There is quite a clear anticorrelation between the intrinsic peak frequency and the overall radio source size in compact steep spectrum (CSS) and gigahertz peaked spectrum (GPS) radio sources. This feature is interpreted in terms of synchrotron self-absorption (although free–free absorption may play a role as well) of the radiation emitted by a small radio source which is growing within the inner region of the host galaxy. This leads to the hypothesis that these objects are young and that the radio source is still developing/expanding within the host galaxy itself.

Very young radio sources must have the peak in their radio spectra occurring above a few tens of gigahertz, and for this reason they are termed high frequency peakers (HFPs). These newly born radio sources must be very rare given that they spend very little time in this stage. Ho = 100 km s−1 Mpc−1 and qo = 0.5 are used throughout this paper.

We present global VLBI and VLBA observations of the compact steep spectrum quasar B1524–136. These observations reveal well-defined radio jets on both sides of the active nucleus. Also, the overall radio structure appears highly distorted and asymmetric with the counter-jet exhibiting several oscillations. A possible scenario is one in which jet and counter-jet are inclined at about 25° and 75° to the line of sight respectively and an environment which is dense on the jet side. Possible implications of these results are discussed.

We have obtained HST/STIS long slit spectroscopy of the aligned emission line nebulae in three compact steep spectrum (CSS) radio sources — 3C 67, 3C 277.1, and 3C 303.1. We find systematic offsets (˜300–500 kms) of the emission line velocities on one or both sides of the radio sources. We also see evidence for broad lines (FWHM ˜500 kms) and complex emission line profiles. In 3C 303.1 the data are consistent with multiple components and possibly split lines. The amplitude of the velocity variations is not so large as to exclude gravitationally-induced motions. However, the complex kinematics, the lack of a signature of Keplerian rotation, and the association of the velocity variations with the radio lobes are consistent with the observed ˜300–500 kms velocities being driven by the expansion of the radio source. Acceleration of the clouds by the bow shock is plausible given the estimated densities in the clouds and the velocities observed in the much smaller compact symmetric objects and with expansion velocities estimated from spectral ageing. This conclusion is unchanged if we consider the scenario in which the cloud acceleration is dominated by the post bow shock flow.

Compact symmetric objects are considered the young counterparts of large doubles according to advance speeds measured or inferred from spectral ageing. Here we present a simple power law model for the CSO/FR II evolution based on the study of sources with well defined hot spots. The luminosity of the hot spots is estimated under minimum energy conditions. The advance of the source is considered to proceed in ram pressure equilibrium with the ambient medium. Finally, we also assume that the jets feeding the hot spots are relativistic and have a time dependent power. Comparison with observational data points to an interpretation of the CSO–FR II evolution in terms of decreasing jet power with time.

We have recently developed a detailed analytical model for powerful radio sources based on the results of a series of 2-D numerical hydrodynamical simulations. Here we make use of the model results to investigate the radio source luminosity evolution. Changes in the radio spectrum due to radiation losses were calculated in two different scenarios for energy gains/losses: Kardashev–h;Pacholczyk (KP) and ‘continuous injection’ (CI). The magnetic field was calculated in two limiting cases: (1) assuming equipartition of energy between relativistic particles and fields and (2) magnetic flux conservation inside the cocoon. The effect of the surrounding medium was taken into account by considering three different ambient density profiles. The evolutionary tracks were plotted in a power–diameter (P-D) diagram and compared with the predictions of self-similar models. In general, the evolutionary tracks cannot be represented by a simple power law and have a complex form that is most probably the result of the nonself-similar evolution of the source.

We review theoretical ideas that seem to be currently important for the physics of GPS and CSS radio sources. These include models for their evolution, the production of emission lines, and the origin of the low frequency turnover. We also describe the initial phases of a program of simulations that is aimed at understanding the radiative interactions between jets, lobes, and dense clouds in the nuclei of these objects.

The anticorrelation between the turnover frequency in the radio spectrum and the projected linear size (indicative of the age in the currently favoured youth model) in powerful and intrinsically compact radio sources suggests that the youngest objects must have the highest turnover frequencies. We present preliminary results of new VLBA observations of a sample of sources peaking at frequencies higher than a few GHz, making them good candidates for being newly born sources. Our initial analysis deals with the morphological classification which allows us to study the contamination from beamed sources dominated by a single unresolved component.

