Free Access
Issue
A&A
Volume 522, November 2010
Article Number A59
Number of page(s) 7
Section Extragalactic astronomy
DOI https://doi.org/10.1051/0004-6361/200913511
Published online 03 November 2010

© ESO, 2010

1. Introduction

Studying the molecular cloud properties in starburst, active and interacting galaxies is important in order to understand the feedback mechanisms between star formation, dynamics and the interstellar medium. Important probes of cloud properties include molecules such as HCN, HCO+ and HNC (e.g. Krips et al. 2008; Graciá-Carpio et al. 2008; Imanishi et al. 2004; Gao & Solomon 2004; Aalto et al. 2002) that trace the dense (n ≳ 104 cm-3) star forming phase of the molecular gas.

Alternatively, to study the bulk properties of the molecular gas one can use the ratio between 12CO and its isotopomer 13CO. Globally, there is a correlation between the 12CO/13CO 1–0 ratio (ℛ) and the FIR f(60 μm)/f(100 μm) flux ratio (e.g. Young & Sanders 1986; Aalto et al. 1995). The extreme values of ℛ (i.e. ℛ  >  20) generally occur in luminous merging galaxies with large dust temperatures (e.g. Aalto et al. 1995; Glenn & Hunter 2001). Within galaxies there is a general trend of increasing ℛ towards the central region where the gas is warmer and denser (e.g. Wall et al. 1993; Aalto et al. 1995; Paglione et al. 2001).

Both high kinetic temperatures and large turbulent line widths will decrease the optical depth (τ), of the 12CO and 13CO 1–0 lines and thus elevate the line ratio ℛ. Therefore, information on the spatial variation of ℛ can be used to identify regions of extreme or unusual physical conditions in the molecular gas – and global values can help classify galaxies. Aalto et al. (1995) suggested some general diagnostics of the cloud conditions and environment based on global values of ℛ:

  • small ratios, ℛ  ≈  6 are an indication of a normal Galactic disc population of clouds dominated by cool, self-gravitating giant molecular clouds (GMCs);

  • intermediate ratios 10 ≲ ℛ ≲ 15 are associated with the inner regions of normal starburst galaxies where the gas is warmer and denser than in the disc;

  • the extreme values ℛ  >  20 seem to originate in warm, turbulent, high pressure gas in the centres of luminous mergers with highly compact molecular distributions and gas surface densities in excess of 104 M pc-2 – two orders of magnitude higher than in typical Milky Way GMCs. Large surface densities require high pressures in hydrostatic equilibrium and low density (n < 103 cm-3) – 12CO emitting – gas must be supported by large turbulent line widths (P ∝ n(δV)2). Thus, τCO can be reduced to moderate (≈ 1) values, resulting in large ℛ.

More recent studies show that high values of ℛ also occur more locally in less extreme galaxies. Tosaki et al. (2002) find this in the spiral arms of M 51 and Meier & Turner (2004) in the centre of the spiral galaxy NGC 6946. Hüttemeister et al. (2000) find large values of ℛ in the large scale bar of NGC 7479 as a result of the dynamical impact of a density wave on gas properties.

The nearby merger NGC 4194 – the Medusa merger – belongs to a class of lower luminosity (LFIR = 8.5 × 1010 L at D = 39 Mpc) E+S mergers (Aalto & Hüttemeister 2000; Manthey et al. 2008), an order of magnitude fainter than well known ultra luminous galaxies (ULIRGs (LFIR ≳ 1012 L)) such as Arp 220. The Medusa has an elevated value of ℛ (≈ 20) (e.g. Aalto et al. 1991; Casoli et al. 1992; Glenn & Hunter 2001), but in contrast to the more luminous high-ℛ mergers, NGC 4194 has a relatively extended (2 kpc) starburst region (e.g. Armus et al. 1990; Prestwich et al. 1994) and Aalto & Hüttemeister (2000) found that the molecular gas was also distributed on the comparatively large scale of 25'' (4.7 kpc), despite its advanced stage of merger. Due to the linear extent of the molecular gas and relative proximity of the Medusa merger, it is possible to spatially resolve the relative distributions of 12CO and 13CO and hence allow the investigation of the underlying causes of the elevated ℛ values observed in this galaxy. The Medusa starburst is particularly interesting to study since its ratio of 163 rivals that of ULIRGs (for example Arp 220 with a of 210), despite the comparatively modest luminosity, indicating an exceptionally high star formation efficiency (SFE) in the Medusa merger. Gao & Solomon (2004) find that for galaxies with LFIR ≲ 1011 L the ratio is typically 33. At higher luminosity they find that instead increases to 100–300. Gao & Solomon (2004) suggest that, in contrast to CO, the ratio is independent of luminosity with . Interestingly the Medusa deviates significantly from this HCN correlation with a based on upper limits from Aalto & Hüttemeister (2000). Further discussion of the underlying causes of this discrepancy are presented in Aalto & Hüttemeister (2000).

