The experimental frequencies were taken from
(1) R. A. Motiyenko, A. Belloche, R. T. Garrod, L. Margulès, H. S. P. Müller, K. M. Menten, and J.-C. Guillemin, 2020, Astron. Astrophys. 642, Art. No. A29.
A small number of low-frequency data was omitted. The uncertainties used were close to those implicitely assumed, that is half of those reported. The parameter set differes slightly from that in (1). The calculation should be sufficiently accurate for all astronomical observations. Caution is advised for transitions with calculated uncertainties exceeding 0.2 MHz. ¹⁴N hyperfine splitting was deemed to be negligible.
The a-dipole moment component was taken from
(2) R. Sugisaki, T. Tanaka, and E. Hirota, 1974, J. Mol. Spectrosc. 49, 241.
The b-dipole moment component was derived from relative intensity measurements in (1).