next up previous


Large-scale HI structure of the Circinus Galaxy

K.L. Jones,1 B.S. Koribalski,2 M. Elmouttie1 and R.F. Haynes2
1Department of Physics, The University of Queensland 4072, Australia
2Australia Telescope National Facility, P.O. Box 76, Epping 2121, Australia

Abstract:

HI observations with the Australia Telescope Compact Array (ATCA) are used to study the neutral gas distribution and kinematics of the Circinus galaxy. One of the most outstanding features of Circinus is its enormous size (at least 80 arcmin or nearly 100 kpc, assuming a distance of 4 Mpc) revealing a beautiful large-scale spiral pattern. The fit of a tilted-ring model to the velocity field reveals a strong kinematic warp of the outer spiral arms. The inner region (radius 250 arcsec or 5 kpc) contains a highly elongated structure with non-circular motions, suggesting that Circinus is a barred spiral galaxy. At the centre of Circinus there is a 50-arcsec or 1-kpc-radius HI ring or disk rotating with a speed of at least 200 kms-1.


\begin{keywords}
galaxies: Seyfert -- galaxies: spiral -- galaxies: structure --...
...cs -- galaxies: individual: (Circinus) -- radio
lines: galaxies
\end{keywords}

Introduction

Circinus, a nearby ($\sim $ 4 Mpc) spiral galaxy with a Seyfert2 nucleus (Moorwood & Oliva 1990; Oliva et al. 1994), was discovered by Freeman et al. (1977) on an optical plate. At a Galactic latitude of $b \sim -4\degr$ it is highly confused by foreground stars. A deep optical picture is displayed by Elmouttie et al. (1995).

Optical observations show a galaxy diameter of several arcminutes (see Table 1), although its location makes it difficult to determine exact values (for a detailed discussion see Freeman et al. 1977). For the same reason the galaxy's morphological type (e.g. SAS3, de Vaucouleurs et al. 1991) remains somewhat uncertain.

Circinus has an active core with strong X-ray, infra-red, and radio emission. Matt et al. (1996) found an X-ray spectrum consistent with Compton scattering and fluorescent emission from cold matter illuminated by an obscured active nucleus. IRAS fluxes of 246 Jy at 60$\mu$ and 314 Jy at 100$\mu$ (Ghosh et al. 1992; Yamada et al. 1993) imply a far-infrared luminosity of 6 109 L$_{\odot}$. Radio continuum observations reveal a bright, compact source (Freeman et al. 1977; Whiteoak & Bunton 1985; Harnett 1987; Davies et al. 1998) as well as prominent lobes roughly perpendicular to the galaxy major axis (Harnett et al. 1990; Elmouttie et al. 1995), similar to those in NGC 3079 (Duric & Seaquist 1988). The nucleus of Circinus is known to contain a highly luminous 22-GHz H2O maser (Gardner & Whiteoak 1982; Whiteoak & Gardner 1986; Nakai et al. 1995; Greenhill et al. 1997), with a spectrum similar to that of the maser in the galaxy NGC 4258, which probably traces a molecular disk rotating around a supermassive object (Greenhill et al. 1995). Not surprisingly, NGC 3079 is also a megamaser galaxy (Baan & Haschick 1996). Harnett et al. (1990) observed the Circinus galaxy in OH absorption using the 64-m Parkes radio telescope at 1665 and 1667 MHz. The broad linewidths ($\sim $180 kms-1) suggest a rapidly rotating ring of clouds surrounding the nucleus or several independent clouds in the nuclear vicinity with inflowing and outflowing motions. Similar linewidths are also observed in HI absorption (Koribalski 1996). The detailed nuclear kinematics are being analysed by Koribalski & Whiteoak (in prep.). A nuclear ring has also been detected in H$\alpha $ and [OIII] as well as a funnel-shaped structure toward the NW in [OIII] indicative of nuclear gas outflow (Marconi et al. 1994). Elmouttie et al. (1998) find outflow velocities in H$\alpha $ of $\sim $150 kms-1.

Early HI observations with the Parkes 64-m radio telescope by Mebold et al. (1976) and Freeman et al. (1977) revealed neutral gas extending over an area more than one degree in diameter (limiting column density: $N_{\rm {H\,{\sc i}}} =
10^{19}$ atoms cm-2) with a half width of about 32 arcmin x 15 arcmin, much larger than the optical extent. The rotation curve appears flat at about 120 kms-1 out to a radius of 32$\pm$2 arcmin, giving a total mass of $\sim1.2~10^{11}$ M$_{\odot}$. The estimated hydrogen mass is about 7.2 109 M$_{\odot}$(see Table 1).



Table 1: Parameters of the Circinus galaxy.
 
