Table 4: Sensitivities of abundance ratios to variations in the atmospheric parameters and to errors in the equivalent widths, as computed for a typical program star with $T_{\rm eff} \sim 4750$ K. The total error is computed as the quadratic sum of the three dominant sources of error, $T_{\rm eff}$, $v_{\rm t}$ and errors in the EWs, scaled to the actual errors as described in the text (Col. 8: tot.1) or as the sum of all contributions (Col. 9: tot.2).
Ratio $\Delta T_{\rm eff}$ $\Delta$ $\log g$ $\Delta$ [A/H] $\Delta v_{\rm t}$ $\langle N_{\rm lines} \rangle$ $\Delta$ EW tot.1 tot.2
  (+50 K) (+0.2 dex) (+0.10 dex) (+0.10 km s-1)     (dex) (dex)
(1) (2) (3) (4) (5) (6) (7) (8) (9)
[Fe/H] I +0.063 -0.007 -0.009 -0.025 23 +0.020 0.039 0.039
[Fe/H] II -0.024 +0.085 +0.023 -0.010 2 +0.068 0.069 0.070
[O/Fe] -0.046 +0.088 +0.026 +0.023 1 +0.096 0.101 0.102
[Na/Fe] -0.023 -0.029 -0.004 +0.018 3 +0.055 0.060 0.060

Source LaTeX | All tables | In the text