A&A 481, 401-410 (2008)
F. Aharonian1,13 - A. G. Akhperjanian2 - A. R. Bazer-Bachi3 - B. Behera14 - M. Beilicke4 - W. Benbow1 - D. Berge1, - K. Bernlöhr1,5 - C. Boisson6 - O. Bolz1 - V. Borrel3 - I. Braun1 - E. Brion7 - A. M. Brown8 - R. Bühler1 - T. Bulik24 - I. Büsching9 - T. Boutelier17 - S. Carrigan1 - P. M. Chadwick8 - L.-M. Chounet10 - A. C. Clapson1 - G. Coignet11 - R. Cornils4 - L. Costamante1,25 - B. Degrange10 - H. J. Dickinson8 - A. Djannati-Ataï12 - W. Domainko1 - L. O'C. Drury13 - G. Dubus10 - J. Dyks24 - K. Egberts1 - D. Emmanoulopoulos14 - P. Espigat12 - C. Farnier15 - F. Feinstein15 - A. Fiasson15 - A. Förster1 - G. Fontaine10 - Y. Fukui26 - Seb. Funk5 - S. Funk1 - M. Füßling5 - Y. A. Gallant15 - B. Giebels10 - J. F. Glicenstein7 - B. Glück16 - P. Goret7 - C. Hadjichristidis8 - D. Hauser1 - M. Hauser14 - G. Heinzelmann4 - G. Henri17 - G. Hermann1 - J. A. Hinton1,14, - A. Hoffmann18 - W. Hofmann1 - M. Holleran9 - S. Hoppe1 - D. Horns18 - A. Jacholkowska15 - O. C. de Jager9 - E. Kendziorra18 - M. Kerschhaggl5 - B. Khélifi10,1 - Nu. Komin15 - K. Kosack1 - G. Lamanna11 - I. J. Latham8 - R. Le Gallou8 - A. Lemière12 - M. Lemoine-Goumard10 - J.-P. Lenain6 - T. Lohse5 - J. M. Martin6 - O. Martineau-Huynh19 - A. Marcowith15 - C. Masterson13 - G. Maurin12 - T. J. L. McComb8 - R. Moderski24 - Y. Moriguchi26 - E. Moulin15,7 - M. de Naurois19 - D. Nedbal20 - S. J. Nolan8 - J.-P. Olive3 - K. J. Orford8 - J. L. Osborne8 - M. Ostrowski23 - M. Panter1 - G. Pedaletti14 - G. Pelletier17 - P.-O. Petrucci17 - S. Pita12 - G. Pühlhofer14 - M. Punch12 - S. Ranchon11 - B. C. Raubenheimer9 - M. Raue4 - S. M. Rayner8 - O. Reimer - M. Renaud1 - J. Ripken4 - L. Rob20 - L. Rolland7 - S. Rosier-Lees11 - G. Rowell1, - B. Rudak24 - J. Ruppel21 - V. Sahakian2 - A. Santangelo18 - L. Saugé17 - S. Schlenker5 - R. Schlickeiser21 - R. Schröder21 - U. Schwanke5 - S. Schwarzburg18 - S. Schwemmer14 - A. Shalchi21 - H. Sol6 - D. Spangler8 - . Stawarz23 - R. Steenkamp22 - C. Stegmann16 - G. Superina10 - T. Takeuchi26 - P. H. Tam14 - J.-P. Tavernet19 - R. Terrier12 - C. van Eldik1 - G. Vasileiadis15 - C. Venter9 - J. P. Vialle11 - P. Vincent19 - M. Vivier7 - H. J. Völk1 - F. Volpe10 - S. J. Wagner14 - M. Ward8
1 - Max-Planck-Institut für Kernphysik, PO Box 103980, 69029 Heidelberg, Germany
2 - Yerevan Physics Institute, 2 Alikhanian Brothers St., 375036 Yerevan, Armenia
3 - Centre d'Etude Spatiale des Rayonnements, CNRS/UPS, 9 Av. du Colonel Roche, BP 4346, 31029 Toulouse Cedex 4, France
4 - Universität Hamburg, Institut für Experimentalphysik, Luruper Chaussee 149, 22761 Hamburg, Germany
5 - Institut für Physik, Humboldt-Universität zu Berlin, Newtonstr. 15, 12489 Berlin, Germany
6 - LUTH, UMR 8102 du CNRS, Observatoire de Paris, Section de Meudon, 92195 Meudon Cedex, France
