The dataset and the fit have been decribed in detail in
(1) S. Brünken, H. S. P. Müller, C. Endres, F. Lewen, and T. Giesen, B. Drouin, J. C. Pearson, and H. Mäder, 2007, Phys. Chem. Chem. Phys. 9, 2103.
The dataset contains extensive v₂ = 0 and 1 pure rotational data both with microwave and with infrared accuracy as well as extensive rovibrational transitions between these states. Details on the Hamiltonian can also be found in that paper.
Besides from (1), ground state transitions measured with microwave accuracy, which were retained in the final line list, were taken from
(2) J. K. Messer, F. C. de Lucia, and P. Helminger, 1984, J. Mol. Spectrosc. 105, 139;
and from
(3) F. Matsushima, M. Matsunaga, G. Qian, Y. Ohtaki, R. Wang, and K. Takagi, 2001, J. Mol. Spectrosc. 206, 41.
Excited state data were taken from (1), (2), and from
(4) O. I. Baskakov, V. A. Alekseev, E. A. Alekseev, and B. I. Polevoi, 1987, Opt. Spectrosc. 63, 1016.
Ground state (far-) infrared transitions were published in
(5) J. W. C. Johns, 1985, J. Opt. Soc. Am. B 2, 1340;
in
(6) R. Paso and V.-M. Horneman, 1995, J. Opt. Soc. Am. B 12, 1813;
and in
(7) G. Mellau, S. N. Mikhailenko, E. N. Starikova, S. A. Tashkun, H. Over, and V. G. Tyuterev, 2004, J. Mol. Spectrosc. 224, 1340.
Excited state data were also taken from that work.
Rovibrational data were provided in
(8) C. Camy-Peyret, J. M. Flaud, A. Mahmoudi, F. Guelachvili, and J. W. C. Johns, 1985, Int. J. Infrared Millimeter Waves 6, 199 ;
and in
(9) R. A. Toth, 1999, J. Mol. Spectrosc. 195, 98;
additional rovibrational data were taken from (1), where also additional assignments were given for transition obtained in the course of the investigations for (9).
With respect to the Jan. 2007 entry, the g_up values have been corrected and a small typographical error in the dipole moment has been corrected. The version number has been retained.
With respect to the Nov. 2007 entry, some highly accurate ground state transition frequencies with D hyperfine splitting were added to the line list. These were published by
(10) G. Cazzoli, L. Dore, C. Puzzarini, J. Gauss, 2010, Mol. Phys. 108, 2335.
Transitions deviating by more than 5 time the experimental uncertainties were omitted from the final fit. These are some FIR and IR lines, some of which may be perturbed. The final parameter set deviates slightly from the ones used in (1) and, for the hyperfine parameters, (4). Furthermore, an updated reference was used for the rotationless dipole moment. Positions of multiply measured lines were averaged for the merged catalog file.
The partition function takes into account both vibrational states. Individual non-zero contributions are given in parentheses.
At lower temperatures, it may be necessary to distinguish between ortho-D₂O and para-D₂O. The ortho and para levels are described by K_a + K_c = even and odd, respectively, with a spin-weight ratio of 2 : 1. The J_KaKc = 1₀₁ level is the lowest para state. It is 12.1170 cm^–1 above ground.
Therefore, separate para and ortho predictions are available up to about 100 cm^–1 along with separate para and ortho partition function values. Note: the spin-weight of 2 for the separate ortho predictions has been eliminated.
The rotationless dipole moment comes from
(11) R. Bhattacharjee, J. S. Muenter, M. D. Marshall, 1991, J. Mol. Spectrosc. 145, 302.
No centrifugal distortion corrections have been employed in the current calculation. This will have small, usually negligible effects for ΔK_a = 1 transitions, in particular for not too high values of J and K_a. In the case of transitions with larger ΔK_a predicted intensities should be viewed with more caution.