A Peculiar Jet and Arc of Molecular Gas Toward the Rich and Young Stellar Cluster Westerlund 2 and a TeV Gamma Ray Source
We have discovered remarkable jet- and arc-like molecular features toward the rich and young stellar cluster Westerlund2. The jet has a length of pc and a width of pc, while the arc shows a crescent shape with a radius of pc. These molecular features each have masses of and show spatial correlations with the surrounding lower density HI gas. The jet also shows an intriguing positional alignment with the core of the TeV gamma ray source HESS J1023-575 and with the MeV/GeV gamma-ray source recently reported by the collaboration. We argue that the jet and arc are caused by an energetic event in Westerlund 2, presumably due to an anisotropic supernova explosion of one of the most massive member stars. While the origin of the TeV and GeV gamma-ray sources is uncertain, one may speculate that they are related to the same event via relativistic particle acceleration by strong shock waves produced at the explosion or by remnant objects such as a pulsar wind nebula or microquasar.
Author(s) in page-headRunning Head \Received2000/12/31\Accepted2001/01/01
ISM: molecules, stars: individual (Westerlund 2), acceleration of particles, stars: supernovae: general
Massive stars have a strong influence on the surrounding interstellar space via their stellar winds and ultraviolet radiation. Moreover, they end their lives in catastrophic supernova (SN) explosions, thus providing more energetic impact on the interstellar medium. Wd2 is a rich star cluster of age Myrs (Piatti et al., 1998). CO observations have allowed a distance estimate to the source of kpc (Furukawa et al. (2009); see also Dame (2007)). The total stellar mass in Wd2 is of the order of , including 12 O stars and 2 WR stars (Rauw et al., 2007). It is also associated with a HII region RCW49, a remarkable infrared nebula as revealed by (Churchwell et al., 2004), and perhaps with the extended TeV gamma-ray source HESS J1023-575, discovered with the HESS telescope array in the direction of the cluster (Aharonian et al., 2007). Recently, low energy (MeV/GeV) gamma-rays have been reported from the same direction by the collaboration (Abdo et al., 2009).
In this Letter we describe the discovery of a spectacular jet and arc of molecular gas detected with the NANTEN telescope in the CO() mm-wave emission line survey (Mizuno & Fukui, 2004), and discuss their possible links to the MeV/GeV and TeV gamma-ray sources.
We show the two new CO features toward Wd2 in Figure 1, where the CO distribution in the velocity range km s is presented in Galactic coordinates. One is a straight feature in the east at (hereafter, the jet) and the other an arc-like feature in the west at (hereafter, the arc). We note that the arc has an intensity peak and a protrusion to the west of HESS J1023-575 at , both of which lie along the jet axis, suggesting the existence of the diametrically opposite part of the jet.
These two features show a remarkable correlation with Wd2 / HESS J1023-575. The jet is extended over to the east of Wd2 / HESS J1023-575, tilted at an angle of to the galactic plane, and is very well aligned with Wd2 / HESS J1023-575 as well as a ridge of the radio continuum corresponding to the HII region, RCW 49(Whiteoak & Uchida, 1997). A least-squares fitting to the jet in Figure 1 yields a regression,
where and are in degrees.
This line passes through the error box of the peak position of HESS J1023-575, which is shifted from the center of the cluster by arcmin. The arc is apparently symmetric with respect to the jet axis and is well centered on HESS J1023-575. It appears somewhat rim-brightened along the outer boundary and shows a marked crescent shape, suggesting part of a swept-up shell. For convenience we define an axis , which runs along the direction of the jet and passes through the peak position of HESS J1023-0575, and an axis which is perpendicular to . Here the origin of both and is .
The velocity distributions of the two features are shown in two overlaid velocity channel maps (Figure 2a) and in a position-velocity map along (Figure 2b).
The jet is generally quiescent, having a velocity width of a few km s between but show fairly large widths of 10 km s at and , if all the features are assumed to be physically associated. We tentatively divide the molecular gas towards the jet into the following four components;
J1: A narrow component at located at km s with a linewidth of 1 2 km s, with little velocity gradient. This component has three local peaks at , and with a winding shape towards .