Table 1: Parameters of the Circinus galaxy.
central position $\alpha,\delta$ (J2000) 14$^{\rm h}$13$^{\rm m}$10$.\!\!^{\rm s}$2,   -652020 (1)
  $\alpha,\delta$ (B1950) 14$^{\rm h}$09$^{\rm m}$18$.\!\!^{\rm s}$0,   -650618  
  l, b 3113, -38 (1,2)
type   SAS3 (1)
adopted distance D 4 Mpc    (-> 1 = 19.4 pc) (2)
optical extent at B25 69 x 27 (PA = 220) (1)
         "      at B25 70 x 35 (       "       ) (3)
optical extent* at B24.5 119 x 46 (2)
         "      at B26.6 172$\pm$17 (extrapolation) (2)
HI extent**   32 x 15 (2)
position angle $PA_{\rm {H\,{\sc i}}}$ 210 $\pm$ 5 (2)
inclination $i_{\rm {H\,{\sc i}}}$ 65 $\pm$ 2 (2)
systemic velocity $v_{\rm sys}$ 439 $\pm$ 2 kms-1   (HI) (2)
   (heliocentric)   380 $\pm$ 47 kms-1  (optical) (1)
velocity width $\Delta v_{20}$ 290 $\pm$ 10 kms-1  (HI) (2)
rotation velocity $v_{\rm rot}$ 152 $\pm$ 7 kms-1   (HI) (2)
far-infrared luminosity $L_{\rm FIR}$ 6 109 L$_{\odot}$ (4,5,6)
HI mass $M_{\rm H\,{\sc i}}$ $7.2 \pm 0.5 \times 10^9$ M$_{\odot}$ (2)
total mass $M_{\rm tot}$ $1.2 \pm 0.2 \times 10^{11}$ M$_{\odot}$ (2)


*extinction-corrected; **half width of the HI column density distribution after Gaussian-correction. References: (1) RC3: de Vaucouleurs et al. 1991, (2) Freeman et al. 1977, (3) ESO-Uppsala: Lauberts & Valentijn 1989, (4) calculated from the 60 and 100$\mu$m fluxes in the IRAS PSC 1988, (5) Ghosh et al. 1992, (6) Lonsdale et al. 1985.




The Australia Telescope Compact Array (ATCA)[*] observations presented here have a much higher angular resolution than the single-dish data and reveal the detailed distribution and kinematics of the HI gas in the disk and along the spiral arms of the Circinus galaxy.

In Table 1 several parameters of the Circinus galaxy are summarized.

In Section 2 we summarize the observations, followed by the results in Section 3, e.g. a detailed description of the HI gas distribution and kinematics. The three-dimensional shape of the Circinus galaxy is described and modelled in Section 4 and the case for Circinus being a barred spiral galaxy is presented in Section 5. The discussion of the HI mass and the total mass is presented in Section 6. In Section 7 the properties of the Circinus HI disk are compared to those of other spiral galaxies. The final section contains our conclusions.

Observations and Data Reduction

The HI observations of the southern Circinus galaxy have been carried out with three configurations (375, 750A and 1.5A) of the ATCA between 1995 February and August. Several observing parameters are summarized in Table 2.



Table 2: Observing Parameters.
 
Table 2: Observing Parameters.
configuration 375 750A 1.5A
baseline range 31-459 m 77-735 m 153-1469 m
  +5.5-6.0 km +3.1-3.8 km +3.0-4.5 km
date 03.8.1995 25.2.1995 31.3.1995
total observing time 13 h 13 h 13 h
total bandwidth 16 MHz
centre frequency 1418 MHz
number of channels 512
channel width 32.7 kHz $\cor$ 6.6 kms-1
angular resolution* 59 x 52
rms noise/channel* $\sim $1.3 mJybeam-1
limit HI column density $\sim7~10^{19}$ atoms cm-2
conversion factor* 1 mJybeam$^{-1} \cor$ 0.196 K
amplitude calibrator 1934-638 (14.9 Jy)
phase calibrator 1329-665 (2.61 Jy)
primary beam 33


*using `natural weighting' of the combined data set with baselines shorter than 3 kilometres.




The data reduction and analysis was carried out with the Miriad and AIPS software packages using mostly standard procedures. The uv-data were inspected to find those channels containing HI emission. As Circinus is near the plane of the Milky Way, Galactic HI emission is present in our observed frequency spectrum at radial velocities between about -280 kms-1 and +50 kms-1. The HI emission from Circinus is observed between about +240 kms-1 and +640 kms-1. The line-free channels were subtracted from the uv-data to form 1) HI line data and 2) narrow-band 20-cm continuum data. Both data sets were Fourier transformed and then CLEANed (Högbom 1974; Clark 1980) as described below. Using `natural weighting' for maximum sensitivity in the HI data resulted in an rms noise of $\sim $1.3 mJybeam-1 per channel and an angular resolution of 59 arcsec x 52 arcsec; ATCA baselines longer than 3 kilometres (see Table 2) were omitted. CLEANing was facilitated by constraining the image to be non-negative. Using `uniform weighting' for maximum angular resolution in the continuum data resulted in an rms noise of 0.13 mJybeam-1 and an angular resolution of 5.2 arcsec x 4.9 arcsec.

To derive the moment maps a smoothing function (width = 5 velocity channels and 3 pixels) and a cut-off level of 5 mJybeam-1 was applied. A smaller cut-off level could have been used, but this resulted only in an increased noise level outside the half width of the primary beam; no primary beam correction was applied.

The HI line data also display a complex absorption structure against the central continuum source which is not discussed here. For further analysis see Koribalski (1996) and Koribalski & Whiteoak (in prep.).

Results

The global HI spectrum and channel maps

The integrated HI spectrum of the Circinus galaxy is shown in Fig. 1 and looks similar to that of Freeman et al. (1977). It shows a double-horn shape, indicating a rather regularly rotating disk, but more gas than expected from a galaxy with a flat rotation curve is seen between the HI peaks. Departure from a plane disk with a flat rotation curve is further indicated by the peculiar shape of the HI emission in the individual channel maps (see Fig. 2). The HI emission extends over a maximum velocity range of $\Delta v \sim$400 kms-1, from about 240 kms-1 to 640 kms-1. The emission at the extreme velocities, mostly present near the galaxy centre, is rather weak. At the 20-per cent level of the peak flux in the HI spectrum the velocity width is $\Delta v_{20}$ = 320 kms-1, at the 50-per cent level we derive $\Delta v_{50}$ = 285 kms-1.

