[Paper Review] A non-grey analytical model for irradiated atmospheres. II: Analytical vs. numerical solutions

Vivien Parmentier, T. Guillot+2 more|arXiv (Cornell University)|Nov 25, 2013

Stellar, planetary, and galactic studies43 references3 citations

TL;DR

This paper develops a fast, accurate analytical model for irradiated exoplanet atmospheres by calibrating non-grey radiative transfer effects using numerical benchmarks. It shows that non-grey thermal opacities—particularly from TiO/VO—dominate temperature inversions more than stellar absorption, and achieves <10% error across 2.5–250 m/s² gravity and 100–3000 K effective temperatures.

ABSTRACT

The recent discovery and characterization of the diversity of the atmospheres of exoplanets and brown dwarfs calls for the development of fast and accurate analytical models. We quantify the accuracy of the analytical solution derived in paper I for an irradiated, non-grey atmosphere by comparing it to a state-of-the-art radiative transfer model. Then, using a grid of numerical models, we calibrate the different coefficients of our analytical model for irradiated solar-composition atmospheres of giant exoplanets and brown dwarfs. We show that the so-called Eddington approximation used to solve the angular dependency of the radiation field leads to relative errors of up to 5% on the temperature profile. We show that for realistic non-grey planetary atmospheres, the presence of a convective zone that extends to optical depths smaller than unity can lead to changes in the radiative temperature profile on the order of 20% or more. When the convective zone is located at deeper levels (such as for strongly irradiated hot Jupiters), its effect on the radiative atmosphere is smaller. We show that the temperature inversion induced by a strong absorber in the optical, such as TiO or VO is mainly due to non-grey thermal effects reducing the ability of the upper atmosphere to cool down rather than an enhanced absorption of the stellar light as previously thought. Finally, we provide a functional form for the coefficients of our analytical model for solar-composition giant exoplanets and brown dwarfs. This leads to fully analytical pressure-temperature profiles for irradiated atmospheres with a relative accuracy better than 10% for gravities between 2.5m/s^2 and 250 m/s^2 and effective temperatures between 100 K and 3000 K. This is a great improvement over the commonly used Eddington boundary condition.

Motivation & Objective

• To develop a fast, accurate analytical model for irradiated giant exoplanets and brown dwarfs with non-grey opacities.
• To quantify the error introduced by common approximations in analytical radiative transfer, especially the Eddington approximation and convective adjustment.
• To calibrate analytical model coefficients using a grid of high-accuracy numerical models for solar-composition atmospheres.
• To determine whether semi-grey or grey approximations are sufficient for predicting temperature profiles in non-grey atmospheres.
• To clarify the physical origin of atmospheric inversions in hot Jupiters, particularly the role of TiO/VO and non-grey thermal effects.

Proposed method

• Compares the analytical solution from Paper I to a state-of-the-art numerical radiative transfer model to quantify errors from approximations.
• Uses a grid of numerical models to calibrate the analytical model’s coefficients for gravity (2.5–250 m/s²) and effective temperature (100–3000 K).
• Applies the Eddington approximation to simplify angular dependency in the radiation field and evaluates its impact on temperature profiles.
• Introduces a hybrid radiative/convective model where convection is applied below the Schwarzschild criterion, assessing its accuracy in non-grey cases.
• Derives functional forms for visible and thermal opacity coefficients based on Rosseland mean opacities and line-by-line opacity data.
• Validates the final analytical model against numerical solutions, achieving <10% relative error across a wide parameter space.

Experimental results

Research questions

• RQ1 How accurate is the Eddington approximation for temperature profiles in non-grey, irradiated atmospheres?
• RQ2 What is the impact of a convective zone on the radiative temperature structure when it extends into the optically thin region?
• RQ3 Can semi-grey or grey approximations adequately reproduce temperature profiles when full non-grey opacities are present?
• RQ4 What is the relative contribution of stellar light absorption versus non-grey thermal effects in driving atmospheric inversions?
• RQ5 Can a fully analytical model be constructed that matches numerical solutions within 10% across a broad range of planetary parameters?

Key findings

• The Eddington approximation introduces relative errors of up to ~5% in non-grey atmospheres and ~2% in grey cases.
• When the radiative/convective boundary lies in the optically thin region, the standard convective adjustment method can produce relative errors of 20% or more in upper atmospheric temperatures.
• Non-grey thermal effects—particularly the reduction of thermal blanketing by broad-band opacities—play a dominant role in creating temperature inversions, more so than enhanced stellar absorption.
• The presence of TiO or VO warms the upper atmosphere and cools the deep atmosphere primarily due to non-grey thermal effects, not just optical absorption.
• The calibrated analytical model achieves a relative accuracy of better than 10% across gravity from 2.5 to 250 m/s² and effective temperatures from 100 to 3000 K.
• Model D, using analytical expressions from Paper I with coefficients from Tables 1 and 3, is recommended for accurate and fast temperature profile computation in solar-composition atmospheres.