J2: A broad component at located over a velocity range of 19 30 km s.
J3: A fairly broad and localized feature at and km s, towards one of the peaks of J1.
J4: The western component partially overlapping with the arc at and km s.
Although it is not conclusive, we suggest that the four components of the jet are physically related as discussed later. The CO linewidth of the arc is km s centered at 26 km s, and we infer that the possible expansion velocity of the arc is not large.
We shall here assume that the jet and arc are at kpc with a 30% error limit, and estimate the mass of each as follows, where an X factor of cm (K km s) is used to convert the CO intensity to molecular mass (Bertsch et al., 1993): , , , , and , respectively. The accuracy in mass is limited to because of the uncertainty in distance.
We show the HI distribution at km s with a width of 0.82 km s in Figure 3 (McClure-Griffiths et al., 2005). The HI shows a clear sign of a hole towards Wd2 and HESS J1023-575. Hereafter we shall call the HI feature surrounding the hole the HI shell. This shell is seen over a velocity range of 18 32 km s. We note that the velocity width of the hole is rather small, a few km s, suggesting that the possible expansion velocity is small at present. The HI hole has an elliptical shape with dimensions of pc by pc, elongated along the direction of S. It is noteworthy that the HI shell exhibits an intensity depression in its northern part, coincident with the molecular arc, suggesting that the arc and shell are physically related. This HI intensity depression is likely due to the conversion of HI into H. The jet is coincident with an elongated spur of HI extended to the east over , from to , while on the west there is little HI beyond the arc. The HI mass of the cloud shown in Figure 3 amounts to in the velocity range 18 32 km s for a conversion factor of cm (K km s (Dickey et al., 1990).
The CO arc and HI shell suggest a spherical shock either due to a SN or stellar winds of WR stars. The secure lower limit for the kinetic energy involved in the arc is estimated to be ergs for a molecular mass of and an expansion velocity of 1 km s. A more likely estimate is perhaps ergs or more if we add the expansion energy of the HI gas to the expansion energy of the gas traced in CO. A SN with mechanical energy ergs is more than sufficient to account for the energetics of the observed molecular and atomic gas, even when taking into account the conversion efficiency into kinetic energy of (Thornton et al., 1998). There are many O stars in Wd2, whose collective mechanical wind energy may be large. For example 10 O-stars each with stellar-wind kinetic energy luminosities of erg s yield erg s. Over a timescale of years the total kinetic power could reach ergs. With the same conversion efficiency, 0.1, a kinetic energy of ergs is available.
The one-sided molecular jet can be explained in terms of in-situ conversion of HI into molecular gas by shock compression as a result of the jet propagating through the atomic medium over a projected length of pc. This is consistent with the huge mass of the molecular jet several times and the considerable amount of HI gas ( ) seen towards it. A lower limit for the time scale of the molecular jet is estimated to be yrs from the timescale of shocked CO formation (Koyama & Inutsuka (2000); Bergin et al. (2004)). There are presently two other known candidates for molecular jets driven by high-energy jets SS433 and MJG348.5 (Yamamoto et al., 2008). The physical parameters of the two molecular jets are estimated as follows; full length pc, molecular mass , and minimum kinetic energy ergs. For the Wd2 clouds, we roughly estimate such kinetic energy to be ergs from the mass, , and the velocity span, km s, of the dominant component J2, which is somewhat larger than the above two cases. We note that an increased velocity dispersion at the far tip of the jet, as seen in J2, is also observed in SS433 and MJG348.5. This is interpreted as a result of the stronger dynamical interaction towards the tip of the jet, where the deceleration of the high-energy jet becomes most significant (Yamamoto et al., 2008).