The velocity field of a uniformly rotating, inclined disk (see, e.g. Giovanelli & Haynes 1988) looks like a `spider'-diagram; lines of equal velocity vary from straight (along the galaxy minor axis) to V-shaped (near the major axis) with the apex at the centre of rotation. In Circinus, the channel maps (see Fig. 2) reveal the typical signs of warped spiral arms, i.e. loops and many systematically kinked lines instead of purely V-shaped structures.

The radio continuum

The narrow-band, 20-cm radio continuum image was convolved with a 20-arcsec beam to emphasize the large-scale structure (see Fig. 3). As found by Elmouttie et al. (1995; and references therein), there is a strong central core (450 mJy) and extended plumes of emission toward the south-east and north-west, these latter being more prominent than the disk emission along the major axis.

Recent H$\alpha $ observations (Elmouttie et al. 1998) revealed a funnel-shaped nuclear gas outflow from the central core into the same direction as the north-western plume, indicating that the two radio plumes are most likely associated with a large-scale bipolar flow of highly ionized gas. Because no effects of this outflow can be seen in the HI data (see, e.g. Fig. 4) we conclude that the plumes are perpendicular to the galaxy plane.

The HI moment maps

The moment maps (HI distribution, velocity field, and velocity dispersion) and a median-filtered optical image of the Circinus galaxy are shown in Fig. 4. The integrated HI distribution (0. moment; Fig. 4a) shows an enormous galaxy. At the first contour level (0.2 Jybeam-1 kms-1 = 39.2 K kms-1) which corresponds to a column density of $N_{\rm {H\,{\sc i}}} = 7.15~10^{19}$ atoms cm-2 or a surface density of 0.6 M$_{\odot}$pc-2 we measure a projected size of about 50 arcmin x 35 arcmin (or 58 kpc x 41 kpc). This is already larger than the ATCA primary beam of $\sim $33 arcmin which clearly restricts our measurements of the outer regions of Circinus.

Starting from the centre of the galaxy and working outward, the following regions are identified:

In Fig. 4a the major-to-minor axis ratio decreases from the centre outwards. For the inner disk (r = 250 arcsec) this ratio is $\sim $4, decreasing to 2 for the prominent inner parts of the great spiral arms (r = 600 arcsec) and to 1.4 at the outer limits of the disk as detected in Fig. 4a. This could either be a projection effect caused by a changing inclination of the HI disk, i.e. a warp, or an intrinsic change in the ellipticity of the orbits, i.e. a bar or oval distortion, or both. The latter is most likely as we have already found independent evidence for a bar in the inner region (see Section 5) and a warp of the outer spiral arms (see Section 4).

Fig. 4b shows an optical image of the Circinus galaxy (taken from the Digitized Sky Survey (DSS), and median filtered to subtract the foreground stars) which is much smaller than the HI image. The optical extent ($\sim $7 arcmin at B25, see Table 1) is about the same as the length of the bar. The slightly extended feature near the bottom of the DSS image is the star HD124197, here saturated.

Fig. 4c displays the mean velocity field (1. moment) of Circinus. Although the field appears quite regular and symmetric, there are strong deviations from plane circular orbits. The line-of-nodes twists in various ways and resembles in general a big, open Z-shape where the major kinks occur again at a radius of $r \sim 600$ arcsec. The largest asymmetry occurs towards the western side of the galaxy, where the velocity contours show a 90 twist which we interpret as two warped and overlapping spiral arms. The minor wiggles ($\sim $5 kms-1) in the velocity contours which occur quite regularly over the whole velocity field are the signature of a spiral density wave (see, e.g. Visser 1980; Canzian 1993; Lindblad, Lindblad & Athanassoula 1996).

Fig. 4d shows the mean velocity dispersion (2. moment) of Circinus. The highest dispersion measured close to the nucleus (up to 145 kms-1) is caused by the fast-rotating nuclear ring. Note also an increase of the velocity dispersion within the bar and near the western edge where two spiral arms appear to overlap.

The position-velocity diagrams

Fig. 5a shows the major axis pv-diagram (PA = 210) of Circinus. Two kinematic components are clearly visible: 1) a quite spectacular, steeply rising nuclear component (r < 50 arcsec) with maximum velocities of $\pm$200 kms-1 and 2) a more slowly rising component which reaches its maximum velocity of $\pm$180 kms-1 at a radius of $\sim $350 arcsec and beyond that gradually drops to $\sim $100 kms-1. The overall shape of the rotation curve is quite similar to that of our own Galaxy. Both components are also seen in the minor axis pv-diagram (Fig. 5b). We interpret the first component as the fast-rotating nuclear ring or disk of radius 50 arcsec or 1 kpc (see also Figs. 5c and d). The second component is quite typical for spiral galaxies of type Sb to Sd (see Giovanelli & Haynes 1988). The strong decline of the measured rotation curve in the outer parts of Circinus is very likely caused by projection effects due to a warp (see Section 4). From the pv-diagrams we obtain a size of 55 arcmin x 40 arcmin or 64 kpc x 47 kpc of the HI disk. The extent along the major and minor axes is quite asymmetric, 29 arcmin NE, 26 arcmin SW (Fig. 5a) and 23 arcmin SE, 17 arcmin NW (Fig. 5b), respectively. Intensity maxima occur at several intersections with the spiral arms.

