Red quasars: the plot thickens

Abstract

It is now well established that many flat-radio-spectrum sources have extremely red colours (Webster et al. 1995, Nature 375, 469). When this result was first reported it ignited considerable controversy: was this a vast, hitherto undetected population of red quasars? Three theories emerged to explain these red objects: They are faint high-redshift radio galaxies, and not quasars at all, They are normal blue quasars partially obscured by dust. The redness is caused by some bizarre red synchrotron emission from jets in these quasars. After several years of observing, we can now confidently say that the red sources are a heterogeneous population: all three theories are correct for some of them. About 10% of the red sources are distant radio galaxies and not quasars at all (Masci et al. 1998, MNRAS 301, 975). Some of these flat-spectrum radio galaxies must lie at very high redshifts (z>3). Another 20% of the red sources appear to be normal blue quasars severely reddened by dust: these dusty quasars will have been missed by virtually all existing surveys. They show heavily reddened emission-line ratios and continuum spectra. The remainder seem to be reddened primarily by a red synchrotron emission component, plus small amounts of dust. This red synchrotron emission component is strongly polarised, but seems to cut off blue-ward of around 1 mu m wavelength, producing the red colours observed. This is theoretically hard to explain: why should this component turn-over always at this wavelength?