The high spatial resolution of the Chandra X-ray Observatory allows us to study the environment of GPS/CSS sources to within an arcsec of the strong compact core. We present the discovery of X-ray jets in two GPS quasars, PKS1127–145 and B2 0738+313, indicating that X-ray emission associated with the relativistic plasma is present at large distances from the GPS nucleus. We also discuss first results from Chandra observations of our GPS/CSS sample. We find that six out of ten sources show intrinsic absorption at a level which may be sufficient to confine the GPS source.

On the basis of the different radio properties (especially different radio morphologies) between GPS galaxies and GPS quasars, I suggest that GPS galaxies are a more homogeneous class mainly consisting of compact symmetric objects, while the majority of GPS quasars are an intrinsically different type of radio source, similar to the common flat spectrum radio sources. For CSS radio sources the difference between galaxies and quasars is less evident. This is interpreted as a bias correlated with size: samples of smaller objects are more likely to be contaminated by unrelated objects. This is consistent also with the high percentage of blazar/quasar objects present in a sample of the even smaller HFP radio sources.

Radio observations of linear polarisation provide a useful probe of the physical conditions in the gaseous environments of compact steep spectrum (CSS) sources. Here, we present polarimetric VLBA observations at 8.4 GHz of the CSS quasar 3C 43. The jet in this source curves gently up to about 220 mas from the core, at which point it bends sharply, probably due to a jet–cloud encounter. Regions of polarised emission are detected along the jet, while the core component shows no significant polarisation.

III Zw 2 shows dramatic radio outbursts roughly every five years. Here we present the light curves and VLBA observations of the latest flare with an excellent time sampling. We have detected superluminal motion with a lower limit for the apparent expansion speed of 1.25±0.09 c at 43 GHz. Spectral and spatial evolution are closely linked and the evolution of III Zw 2 in a turnover frequency vs. linear size diagram is similar to the correlation for GPS and CSS sources. Before and after this rapid expansion we observe a period of virtually no expansion. However, at 15 GHz III Zw 2 shows a constant slow expansion (∼0.6 c). The difference is qualitatively explained by optical depth effects in an ‘inflating-balloon model’, describing the evolution of radio lobes on an ultra-compact scale. The stop-and-go behaviour could be explained by a jet interacting with a molecular cloud or the molecular torus.

The study of both neutral and ionised gas in young radio sources is providing key information on the effect the radio plasma has on the ISM of these objects. We present results obtained for the compact radio sources PKS 1549–79, 4C 12.50 and PKS 1814–63 and for the intermediate-size radio galaxy 3C 459. At least in the first two, low ionisation optical emission lines and HI absorption appear to be associated with the extended, but relatively quiescent, dusty cocoon surrounding the nucleus. The [OIII] lines are, on the other hand, mostly associated with the region of interaction between the radio plasma and the ISM, indicating a fast outflow from the centre. A case of fast outflow (up to ∼1000 km s-1) is also observed in HI in the radio source 4C 12.50. As the radio source evolves, any obscuring material along the radio axis is swept aside until, eventually, cavities (of the same kind as observed e.g. in Cygnus A) are hollowed out on either side of the nucleus. We may witness this phase in the evolution of a radio source in the radio galaxy 3C 459.

We report results from nearly simultaneous pentachromatic VLBI observations towards a nearby GPS galaxy NGC 1052. The observations at 1.6 and 4.8 GHz with the VSOP, and at 2.3, 8.4, and 15.4 GHz with VLBA, provide linear resolutions of ∼0.1 pc. Convex spectra of a double-sided jet imply that synchrotron emission is obscured through foreground cold dense plasma, in terms of free–free absorption (FFA). We found a central condensation of the plasma which covers about 0.1 and 1 pc of the approaching and receding jets, respectively. A simple model with a geometrically thick plasma torus perpendicular to the jets is established to explain the asymmetric distribution of FFA opacities.

Data from new Australia Telescope Compact Array observations of a nearby gigahertz peaked spectrum radio source, PKS 1718–649, show significant variability both below and above the spectral turnover frequency during a nine month period. Simple models of synchrotron self-absorption (SSA) and free–free absorption (FFA) cannot explain the individual spectra or the spectral variability. Novel SSA or FFA models will need to be investigated in order to explain the complex variability in this source. Some potential explanations for the spectral variability are explored here but we can make no strong conclusion about which of SSA or FFA (or alternative mechanisms) causes the peaked spectrum in PKS 1718–649.

