Atmospheric Circulation of Hot Jupiters: Three-dimensional circulation models of HD 209458b and HD 189733b with Simplified Forcing

TL;DR

This study uses 3D numerical simulations to model atmospheric circulation on hot Jupiters HD 209458b and HD 189733b, predicting temperature patterns, spectra, and light curves that align with some observations but highlight the need for more realistic radiative transfer models.

Contribution

The paper introduces a simplified 3D circulation model for hot Jupiters that captures key flow patterns and spectral features, providing a foundation for future detailed radiative transfer studies.

Findings

Eastward equatorial jet with 3-4 km/sec speed at deeper levels

Strong day-night temperature contrasts varying with pressure

Predicted flux ratios match observed phase offset but not flux minima

Abstract

We present global, three-dimensional numerical simulations of the atmospheric circulation on HD 209458b and HD 189733b and calculate the infrared spectra and light curves predicted by these simulations, which we compare with available observations. Radiative heating/cooling is parameterized with a simplified Newtonian relaxation scheme. Our simulations develop day-night temperature contrasts that vary strongly with pressure. At low pressure (<10 mbar), air flows from the substellar point toward the antistellar point, both along the equator and over the poles. At deeper levels, the flow develops an eastward equatorial jet with speeds of 3-4 km/sec, with weaker westward flows at high latitudes. This basic flow pattern is robust to variations in model resolution, gravity, radiative time constant, and initial temperature structure. Nightside spectra show deep absorption bands of H2O, CO, and/or CH4, whereas on the dayside these absorption bands flatten out or even flip into emission. This results from the strong effect of dynamics on the vertical temperature-pressure structure; the temperature decreases strongly with altitude on the nightside but becomes almost isothermal on the dayside. In Spitzer bandpasses, our predicted planet-to-star flux ratios vary by a factor of 2-10 with orbital phase, depending on the wavelength and chemistry. For HD 189733b, where a detailed 8-micron light curve has been obtained, we correctly produce the observed phase offset of the flux maximum, but we do not explain the flux minimum and we overpredict the total flux variation. This discrepancy likely results from the simplifications inherent in the Newtonian relaxation scheme and provides motivation for incorporating realistic radiative transfer in future studies.