2014A&A...568A.114M

A tiny fraction of the quasar population shows remarkably weak emission lines. Several hypotheses have been developed, but the weak line quasar (WLQ) phenomenon still remains puzzling. The aim of this study was to create a sizeable sample of WLQs and WLQ-like objects and to evaluate various properties of this sample. We performed a search for WLQs in the spectroscopic data from the Sloan Digital Sky Survey Data Release 7 based on Kohonen self-organising maps for nearly 10⁵ quasar spectra. The final sample consists of 365 quasars in the redshift range z=0.6-4.2 (<z≥1.50±0.45) and includes in particular a subsample of 46 WLQs with equivalent widths W_MgII<11Å and W_CIV<4.8Å. We compared the luminosities, black hole masses, Eddington ratios, accretion rates, variability, spectral slopes, and radio properties of the WLQs with those of control samples of ordinary quasars. Particular attention was paid to selection effects. The WLQs have, on average, significantly higher luminosities, Eddington ratios, and accretion rates. About half of the excess comes from a selection bias, but an intrinsic excess remains probably caused primarily by higher accretion rates. The spectral energy distribution shows a bluer continuum at rest-frame wavelengths >1500Å. The variability in the optical and UV is relatively low, even taking the variability-luminosity anti-correlation into account. The percentage of radio detected quasars and of core-dominant radio sources is significantly higher than for the control sample, whereas the mean radio-loudness is lower. The properties of our WLQ sample can be consistently understood assuming that it consists of a mix of quasars at the beginning of a stage of increased accretion activity and of beamed radio-quiet quasars. The higher luminosities and Eddington ratios in combination with a bluer spectral energy distribution can be explained by hotter continua, i.e. higher accretion rates. If quasar activity consists of subphases with different accretion rates, a change towards a higher rate is probably accompanied by an only slow development of the broad line region. The composite WLQ spectrum can be reasonably matched by the ordinary quasar composite where the continuum has been replaced by that of a hotter disk. A similar effect can be achieved by an additional power-law component in relativistically boosted radio-quiet quasars, which may explain the high percentage of radio quasars.