We have imaged NGC 4194 in the 13CO 1–0 line with the Owens Valley Radio Observatory (OVRO) millimetre array. Our aim was to compare the distribution of the 12CO and 13CO emission within the galaxy and to see how it relates to star formation, gas surface density and dynamics. In Sects. 2 and 3 we discuss the observations and results. In Sect. 4 we discuss the ℛ line ratio variations within the Medusa merger in terms of warm gas in the starburst region and possible diffuse gas in the central dust lane. Future observational tests for these proposed scenarios are suggested.

2. Observations

We have obtained maps of 13CO 1–0 using the Caltech six-element OVRO millimeter array. Two tracks were taken in the low resolution configuration in April 2000. The naturally weighted synthesised beam size is 4'̣56 × 3'̣98 (861 × 752 pc for D = 39 Mpc) and the beam position angle (BPA) is –40°.

System temperatures were 500–600 K. The quasar 1150+497 was used for phase calibration and Neptune and Uranus for absolute flux calibration. The digital correlator, centred at VLSR = 2560 kms-1, provided a total velocity coverage of 1170 kms-1. Data were binned to 4 MHz resolution, corresponding to 11 kms-1, and to construct the map, the resolution was reduced to 44 kms-1. The sensitivity of this map is 3 mJy beam-1 channel-1, corresponding to 0.014 K channel-1.

thumbnail Fig. 1

The 110 GHz continuum of NGC 4194 overlaid on an HST WFPC2 image. The contours are: 0.9 × (−3,3,4,5) mJy beam-1. Rms in the map is 1 mJy beam-1 thus the first contour is at 2.7σ. The peak flux density is 4.6 mJy beam-1. The cross marks the position of the 1.4 GHz continuum (Condon et al. 1990). Hatched ellipse indicate beam size and orientation.

Open with DEXTER

2.1. Continuum subtraction

Using the aips task imlin we fitted continuum to the line-free channels and subtracted from the cleaned map. The resulting continuum map is presented in Fig. 1 and the flux density and position in Table 1. The 1.4 GHz continuum position of Condon et al. (1990) agree within 0''̣4 with the 110 GHz position which is within the positional errors of the 13CO map. The 1.4 GHz radio continuum emission is dominated by synchrotron emission resulting primarily from star-formation, along with a small contribution from the compact core (Beswick et al. 2005; Condon et al. 1990). The 110 GHz emission (see Fig. 1) is coincident with the area of highest dust obscuration observed in the optical. This emission is tracing the thermal emission from dust in this region illuminated by ongoing star-formation. This is consistent with the alignment in the 110 and 1.4 GHz continuum positions observed.

2.2. Alignment of the data

Both the 12CO and 13CO observations were made using the same instrument and employing identical gain and passband calibrators, thus minimising potential misalignments due to instrumental effects. Furthermore, the 110 GHz continuum extracted from the 13CO data set is positionally coincident with the 1.4 GHz continuum (see Sect. 2.1) suggesting that there are no systematic errors in the positions of the 13CO data set. We also note that the shift between 12CO and 13CO is partially caused by missing 13CO emission in some velocity channels (see Sect. 3.3).

Table 1

Continuum and 13CO line results.

