Assigned ID: 435 ----- Attempting Submission 1 (Fri Feb 26 17:22:28 GMT 2010) ----- The masses of the black holes that power quasars can be estimated from individual emission-line spectra by using the quasar luminosity to infer the size of the line-emitting region and the emission-line width to infer the velocity dispersion of the line-emitting gas. This seems to be well-established, at least for certain emission lines. However, whether or not the C IV 1549 emission line can be used for such purposes is controversial. It is our contention that studies in which the masses estimated from C IV do not agree with those inferred from other emission lines are based on spectra with insufficient signal-to-noise ratios for precision line width measurements. Nearly every study that challenges the use of C IV for black hole mass estimation is based on spectra from large surveys, such as SDSS, and the quality of many of these spectra is too low to measure the chracteristics of spectral features in individual objects with the necessary precision. We propose to demonstrate this by taking a sample of high-quality HST spectra of AGNs for which masses inferred from C IV and Hbeta are consistent and then adding random noise to the UV spectra to replicate SDSS-quality data and remeasuring the lines. This apparently trivial exercise has significant consequences for studies of the evolution of black hole masses with host galaxy properties because use of low signal-to-noise spectra for black hole mass measurement not only increases random errors but also introduces large systematic errors.

AGN/QUASARS 12 25094 false 0.0 0.0 false false false false false false SUPPORTED