The Unusual Tidal Dwarf Candidate in the Merger System NGC 3227/3226: Star Formation in a Tidal Shock?

Abstract

We report the discovery of active star formation in the H i cloud associated with the interacting Seyfert system NGC 3227/NGC 3226 that was originally identified as a candidate tidal dwarf galaxy (TDG) by Mundell et al. and that we name J1023+1952. We present broad-band optical B, R, I (from the INT) and ultraviolet images (from XMM-Newton) that show the H i cloud is associated with massive on-going star formation seen as a cluster of blue knots (M B ∼ <−15.5 mag) surrounded by a diffuse ultraviolet halo and co-spatial with a ridge of high neutral hydrogen column density (N H ∼3.7 × 10 21 cm −2 ) in the southern half of the cloud. We also detect Hα emission from the knots with a flux density of F Hα ∼2.55 × 10 −14 erg s −1 cm −2 corresponding to a star-formation rate of SFR(H Hα ) ∼10.6 ×10 −3 M ⊙ yr −1 . J1023+1952 lies at the base of the northern tidal tail, and, although it spatially overlaps the edge of the disk of NGC 3227, Mundell et al. showed that the H i cloud is kinematically distinct with an H i mean velocity 150 km s −1 higher than that of NGC 3227. Comparison of ionized (Hα) and neutral (H i) gas kinematics of the cloud show closely matched recessional velocities, providing strong evidence that the star-forming knots are embedded in J1023+1952 and are not merely optical knots in the background disk of NGC 3227, thus confirming J1023+1952 as a gas-rich (M H /L B > 1.5) dwarf galaxy. No star formation is detected in the northern half of the cloud, despite similar H i column densities; instead, our new high resolution H i image shows a ridge of high column density coincident with the reddest structures evident in our B−I image. We suggest these structures are caused by the background stellar continuum from the disk of NGC 3227 being absorbed by dust intrinsic to J1023+1952, thus placing J1023+1952 in front of NGC 3227 along the line of sight. We discuss two scenarios for the origin of J1023+1952; as a third, pre-existing dwarf galaxy involved in the interaction with NGC 3227 and NGC 3226, or a newly-forming dwarf galaxy condensing out of the tidal debris removed from the gaseous disk of NGC 3227. The first scenario is feasible given that NGC 3227 is the brightest member of a galaxy group, an environment in which pre-existing dwarf galaxies are expected to be common. However, the lack of a detectable old stellar population in J1023+1952 makes a tidal origin more likely. If J1023+1952 is a bound object forming from returning gaseous tidal tail material, its unusual location at the base of the northern tail implies a dynamically young age similar to its star-formation age, and suggests it is in the earliest stages of TDG evolution. Whatever the origin of J1023+1952 we suggest that its star formation is shock-triggered by collapsing tidal debris.

The SKA view of the Neutral Interstellar Medium in Galaxies

“…Oosterloo et al 2007;Mapelli et al 2008). Clouds and filaments at similar column densities are observed also around other galaxies (Fraternali et al 2002;Mundell et al 2004;Westmeier et al 2005;Sancisi et al 2008). As an aside, a similar increase in the prominence of H I when going to lower column densities is also found for early-type galaxies.…”

Molecular Gas and Dust in Arp 94: The Formation of a Recycled Galaxy in an Interacting System

“…The subsequent discovery of a region of very blue star-forming knots with luminosities between 10 6 and 5 ; 10 6 L , embedded in a high H i column density ridge in the southern half of J1023+1952 (Mundell et al 2004), confirmed its classification as a dwarf galaxy; the inferred SFR, UV luminosities, and H equivalent widths of the knots suggested a starburst age less than 10 Myr. Near-infrared imaging provided further evidence for the youth of J1023+1952; no additional embedded star formation or old stellar population were found ( Mundell et al 2004).…”

“…Although the edges of the galactic disks of NGC 3227 and J1023+ 1952 spatially overlap in projection (Mundell et al 1995(Mundell et al , 2004, it is possible to disentangle their different kinematic components in the current 12 CO(1Y0) data set. On average, the emission from NGC 3227 is blueshifted by approximately 200 km s À1 with respect to that of J1023+1952, with line widths not in excess of 100 km s À1 , making it possible to separate them by fitting a double Gaussian line profile.…”

“…Spitzer Mid-IR Emission At all observed wavelengths, the only obvious emission within J1023+1952 is from the SF knots in the south. Figure 8a shows the optical B-band image with the linear arrangement of bright blue knots labeled 1, 2, 3, 4, and 5, according to Mundell et al (2004); we also label additional knots, 1a and 2a, which were visible in the images of Mundell et al (2004), but not specifically identified by them. In Figs.…”

“…We carried out aperture photometry of the SF knots on images at optical wavelengths (B, V, and I bands), H (data from Mundell et al 2004), and the Spitzer IRAC bands and 15 m emission. Due to the poorer spatial resolution, we could not derive the 24 m fluxes for the individual knots.…”