3. Results: morphology of the 13CO emission

3.1. Previous observations: the 12CO distribution

High resolution OVRO aperture synthesis maps of the 12CO 1–0 emission in the “Medusa” galaxy merger (NGC 4194) are presented in Aalto & Hüttemeister (2000). It was found that the molecular emission is surprisingly extended. The 12CO emission is distributed on a total scale of 25'' (4.7 kpc) – despite the apparent advanced stage of the merger. The complex, striking 12CO morphology occupies the centre and the north-eastern part of the main optical body (see Fig. 2 (left panel)). The extended 12CO flux density traces the two prominent dust lanes: one of which crosses the central region at right angles to the optical major axis, and the second which curves to the north-east and joins the base of the northern molecular tidal tail. The total molecular mass in the system is estimated to be at most 2 × 109 M, depending on which 12CO–H2 conversion factor is applicable.

3.2. Integrated 13CO 1–0 intensity

The naturally weighted, 4''̣5 resolution, 13CO emission is distributed on a scale of 8'' (1.5 kpc) in an oval structure centred on the optically bright centre of the galaxy. There is also an off-centre 13CO concentration 10'' north of the centre. In Fig. 2 (right panel) we show the integrated 13CO map overlaid on a 12CO map smoothed to the 13CO resolution. Integrated- and peak flux densities are listed in Table 1.

The 13CO emission in the inner 8'' is shifted towards the north-west as compared to the 12CO distribution. The shift is 2''̣3 (430 pc). Also the northern 13CO emission peak is shifted off the corresponding 12CO peak to the south-west by 2''̣5. In Fig. 3 (centre panel) the 13CO 1–0 map is overlaid on an HST WFPC2 image. This shows that the 13CO emission is not tracing the central dust lane (which the 12CO does (left panel in Fig. 3)) but instead is shifted towards the north-west part of the starburst and towards the western dust features. The moment maps were produced through smoothing in velocity and space (3 channels, 3 pixels boxcar smoothing) and then applying a 2σ flux cut-off (allowing both positive and negative flux levels).

thumbnail Fig. 2

Left: overlay of the 12CO contours (white, apart from clouds in the tail marked in black) on a greyscale, overexposed optical R-band image (Mazzarella & Boroson 1993). The dashed curve marks the edge of the OVRO primary beam (from Aalto & Hüttemeister 2000). Right: integrated 13CO 1–0 line emission in greyscale overlaid on a 12CO contour map smoothed to the 13CO resolution of 4''̣5. The greyscale range is 0.9–2.0 Jy kms-1 beam-1. The CO contours are 5.3  ×  (1, 2, 3, 4, 5, 6, 7, 8) Jy kms-1beam-1 (starting at 6σ). Cross indicate position of 1.4 GHz radio continuum peak (as in Fig. 1).

Open with DEXTER

3.3. The channel map

In Fig. 4 the channel maps of 12CO and 13CO 1–0 are presented, smoothed to a resolution of 5''. Compared to 12CO, 13CO-emission is “missing” primarily in three velocity bins in the centre: 2495, 2538 and 2620 kms-1. In the first case we find 13CO emission 10'' north of the centre – despite 12CO in the same bin having a peak in the centre. In the second case 13CO emission is found only west of the centre while 12CO is distributed east and west of the radio peak. In the 2620 bin there is a bright 12CO peak in the centre which is missing in 13CO. In general, the channel maps show that there is no 13CO emission east, north-east of the radio continuum peak apart from a small region of emission 10'' north of the central region.

3.4. Line ratios

We find that ℛ is 7–10 in the region 10'' north of the centre – while in the centre it is significantly greater: ranging from 16 in the western region of the centre to >45 (3σ) in the eastern part of the centre where the central dust lane is making a 90 degree turn. Over a distance of 5'' (945 pc) the line ratio, , changes by more than a factor of 3. A map of the varying ℛ line ratios is displayed in Fig. 3 (right panel).

thumbnail Fig. 3

Left: 12CO contours on a WFPC2 image zoomed in on the central region. Centre: 13CO contours on WFPC2 image. Both 12CO and 13CO contours are 40% of peak flux  ×  (1, 1.5, 2, 2.25). For 13CO the first contour starts at 2.5σ, for 12CO the first contour starts at 24σ. Right: line ratio map where the greyscale ranges from light (ℛ = 10) to dark (ℛ = 60). Both limits and real values in greys (same scale), and the area where there are actual measures overlaid with contours (yellow). Contours are ratios ℛ = 15, 25, 35, 45. Limits are at the 3σ13CO detection threshold. Please note the small pixel-size (0''̣5) of the 12CO and 13CO maps which may give the impression that there is structure beyond the actual angular resolution of the 12CO and 13CO maps.

