The experimental lines are from
(1) F. X. Brown, S. Yamamoto, and S. Saito, 1997, J. Mol. Struct. 413–414, 537.
Most of the lines weighted out in (1) were also weighted out here. Two lines fit quite well and were weighted in, though one was given a larger uncertainty together with the adjacent HFS component. One typograpohical error was corected. The present fit employs Watson's S-reduction. Similar to (1), D_K and H_K were kept fixed to values from HS₂ summarized in
(2) M. Tanimoto, T. Klaus, H. S. P. Müller, and G. Winnewisser, 2000, J. Mol. Spectrosc. 199, 73;
in addition, a value for H_K^S was guessed. The hyperfine parameter T_bb was locked to T_aa, a constraint reasonable with the present data set, but unlikely to hold strictly.
The b-type transitions can not be predicted with confidence, and they are rather weak. Therefore, their prediction was omitted. Predictions with calculated uncertainties larger than 0.5 MHz should be viewed with caution, in particular at low frequencies at which the hyperfine splitting is less certain. These limitations are deemed to be more important for astronomical observations than the increase in uncertainty at higher K_a or higher frequencies.
The dipole moment is from an ab initio calculation from
(3) D. E. Woon and E. Herbst, 2009, Astrophys. J. Suppl. Ser. 185, 273.