7 - DAPNIA/DSM/CEA, CE Saclay, 91191 Gif-sur-Yvette, Cedex, France
8 - University of Durham, Department of Physics, South Road, Durham DH1 3LE, UK
9 - Unit for Space Physics, North-West University, Potchefstroom 2520, South Africa
10 - Laboratoire Leprince-Ringuet, IN2P3/CNRS, École Polytechnique, 91128 Palaiseau, France
11 - Laboratoire d'Annecy-le-Vieux de Physique des Particules, IN2P3/CNRS, 9 Chemin de Bellevue, BP 110, 74941 Annecy-le-Vieux Cedex, France
12 - APC, 11 place Marcelin Berthelot, 75231 Paris Cedex 05, France,
13 - Dublin Institute for Advanced Studies, 5 Merrion Square, Dublin 2, Ireland
14 - Landessternwarte, Universität Heidelberg, Königstuhl, 69117 Heidelberg, Germany
15 - Laboratoire de Physique Théorique et Astroparticules, IN2P3/CNRS, Université Montpellier II, CC 70, Place Eugène Bataillon, 34095 Montpellier Cedex 5, France
16 - Universität Erlangen-Nürnberg, Physikalisches Institut, Erwin-Rommel-Str. 1, 91058 Erlangen, Germany
17 - Laboratoire d'Astrophysique de Grenoble, INSU/CNRS, Université Joseph Fourier, BP 53, 38041 Grenoble Cedex 9, France
18 - Institut für Astronomie und Astrophysik, Universität Tübingen, Sand 1, 72076 Tübingen, Germany
19 - LPNHE, Université Pierre et Marie Curie Paris 6, Université Denis Diderot Paris 7, CNRS/IN2P3, 4 place Jussieu, 75252 Paris Cedex 5, France
20 - Institute of Particle and Nuclear Physics, Charles University, V Holesovickach 2, 180 00 Prague 8, Czech Republic
21 - Institut für Theoretische Physik, Lehrstuhl IV: Weltraum und Astrophysik, Ruhr-Universität Bochum, 44780 Bochum, Germany
22 - University of Namibia, Private Bag 13301, Windhoek, Namibia
23 - Obserwatorium Astronomiczne, Uniwersytet Jagiellonski, Kraków, Poland
24 - Nicolaus Copernicus Astronomical Center, Warsaw, Poland
25 - European Associated Laboratory for Gamma-Ray Astronomy, jointly supported by CNRS and MPG
26 - Department of Astrophysics, Nagoya University, Chikusa-ku, Nagoya 464-8602, Japan
Received 1 May 2007 / Accepted 16 January 2008
Aims. Observations of shell-type supernova remnants (SNRs) in the GeV to multi-TeV -ray band, coupled with those at millimetre radio wavelengths, are motivated by the search for cosmic-ray accelerators in our Galaxy. The old-age mixed-morphology SNR W 28 (distance 2 kpc) is a prime target due to its interaction with molecular clouds along its northeastern boundary and other clouds situated nearby.
Methods. We observed the W 28 field (for 40 h) at very high energy (VHE) -ray energies (E>0.1 TeV) with the HESS. Cherenkov telescopes. A reanalysis of EGRET E>100 MeV data was also undertaken. Results from the NANTEN 4 m telescope Galactic plane survey and other CO observations were used to study molecular clouds.
