1 Introduction

Radio emission from Seyfert galaxies traces energetic, mechanical processes occurring in the active nucleus. The radio continuum spectra of Seyferts are almost invariably steep, consistent with optically-thin synchrotron emission. In those sources where the radio continuum has been resolved, the morphology is commonly linear, interpreted to trace a stream of ejected plasma, or a jet, originating from the central engine. For example, most well-resolved Seyfert radio sources are double or triple sources straddling the active nucleus. The radio emission of the nearest and brightest Seyferts resolves into tightly collimated structures originating from the active nucleus, consistent with a jet interpretation (e.g., NGC 1068: Gallimore et al. 1996; Muxlow et al. 1996; Wilson & Ulvestad 1983; NGC 4151: Pedlar et al. 1993; Mrk 3: Kukula et al. 1993). There is evidence that Seyfert radio jets impact the dense gas in the near-nucleus environment and affect the gas distribution, morphology, and ionization of the narrow-line region (NLR). The orientation of the radio jet is also the only straightforward measure of the symmetry axis of the active nucleus, useful for testing orientation-based unifying schemes for AGNs.

Early studies searching for differences between Seyfert 1's and 2's (de Bruyn & Wilson 1978; Meurs & Wilson; Ulvestad & Wilson 1984a; Ulvestad & Wilson 1984b) concluded that Seyfert 2's have stronger and larger radio sources than Seyfert 1's, however, these studies were biased by optical selection: weaker Seyfert 2 galaxies were omitted from the samples. There was found to be little or no difference between Seyfert 1 and 2 radio sources in follow-up studies which properly considered relatively unbiased, volume-limited samples (Edelson 1987; Ulvestad & Wilson 1989; Giuricin et al. 1990). More specifically, there is no statistically significant difference in the distribution of radio luminosity, and only a marginal difference in the distribution of radio source sizes, with Seyfert 2's tending to be slightly larger than Seyfert 1's (although only at <90% significance, Wilson 1991). In contrast, unifying schemes predict that the radio jets in Seyfert 2's should appear larger in projection, since the unifying model orients the collimating disk more nearly edge-on in narrow-line AGNs. It is not simple to reconcile the statistics of radio sources with the unified scheme hypothesis.

While there is no tendency for the radio sources to orient in any preferred direction with respect to the plane of the host galaxy (Ulvestad & Wilson 1984a), the radio sources are commonly elongated in the same direction as the NLR (Haniff et al. 1988; Pogge 1989; Wilson et al. 1988), i.e. the inner part of the optical emission regions (extending up to few kpc). In some Seyfert 2s the NLR has a conical or bi-conical shape with the nucleus at the apex (e.g. Pogge 1989, 1997). Wilson & Tsvetanov (1994) showed that the radio axis is invariably co-aligned with the ionization cone axis for the 11 ionization cones known at that time.

The alignment of the radio jet and NLR raises a question of energetics -- does the radio jet ionize and heat the NLR significantly compared to ionizing radiation from the AGN? In several well-studied cases, there is a detailed morphological association between the NLR and the radio emission (e.g. Whittle et al. 1988; Whittle 1989; Capetti et al. 1996; Gallimore et al. 1996). Currently, the numbers of such well-studied cases is too small to address questions of interaction and energetics in a statistically complete and meaningful sense.

To this end, Tsvetanov et al. (in preparation) have recently assembled a volume-limited sample (cz< 3600 km s^-1) of well-classified Seyfert galaxies. The main advantage of this sample is that all of the sources have been extensively observed in the optical with narrow-band imaging done with the ESO NTT, 3.6 m and 2.2 m telescopes. This survey provides high-quality emission-line maps (in [O III] $\lambda$5007 and H$\alpha$+[N II]) with a typical resolution of $\sim1$$^{\prime\prime}$. The sample includes objects with and without known extended emission line regions, providing detailed information about the extent, morphology, and degree of ionization (obtained from the [O III] $\lambda$5007/H$\alpha$+[N II] ratio) of the emission-line regions. It is therefore important to obtain for such a sample detailed radio images in order to perform a detailed comparison of the radio and optical morphology on arcsecond (hundred-pc) scales.

Here we present a radio imaging survey of the Tsvetanov et al. sample, made using the Very Large Array (VLA) and the Australia Telescope Compact Array (ATCA) aperture synthesis telescopes. Observing configurations and frequencies were chosen to match the $\sim$ 1$^{\prime\prime}$ resolution of the optical narrow-band observations. We discuss the radio properties and statistics for the 29 surveyed sources, and a more detailed comparison of the radio and optical properties will be presented in a forthcoming paper.