Strong infrared lines among the lowest three even configurations of Fe II

Two recent observational studies [1-2] of the infrared [Fe II] $\lambda$12567/$\lambda$16435 line ratio suggest that the available theoretical A-values [3-4] for these transitions contains errors of up to 40\%, although the two observations also differ to a similar extent. The empirically derived A-values [1] generally agree much more closely with [3] than with [4]. We have carried out a large scale CI calculation of [Fe II] lines covering the near infrared to far infrared spectrum, with a total of 72,492 configuration state functions (CSFs) for the 100 fine-structure levels belonging to the $3d^64s$, $3d^7$ and $3d^54s^2$ configurations. Our results also tend to show much better agreement with [3] than with [4], and therefore with the empirical A-values of [1]. In particular, our calculated result for the ratio of the A-values of $\lambda$12567 and $\lambda$16435 is 1.04, in agreement with [3], while [1] obtain 1.13, only 9\% higher, whereas [4] obtain 0.79, similar to the result implied in [2]; it is argued in [1] that such a low value is inconsistent with their observations. We anticipate that our new results will have a significant effect on the interpretation of other astrophysical observations.\\ $[1]$ Smith N \& Hartigan P, ApJ {\bf{638}} (2006) 1045\\ $[2]$ Rodriguez-Ardila A {\it et al} A \& A {\bf{425}} (2004) 457\\ $[3]$ Nussbaumer H \& Storey P J, A \& A {\bf{193}} (1988) 327\\ $[4]$ Quinet P {\it et al} A \& AS {\bf{120}} (1996) 361\\