ON THE SCATTERING CONTRIBUTIONS TO N-V-LAMBDA-1240 AND C-IV-LAMBDA-1549 IN QSOS

The large relative strength of N V lambda 1240 emission from QSOs has gained new importance since Hamann and Ferland used it to infer nitrog en overabundances and supersolar metallicities. However, Hamann acid F erland did not consider that N V might be selectively enhanced by Ly a lpha photons scattered in the broad absorption line regions (BALRs). H ere we calculate the total BALR scattering contributions to N V lambda 1240 and C IV lambda 1549 for cases with average and extreme (very fa vorable) scattering conditions. We use a large sample of observed BALs to show that the average BALR, with mean covering factor (q) over bar = 0.12, scatters, less than or similar to 5.6 Angstrom equivalent wid th in N V and less than or similar to 4.0 Angstrom in C IV in non-BALQ SOs. These upper limits correspond to less than or similar to 31% and less than or similar to 13% of the measured averages for the N V and C IV emission lines, respectively. The scattering of Ly alpha emission nearly doubles the mean-scattered flux in N V. However, wind models of the BALR show that the narrowest mean scattering profiles are nearly twice as bread as the average observed emission lines. Therefore, only a fraction of the scattered pur contributes to the measured emission fines, especially for N V, where line blending severely limits our abi lity to measure broad-emission components. From our narrowest scatteri ng profiles, we estimate that BALR scattering accounts for no more tha n 18% of the measured average N V emission. The wind models also show that any nonspherically symmetric BALR geometry produces broader mean profiles in BAL versus non-BALQSO samples, which contradicts recent ob servations if scattering is assumed to dominate the emission lines. We conclude that scattering in the BALR does not contribute significantl y to either N V or C IV. In particular, it does not account for the la rge observed strength of N V emission. This result is supported by oth er evidence, including the observed similarity of the N V and C IV emi ssion profiles and the lack of correlation between the N V and Ly alph a emission strengths. Therefore, overabundant nitrogen remains the mos t viable interpretation of the strong N V emission line.