The transition frequencies were taken from
(1) J. Casado, L. Nygaard, and G. O. Sørensen, 1971, J. Mol. Struct., 8, 211;
from
(2) E. Fliege, G. Bestmann, R.Schwarz, and H. Dreizler, 1981, Z. Naturforsch. A, 36, 1124;
from
(3) K. Vormann, U. Andresen, N. Heineking, and H. Dreizler, 1988, Z. Naturforsch. A, 43, 283;
and from
(4) G. Wlodarczak, J. Burie, J. Demaison, K. Vormann, and A. G. Császár, 1989, J. Mol. Spectrosc., 134, 297;
and from
(5) K. Wohlfahrt, M. Schnell, J.-U. Grabow, and J. Küpper, 2008, J. Mol. Spectrosc., 247, 119.
Uncertainties were used as published, except for those in 4, which were reduced from 50 to 25 kHz. Two lines from that work with large residuals were omitted. Uncertainties reported in (2), (3), and (5) are somewhat conservative.
Predictions with uncertainties exceeding 500 kHz should be viewed with caution.
Jan. 2018: ¹⁴N hyperfine splitting was resolved in part in the detected lines. Therefore, a separate hyperfine calculation is provided for J' ≤ 20 and frequencies below 54 GHz along with separate partition function values.
Spin-statistics have been considered in the calculation of the intensities. They are 5 : 3 for transitions with even and odd K_a, respectively.
The dipole moment was determined in (5).