We now consider the origin of the jet and arc. Presently, there are two scenarios in which highly collimated flows of sufficient energy and length may be formed: (i) a highly anisotropic supernova explosion; (ii) a high-energy accretion-powered jet from a compact object such as in a microquasar. Anisotropic energy output is thought to have occurred in several known SNe including Cas A and W49B (Hwang et al. (2004); Miceli et al. (2008)). The expected speed of the jets is km s, suggesting a travel time of yrs over pc if the jet axis is nearly perpendicular to the line of sight. Theoretical calculations indicate that a strongly magnetized and rotating neutron star formed in a SNe results in highly collimated jet (Burrows et al., 2007). Such explosions are able to release ergs in energy (Komissarov & Barkov (2007)), and the compact remnant is expected to be observable as a magnetar. The frequency of known magnetars is low, with only 12 magnetars (Mereghetti, 2008) amongst the 1500 or so known pulsars (Manchester, 2004), while jet-like features in historical SNRs such as Cas A suggest that the anisotropic SNe may not be very rare. To summarize, the anisotropic SNe scenario may well explain the jet and arc as caused by a single collimated SNe.
An alternative is that a conventional, isotropic SNe occurred in a binary system, leading to the formation of an accretion-powered jet. For example, the microquasar SS433, exhibits an X-ray jet of 150 pc and molecular jet of 400 pc in length (Kotani (1998); Yamamoto et al. (2008)). In Wd2 the arc and shell might be formed by the SN explosion, and the molecular jet by a long-lived microquasar jet. The SS433 jet has a momentum flow rate of yr and can supply ergs in kinetic energy in yrs (Kotani (1998); Marshall et al. (2002)), sufficiently large to supply the kinetic energy above.
Concerning the origin of the gamma-ray emission, the intriguing alignment of the gravity centers of both the and HESS source positions with the molecular jet and arc suggests a common (initial) event, e.g., a supernova explosion. The link could be direct, via radiation of relic particles produced at the SN explosion, or via on-going acceleration of particles by a relativistic object such as a remnant of the explosion, e.g. supernova remnant shell, a pulsar wind nebula (PWN) or a microquasar jet. The first scenario favors a hadronic origin for the gamma-ray emission. Because of severe radiative (synchrotron and inverse Compton) losses, the relic ultrarelativistic electrons hardly could survive to the present epoch. Indeed, for any reasonable parameters characterizing the ambient medium, the lifetime of multi-TeV electrons is significantly shorter than the age of the system. On the other hand, inverse Compton radiation of gamma-rays by continuously accelerated electrons, e.g., by a PWN, is a viable option. The gamma-ray luminosity is about erg s. If the main target for IC gamma-rays is the 2.7 K CMBR, this mechanism would require acceleration of TeV electrons at a rate of erg s, or an order of magnitude less, if the optical radiation of stars plays a dominant role. The latter case can be realized when acceleration and radiation of electrons take place not far from the stars. The extended character of the gamma-ray emission excludes a single star origin for the observed TeV emission. In this case the extension of the gamma-ray source is effectively determined by the volume occupied by stars, and consequently the size of the source is expected to be energy-independent. In contrast to this scenario, for a single PWN we expect an energy-dependent morphology caused by radiative energy losses which do not allow the highest energy electrons to propagate to large distances. Such an effect is observed in the PWN HESS J1825-137 (Aharonian et al., 2006). In the case of a hadronic scenario we expect just the opposite dependence. Since generally the higher energy protons propagate faster than low energy protons, we might expect an increase of the angular size of the source with energy (Aharonian, 2004). Another test to distinguish between the hadronic and electronic models is the ratio of GeV and TeV gamma-ray fluxes. In the case of the IC mechanism, we expect suppression of low energy gamma-rays and a rather low GeV/TeV ratio. It is expected that future higher quality GeV and TeV data should allow us to conduct detailed a quantitative study of the morphological and spectral characteristics of the radiation in different energy bands, which could lead to definite conclusions. Here we would prefer to limit the discussion by noting that the total energy required to explain the TeV gamma-ray data reported by the HESS collaboration for a source located at a distance of 5 kpc is about cm erg. The CO observations indicate a rather low density of gas in the region that coincides with the location of the TeV gamma-ray source (see above), most likely, cm. Then the total budget in relativistic protons is required as large as erg. This can be considered as an argument in favor of a very strong SN explosion with total energy exceeding erg.
