Knowledge Bank

The auroral emission spectrum of Jupiter is normally confined to small areas near the north and south poles, but on two occasions in September and November 1988 much more widespread activity was observed in the 2-$\mu m$ wavelength region. The lines observed in this global-scale activity are different from those of $H2$ and $H3+$ in the normal auroral emission at this wavelength, and were different on the two different occasions in 1988. From comparisons with laboratory absorption $spectra1,2$, it is suggested that these lines are due to the dimer $(H2)2$, observed in emission for the first time. The November 1988 lines have unstable upper levels with $l\ge 2$, while the September 1988 lines have bound upper levels with $l\le 1$. The proposed explanation of these unusual relative intensities is that the November lines were formed by inverse predissociation $H2\ast +H2\to [(H2)2\ast ]\prime \to (H2)2+h\nu$ at a level in the atmosphere where the pressure is sufficient to favor the two-body process over normal auroral quadrupole emission from the excited monomer $H2\ast$, whereas the September lines were from a higher-pressure level which allows the three-body process $H^{\prime}{2} + H{2} + M \rightarrow (H_{2}){2}^{*} + M \rightarrow (H{2})_{2} + h\nu + M $, where M is probably $H2$, to be dominant. This interpretation suggests that widespread auroral activity can be confined to different pressure levels on different occasions.