The gigahertz peaked-spectrum (GPS) and compact steep-spectrum (CSS) radio sources make up significant fractions of the bright (centimeter-wavelength-selected) radio source population (~ 10% and ~ 30%, respectively) but are not well understood. The GPS sources are powerful (log P1.4 25 W Hz-1), compact ( 1 kpc), and have a convex radio spectrum that peaks between about 500 MHz and 10 GHz (observer's frame). Any population of sources that peaks above 10 GHz would be strongly underrepresented in the present centimeter-wavelength-selected samples. The CSS sources are just as powerful, are larger (between 1 and 20 kpc in size), and have convex spectra that tend to peak at lower frequencies (< 500 MHz) where the practical lower limit in measuring the spectral turnover frequency of ~ 30 MHz is given by the ionospheric cutoff at about 10 MHz and the difficulty of making low-frequency measurements (including source confusion). As we shall see below, there is a great deal of overlap in the properties of CSS and GPS sources and the definition given above is somewhat arbitrary, though it is a useful working definition.
Previous general discussions and reviews of the CSS sources have been presented by Saikia (1988), Fanti et al. (1990b), Fanti & Fanti (1994), Dallacasa, Fanti, & Fanti (1993), Fanti & Spencer (1995), and for GPS sources by O'Dea, Baum, & Stanghellini (1991). Other useful references are the proceedings of the Dwingeloo and Leiden Workshops on Compact Steep-Spectrum and GHz Peaked-Spectrum Radio Sources (Fanti et al. 1990a; Snellen et al. 1996d). We adopt a Hubble constant of H0 = 75 km s-1 Mpc-1 and q0 = 0.0.
GPS and CSS sources are important for two reasons. First, because of their bright and relatively symmetric radio structure on these scales, they probe the narrow-line region (NLR) and interstellar medium (ISM) of the host galaxy via, e.g., Faraday rotation/depolarization and HI absorption, as well as through their interaction with the emission-line gas. On the other hand, in the large-scale sources, the radio emission extends beyond these scales and is not able to serve as a probe. In addition, the core-jet nuclear radio structure of the larger sources does not generally provide a useful probe of the small scales because of its asymmetry and peculiar geometry. Second, the GPS and CSS sources may be the younger stages of powerful large-scale radio sources - giving us insight into radio source genesis and evolution. In this paper, I review the radio, infrared, optical, and X-ray properties, and the current hypotheses for their origin and evolution.