Radial variation of the HI flux density

The approximate radial flux density distribution (Fig. 7a) has been obtained from the integrated HI distribution (Fig. 4a) assuming circular symmetry as well as an average inclination (i = 55) and position angle (PA = 210). Please note that parts of the HI disk have non-circular orbits (see Section 5) and both the inclination and position angle vary by more than $\pm$10 (see Section 4). The three regions previously identified from the HI moment maps (Fig. 4) are briefly discussed in Fig. 7a, and as follows:

Fig. 7b shows a comparison of the total mass and the HI surface density. The ratio is approximately constant, at least out to a radius of r = 1500 arcsec, beyond that the HI flux density represents only a lower limit. One possible explanation of the constant ratio is that the dark matter is distributed similar to the HI (see, e.g., Pfenniger & Combes 1994). For a detailed comparison with other spiral galaxies see Bosma (1981) and Broeils & van Woerden (1994).

The HI spectra

Over most of the disk the HI spectra are single peaked with a FWHM of 20-30 kms-1 (3-5 channel widths). More complex spectra are seen only in a few regions, e.g. within $\sim $150 arcsec of the galaxy centre (Fig. 3). The narrow-band 20-cm continuum image shows which spectra may be affected by HI absorption. The radio lobes toward the south-east and north-west are projections of plumes directed out of the plane of the disk (Elmouttie et al. 1995). The spectra at the position of the radio lobes do not show any significant high-velocity wings or other peculiarities indicating that there is no substantial amount of neutral hydrogen gas within the lobes and no interaction of the lobes with the HI disk (supporting their orientation out of the plane).

Double-peaked spectra are seen within the bar and in a few regions close to the spiral arms, the latter possibly being a result of the warp and some overlapping spiral arms. The separation of the peaks in the spectra is up to $\sim $40 kms-1.

The warp of the Circinus HI disk

There are numerous features which hint at a significant warp of the outer spiral arms in Circinus; the unusual loops and kinks in several channel maps (Fig. 2), the changing major-to-minor axis ratio of the overall HI distribution (Fig. 4a), and the general appearance and, in particular, the Z-shape of the velocity field (Fig. 4c). Any of these features alone could possibly be explained by a different mechanism. For example, spiral density waves produce a twisted velocity field (Walsh, Staveley-Smith & Oosterloo 1997), where individual channel maps might resemble those observed in Circinus. And, in the inner disk, the changing major-to-minor axis ratio is almost certainly caused by a change in orbit shape (see Section 5). But we are quite confident that the outer disk of Circinus is warped. Assuming circular orbits, which is not true for all parts of the disk, we fit a tilted-ring model to the mean velocity field to derive the rotation curve, and the radial variation of the inclination and position angle.

Fitting a tilted-ring model

A tilted-ring model was fitted to the Circinus velocity field (Fig. 4c) using the AIPS task Rocur according to the method described by Begeman (1989). Within the central 10-arcmin radius we found a systemic velocity of $v_{\rm sys}$ = 441.5$\pm$1.5 kms-1 and a centre position of $\alpha,\delta$(J2000) = $14^{\rm h}\,13^{\rm m}\,12^{\rm s}$, -652026, both consistent with previous data (see, e.g. Table 1). The deviations are well within the velocity and angular resolution of our data. Keeping those parameters fixed, we then derive the inclination angle (i), the position angle (PA) and the rotation velocity ($v_{\rm rot}$) of Circinus as a function of radius (r). For the tilted-ring fit we used a ring width of 30 arcsec or a half beam-width (in the plane of the galaxy). Fig. 8 shows the best fit obtained for the overall disk as well as for the approaching and receding sides separately.

The main results are as set out below.

Out to a radius of about 900 arcsec the fits of the approaching and receding sites of the galaxy agree quite well. Some deviations are found further out, generally increasing with radius. This possibly indicates an asymmetry of the Circinus disk in the outer regions. Beyond about 1700 arcsec the tilted-ring fit is rather unreliable, mainly because of a lack of data along the major axis. A future ATCA mosaic of the whole Circinus galaxy will allow us to study its outer regions in more detail.

The model and the residual velocity fields obtained from the tilted-ring fit of Circinus are shown in Figs. 9a and b, respectively. Whereas the model nicely reproduces the main features of the observed velocity field, the residual field indicates various deviations over the whole disk where the tilted-ring model was less successful. The bar, asymmetries of the disk and spiral density waves are likely to be the main reasons for the small discrepancies.

The three-dimensional shape of the warped disk is illustrated in Fig. 10, taking into account the slightly different results for the approaching and receding side of Circinus.

Circinus is a barred spiral galaxy

The HI gas dynamics in the inner disk of Circinus (see Fig. 6) suggest that it is a barred galaxy. All previous classifications described it as an unbarred galaxy (see Table 1). The bar has a length of $\sim $500 arcsec or roughly 10 kpc. Its position angle of $PA_{\rm bar}$ = 225 is similar to that of the major axis (PA = 210). The two great spiral arms originate at the ends of the bar as is commonly observed in barred galaxies. The geometry of Circinus, its velocity field, and various other indicators (e.g., the shape and polarization of the radio lobes and the dust lane) show that it is rotating in an anti-clockwise sense and that the spiral arms observed in HI are trailing. The orbits within the bar are elliptical as can be shown by measuring the angle between the major and minor axes which would be 90 for purely circular motion. Following the contour of the systemic velocity in Fig. 6 we find a much smaller angle of $\sim $60 within the bar. A similar analysis is shown by Lindblad et al. (1996; their Fig. 20) on the galaxy NGC 1365.

