FR-type radio sources at 3 GHz VLA-COSMOS: relation to physical properties and large-scale environment

Aims We probe the physical properties and large-scale environment of radio AGN in the faintest FR population to-date, and link them to their radio structure. We use the VLA-COSMOS Large Project at 3 GHz, with resolution and sensitivity of 0".75 and 2.3 µ Jy/beam, respectively, to explore the FR dichotomy down to µ Jy levels. Methods We classify objects as FRIs, FRIIs or hybrid FRI/FRII based on the surface-brightness distribution along their radio structure. Our control sample is the jet-less/compact radio AGN (COM AGN) which show excess radio emission at 3 GHz VLA-COSMOS exceeding what is coming from star-formation alone; this sample excludes FRs. Largest angular projected sizes of FR objects are measured by a machine-learning algorithm and also by hand, following a parametric approach to the FR classification. Eddington ratios are calculated using scaling relations from the X-rays, while we include the jet power by using radio luminosity as a probe. We investigate their host properties (star-formation ratio, stellar mass, morphology), and we explore their incidence within X-ray galaxy groups in COSMOS, as well as in the density fields and cosmic-web probes in COSMOS. Our sample is composed of 59 FRIIs, 32 FRI/FRIIs, 39 FRIs, and 1818 COM AGN at 0.03 =z= 6. FR objects have on average similar radio luminosities ( L3 GHz~1023 W Hz-1 sr-1 ), spanning a range of 1021-26 W Hz-1 sr-1 , and lie at a median redshift of z ~ 1 . FRs reside in their majority in massive quenched hosts ( M* >1010.5M? ), with older episodes of star-formation linked to lower X-ray galaxy group temperatures, suggesting radio-mode AGN quenching. Irrespective of their radio structure, FRs and COM AGN are found in all types and density environments (group or cluster, filament field). Conclusions By relating the radio structure to radio luminosity, size, Eddington ratio, and large-scale environment, we find a broad distribution and overlap of FR and COM AGN populations. We discuss the need for a different classification scheme, that expands the classic FR classification by taking into consideration the physical properties of the objects rather than their projected radio structure which is frequency-, sensitivity- and resolution-dependent. This point is crucial in the advent of current and future all-sky radio surveys.