Results. We have discovered VHE -ray emission (HESS J1801-233) coincident with the northeastern boundary of W 28 and a complex of sources (HESS J1800-240A, B and C) 0.5 south of W 28 in the Galactic disc. The EGRET source (GRO J1801-2320) is centred on HESS J1801-233 but may also be related to HESS J1800-240 given the large EGRET point spread function. The VHE differential photon spectra are well fit by pure power laws with indices to 2.7. The spectral indices of HESS J1800-240A, B, and C are consistent within statistical errors. All VHE sources are 10 in intrinsic radius except for HESS J1800-240C, which appears pointlike. The NANTEN 12CO(J=1-0) data reveal molecular clouds positionally associating with the VHE emission, spanning a 15 km s-1 range in local standard of rest velocity.
Conclusions. The VHE/molecular cloud association could indicate a hadronic origin for HESS J1801-233 and HESS J1800-240, and several cloud components in projection may contribute to the VHE emission. The clouds have components covering a broad velocity range encompassing the distance estimates for W 28 (2 kpc) and extending up to 4 kpc. Assuming hadronic origin and distances of 2 and 4 kpc for cloud components, the required cosmic-ray density enhancement factors (with respect to the solar value) are in the range 10 to 30. If situated at 2 kpc distance, such cosmic-ray densities may be supplied by SNRs like W 28. Additionally and/or alternatively, particle acceleration may come from several catalogued SNRs and SNR candidates, the energetic ultra compact HII region W 28A2, and the HII regions M 8 and M 20, along with their associated open clusters. Further sub-mm observations would be recommended to probe in detail the dynamics of the molecular clouds at velocites >10 km s-1 and their possible connection to W 28.
Key words: gamma rays: observations
The study of shell-type supernova remnants (SNRs) at -ray energies is motivated by the long-held idea that they are the dominant sites of hadronic Galactic cosmic-ray (CR) acceleration to energies approaching the knee (1015 eV) (e.g. Ginzburg & Syrovatskii 1964; Blandford & Eichler 1987). CRs (hadrons and electrons) are injected into the SNR shock front, and are then accelerated via the diffusive shock acceleration (DSA) process (for a review see Drury 1983). Subsequent -ray production from the interaction of these CRs with ambient matter and/or electromagnetic fields is a tracer of such non-thermal particle acceleration, and establishing the hadronic or electronic nature of the parent CRs in any -ray source remains a key issue. Two SNRs, RX J1713.7-3946 and RX J0852.0-4622, have so far established shell-like morphology in VHE -rays (Aharonian et al. 2004a,2005c,2006b,2007a,2007b), with spectra extending to 20 TeV and beyond. In particular for RX J1713.7-3946, particle acceleration up to at least 100 TeV is inferred from the HESS observations. Although a hadronic origin of the VHE -ray emission is highly likely in the above cases (Aharonian et al. 2006b,2007b; Berezhko & Völk 2006; Berezhko et al. 2007), an electronic origin is not ruled out.
Disentangling the electronic and hadronic components in TeV SNRs may be made easier by studying: (1) SNR -ray spectra well beyond 10 TeV, an energy regime where electrons suffer strong radiative energy losses and due to Klein-Nishina effects the resulting inverse-Compton spectra tend to show a cut-off; (2) older SNRs (age approaching 105 yr) in which accelerated electrons have lost much of their energy through radiative cooling and do not reach multi-TeV energies; (3) SNRs interacting with adjacent molecular clouds of very high densities n> 103 cm-3. It is the latter regard especially (and to a certain degree the second) which makes the SNR W 28 (G6.4-0.1) an attractive target for VHE -ray studies. In this paper we outline the discovery of VHE -ray emission from several sites in the W 28 field and briefly discuss their relationship with molecular clouds, W 28, and other potential particle accelerators in the region.
