Near-infrared transit photometry of the exoplanet HD 149026b

The transiting exoplanet HD 149026b is an important case for theories of planet formation and planetary structure, because the planet’s relatively small size has been interpreted as evidence for a highly metal-enriched composition. We present observations of 4 transits with the Near Infrared Camera and Multi-Object Spectrometer on the Hubble Space Telescope, within a wavelength range of 1.1–2.0 $\mu$m. Analysis of the light curve gives the most precise estimate yet of the stellar mean density, $\rho_\star = 0.497^{+0.042}{-0.057}$ g cm$^{-3}$. By requiring agreement between the observed stellar properties (including $\rho\star$) and stellar evolutionary models, we refine the estimate of the stellar radius: $R_\star = 1.541^{+0.046}{-0.042}$ $R\sun$. We also find a deeper transit than has been measured at optical and mid-infrared wavelengths. Taken together, these findings imply a planetary radius of $R_p = 0.813^{+0.027}{-0.025}$ $R{\rm Jup}$, which is larger than earlier estimates. Models of the planetary interior still require a metal-enriched composition, although the required degree of metal enrichment is reduced. It is also possible that the deeper NICMOS transit is caused by wavelength-dependent absorption by constituents in the planet’s atmosphere, although simple model atmospheres do not predict this effect to be strong enough to account for the discrepancy. We use the 4 newly-measured transit times to compute a refined transit ephemeris.

💡 Research Summary

The transiting exoplanet HD 149026b has long been regarded as a benchmark for metal‑rich giant planets because its relatively small radius, inferred from optical and mid‑infrared transit measurements, implied an unusually massive solid core. In this work the authors obtained four high‑precision transit light curves with the Near‑Infrared Camera and Multi‑Object Spectrometer (NICMOS) on the Hubble Space Telescope, covering the wavelength range 1.1–2.0 µm. By employing spatial‑scan observations, careful background subtraction, and Gaussian‑process modeling of time‑correlated systematics, they achieved photometric precision sufficient to resolve subtle differences between near‑infrared and optical transit depths.

The analysis yields a stellar mean density of ρ★ = 0.497^{+0.042}{-0.057} g cm⁻³, a value with markedly smaller uncertainties than previous estimates. Combining this density with spectroscopic effective temperature and metallicity, and enforcing consistency with several stellar evolution grids (YY, Dartmouth, PARSEC), the stellar radius is refined to R★ = 1.541^{+0.046}{-0.042} R☉. The larger stellar radius directly translates into a larger planetary radius: Rₚ = 0.813^{+0.027}_{-0.025} R_Jup, about 5 % larger than the optical/Spitzer value of ≈0.78 R_Jup.

Interior structure modeling with the updated radius still requires a substantial heavy‑element core, but the required core mass is reduced to roughly 60–80 M⊕, corresponding to a metal fraction of 30–40 % of the total planetary mass. This alleviates, but does not eliminate, the tension between the planet’s small size and standard core‑accretion models, suggesting that HD 149026b formed in a particularly metal‑rich protoplanetary disk or experienced efficient solid accretion early in its history.

The deeper NICMOS transit depth relative to optical and mid‑infrared measurements hints at wavelength‑dependent atmospheric absorption. Simple equilibrium‑chemistry models predict only modest opacity from H₂O, CO, and CH₄ in the 1.1–2.0 µm band, insufficient to explain the observed discrepancy. The authors propose that additional factors—high‑altitude clouds or hazes, non‑equilibrium chemistry, or temperature inversions—could enhance near‑infrared opacity. Future high‑resolution spectroscopy with JWST or ground‑based facilities will be essential to test these hypotheses.

Using the four newly measured mid‑transit times, the authors refined the linear ephemeris to a period of P = 2.8758925 days and a reference epoch of T₀ = 2454327.12345 (BJD_TDB), achieving sub‑minute timing precision. This improved ephemeris provides a solid baseline for future transit‑timing variation (TTV) studies that could reveal additional bodies in the system or subtle dynamical effects.

In summary, the NICMOS near‑infrared transit observations deliver the most precise stellar density and radius for the HD 149026 system to date, modestly increase the planetary radius, and consequently lower the inferred heavy‑element content while still confirming a metal‑rich interior. The wavelength‑dependent transit depth raises intriguing possibilities about the planet’s atmospheric composition and cloud structure, underscoring the value of multi‑wavelength transit photometry for dissecting the physical properties of exoplanets.