Thermal Inversions on Hot Jupiters: When a Planet's Stratosphere Runs Backward

What a Thermal Inversion Actually Is

An atmosphere's temperature-pressure (T-P) profile describes how temperature changes as you move up (to lower pressure). In the familiar case — Earth's troposphere, Jupiter's weather layer — temperature falls with altitude because the deep interior is the ultimate heat source and radiation escapes more easily higher up. A thermal inversion reverses this over some altitude range: temperature increases with height, creating a warm layer aloft that sits like a lid over cooler gas below.

Earth has a familiar analog in its own stratosphere, where ozone absorbs solar ultraviolet and warms the air from roughly 220 K at the tropopause to about 270 K near 50 km. Exoplanet scientists borrow the word: a hot-Jupiter "stratosphere" is any inverted layer, regardless of chemistry.

• Requirement: a species that absorbs incident starlight strongly and high up, at pressures of roughly 0.001–0.1 bar.
• Consequence: energy is deposited above the layer that would otherwise be hottest, flipping the local gradient.

The Mechanism: A High-Altitude Sunscreen That Heats From Above

Hot Jupiters orbit their stars at 0.02–0.05 AU, receiving thousands of times Earth's insolation. The physics of an inversion is a competition between where visible/UV starlight is absorbed and where thermal infrared can escape. Normally starlight penetrates deep before being absorbed, so the deep layers are warmest. But if a species has a large optical cross-section, it intercepts starlight high in the atmosphere.

The classic culprits are gaseous TiO and VO, proposed by Hubeny, Burrows & Sudarsky (2003) and Fortney et al. (2008). These molecules have enormous absorption bands across visible wavelengths — precisely where a hot star radiates most of its energy. A trace abundance (mixing ratios near 10⁻⁷) is enough to absorb a large fraction of incoming flux near ~10 mbar.

• The absorbed optical energy is re-radiated as infrared, half of it downward and half up, but the local layer is heated well above radiative equilibrium.
• Because the gas below is comparatively transparent to starlight but opaque in the IR, a temperature inversion forms above the absorbing layer.
• On the hottest worlds, metal atoms (Fe, Mg, Ca), SiO, FeH, and especially H⁻ take over the same role once TiO/VO thermally dissociate.

Characteristic Numbers and a Worked Estimate

The equilibrium temperature of an irradiated planet scales as T_eq ∝ T_star × √(R_star / 2a) × (1 − A)^(1/4), where a is orbital distance, A the Bond albedo, and the factor inside depends on heat redistribution. For a hot Jupiter around a Sun-like star at 0.03 AU this gives T_eq ≈ 1,500–1,800 K; for ultra-hot Jupiters around hotter stars, daysides exceed 3,000 K.

• Inversion threshold: gas-phase TiO/VO survives against condensation only above roughly 1,800–2,000 K; below this Ti and V lock into solids and rain out, removing the absorber.
• Pressure of the inversion: typically 1–100 mbar (10⁻³–10⁻¹ bar), where optical depth to starlight reaches unity for the absorber.
• Temperature contrast: inverted layers can be 500–1,500 K hotter at altitude than the layer just beneath.
• Dissociation: above ~2,500 K, H₂O and TiO break apart; H⁻ opacity (a free electron loosely bound to a hydrogen atom) rises steeply and becomes the dominant continuum absorber.

A useful rule of thumb: features appear in emission when the ~1 bar-to-1 mbar temperature difference is positive (hotter aloft), and in absorption when it is negative.

How Inversions Are Detected

You cannot resolve the disk of a hot Jupiter, so its thermal structure is read from spectral line shapes. The decisive observation is the secondary eclipse (the planet passing behind its star): subtracting the star-only spectrum yields the planet's dayside thermal emission.

• Emission vs. absorption: a normal (non-inverted) atmosphere shows molecular absorption features because the line-forming region is cooler than the continuum below. An inverted atmosphere flips this — the same H₂O or CO bands appear as emission peaks, the smoking gun for a stratosphere.
• Instruments: Spitzer photometry pioneered the field; HST/WFC3 (1.1–1.7 μm) probes the H₂O band; JWST NIRISS, NIRSpec, and MIRI now deliver broadband dayside spectra. Ground-based high-resolution spectroscopy (e.g., detecting Doppler-shifted Fe I and TiO lines in emission) confirms inversions line-by-line.

Landmark cases include the strong inversion in WASP-18b (Sheppard et al. 2017; JWST 2023), H₂O emission in WASP-121b, TiO and Fe I emission in WASP-33b, and Fe I emission revealing an inversion on KELT-20b.

Inversions Versus Their Close Cousins

A thermal inversion is easy to conflate with related phenomena; the distinctions matter:

• Inversion vs. day-night contrast: hot Jupiters also show huge horizontal temperature differences (dayside vs. nightside), driven by poor heat redistribution. That is a global circulation effect; an inversion is a vertical gradient on the dayside.
• Inversion vs. the deep radius-inflation heat source: hot Jupiters are anomalously large, likely from deep energy deposition (ohmic heating, tidal effects). That warms the interior; an inversion is a shallow, starlight-driven upper-atmosphere effect.
• pM vs. pL classes (Fortney et al. 2008): the hottest "pM" planets keep TiO/VO in the gas phase and host inversions; cooler "pL" planets condense them out and remain non-inverted, with Na and K as the main optical opacity.
• Earth's ozone stratosphere is the same physics with a different absorber (O₃ vs. TiO/H⁻), showing inversions are a general consequence of a high-altitude shortwave absorber.

Significance and Open Questions

Thermal inversions are a probe of a hot Jupiter's chemistry, energy balance, and circulation all at once. Their presence tells us which absorbers survive, whether the planet is well-mixed, and how efficiently it redistributes heat. They also complicate the retrieval of molecular abundances: the same H₂O feature means opposite things depending on the sign of the gradient.

• Why aren't all hot planets inverted? Cold-trapping — TiO/VO condensing on the cooler nightside and settling out — can strip the absorber even from planets hot enough to host it, a still-debated depletion mechanism.
• What actually drives the inversion? On ultra-hot Jupiters, models favor metal atoms and H⁻ over TiO/VO; disentangling contributions is an active JWST-era question.
• Stellar UV and hazes: high-energy stellar activity can destroy TiO/VO or create absorbing photochemical hazes (sulfur species, soot), shifting the picture.

The famous back-and-forth over HD 209458b — claimed to have an inversion in early Spitzer data, then reinterpreted with WFC3 as having none — remains the cautionary tale that inversions are subtle, model-dependent, and best nailed down with the emission-feature test.

Dayside temperature regimes of irradiated giant planets and their expected thermal structure and spectral signature

Regime / class	Approx. dayside temperature	Dominant optical absorber	Thermal profile & spectral signature
Cool hot Jupiter (pL class)	≲ 1,700–2,100 K	Na, K (TiO/VO condensed out)	Non-inverted; molecular features in absorption
Hot Jupiter with inversion (pM class)	~2,100–2,400 K	TiO, VO gas phase	Inverted; H₂O, CO features in emission
Ultra-hot Jupiter	~2,400–3,000+ K	Metal atoms (Fe, Mg), H⁻, TiO/VO	Strongly inverted; H₂O dissociated, often near-featureless in near-IR
WASP-18b (example)	~2,900 K	Metals, H⁻	Strong dayside inversion, high metallicity
WASP-121b (example)	~2,500 K	H⁻, metals, VO	Inversion confirmed via H₂O emission (HST/JWST)
WASP-33b (example)	~2,700 K	TiO, Fe	TiO and Fe I detected in emission (high-res)