Plots of abundance vs. equivalent width are commonly used in abundance studies to determine the microturbulence. In the case of HD 101412, plots of this kind show that the stronger lines give lower abundances even when zero microturbulence is used in the abundance calculation. The plot below is for Fe I lines in the atmosphere adopted for HD 101412, which has Teff = 8300K, log(g) = 3.8. Zero microturbulence is assumed.

Note that the lines with Log(W) greater than about 1.3 (equivalent widths greater than about 20mA) give lower abundances than the weaker lines.

A similar plot for Fe II lines is shown below. Again, zero microturbulence was used in calculating the abundances.

The trend for strong lines to give lower abundances is plain in this plot also. The red star is for the line 4413.60 Fe II-32, in Multiplet 32, a⁴H - z⁴F^o. This line could have been excluded from the abundances, as the transition violates the LS-coupling selection rule Delta-L = 0, +/- 1. For such lines, calculated oscillator strengths are known to be less reliable than those that are allowed by LS rules.