In summary, our observations conducted with the NANTEN telescope in the CO() mm-wave emission led to the discovery of a spectacular jet and arc of molecular gas toward the young star cluster Westerlund 2 which may be the result of a powerful supernova explosion. Another consequence of this explosion could be the GeV and TeV gamma-ray sources located in the same region as reported recently by the and HESS collaborations. Further detailed studies of spectral and morphological features of gamma-ray emission are requited to explore the links between these two phenomena.
The original NANTEN telescope was operated based on a mutual agreement between Nagoya University and the Carnegie Institution of Washington. This work is financially supported in part by a Grant-in-Aid for Scientific Research (KAKENHI) from the MEXT and from JSPS, in part, through the core-to-core program and by the donation from individuals and private companies.
- Abdo et al. (2009) Abdo, A. A. et al. 2009, submitted.
- Aharonian (2004) Aharonian, F. et al. 2004, Very high energy cosmic gamma radiation : A crucial window on the extreme Universe (River Edge, NJ: World Scientice Publishing), Chapter 4
- Aharonian et al. (2006) Aharonian, F. et al. 2006, \aap, 460, 365
- Aharonian et al. (2007) Aharonian, F. et al. 2007, \aap, 467, 1075
- Bergin et al. (2004) Bergin, E. A., Hartmann, L. W., Raymond, J. C., & Ballesteros-Paredes, J., 2004, \apj, 612, 921
- Bertsch et al. (1993) Bertsch, D. L., Dame, T. M., Fichtel, C. E., Hunter, S. D., Sreekumar, P., Stacy, J. G., & Thaddeus, P. 1993, \apj, 416, 587
- Burrows et al. (2007) Burrows, A., Dessart, L., Livne, E., Ott, C. D., & Murphy, J. 2007, \apj, 664, 416
- Churchwell et al. (2004) Churchwell, E. et al. 2004, \apjs, 467, 1075
- Dame (2007) Dame, T. M. 2007, \apj, 665, 163
- Dickey et al. (1990) Dickey, J. M., Lockman, F. J. 1990, \aapr, 28, 215
- Furukawa et al. (2009) Furukawa, N. et al. 2009, submitted.
- Hwang et al. (2004) Hwang et al. 2004, \apj, 615, 117
- Komissarov & Barkov (2007) Komissarov, S. S., & Barkov, M. V. 2007, \mnras, 382, 1029
- Kotani (1998) Kotani, T. 1998, Doctoral Thesis of University of Tokyo
- Koyama & Inutsuka (2000) Koyama, H., & Inutsuka, S. 2000, \apj, 532, 980
- Manchester (2004) Manchester, R. N. 2004, science, 304, 542
- Marshall et al. (2002) Marshall, H. L., Canizares, C. R., & Schulz, N. S. 2002, \apj, 564, 941
- McClure-Griffiths et al. (2005) McClure-Griffiths, N. M., Dickey, J. M., Gaensler, B. M., Green, A. J., Haverkorn, M., & Strasser, S. 2005, \apjs, 158, 178
- Mereghetti (2008) Mereghetti, S. 2008, \aapr15, 225
- Miceli et al. (2008) Miceli, M., Decourchelle, A., Ballet, J., Bocchino, F., Hughes, J., Hwang, U., & Petre, R. 2008, Adv. Space Research, 41, 390
- Mizuno & Fukui (2004) Mizuno, A., & Fukui, Y. 2004, in ASP Conf. Ser., 317, Milky Way Surveys: The Structure and Evolution of our Galaxy, ed. D. Clemens, R. Y. Shah, & T. Brainerd, 59
- Piatti et al. (1998) Piatti, A. E., Bica, E., & Claria, J. J. 1998, \aaps, 127, 423
- Rauw et al. (2007) Rauw, G., Manfroid, J., Gosset, E., Naze, Y., Sana, H., De Becker, M., Foellmi, C., & Moffat, A. F. J. 2007, \aap, 463, 981
- Thornton et al. (1998) Thornton, K., Gaudilitz, M., Janka, H.-Th., & Steinmetz, M. 1998, \apj, 500, 95
- Whiteoak & Uchida (1997) Whiteoak, J. B. Z., & Uchida, K. I. 1997 \aap, 317, 563
- Yamamoto et al. (2008) Yamamoto, H. et al. 2008 PASJ, 60, 715