JWST finds changing gases and patchy water clouds on the coldest known brown dwarf
Astronomers used the James Webb Space Telescope to watch WISE 0855, the coldest known brown dwarf, as it rotated. This object has a temperature near 265 K (about -8 °C). The team took medium-resolution spectra (R ≈ 1000) every 15 minutes for 11 hours, covering infrared wavelengths from 2.87 to 5.27 micrometers with signal-to-noise ratios around 80–100.
The spectra changed with time. The strongest change came from absorption by carbon monoxide (CO) gas. At some wavelengths the CO signal varied by as much as 10% from peak to trough. The authors applied a statistical tool called principal component analysis and found that variations in CO track variations in phosphine (PH3), a phosphorus-bearing gas.
To interpret those changes, the researchers compared the time-series spectra to atmosphere and structure models. They argue the correlated CO and PH3 variations come from gas that has been mixed up from deeper layers and then “quenched,” meaning the chemical mix was set at higher pressure and carried upward by vertical mixing. In contrast, changes in the spectrum that point to water clouds are best explained by changes in cloud thickness higher in the atmosphere.
The data provide evidence for patchy water clouds in WISE 0855’s upper atmosphere. In planetary atmospheres, lower pressure means higher altitude, so the authors infer that cloud thickness changes happen at lower pressures (higher up) than the chemical changes tied to CO and PH3. This spatial separation helps disentangle the roles of chemistry, temperature and clouds in a very cold, hydrogen-dominated atmosphere.
Important caveats remain. Earlier studies of WISE 0855 showed that different models can fit parts of the spectrum, but no single model matched everything. Other recent retrievals found pressure–temperature profiles that differ from standard radiative–convective expectations and suggested missing sources of opacity. Some analyses place clouds much deeper (around 10 bars) than where water is expected to condense (about 1–2 bars), so the precise cloud locations and some model details are still uncertain.