In 46 of 72 young disks, the gas is already on its way out

What the instrument actually saw

Naman Bajaj and colleagues went back to archival data from JWST's Mid-Infrared Instrument and pulled out the integral field spectra of 72 disks tilted more than 40 degrees to our line of sight, 67 of them Class II and 5 still Class I. Extended emission, meaning light that does not sit on the star, appeared toward 66. Sorting that emission by shape and by line, they counted conical winds traced in molecular hydrogen toward 46 disks and fast jets traced in singly ionised neon at 12.81 micrometres toward 40. Every source carrying a neon jet carried a wind as well: molecular hydrogen in 85 per cent of them, atomic oxygen in the rest.^[1]

That inventory is the part of the result resting on measurement. The shapes are resolved on the sky, the lines are identified by wavelength, and the counts are counts. Everything past this point is an argument about what produces them, and the two are worth keeping apart.^[1]

The trend that does the work

The detection fractions for both the neon jets and the molecular winds rise with the star's mass accretion rate, and show no dependence on how tilted the disk is or how massive the star is. The authors read that pattern as favouring winds and jets launched magnetically from the disk, dominant while the disk is still feeding the star, over photoevaporation driven by the star's high-energy radiation.^[1]

The load-bearing check here is that the fractions stay flat against inclination. If these structures were showing up mainly because a favourable viewing angle made them easier to resolve, the fractions would climb with tilt; they do not. A quieter alternative survives all the same: accretion rate sets line brightness, so a more strongly accreting disk is a disk whose wind is easier to detect at a fixed exposure. The trend would then be about what the instrument can reach as much as about how often winds exist. Telling the two apart takes a set of disks selected on accretion rate with matched exposure depth, which archival data cannot supply.^[1]

The clock a planet runs against

Bajaj puts the consequence plainly: planet formation is a race against time, and a gas giant like Jupiter has to assemble its massive atmosphere while the disk still holds enough gas to give. The survey leaves that race untimed for any particular system. Its contribution is to show that the dispersal channel is already running during the accreting phase, and running in most of the disks we can see well enough to check.^[1]

The number to watch is the molecular-wind detection fraction in a set of disks chosen the other way round. If weakly accreting Class II disks are observed with the same instrument at exposure depths matched to this survey by the end of 2027, and the conical hydrogen winds turn up in a clearly smaller share than the 46 of 72 counted here, the link to accretion is physical. If the share holds up, the archival trend was reading the instrument's reach. Either answer narrows the search, which is the ordinary way a survey earns its place.^[1]