Based on the free–free absorption (FFA) model of gigahertz peaked spectrum (GPS) sources, we explain both Faraday rotation asymmetry and opacity asymmetry consistently between two components of a GPS quasar 2134+004. The FFA model assumes dense plasma around the central core to produce FFA, and the difference of path lengths in the plasma toward each component could cause these asymmetries. The component that is closer to the observer has a shorter path length, and consequently, smaller opacity of FFA and lower Faraday rotation. In a simple case, the ratio of Faraday rotation between two components is a function of the ratio of path length, and is the same as the ratio of opacity. Then these two ratios are shown to be essentially the same by our observations. We could thus distinguish between near-side and far-side components by the asymmetries.

We present initial results from a program that we have begun, aimed at increasing the number of GPS and CSS sources for which good spectroscopic data are available for a variety of emission lines. We combine new observational data with previously published results to investigate whether there are any relationships between the excitation state of the gas and the radio properties of the sources, retaining distinctions between GPS and GSS galaxies and quasars. In scenarios where the youngest, most compact sources interact most strongly with the gas, it is perhaps to be expected that this can have an effect on the overall excitation state of the gas, which could therefore correlate with radio properties such as turnover frequency that plausibly trace the source size and age. However, we find a degree of scatter between the turnover frequency and gas excitation state that is several times larger than the characteristic degree of uncertainty in our line ratio measurements. This would suggest that either there is no underlying relationship between turnover frequency and excitation state, or alternatively that any potential relationship is completely diluted by variations in the intrinsic properties of the gas from one source to the next, which would thus be an important factor in determining the source properties.

Flux density monitoring data at 2.3 and 8.4 GHz is presented for a sample of 33 southern hemisphere GPS sources, drawn from the 2.7 GHz Parkes survey. This monitoring data, together with VLBI monitoring data, shows that a small fraction of these sources, ∼10%, vary. Their variability falls into several categories: sources whose spectral classification is, at best, ephemeral on a timescale of years; sources with a stable GPS classification that vary, but retain their classification; and a small number of sources that exhibit interstellar scintillation, but that maintain a mean GPS spectrum. Existing data on GPS sources with higher frequency peaks, ≥3 GHz, reveals that many such sources vary. However, the majority of these sources possess a GPS spectrum only during outbursts, and hence should perhaps be classified as ephemeral GPS sources. In addition, significant levels of circular polarisation have been found in a number of GPS sources, both amongst the variables and non-variables. Remarkable amongst these is PKS 1519–273, which possesses strong and variable circular polarisation, and which exhibits IDV in all Stokes parameters.

X-ray observations of PKS 1934–63 and S5 1946+708 have been made with Beppo SAX lasting 100 and 40 ksec, respectively. Both sources were detected, and in both there is evidence (at 90% confidence) of a strong iron Kα line which would indicate that the nuclei are surrounded by Compton thick material (˜1 g cm−2).

For 1946+708 VLBI H I absorption data are available from the literature. With the condition that free–free absorption should be modest for the source to be observable a minimum radius of around 20 pc is derived for the absorbing torus. The torus is predominantly molecular with a density of 108 cm−3. The corresponding pressure is rather high. Observations with XMM have been requested to verify the large equivalent width of Kα with a better confidence level.

In 1934–63 no important H I absorption was previously detected and no nucleus has been seen in radio data at 8 GHz. If this is due to free–free absorption a molecular torus with a radius less than 5 pc could be responsible. The density would be 108 cm−3 or more.

The relatively low S/N ratio of our observations does not allow a fully convincing conclusion to be reached. But our discussion shows that the combination of X-ray and high resolution 21-cm absorption data can provide important information on the physical parameters in the absorbing ‘torus’.

Paper to appear in full in Astronomy and Astrophysics.

We report very long baseline array (VLBA) observations at 2.3, 8.4, and 15.4 GHz towards nine gigahertz peaked spectrum (GPS) sources. One Seyfert 1 galaxy, one Seyfert 2 galaxy, three radio galaxies, and four quasars were included in our survey. We obtained spatial distributions of the free–free absorption (FFA) opacity with milliarcsecond resolution for all sources. It is found that type 1 (Seyfert 1 and quasars) and type 2 (Seyfert 2 and radio galaxies) sources showed different distributions of the FFA opacities. The type 1 sources tend to show more asymmetric opacity distributions towards a double lobe, while those of the type 2 sources are rather symmetric. Our results imply that the different viewing angle of the jet causes the difference of FFA opacity along the external absorber. This idea supports the unified scheme between quasars and radio galaxies, proposed by Barthel (1989).