The bar pattern speed is about 15 kms-1kpc-1 assuming a co-rotation radius of 600 arcsec, which corresponds to 1.2 x the bar radius (Contopoulos & Grosbøl 1989; Athanassoula 1992).

The nuclear ring is very likely located close to the inner Lindblad resonance. Although the ring is clearly seen in HI (e.g. Fig. 5), it is much better resolved in H$\alpha $ images of Circinus (Marconi et al. 1994; Elmouttie et al. 1998). The H$\alpha $ ring ($\sim $40 arcsec diameter) is about half the size of the HI ring ($\sim $100 arcsec diameter) and is inclined at an angle of $\sim $40. CO spectra from the inner part of Circinus also suggest a circum-nuclear disk or torus (Johansson et al. 1991; Israel 1992; Curran et al. 1998).

In addition to the nuclear ring the H$\alpha $ images clearly show two short, symmetric extensions starting at the southern and northern part of the nuclear ring and curving in an anti-clockwise direction (see Fig. 11; also Marconi et al. 1994). They look very much like spiral arms, but are curved in the opposite direction from the large-scale HI arms. Since the HI and H$\alpha $ velocity fields do not indicate any counter-rotation, we have to conclude that these features are two leading spiral arms. Thus, Circinus has both a set of trailing and leading spiral arms. This is certainly a rather unique phenomenon and difficult to explain. One possible explanation has been presented by Yuan & Kuo (1997). They studied ``the non-linear response to a periodic bar potential in a differentially rotating gaseous disk'' and find that ``major morphological differences exist between spiral waves excited at the three types of Lindblad resonance.'' In particular, they find that ``spirals associated with the outer Lindblad resonance are tightly wound, while those associated with the inner Lindblad resonances are relatively open. In general spirals are trailing. However, those excited at the inner inner Lindblad resonance are leading.'' Yuan & Kuo also examine the process of fueling of AGNs and starburst rings in the various spiral wave modes.

The mass of Circinus

The total HI flux density, $F_{\rm {H\,{\sc i}}}$, of Circinus detected by the ATCA out to a radius of $\sim $30 arcmin is 1000 Jy kms-1. Using the expression $2.36~10^5~D^2~F_{\rm {H\,{\sc i}}}$ (Roberts 1962) we derive an HI mass of $M_{\rm {H\,{\sc i}}}~\ga~3.8~10^9$ M$_{\odot}$. As expected, this value is considerably smaller than the $7.2\,\pm\,0.5~10^9$ M$_{\odot}$ derived from the single-dish map by Freeman et al. (1977). The higher flux value is more realistic and can be explained by (a) the enormous extent of Circinus which is much larger than the primary beam of the 22-m dishes of the ATCA and (b) the nature of the interferometer to filter diffuse extended emission (although (a) is very likely the dominating reason).

The total mass of Circinus within the same radius ($R_{\rm max}$ = 35 kpc) is $M_{\rm tot} = 2.31~10^5~R_{\rm max}~v_{\rm rot}^2
= 1.9\,\pm\,0.3~10^{11}$ M$_{\odot}$. Because the extent of Circinus is much larger than measured here, this value for the total mass should be regarded as a lower limit. Freeman et al. (1977) find $M_{\rm tot} = 1.2~10^{11}$ M$_{\odot}$; this is smaller than our value because they used an inclination-corrected rotational velocity of only 118 kms-1. The hydrogen to total mass ratio is about 0.02-0.04.

Comparison with other galaxies

Rules for galactic warps

Briggs (1990) gives the rules of behaviour for galactic warps from a detailed study of 12 galaxies with extended warped disks as follows.

We conclude that Circinus follows Brigg's rules for the behaviour of warps but find that an additional change in the major axis position angle occurs at nearly twice the Holmberg radius.

Total mass and HI extent

An HI mass of 7.2 109 M$_{\odot}$ places Circinus as greater than average (2.5-4.0 109 M$_{\odot}$) for type Sa-Sc spirals (Haynes & Giovanelli 1984). Circinus is one the largest and most gas-rich Seyfert galaxies. In a single-dish survey of 91 Seyfert or Seyfert-like galaxies, Mirabel & Wilson (1984), found seven very gas-rich galaxies. There is however a great range of HI properties of Seyfert galaxies. Of the 91 galaxies, HI was detected in only 39.

