Polarimetry is a powerful technique for revealing two- and three-dimensional structures in astrophysical objects beyond the spatial resolution provided by direct imaging at any telescope. The light becomes polarized when interacting with an anisotropic medium, e.g., in the presence of magnetic fields or directed radiation beams. This anisotropy becomes imprinted into polarization properites of the light and can be inferred through polarized radiative transfer. I will demonstrate how such an analysis reveals 3D structure of starspots, presence of exoplanets and their properties as well as potentially inhomogeneities in exoplanetary atmospheres.

Motivated by the recent observation on the light curve asymmetry of the transit hot-Jupiter WASP-12b, we propose a model where the interaction of the stellar plasma with the planet is able to modify the structure of the outer atmosphere of WASP-12b. The product of such interaction is the formation of a bow-shock ahead of the planetary orbital motion, which could explain the early transit ingress observed in the near-UV. We investigate the physical conditions of the external ambient medium around the planet that could allow for the formation of a bow-shock, and conclude that the compression of the planetary atmosphere should be a common effect in close-in planetary systems. In addition, our model provides a new technique to probe for the presence of a planetary magnetic field. In the case of WASP-12b, an upper limit for the planetary magnetic field strength ranges from 15 to 22G, comparable to the intensity of Jupiter's magnetic field. We discuss whether this effect could be detected by other techniques, such as interferometry and polarimetry.

Cool main-sequence stars of spectral types F through L have a thick convective envelope or are even fully convective. In many of such stars, magnetic fields of various strengths have been detected. In the Sun, the surface magnetic field is observed to be highly structured owing to its interaction with the convective flows. This has significant impact on the magnetic signatures in spectral lines that are used to detect and measure the field. In contrast to the Sun, the structure and properties of magnetic fields on other cool stars are unknown. In the absence of spatially resolved observations, the effect of the magnetic structure on signatures of the magnetic field can be evaluated by numerical simulations of the magneto-convective processes including the relevant physical processes, such as compressibility, partial ionization, and radiative energy transport. We carried out 3D radiative magnetohydrodynamic simulations of the convective and magnetic structure in the surface layers (uppermost part of the convection zone and photosphere) of main-sequence stars of spectral types F3 to M4. The simulation results were analyzed in terms of sizes and properties of the convection cells (granules) and magnetic flux concentrations as well as velocity, pressure, density, and temperature profiles, including their impact on observable quantities like spectral line shapes and micro-variability. Our numerical simulations show for the first time a qualitative difference in the magneto-convection between solar-like stars and M dwarfs. Owing to higher surface gravity, lower opacity (resulting in higher density at optical depth unity), and more stable downflows, small-scale magnetic structures concentrate into pore-like configurations of reduced intensity. This implies that in very cool stars magnetic surface structures like plage regions and starspots significantly differ from the solar example. Such a difference would have major impact on the interpretation of Doppler imaging data and the analysis of M dwarf spectra.

I describe the development of our generalized stellar atmosphere code PHOENIX-3D. I describe our full characteristics method of solving the radiative transfer problem and show that it provides excellent results for both static and moving atmospheres in test problems. I discuss the implementation of both Eulerian and co-moving frame solutions for the case of moving atmospheres and discuss the computational advantages and disadvantages of both choices.

Of the many known extrasolar planets, over 100 have orbital semi-major axes less than 0.1 AU, and a significant fraction of these hot Jupiters and Neptunes are known to transit their stars, allowing them to be characterized with the Spitzer, Hubble, and groundbased telescopes. The stellar flux incident on these planets is expected to drive an atmospheric circulation that shapes the day-night temperature difference, infrared light curves, spectra, albedo, and atmospheric composition, and recent Spitzer infrared light curves seem to show evidence for dynamical meteorology in these planets' atmospheres. Here, I will survey basic dynamical ideas and detailed 3D numerical models that illuminate the atmospheric circulation of these exotic planets. I will describe the dynamical mechanisms for pumping and maintaining the fast jets that develop in these models, particularly the broad eastward (superrotating) equatorial jet that seems to be a near-universal feature of 3D models of synchronously rotating hot Jupiters on 2-4 day orbits. The role of friction in affecting the atmospheric circulation, and the conditions that promote time variability, will also be discussed.

We observed seven transits and seven eclipses of the planet system HD 189733 at 8 microns with the IRAC camera on the Spitzer Space Telescope. With a new correction for the instrumental 'ramp' and for stellar variability, we are able to achieve relative photometry precise to 0.35 mmag over 590 days. With these advances, we are able to place limits on the dayside and nightside variability of the exoplanet, and show that the night side is 2/3 as bright as the day side, consistent with phase-function measurements for this system. We show that the secondary eclipse is offset from half of an orbital phase after transit due to the non-uniformity of the planet surface brightness (after correcting for light travel time across the system); this is in effect the first eclipse map of an exoplanet. We show that the system is consistent with zero eccentricity. We compute an updated ephemeris from the transit times which are measured to a precision of 3 seconds, and place limits on the presence of low mass companions in mean-motion resonances. We derive more precise parameters of the system, we show that the star is limb-darkened at this wavelength, consistent with stellar models, and we argue that the transits depths may be contaminated by star spots.

Hydrodynamic turbulence is an important process shaping the conditions of atmospheres of brown dwarfs and young gas giant planets. Convection dominates the energy transport and thermal profile in the optically thick layers, but also drives turbulent mixing of the atmosphere inside the radiative regime, which in turn is responsible for the distribution of condensible material and departures from chemical equilibrium in the gas phase. 2D and 3D local radiation hydrodynamics simulations with the CO5BOLD code demonstrate that this turbulent velocity field can be described by a region of exponentially decreasing overshoot above the Schwarzschild boundary and the formation of gravity waves in the higher layers, with amplitude increasing with height. While the overshoot regime can contribute to the replenishment and support of low cloud layers, dust distribution at higher altitudes several non-equilibrium chemistry effects are primarily driven by gravity wave mixing. Based on the gravity wave model from these CO5BOLD simulations, we present calculations for the upmixing of carbon monoxide, carbon dioxide, methane, nitrogen and ammonia under non-equilibrium conditions inside classical 1D PHOENIX stellar atmosphere models including detailed radiative transfer and high resolution synthetic spectra. The results reproduce well the observed mid-infrared spectra and photometry of cool T dwarfs and can also be applied to the first directly imaged extrasolar planets of the HR 8799 system.