This paper reports new James Webb Space Telescope (JWST) observations of the brown dwarf companion GJ 758 B. The team used JWST’s Near Infrared Spectrograph (NIRSpec) with an integral field unit (IFU), which produces a spectrum at each point in a small field, to take high-contrast spectra from 2.87 to 5.27 micrometers (µm). A template-matching analysis finds the companion at the expected orbital position with very high significance (265.2σ). Extensive radial velocity monitoring of the primary star gives an independent, model-free measurement of the companion’s mass from their joint orbital motion, making this system a useful benchmark for testing atmosphere and evolution models.

The researchers combined the new NIRSpec data with existing space- and ground-based photometry and spectra. They ran a broad set of spectral fits using several families of atmospheric models. At a high level, these models try to reproduce the observed light by adjusting physical properties such as chemical composition (metallicity), surface gravity, cloud properties, and temperature structure. The goal is to infer the brown dwarf’s mass, radius and atmospheric composition from the spectrum and photometry.

A key finding is that simple, uninformed fits can give misleading results. When the authors used an uninformative (uniform) prior on radius, the fits were pulled into an unphysical region of parameter space. That happened because metallicity, gravity and clouds trade off against each other: different combinations can produce similar spectra, so the model chosen can push the inferred mass, gravity and radius to unrealistic values. Applying a Gaussian prior on the mass, informed by the dynamical (radial velocity) measurement, reduces this problem. Even so, models that omit clouds still struggle to match the K-band photometry and return masses inconsistent with the dynamics.