High Frequency GPS sources in the AT20G Survey

The Australia Telescope 20GHz (AT20G) survey was used to select a complete sample of 656 Gigahertz Peaked Spectrum (GPS) sources with spectral turnovers above 5GHz. The AT20G has near simultaneous observations at 4.8, 8.6 and 20GHz, which makes it possible to exclude flat spectrum variability as a cause of a source’s peaked spectrum. Optical identification of the sample results in 361 QSOs and 104 galaxies and 191 blank fields. Redshifts are known for 104 of the GPS sources. The GPS sources from the AT20G are discussed and compared to previously known samples. The new sample of high frequency peaking GPS sources is found at a lower redshift than previous samples and to also have a lower 5GHz radio power. Evidence is found to support the idea that the origin of the GPS spectral shape are intrinsically different for galaxies and QSOs. This paper is an elaboration and extension of the talk given at the $4^{th}$ CSS/GPS conference in Riccione in May 2008.

💡 Research Summary

The paper presents a comprehensive study of high‑frequency Gigahertz‑Peaked Spectrum (GPS) sources selected from the Australia Telescope 20 GHz (AT20G) survey. By exploiting the near‑simultaneous measurements at 4.8, 8.6 and 20 GHz, the authors constructed a clean, flux‑limited sample of 656 objects whose radio spectra peak above 5 GHz. This simultaneous multi‑frequency approach is crucial because it eliminates the possibility that a flat‑spectrum source appears peaked simply due to variability, a common contaminant in earlier GPS samples that relied on non‑simultaneous data.

Optical cross‑identifications were performed using major catalogues (SDSS, 6dFGS, SuperCOSMOS). The resulting breakdown consists of 361 quasars (QSOs), 104 galaxies, and 191 fields with no obvious optical counterpart (“blank fields”). Redshifts are available for only 104 sources, yielding a median redshift of z ≈ 0.5—significantly lower than the typical z ≈ 1–2 found in classic GPS samples selected at lower frequencies. This indicates that the high‑frequency GPS population resides preferentially in the relatively nearby Universe.

Radio power analysis at 5 GHz shows that the AT20G GPS objects have, on average, about one order of magnitude lower luminosity than their low‑frequency counterparts. The distribution of power differs between the two optical classes: QSOs tend to be more luminous and exhibit broader spectral peaks, whereas galaxies display lower luminosities and sharper, more symmetric peaks. These differences are mirrored in variability properties; QSOs show larger flux density variations across the three AT20G bands, consistent with active jet activity, while galaxies are comparatively stable.

The authors compare their sample with previously published GPS catalogs and find three key distinctions: (1) a lower median redshift, (2) reduced 5 GHz radio power, and (3) a clear dichotomy in spectral shape and variability between QSOs and galaxies. They argue that these observations support the hypothesis that the GPS phenomenon does not arise from a single physical mechanism. For galaxies, the peaked spectra are likely dominated by free‑free absorption (FFA) in a dense interstellar medium surrounding a young, compact radio source. In contrast, the QSO spectra may be shaped primarily by synchrotron self‑absorption (SSA) within relativistic jets, leading to broader peaks and higher variability.

Statistically, the high‑frequency GPS sources constitute roughly 10 % of the overall AT20G radio population, a proportion higher than that reported for low‑frequency GPS samples. Moreover, there is an apparent trend that sources with higher turnover frequencies tend to have lower redshifts and lower radio powers, suggesting a possible evolutionary sequence where the spectral peak shifts to lower frequencies as the source expands and ages.

The paper concludes that the AT20G high‑frequency GPS sample provides a valuable, previously under‑explored window onto the early stages of radio‑AGN evolution. The distinct properties of the QSO and galaxy subsamples imply that GPS spectra can arise from fundamentally different environments and physical processes. The authors recommend follow‑up very‑long‑baseline interferometry (VLBI) imaging to resolve source sizes, as well as deeper optical/infrared spectroscopy to increase redshift completeness, in order to refine models of GPS source formation and evolution.