The Circinus HI extent is at least 80 arcmin or 96 kpc (at a distance of 4 Mpc), a factor of $\sim $5 larger than the estimated Holmberg diameter (De26.6) of $\sim $17 arcmin (Freeman et al. 1977; see Table 1). Bosma (1981), in a study of the HI distribution and kinematics of 20 spiral galaxies, found typical HI diameters of 20-50 kpc at a limiting column density of 1.82 1020 atoms cm-2. Three of them had diameters larger then 96 kpc (M81: 100 kpc; M33: 130 kpc; and M31: 250 kpc). In a Westerbork Synthesis Radio Telescope search for spiral galaxies with extended HI disks (sensitivity 1020 atoms cm-2), Broeils & van Woerden (1994) found 5 out of 49 spiral galaxies had an HI extent > 96 kpc (125 kpc maximum). For a summary of high-sensitivity observations of extended HI galaxy envelopes see Huchtmeier & Richter (1982). At a sensitivity of 4-10 1018 atoms cm-2 they list eight examples of galaxies with HI extent 150-300 kpc. Some of the galaxies with the greatest HI diameters are: Mkn 348 (Morris & Wannier 1980; Heckman et al. 1982), NGC 628 (Briggs et al. 1980), M101 (Huchtmeier & Witzel 1979), and M83 (Huchtmeier & Bohnenstengel 1981). We conclude that the Circinus HI disk is among the most extensive spiral galaxies yet discovered.

Conclusions

Circinus is truly a remarkable galaxy. Its nuclear region is very active, revealing a starburst and an AGN as well as a fast-rotating ring and gas outflow; the prominent, large-scale radio lobes are possibly related to the nuclear outflow. Further kinematic features are the 5-kpc bar and the warp of the extended HI layer. Circinus is huge compared to most galaxies and is probably the closest AGN spiral galaxy known. It also appears quite isolated as its closest known neighbours are more than 15 away. No signs of tidal interactions have been found. A survey of the southern sky and in particular the ``zone-of-avoidance'' with the newly installed multibeam system at the 64-m Parkes telescope (Staveley-Smith et al. 1996) will provide more details about the large-scale environment of the Circinus galaxy (Koribalski et al., in prep.). The galaxy ESO097-G012, whose velocity is so far unknown, lies at a projected distance of only 21.6 arcmin from the centre of Circinus (close to the bottom right of Fig. 4) but must be much further away as it has not been detected within the observed velocity range.

Previous single-dish HI observations by Freeman et al. (1977) showed that the Circinus galaxy has an enormous extent of at least 80 arcmin or nearly 100 kpc, assuming a distance of 4 Mpc. This is much larger than the primary beam of the ATCA antennas and our HI study of the outer parts of the disk is therefore quite limited. To overcome this deficit we will mosaic the whole Circinus galaxy with the ATCA, which will also enable us to combine the interferometer data with the single-dish multibeam data described above.

The biggest advantage of the ATCA data presented here is their sensitivity and high angular resolution, which allows us to analyse in detail the enormous HI disk of Circinus. Because of its location near the Galactic Plane, Circinus is very difficult to study in the optical regime, and its optical size and type remain uncertain. Most other studies have concentrated on the bright nuclear region, and the existing single-dish HI observations have not revealed any significant structure. The most important results obtained from the ATCA HI data are summarized in the following.

Acknowledgments

Bibliography

1992
Athanassoula E., 1992, MNRAS, 259, 328 and 345

1996
Baan W.A., Haschick A., 1996, ApJ, 473, 269

1989
Begeman K.G., 1989, A&A, 223, 47

1981
Bosma A., 1981, AJ, 86, 1825

1990
Briggs F.H., 1990, AJ, 352, 15

1980
Briggs F.H., Wolfe A.M., Krumm N., Salpeter E.E., 1980, ApJ, 238, 510

1994
Broeils A.H., van Woerden H., 1994, A&AS, 107, 129

1993
Canzian N., 1993, ApJ, 414, 487

1980
Clark B.G., 1980, A&A, 89, 377

1989
Contopoulos G., Grosbøl P., 1989, A&AR, 1, 261

1998
Curran S.J., Johansson L.E.B., Rydbeck G., Booth R.S., 1998, A&A, submitted

1998
Davies R.I., Forbes D.A., Ryder S., Ashley M.C.B., Burton M., Storey J.W.V., Allen, L.E., Ward M.J., Norris R.P., 1998, MNRAS, 293, 189

1991
de Vaucouleurs G., de Vaucouleurs A., Corwin Jr. H.G., Buta R.J., Paturel G., Fouqué P., 1991, ``Third Reference Catalogue of Bright Galaxies'' (New York: Springer Verlag), [RC3]

1988
Duric N., Seaquist E.R., 1988, ApJ, 326, 574

1995
Elmouttie M., Haynes R.F., Jones K.L., Ehle M., Beck R., Wielebinski R., 1995, MNRAS, 275, L53

1997
Elmouttie M., Koribalski B., Gordon S., Taylor K., Houghton S., Lavezzi T., Haynes R.F., Jones K.L., 1998, MNRAS, 297, 49

1977
Freeman K.C., Karlsson B., Lynga G., Burrell J.F., van Woerden H., Goss W.M., 1977, A&A, 55, 445

1982
Gardner F.F., Whiteoak J.B., 1982, MNRAS, 201, 13P

1992
Ghosh S.K., Bisht R.S., Iyengar K.V.K., Rengarajan T.N., Tandoon S.N., Verma R.P., 1992, ApJ, 391, 111

1988
Giovanelli R., Haynes M.P., 1988, in ``Galactic and Extragalactic Radio Astronomy'', eds. G.L. Verschuur and K.I. Kellermann, Springer-Verlag, p.522ff

1995
Greenhill L.J., Jiang D.R., Moran J.M., Reid M.J., Lo K.Y., Claussen M.J., 1995, ApJ, 440, 619

1997
Greenhill L.J., Ellingsen S.P., Norris R.P., Gough R.G., Sinclair M.W., Moran J.M., Mushotzky R., 1997, ApJ, 474, L103