Open with DEXTER

4. Discussion

The line ratio change in the Medusa occurs on scales of 0.5 to 1 kpc where the ratio goes from extreme ratios >40 to more normal starburst ratios of 16. This observed gradient in ℛ is likely to be caused either by the presence of diffuse, unbound molecular gas or a temperature gradient in the ISM. In the following section we discuss each of these scenarios and compare these with observations in other galaxies.

4.1. Diffuse gas in the central dust lane?

Optical images of the Medusa merger are dominated by the strong absorption from several dust lanes – the most prominent one crossing the central starburst region. We suggest that the intense starburst activity is being fed by gas funnelled along the dust lane – supported by the bright 12CO emission associated with the dust lane(s), and the distribution and dynamics seen in the Hi absorption at high resolution (Beswick et al. 2005). The central dust lane is in front of the main burst region in the east and then appears to curve into the burst region itself at the centre. A possible scenario is then that the molecular gas is diffuse (non self-gravitating) in the dust lane region – creating the general east-west gradient in the line ratio ℛ.

Interestingly, a comparison of the Medusa 13CO image and the Hα images of Hattori et al. (2004) show that the position of the Hα equivalent width (EW) peak coincides with the regions of lower values of ℛ. The large EW is consistent with the region being an extra-nuclear star forming region. It coincides with the western radio continuum arm and the Giant Molecular Association (GMA) feature c in the high-resolution 12CO map (see Aalto & Hüttemeister 2000). Thus, it seems that the 13CO emission correlates spatially with more quiescent, spiral-arm like star formation, shifted from the dust lane and the central starburst. Note that if the 12CO 1–0 is mainly tracing diffuse gas the 13CO 1–0 emission is showing the real mass distribution of the molecular gas. Figure 10 of Aalto & Hüttemeister (2000) shows that the correlation between 1.4 GHz radio continuum and the central dust lane is poor – this could be a further indication of both a lack of ongoing star formation in the dust lane as well as the molecular gas potentially being diffuse in the dust lane (see for instance the discussion of 12CO and radio continuum in the spiral arm of M 83 by Rand et al. 1999).

thumbnail Fig. 4

Channel maps of 12CO (left) and 13CO (right). Boxes indicate channels of “missing” 13CO emission compared to 12CO. The 12CO map has been smoothed to the same resolution as 13CO. Intensity levels for 12CO and 13CO are same fraction of peak: 35% (–1.5, –1, 1.5, 2, 2.5) – where the peak is 360 and 25 mJy beam-1 for the 12CO and 13CO respectively. Note that for 12CO a big fraction of the detected, extended flux is not shown due to the cut-off. The comparative channels maps show the position of the peak emission for 12CO and 13CO. The velocity in panel 18 = 2714 kms-1, panel 14 is 2538 kms-1, and panel 10 = 2362 kms-1. Increment between panels is 44 kms-1. Crosses indicate the 1.4 GHz radio continuum position. The beam is marked in the upper right corner of the first panel for the 12CO and 13CO channel maps. Greyscale for the 12CO map ranges from 50 to 400 mJy beam-1 and for 12CO from 7.5 to 26 mJy beam-1. For the 13CO map the first contour corresponds to 2.5σ.

Open with DEXTER

4.1.1. Spatial shifts in ℛ in other galaxies

Even if the central region of the Medusa is more chaotic than that of a density wave spiral galaxy, and the extinction structure suggests a three-dimensional central region, it is tempting to compare the shift in 12CO and 13CO to that found in other galaxies. Tosaki et al. (2002) found a spatial shift between 12CO and 13CO 1–0 emission on similar spatial scales in the southern spiral arm of M 51. The shift seems to separate diffuse (unbound) lower density gas (as traced by 12CO) from self-gravitating gas emitting the bulk of the 13CO line emission. Tosaki et al. (2002) suggested that there is a 107 yr time delay between the accumulation of gas by the M 51 spiral density wave and the formation of self-gravitating clouds, resulting in the 13CO emission being found downstream in the spiral arm where it is also spatially correlated with Hα emission. Meier & Turner (2004) find for NGC 6946 that the value of ℛ reaches high values of 40 away from the central starburst which they attribute to diffuse, low-density molecular gas in and behind the molecular arms.

