission from metals

ModellingModelling

Model of the HD 189733b atmosphere

The physical properties of the upper atmosphere of HD 189733b can be constrained with transmission spectroscopy of HD 189733A (atmospheric model from Salz et al. 2015)

The gas of the HD 189733b atmosphere is photoionized by the UV emission of HD 189733A (e.g., Sanz-Forcada et al. 2011) and cools by radiation from collisionally excited atomic hydrogen

temperature of ~ 10000 K

→ this high temperature produces a (slow) evaporative-wind

ResultsResults

Despite HD 189733b extended evaporating-atmosphere, we find that its X-ray absorption radius at 0.7 keV is ∼ 1.01x the planetary radius for an atmosphere of atomic H and He (including ions), and produces a maximum depth of ∼ 2.1% at ±46 min from the center of the planetary transit on the geometrically thick and optically thin corona

In the 0.25 – 2 keV energy band for XMM-Newton pn, we numerically compute that this maximum depth is only ∼ 1.6% at ±47 min from the transit center, and little sensitive to the metal abundances assuming that the addition of metals in the atmosphere does not dramatically change the density-temperature profile

ResultsResults

Marin & Grosso (2017) ResultsResults

The exoplanet’s flux is 3 to 5 orders of magnitude fainter than the host star’s one (with maximums at egress)

At most, the reprocessed flux is lower than 10-16 erg/cm²/s

Marin & Grosso (2017) ConclusionsConclusions

Transit

Despite HD 189733b extended evaporating-atmosphere, its X-ray absorption radius at 0.7 keV is ∼ 1.01x the planetary radius (for an atmosphere of atomic H and He including ions)

In the 0.25 - 2 keV energy band observed with XMM-Newton pn, the predicted maximum depth is only of 1.6% → far from the 6 - 8% transit depth observed by Poppenhaeger et al. (2013)

Direct detection

Both the modulation of the X-ray flux with the orbital phase and the scattered-induced continuum polarization cannot be observed with the current X-ray facilities

Future ?

The direct detection of the X-rays scattered by HD 189733b might be considered with the possible advent of interferometric facilities in X-rays, e.g., the Black Hole Mapper visionary- mission with (sub)micro-arcsecond resolution (Kouveliotou et al. 2014)