Brown dwarfs — objects between 13 and 80 times Jupiter's mass, too heavy to be planets and too light to fuse hydrogen like stars — are rare as companions to sun-like stars. This scarcity has been called the brown dwarf desert. Its existence has been known for decades, but whether it extends beyond close-in orbits, and whether it has a precise location in mass, has remained uncertain.
Cui, Feng, and colleagues (PNAS, 2026) measured it. Using 55 companions detected through combined radial velocity and Gaia astrometry around 790 sun-like stars, they placed the desert minimum at 30.9 Jupiter masses, with 3-sigma significance. Giant planets in the 5–14 Jupiter mass range occur around 12.6% of stars. Stellar-mass companions (71–120 Jupiter masses) occur around 4.2%. But companions in the 24–42 Jupiter mass range — the center of the desert — occur around only 1.6% of stars. The desert extends from 2 to 20 astronomical units with no sign of thinning at wider separations.
Two formation mechanisms make companions. Core accretion builds planets from the bottom up: dust grains aggregate into pebbles, pebbles into cores, cores capture gas envelopes. This process has a mass ceiling — the available solid material in a protoplanetary disk limits how massive the final planet can become. Gravitational instability works from the top down: a dense region of the disk collapses directly into a massive companion. This process has a mass floor — a collapsing fragment must exceed a threshold density to overcome thermal pressure.
The desert sits between the ceiling and the floor. Core accretion rarely produces objects above roughly 14 Jupiter masses. Gravitational instability rarely produces objects below roughly 70 Jupiter masses. Between these limits, neither mechanism operates efficiently. The rarity of brown dwarf companions is not caused by a single process that avoids certain masses. It is caused by two processes, each effective in its own range, whose ranges do not overlap. The gap is the seam between two kinds of construction.
The lower-mass companions (below 30 Jupiter masses) show lower eccentricities and lower host-star metallicities than higher-mass companions — the signatures expected from core accretion versus disk instability. These statistical differences confirm that the two populations flanking the desert are genuinely distinct: built by different mechanisms, with different sensitivities to initial conditions. A five-AU gap within the low-mass population itself suggests further substructure — possibly two sub-populations within the core-accretion regime.
The structural observation: an absence in nature can be the signature of two presences. The desert is not where formation fails. It is where two formation mechanisms, each successful in its own domain, leave a margin between their operating ranges. The emptiness is not caused by something that prevents brown dwarfs. It is caused by the geometry of two processes whose reach does not quite connect.