2019A&A...631A.132M

Using a sample of 2922 z<0.2, spectroscopically identified active galactic nuclei (AGN), we explore the relationship between radio size and the prevalence of extreme ionised outflows, as traced using broad [OIII] emission-line profiles in spectra obtained by the Sloan Digital Sky Survey (SDSS). To classify radio sources as compact or extended, we combined a machine-learning technique for morphological classification with size measurements from two-dimensional Gaussian models to data from all-sky radio surveys. We find that the two populations have statistically different [OIII] emission-line profiles; the compact sources tend to have the most extreme gas kinematics. When the radio emission is confined within 3'' (i.e. within the spectroscopic fibre or ≤5kpc at the median redshift), the chance of observing broad [OIII] emission-line components, which are indicative of very high velocity outflows and have a full width at half-maximum >1000km/s, is twice as high. This difference is greatest for the highest radio luminosity bin of log[L_1.4GHz/(W/Hz)]=23.5-24.5 where the AGN dominate the radio emission; specifically, >1000km/s components are almost four times as likely to occur when the radio emission is compact in this subsample. Our follow-up ~=0.3''-1'' resolution radio observations for a subset of targets in this luminosity range reveal that radio jets and lobes are prevalent, and suggest that compact jets might be responsible for the stronger outflows in the wider sample. Our results are limited by the available relatively shallow all-sky radio surveys, but forthcoming surveys will provide a more complete picture of the connection between radio emission and outflows. Overall, our results add to the growing body of evidence that ionised outflows and compact radio emission in highly accreting "radiative" AGN are closely connected, possibly as a result of young or weak radio jets.