The existence of mass components having opposite angular momentum in disk galaxies has been reported mainly for early type spirals. The origin of the counterrotation phenomenon is likely external, especially in those cases where large masses are involved and for galaxies where counterrotation in the stars and/or the gas is present within a large radial extent. Different types of counterrotation have been observed in stellar disks, either in the form of two well mixed disks of comparable masses (as in NGC 4550: Rubin et al. 1992; Rix et al. 1992), or revealed by the kinematical decoupling of nuclear disks or bulges (as in NGC 3593: Bertola et al. 1996, hereafter B96). Counterrotating gaseous disks can also adopt different configurations: whereas the over-all gas component of NGC 3626 counterrotates with respect to the stars (Ciri et al. 1995; García-Burillo et al. 1998), the nuclear gas counterrotates with respect to the outer disk gas in NGC 4826 (Braun et al. 1992; Rubin 1994).
Thakkar & Ryden (1996, 1998) have tested two plausible external origins of counterrotating systems using numerical simulations: a gas accretion infall process, or the result of a retrograde merger with a gas-rich dwarf galaxy. Their simulations succeed in building up thin and dynamically relaxed gaseous disks. However there remains a wealth of unanswered questions. In particular, so far there have been few studies on the long-standing evolution of counterrotating gaseous disks. Results based on the analytical approach of Lovelace et al. (1997) indicate that the main dynamical instabilities governing two-stream flows in galactic disks (containing both stars and gas) consist of one-arm spirals. More precisely, the spirals should be leading with respect to the most massive disk. This is in contrast with the observed preponderance of two-arm trailing spirals and bars, typical of disks that host no counterrotating component. The first numerical simulations of counterrotating disks by Comins et al. (1997) dealt only with the stellar component. Their results indicate that leading one-arm spirals do appear at the beginning of the run (after 1-2 rotation periods), but they vanish and transform into one-arm trailing spirals.
Observational constraints on the distribution and kinematics of the neutral gas (either atomic or molecular) in counterrotating galaxies can help us to elucidate the origin of the counterrotation, as well as test and refine the predictions of the different theoretical scenarios. A few HI maps exist of counterrotating galaxies (van Driel et al. 1989; Braun et al. 1994; Thakkar et al. 1997). However, for the small number of observed counterrotators, HI gas is seen to reside in the outer disk in the form of decoupled/inclined rings or even reflect the geometry of large scale cooling flows. Molecular gas is the privileged tracer of neutral gas in the inner disk, where dynamical instabilities, to which the gas is particularly sensitive, might develop. In addition, the study of the molecular gas phase and its relation to star formation tracers can be used to infer the star formation history of the accreted gas. The first massive counterrotating molecular gas disk (Mgas0.3109M) was found in the Sa spiral NGC 3626 by García-Burillo et al. (1998). In this galaxy, atomic and ionized gas share counterrotation with the molecular gas (Ciri et al. 1995; van Driel et al. 1989)
In this paper we study the morphology and dynamics of molecular gas in the inner disk of the starburst Sa spiral NGC 3593, using high resolution (43") interferometer observations made in the 1-0 line of 12CO. The present observations resolve the gas distribution in the disk and can be used to model the kinematics of the gas. The evolution of the gas response is studied by self-consistent numerical simulations, including the stars and the gas, adapted to the precise physical parameters of this galaxy. NGC 3593 is known to possess two counterrotating exponential stellar disks although with different scale lengths and total masses (rI=40", MI=1.21010M and rII=10", MII=2.7109M; B96, D=12.4Mpc). Ionized gas corotates with the less massive stellar disk (disk II ). 12CO single-dish observations (Wiklind & Henkel 1992) show that molecular gas emission comes mainly from a compact nuclear disk of radius r=350-620 pc. Although Wiklind & Henkel (1992) used a deconvolution algorithm to improve spatial resolution, the coarse resolving power of the original 30m data (22" and 13" in the 1-0 and 2-1 lines of 12CO, respectively) hampered a detailed study of the distribution and kinematics of the nuclear gas. A mere inspection of the major axis position-velocity (p-v) plot already indicates that molecular gas counterrotates with respect to disk I, similarly to HI (Krumm & Salpeter 1979).
Particular attention will be devoted to the comparison of the CO maps with other gaseous/stellar tracers, the objective being to analyze the gas response to the stellar potential and also the nature of the nuclear starburst.
© European Southern Observatory (ESO) 2000
Online publication: December 5, 2000