Transition frequencies with microwave accuracy were determined by
(1) D. Dangoisse, E. Willemot, and J. Bellet, 1978, J. Mol. Spectrosc. 71, 414.
The lowest frequency line as well as measurements from earlier sources have been omitted. Extensive far-infrared data were reported in
(2) J. Lohilahti, H. Mattila, V.-M. Horneman, and F. Pawlowski, 2005, J. Mol. Spectrosc. 234, 279.
The extensive data set of the first entry from Sep. 2008 was supplemented recently by an extensive set of lines recorded betwenn 1.1 and 1.52 THz and reported by
(3) O. Zakharenko, R. A. Motiyenko, L. Margulès, and T. R. Huet, 2015, J. Mol. Spectrosc. 317, 41.
Some higher order parameters were estimated from the D₂CO parameters.
The uncertainties of (3) appeared to be quite conservative. Therefore, we decreased these uncertainties by a factor of 1.5. The far-infrared data showed several residuals larger than three times the reported uncertainties. The majority of the lines was higher in frequency than calculated. Even after omission of these lines, the far-infrared data were reproduced just barely within the reported 3 MHz. Moreover, the lines are on avarage still 1.5 MHz high. It may be advisable to set the uncertainties to 4.5 MHz. The data from (1) and (2) have not been merged.
The predictions of very high lying rotational states have been improved with respect to the first entry. The predictions should be adequate throughout not only for astronomical observations but also for laboratory spectroscopy.
At low temperatures, it may be necessary to discern between ortho-D₂CO and para-D₂CO. The ortho states are described by K_a even, the para states by K_a odd. The nuclear spin-weight ratio is 2 : 1 for ortho-D₂¹³CO : para-D₂¹³CO. The J_KaKc = 1₁₁ is the lowest para state. It is 5.5867 cm^–1 above ground.
The dipole moment was assumed to agree with that of D₂CO, see e032502.cat.