Within the barred galaxy NGC 7479 spatial shifts between 13CO and 12CO have also been found (Hüttemeister et al. 2000). These shifts have been attributed to density-wave or bar-like dynamical effects. The similarity of these spatial shifts with those found in the Medusa may suggest similar dynamical effects may be important. In particular, the transition region between diffuse inter arm gas and self-gravitating gas in the arms of M 51 suggesting that the similar shift in the Medusa is caused by the funnelling and compression of gas towards the starburst region.

4.2. Temperature gradient in the gas?

In this scenario, the elevated value of ℛ in the eastern part of the central region is caused not by diffuse molecular gas but by the gas there being warmer and denser than in the north-west. For a temperature gradient to result in elevated values of ℛ it is required that the average gas densities exceed 3 × 103 cm-3 (since the 1–0 transition of 12CO and 13CO must be thermalised). A change of ℛ from 16 to >40 can be caused by a temperature increase from 40 to 150 K in the gas (for a constant column- and gas density per cloud). If both the temperature and gas density is increasing towards the east then a smaller change in temperature could explain the change in ℛ. If it is filled with warm and dense gas, the dust lane would be more likely to harbour intense star formation than the rest of the central region. However, as we mentioned above, the dust lane shows no sign of elevated star formation rates, although it is potentially hidden behind large masses of dust. High resolution imaging of higher J-transition (e.g. 2–1) emission from 13CO and 12CO will allow these diffuse gas and temperature effects to be distinguished. If the faint 13CO 1–0 emission in the eastern central part is caused by diffuse gas the 13CO 2–1 emission should be even fainter there while the opposite is true if it is caused by a high gas temperature. This is because the 2–1 level would be more populated than the 1–0 for 13CO if the gas is warm – while the opposite is true if the gas is diffuse. One example where faint 13CO 1–0 is accompanied by bright 13CO 2–1 is the merger Arp 299 where it was suggested that the elevated value of ℛ is due to temperature effects (Aalto et al. 1999). Other studies of multi-transition 13CO can be found in Aalto et al. (e.g. 1995); ); )Glenn & Hunter (e.g. 2001); )Israel (e.g. 2009). Additionally, high density tracer species such as HCN or HCO+ will reveal where gas of densities n > 104 cm-3 is located in relation to the starburst and dust lane. If the dust lane is filled with diffuse gas very little dense gas will be found there. A high resolution dust SED will reveal where hot and cold dust is in relation to the dust lane and central- and off-nuclear star formation.

An abundance gradient: if the physical conditions in the gas are kept constant, an elevated ℛ may be caused by a change in the 12C/13C abundance ratio. This possibility has been discussed by Casoli et al. (e.g. 1992) as an effect of low metallicity gas being transported from the outskirts of the merger to its centre. The starburst would then contribute to enriching the gas in 13C. In this scenario the gas in the dust lane has then not yet been enriched by the starburst and there is therefore an east-west and north-south age-gradient. However, abundances in NGC 4194 are found to be close to solar, and no significant metallicity gradient can be found from a study of the properties of the star forming regions (Hancock et al. 2006).

4.3. The large scale gradient in ℛ

In the Medusa, the lowest ratios of 7–10 are found in the tail-like region north of the centre. It is not unusual to find that ℛ is decreasing away from the central region. This has been seen towards many starburst galaxies including NGC 3256, NGC 1808 and large spiral galaxies such as M 51. The gas 10'' north of the centre may therefore consist of rather ordinary cool Galactic-type GMCs. This is interesting since the gas there does not appear to be in a normal disc distribution, but rather in a shell or a tidal tail. There is no apparent star formation going on in this region and the line widths are narrow. If the gas in the central region is accumulating from the outer regions of the Medusa via the central dust lane, then our results suggest that the molecular cloud properties are normal in the outskirts, before becoming diffuse due to the funnelling of gas towards the centre and then accumulates as self-gravitating, star-forming clouds in the centre of the Medusa.