W 28 (G6.4-0.1) is a mixed-morphology SNR, with dimensions and an estimated distance between 1.8 and 3.3 kpc (e.g. Goudis 1976; Lozinskaya 1981). It is an old-age SNR (age 35 000 to 150 000 yr; e.g. Kaspi et al. 1993), thought to have entered its radiative phase of evolution (e.g. Lozinskaya 1981) in which much of its CRs have escaped into the surrounding interstellar medium (ISM). We note also that the evolutionary status (Sedov and/or radiative) of shell-type SNRs may depend on the density of their surroundings (see e.g. Blondin et al. 1998).
W 28 is distinguished by its interaction with a molecular cloud (Wootten 1981) along its north and northeastern boundaries. This interaction is traced by the high concentration of 1720 MHz OH masers (Frail et al. 1994; Claussen et al. 1997,1999), and also the location of very high-density (n>103 cm-3) shocked gas (Arikawa et al. 1999; Reach et al. 2005). The shell-like radio emission (Long et al. 1991; Dubner et al. 2000) peaks at the northern and northeastern boundaries where interaction with the molecular cloud is established. Further indication of the influence of W 28 on its surroundings is the expanding HI void at a distance 1.9 kpc (Velázquez et al. 2002). The X-ray emission, which overall is well-explained by a thermal model, peaks in the SNR centre but has local enhancements in a region overlapping the northeastern SNR/molecular cloud interaction (Rho & Borkowski 2002).
In the neighbourhood of W 28 are the radio-bright HII regions M 20 (Trifid Nebula at kpc Lynds et al. 1985 - with open cluster NGC 6514), M 8 (Lagoon Nebula at kpc Tothill et al. 2002 - containing the open clusters NGC 6523 and NGC 6530) and the ultra-compact HII region W 28A2, all of which are representative of the massive star formation taking place in the region. Further discussion concerning the active star formation in this region may be found in van den Ancker et al. (1997) and references therein. Additional SNRs in the vicinity of W 28 have also been identified: G6.67-0.42 and G7.06-0.12 (Yusef-Zadeh et al. 2000), G5.55+0.32, G6.10+0.53 and G7.20+0.20 (Brogan et al. 2006). The pulsar PSR J1801-23 spin-down luminosity erg s-1 and distance d = 13.5 kpc (based on its dispersion measure) is at the northern radio edge (Kaspi 1993). More recent discussion (Claussen et al. 2002) assigns a lower limit of 9..4 kpc for the pulsar distance.
W 28 has also been linked to -ray emission detected at E>300 MeV by COS-B (Pollock 1985) and E>100 MeV by EGRET (Sturner & Dermer 1995; Esposito et al. 1996; Zhang et al. 1998). The EGRET source, listed in the 3rd catalogue (Hartman et al. 1999) as 3EG J1800-2338, is positioned at the southern edge of the radio shell. We have also performed an analysis of EGRET data, with additional data not included in the 3rd catalogue, and results are discussed later in this paper.
Previous observations of the W 28 region at VHE energies by the CANGAROO-I telescope revealed no evidence for such emission (Rowell et al. 2000) and upper limits at the 0.2 to 0.5 Crab-flux level for energies E>1.5 TeV (1.1 to erg cm-2 s-1) were set for various regions.
The total observation time covering the W 28 region amounts to 42 h in a series of runs (with typical duration 28 min) spread over the 2004, 2005 and 2006 seasons. Runs were accepted for analysis if they met quality control criteria based on the recorded rate of isotropic CR background events, the number of malfunctioning pixels in each camera, the calibration and the tracking performance (see Aharonian et al. 2004b for details).
Data were analysed using the moment-based Hillas analysis procedure, the same used in the analysis of the
inner Galactic Plane Scan datasets (Aharonian et al. 2005a,2006a). Observations covered a range
of zenith angles leading to energy thresholds of 320 GeV with hard cuts (Cherenkov image integrated intensity or size
and 150 GeV for standard cuts (size >80 photoelectrons). Hard cuts were used in VHE -ray images,
source location studies and energy spectra. In addition, Standard cuts were used in energy spectra in order to increase the
energy coverage of extracted spectra.
