NIST can be outdated.

[aricciv]

I found that some of the values from the NIST database are outdated. In particular, the A coefficient for the 40Ca+ ion D5/2 (729 nm) transition is pretty off. The value from NIST is a theoretical value of 1.3 s^-1 , whereas more recent measurements (i.e. https://journals.aps.org/pra/abstract/10.1103/PhysRevA.62.032503) give 0.86 s-1. Dipole transitions have less discrepancies, but are still outdated.

[mgrau]

Thanks. This isn't really an issue for this library. You could contact the NIST ASD team about it, but I don't think their goal is to provide high precision matrix elements for particular atoms. The Portal for High-Precision Atomic Data and Computation does aspire to do this, and you can see that the calcium quadrupole transition matrix element is accurate there (in fact it's a more recent theory estimate).

Regarding a solution for atomphys, there are two. One is adapting the library to import from the high precision portal, or the NIST ASD, or both. Another option is to by hand modify the transition objects to reference already known, more accurate values. The latter is possible to do right now, although there isn't any documentation about how to do it.