4.4. The nature of the Medusa starburst

4.4.1. Comparing with local minor mergers

One of the few E+S minor mergers to have its molecular medium studied is the Medusa “look-alike” NGC 4441. Based on single dish observations of 12CO and 13CO 2–1 and 1–0 Jütte et al. (2010) suggest that the molecular gas is in an unbound, diffuse state. Interestingly NGC 4441 appears to be an evolved version of the Medusa with similar morphology and luminosity and high resolution 12CO 1–0 observations of NGC 4441 show that it traces a centrally crossing dust lane. The star formation rate of NGC 4441 is much lower than that of the Medusa – likely due to the diffuse state of its molecular medium. In this paper we suggest that a fraction of the Medusa molecular ISM may be the diffuse – it is possible that the entire molecular gas content of NGC 4441 is in this state. This raises interesting questions on the nature of the Medusa starburst and what it will evolve into. Perhaps it will be left with a remnant, feeble dust lane, a small reservoir of diffuse gas and only low level star formation, like NGC 4441.

4.4.2. Is the star formation out of equilibrium?

Despite being a minor merger of comparatively moderate luminosity, the Medusa has an LFIR/LCO ratio (163) similar to those of typical ULIRGs such as Arp 220 (210) – suggesting that its SFE rivals that of the compact ULIRGs. However, the fraction of dense (n > 104 cm-3) gas is significantly lower in the Medusa despite its similar SFE to ULIRGs (Aalto & Hüttemeister 2000). Aalto & Hüttemeister (2000) estimate the gas consumption time for the Medusa starburst to 40 Myr and, if a significant fraction of the molecular medium is diffuse, this time could be even shorter.

The picture becomes even more interesting when one considers that most of the ongoing star formation in the Medusa seems not to be traced by FIR or radio emission. The Hα SFR is found to be  ≈ 46 M yr-1 (Hancock et al. 2006), while the FIR estimated SFR is 6–7 M yr-1 (Aalto & Hüttemeister 2000). (Note, that the SFR estimated from Hα would suggest an even more extreme SFE than the already high SFE derived from the FIR emission.) It is possible that variations in the initial mass function (IMF) can account for some of these differences (e.g. Wilkins et al. 2008), although the issue is far from resolved. Such an anomaly in the form of a flatter upper IMF could lead to enhanced Hα emission. The molecular morphology is also quite different from that of 1.4 GHz radio continuum – which is also quite different from that found from other luminous galaxies where the two distributions are usually found to coincide. The optical star formation is concentrated in knots that could be precursors to globular clusters. These knots have no clear spatial relation to either the molecular gas distribution or the continuum. This indicates a local deviation from the Kennicutt-Schmidt relation (KS) in the Medusa that warrants further study. We suggest that the starburst of the Medusa is in an extreme transient phase of very high efficiency.

4.4.3. Comparing the Medusa to high-redshifts extreme starbursts

The Medusa merger has a number of features in common with some extreme starbursts at high-z (despite its significantly lower luminosity). The starburst regions in high-z systems often seem extended (>3 kpc), gas consumption times are similarly short and there is a combination of ordered rotation and merger-driven random motions and inflows. More importantly, Bothwell et al. (2010) find for three z = 3 ULIRGs a significant size difference between the CO distribution and star formation tracers. They find that this size difference results in the SFEs within systems to vary by up to a factor of five. As a consequence to their results they conclude that SMGs lie significantly above the KS relation, indicating that stars may be formed more efficiently in these extreme systems than in other high-z starburst galaxies. The Medusa merger is (by far) less luminous than the high-z systems studied by Bothwell et al. (2010) but a careful study of the extreme starburst of the Medusa merger may lead to insights that can be applied to more distant galaxies.

5. Conclusions

We have obtained a high resolution map of the Medusa merger, NGC 4194, in the 13CO 1–0 line with the OVRO array. The main conclusions we draw from this map are:

  • 1.

    The 13CO 1–0 emission is mainly distributed in the north-western part of the inner 8'' region. There is also an off-centre 13CO peak in the tail-like distribution to the north, where the CO/13CO line ratios are similar to those of galactic type GMCs

  • 2.

    The 12CO/13CO 1–0 line ratio, ℛ, varies on a scale of 954 pc from 16 (typical values for starburst galaxies) in the north-west to >40 in the eastern part of the central region.

  • 3.

