What are the gigahertz peaked-spectrum radio sources?

Abstract

We discuss the astrophysical implications of recent radio and optical observations of the powerful, compact gigahertz peaked-spectrum (GPS) radio sources. The defining characteristics of these objects are (1) their peaked radio spectra with narrow spectral shape and steep high- frequency spectrum, (2) low radio (and optical?) polarization, (3) large radio luminosities, and (4) mostly compact radio structure. The nature of these enigmatic objects is still not understood. However, we present some tentative conclusions and suggest a scenario based on the existing data. The turnover in the spectrum is probably due to synchrotron self- absorption, and we assume this as the basis for our inferences from the radio spectral shape. However, as we discuss below, these sources may have higher than normal electron densities in their nuclei, and it is possible that free-free absorption plays a role in at least some sources. The result that some GPS sources have a very narrow spectral shape is consistent with the hypothesis that there is a narrow range of size scales which dominate the radio luminosity; i.e., very extended (~>100 pc) and very compact (<~0.1 pc) emission is apparently very weak if it exists at all. Based on the steep high-frequency radio spectrum and the low variability, there does not appear to be a strong beamed "jetlike" component in these objects. Instead the radio luminosity is dominated by compact (i.e., parsec scale), steep-spectrum emission ("microlobes"?). The highly inverted low-frequency radio spectrum is consistent with the hypothesis that the radio source is tightly confined. We put on a firm foundation the result that the radio polarization of these sources is systematically low. This is consistent either with a very tangled magnetic field or very large Faraday rotation measures (implying high magnetic fields and or very large electron densities). The existence of high rotation measures is tentatively found in two objects (both high-redshift quasars) but it is premature to conclude that this is a general trend in GPS sources. The optical identifications are a mixture of galaxies and quasars. As previously noted (e.g., Phillips & Mutel; Pearson & Readhead) we find that the galaxies tend to have more symmetric milliarcsecond radio structure than the quasars. CCD imaging of small samples of GPS galaxies suggest that they tend to be interacting/merging. Spectroscopy of two GPS galaxies is consistent with the hypothesis that they have unusually dense, highly reddened narrow-line regions. We suggest that GPS radio sources are formed when the radio plasma is confined on the scale of the narrow-line region by an unusually dense and clumpy ISM. In this scenario, the high electron densities are responsible for depolarizing the radio source and in some cases producing very high Faraday rotation measures. The existing (sparse) optical spectroscopic results are also consistent with the existence of a dense and dusty nuclear ISM. The GPS quasars tend to have very high redshifts (half with z ~> 3). The hosts of these radio sources may be protogalaxies with very dense and clumpy interstellar media. The lower redshift GPS galaxies may have recently cannibalized a gas-rich companion, resulting in a very dense and clumpy nuclear ISM.