1987
Harnett J.I., 1987, MNRAS, 227, 887

1990
Harnett J.I., Whiteoak J.B., Reynolds J.E., Gardner F.F., Tzioumis A., 1990, MNRAS, 244, 130

1984
Haynes M.P., Giovanelli R., 1984, AJ, 89, 758

1982
Heckman T.M., Sancisi R., Balick B., Sullivan W.T., 1982, MNRAS, 199, 425

1974
Högbom J.A., 1974, A&AS, 15, 417

1981
Huchtmeier W.K., Bohnenstengel H.-D., 1981, A&A, 100, 72

1982
Huchtmeier W.K., Richter O.-G., 1982, A&A, 109, 331

1979
Huchtmeier W.K., Witzel A., 1979, A&A, 74, 138

1992
Israel F.P., 1992, A&A, 265, 487

1988
IRAS, 1988, IRAS Point Source Catalogue, Vol. 6, NASA RP-1190, Washington DC.

1991
Johansson L.E.B., Aalto S., Booth R.S., Rydbeck G., 1991, in ``Dynamics of Disk Galaxies'', p.249

1996
Koribalski B., 1996, in ``The Minnesota Lectures on Extragalactic Neutral Hydrogen'', ASP Conf. Series Vol. 106, ed. Evan D. Skillman, p.238

1989
Lauberts A., Valentijn E.A., 1989, ``The Surface Photometry Catalogue of the ESO-Uppsala galaxies'', University of Minnesota

1996
Lindblad P.A.B., Lindblad P.-O., Athanassoula E., 1996, ApJ, 313, 65

1985
Lonsdale C.J., Helou G., Good J.C., Rice W.L., 1985, ``Catalogued Galaxies and Quasars Observed in the IRAS Survey'', Appendix B (Washington)

1994
Marconi A., Moorwood A.F.M., Origlia L., Oliva E., 1994, ESO Messenger, 78, 20

1996
Matt G., Fiore F., Perola G.C., Piro L., Fink H.H., Grandi P., Matsuoka M., Oliva E., Salvati M., 1996, MNRAS, 281, L69

1976
Mebold U., Goss W.M., van Woerden H., 1976, Proc. ASA, 3, 72

1984
Mirabel I.F., Wilson A.S., 1984, ApJ, 277, 92

1990
Moorwood A.F.M., Oliva E., 1990, A&A, 239, 78

1980
Morris M., Wannier P.G., 1980, ApJ, 238, L7

1995
Nakai N., Inoue M., Miyazawa K., Hall P., 1995, PASJ, 47, 771

1994
Oliva E., Salvati M., Moorwood A.F.M., Marconi A., 1994, A&A, 288, 457

1995
Ott, M. 1995, PhD Thesis, University of Bonn

1994
Pfenniger D., Combes F. 1994, A&A, 285, 79

1962
Roberts M.S., 1962, AJ, 67, 437

1995
Staveley-Smith L., Wilson W., Bird T., Disney M., Ekers R.D., Freeman K., Haynes R., Sinclair M., Vaile R., Wright A., 1996, Publ. ASA, 13, 243

1980
Visser H.C.D., 1980, A&A, 88, 149 and 159

1997
Walsh W., Staveley-Smith L., Oosterloo T., 1997, AJ, 113, 1591

1985
Whiteoak J.B., Bunton J.D., 1985, Proc. ASA, 6, 171

1986
Whiteoak J.B., Gardner F.F., 1986, MNRAS, 222, 513

1993
Yamada T., Takata T., Djamaluddin T., Tomita A., Aoki K., Takeda A., Saito M., 1993, ApJS, 89, 57

1997
Yuan C., Kuo C.-L., 1997, ApJ, 486, 750


 
Figure 1: The integrated HI spectrum of the Circinus galaxy shows a double-horn shape, indicating a rather regularly rotating disk. The emission extends over $\Delta v \sim 400$ kms-1, from about 240 kms-1 to 640 kms-1. The relative large amount of gas near the systemic velocity ( $v_{\rm sys} \sim 440$ kms-1) indicates a falling rotation curve, which is caused by the warping of the outer spiral arms. - Only those channels containing line emission have been CLEANed and are displayed here; the velocity resolution is 6.6 kms-1.
\begin{figure}
\begin{tabular}{l}
\mbox{\psfig{figure=FIG1.PS,width=8.3cm} }
\end{tabular}
\end{figure}


 
Figure 2: The HI channels maps of Circinus. These maps have been obtained using `natural weighting' of the uv-data. The centre velocity of the channel is indicated in each panel; the channel width is 6.6 kms-1. The first and last channel maps are unCLEANed. The contour levels are -4.5 , 4.5 ($\sim $3$\sigma $), 20, 50, and 100 mJybeam-1; negative contours are dashed. The cross marks the galaxy centre (see Table 1).
\begin{figure*}
\begin{tabular}{l}
\mbox{\psfig{figure=FIG2_1.PS,width=17.6cm} }
\end{tabular}
\end{figure*}


 
Figure 2: continued.
\begin{figure*}
\begin{tabular}{l}
\mbox{\psfig{figure=FIG2_2.PS,width=17.6cm} }
\end{tabular}
\end{figure*}


 
Figure 2: continued.
\begin{figure*}
\begin{tabular}{l}
\mbox{\psfig{figure=FIG2_3.PS,width=17.6cm} }
\end{tabular}
\end{figure*}