    We suggest that this shift between 12CO and 13CO emission is caused by the presence of diffuse gas in a prominent dust lane crossing the eastern part of the centre. A strong density wave-like phenomenon may cause shocks (hence the dust lane) and the 13CO is tracing more self-gravitating gas in the post-shock region in the starburst. In this scenario, the actual molecular mass distribution is better traced by the 13CO 1–0 emission than the 12CO. Alternatively, the line ratio shift is caused by temperature and density gradients in the gas. These two scenarios can be tested through imaging of 13CO 2–1 and high density tracer molecules such as HCN and HCO+.

  • 4.

    The extreme star formation efficiency and the lack of correlation between various star formation tracers suggest that the inner region of the Medusa merger is in a highly transient phase. We predict that it will evolve into a quiescent object similar to NGC 4441, but that in its current evolutionary stage it may have some features in common with high redshift starburst galaxies.

Acknowledgments

This research has made use of the NASA/IPAC Extragalactic Database (NED) which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration.

References

All Tables

Table 1

Continuum and 13CO line results.

All Figures

thumbnail Fig. 1

The 110 GHz continuum of NGC 4194 overlaid on an HST WFPC2 image. The contours are: 0.9 × (−3,3,4,5) mJy beam-1. Rms in the map is 1 mJy beam-1 thus the first contour is at 2.7σ. The peak flux density is 4.6 mJy beam-1. The cross marks the position of the 1.4 GHz continuum (Condon et al. 1990). Hatched ellipse indicate beam size and orientation.

Open with DEXTER
In the text
thumbnail Fig. 2

Left: overlay of the 12CO contours (white, apart from clouds in the tail marked in black) on a greyscale, overexposed optical R-band image (Mazzarella & Boroson 1993). The dashed curve marks the edge of the OVRO primary beam (from Aalto & Hüttemeister 2000). Right: integrated 13CO 1–0 line emission in greyscale overlaid on a 12CO contour map smoothed to the 13CO resolution of 4''̣5. The greyscale range is 0.9–2.0 Jy kms-1 beam-1. The CO contours are 5.3  ×  (1, 2, 3, 4, 5, 6, 7, 8) Jy kms-1beam-1 (starting at 6σ). Cross indicate position of 1.4 GHz radio continuum peak (as in Fig. 1).

Open with DEXTER
In the text
thumbnail Fig. 3

Left: 12CO contours on a WFPC2 image zoomed in on the central region. Centre: 13CO contours on WFPC2 image. Both 12CO and 13CO contours are 40% of peak flux  ×  (1, 1.5, 2, 2.25). For 13CO the first contour starts at 2.5σ, for 12CO the first contour starts at 24σ. Right: line ratio map where the greyscale ranges from light (ℛ = 10) to dark (ℛ = 60). Both limits and real values in greys (same scale), and the area where there are actual measures overlaid with contours (yellow). Contours are ratios ℛ = 15, 25, 35, 45. Limits are at the 3σ13CO detection threshold. Please note the small pixel-size (0''̣5) of the 12CO and 13CO maps which may give the impression that there is structure beyond the actual angular resolution of the 12CO and 13CO maps.

Open with DEXTER
In the text
thumbnail Fig. 4

Channel maps of 12CO (left) and 13CO (right). Boxes indicate channels of “missing” 13CO emission compared to 12CO. The 12CO map has been smoothed to the same resolution as 13CO. Intensity levels for 12CO and 13CO are same fraction of peak: 35% (–1.5, –1, 1.5, 2, 2.5) – where the peak is 360 and 25 mJy beam-1 for the 12CO and 13CO respectively. Note that for 12CO a big fraction of the detected, extended flux is not shown due to the cut-off. The comparative channels maps show the position of the peak emission for 12CO and 13CO. The velocity in panel 18 = 2714 kms-1, panel 14 is 2538 kms-1, and panel 10 = 2362 kms-1. Increment between panels is 44 kms-1. Crosses indicate the 1.4 GHz radio continuum position. The beam is marked in the upper right corner of the first panel for the 12CO and 13CO channel maps. Greyscale for the 12CO map ranges from 50 to 400 mJy beam-1 and for 12CO from 7.5 to 26 mJy beam-1. For the 13CO map the first contour corresponds to 2.5σ.

Open with DEXTER
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.