 
Figure 2: continued.
\begin{figure*}
\begin{tabular}{l}
\mbox{\psfig{figure=FIG2_4.PS,width=17.6cm} }
\end{tabular}
\end{figure*}


 
Figure 3: HI spectra near the centre of Circinus. The position at which the spectrum was taken corresponds to the bottom left corner of the corresponding subplot. The overlaid narrow-band 20-cm continuum image (contours) shows the strong central source, the weak disk and the prominent radio lobes.
\begin{figure}
\begin{tabular}{l}
\mbox{\psfig{figure=FIG3.PS,width=8.3cm} }
\end{tabular}
\end{figure}


 
Figure 4: Moment maps of the Circinus galaxy. -(a) HI distribution. The contour levels are 0.2, 0.5, 1, 2, 3, 4, and 5 Jybeam-1 kms-1. -(b) Median-filtered optical image from the Digitized Sky Survey (DSS). -(c) Mean HI velocity field. The contour levels range from 291 to 581 kms-1in steps of 10 kms-1. -(d) HI velocity dispersion (contours) overlaid onto the HI distribution (grey scale). The contour levels are 5, 10, 15, 20, 30, 40 and 50 kms-1. The beam size is displayed in the bottom left corner.
\begin{figure*}
\begin{tabular}{ll}
\mbox{\psfig{file=FIG4A.PS,width=8.3cm} }&
...
...8.3cm} }&
\mbox{\psfig{file=FIG4D.PS,width=8.3cm} }
\end{tabular}
\end{figure*}


 
Figure: Position-velocity diagrams of the Circinus galaxy along (a) the major axis (PA = 210, width = 150 arcsec) and (b) the minor axis. Figures (c) and (d) show the nuclear region only.
\begin{figure*}
\begin{tabular}{l}
\mbox{\psfig{file=FIG5A.PS,width=16cm} }
\en...
...8.3cm} }&
\mbox{\psfig{file=FIG5D.PS,width=8.3cm} }
\end{tabular}
\end{figure*}


 
Figure 6: The HI distribution (grey scale) and mean velocity field (contours) of the inner disk of Circinus. The contour levels are nearly the same as those in Fig. 4c; to emphasize the systemic velocity the two contours next to it have been omitted. The beam size is displayed in the bottom left corner.
\begin{figure}
\begin{tabular}{ll}
\mbox{\psfig{file=FIG6.PS,width=8.3cm} }&
\end{tabular}
\end{figure}


 
Figure 7: (a) The HI column density, $N_{\rm H\,{\sc i}}(r)$, of Circinus as a function of projected radius assuming circular orbits and an average inclination and position angle. (b) A comparison of the surface density of the total mass, $M_{\rm tot}$ (solid line) and the HI mass, $M_{\rm H\,{\sc i}}$, (open circles).
\begin{figure*}
\begin{tabular}{ll}
\mbox{\psfig{file=FIG7A.PS,width=8.3cm} }&
\mbox{\psfig{file=FIG7B.PS,width=8.3cm} }
\end{tabular}
\end{figure*}


 
Figure 8: Results of a three parameter fit to the mean velocity field of the Circinus galaxy as a function of radius: -(a) the rotation velocity, $v_{\rm rot}(r)$, -(b) the position angle, PA(r) and -(c) the inclination angle, i(r).
\begin{figure*}
\begin{tabular}{l}
\mbox{\psfig{file=FIG8.PS,width=13cm} }
\end{tabular}
\end{figure*}


 
Figure 9: (a) Model and (b) residual velocity field for Circinus using the parameters of the tilted-ring model (see Fig. 8). Overlaid onto the residual velocity field (-50 to +22 kms-1; grey scale) is the observed velocity field (contours). The contour levels in both maps are the same as in Fig. 4c. The beam size is displayed in the bottom left corner.
\begin{figure*}
\begin{tabular}{ll}
\mbox{\psfig{file=FIG9A.PS,width=8.3cm} }&
\mbox{\psfig{file=FIG9B.PS,width=8.3cm} }
\end{tabular}
\end{figure*}


 
Figure 10: Projection of the 3-dimensional shape of the Circinus HI disk derived from the tilted-ring model (see Fig. 8).
\begin{figure*}
\begin{tabular}{l}
\mbox{\psfig{file=FIG10.PS,width=14cm} }
\end{tabular}
\end{figure*}


 
Figure 11: H$\alpha $ image of the inner region of the Circinus galaxy. In addition to the bright nuclear emission two short, symmetric extensions are visible. They start at the southern and northern part of the nuclear ring and curve in an anti-clockwise direction, opposite to that of the main spiral arms seen in HI.
\begin{figure}
\begin{tabular}{l}
\mbox{\psfig{file=FIG11.PS,width=8.3cm} }
\end{tabular}
\end{figure}

About this document ...

Large-scale HI structure of the Circinus Galaxy

This document was generated using the LaTeX2HTML translator Version 98.1p5 (May 15th, 1998)

Copyright © 1993, 1994, 1995, 1996, 1997, Nikos Drakos, Computer Based Learning Unit, University of Leeds.

The command line arguments were:
latex2html -split 0 circ_hi.tex

The translation was initiated by Baerbel Koribalski on 1998-09-01


Footnotes

... (ATCA)[*]
The Australia Telescope is funded by the Commonwealth of Australia for operation as a National Facility managed by CSIRO.

next up previous
Baerbel Koribalski
1998-09-01