Generally consistent results were obtained using an alternative analysis based on a model of Cherenkov image parameters
(de Naurois 2006), which also utilises an independent calibration and lower cut on image size of >60 photoelectrons. A forthcoming paper will highlight results in detail from this analysis, which achieves improved sensitivities
at lower thresholds compared to the pure Hillas-based analysis.
|Figure 1: Image (1.5 ) of the VHE -ray excess counts (events), corrected for exposure and smoothed with a Gaussian of radius 4.2 (standard deviation). Overlaid are solid green contours of VHE excess (pre-trial) significance levels of 4, 5, and 6, after integrating events within an oversampling radius appropriate for pointlike sources. The thin-dashed circle depicts the approximate radio boundary of the SNR W 28 guided predominantly by the bright northern emission (see Dubner et al. 2000; Brogan et al. 2006). Identified here are VHE source regions HESS J1801-233 to the northeast, and a complex of sources HESS J1800-240 (A, B & C) to the south of W 28. Also indicated are: HII regions (black stars); W 28A2 (see text), G6.1-0.6 (Kuchar & Clark 1997), 6.225-0.569 (Lockman 1989); The 68% and 95% location contours (thick-dashed yellow lines) of the E>100 MeV EGRET source GRO J1801-2320; the pulsar PSR J1801-23 (white triangle). The inset to the bottom left depicts a pointlike source for this analysis after the Gaussian smoothing applied to the main image.|
|Open with DEXTER|
The VHE -ray image (Fig. 1) reveals two sites of VHE -ray emission in the direction of the northeastern and southern boundaries of the W 28 SNR. The colour scale in this figure depicts the Gaussian-smoothed VHE excess counts above a CR background estimate according to the template model (Rowell 2003), along with significance contours obtained after integrating events within a radius of 0.1 from each bin centre (appropriate for pointlike source searching). Similar images were obtained using alternative CR background models. A smoothing radius of was used to sufficiently smooth out random fluctuations in the image. An assessment of the VHE post-trial significances was made from our original search for marginally extended sources, which employed an a priori integration radius . Under this scheme we applied trials (a very conservative value applied to these data) accumulated in searching for sources in the inner Galactic Plane (as in Aharonian 2005a). The pre-trial significance of the VHE sources, at , is therefore converted to a post-trial significance of .
Based on the significance contours in Fig. 1, we assign labels to the northeastern source, HESS J1801-233, and to the complex of sources to the south, HESS J1800-240, according to their best fit positions (fitting a 2D Gaussian and ellipse respectively to the unsmoothed excess map). Three components of HESS J1800-240 are identified, labeled here A, B and C from East to West. These components represent local peaks above their surrounds. Although not convincingly resolved under this analysis these components may comprise separate sources (or at least in part) due to their possible relationship with distinct multiwavelength counterparts (discussed later).
Differential photon energy spectra were extracted from HESS J1801-233 and all three components of HESS J1800-240. Spectra were well-fit by pure power laws ( ) with photon indices to 2.7 in the energy range 0.3 to 5 TeV (see Table 1 for results). Spectral fits were obtained using fluxes from a combination of hard and standard cuts to maximise the energy coverage. Spectral analysis employed the reflected background model (Berge et al. 2007), in which control regions reflected through each tracking position (taking care to avoid known VHE -ray sources) were used to estimate the CR background. Within the statistical and systematic errors, the photon indices appear consistent throughout HESS J1800-240. Except for HESS J1800-240C, all of the VHE sources appear extended with intrinsic radii of 10. At a distance of 2 kpc, the VHE source luminosities in the energy range 0.3 to 3 TeV would be on the order of 1033 erg s-1.
Table 1: Numerical summary for the VHE and E>100 MeV sources in the W 28 region including positional and spectral information.
In searching for molecular cloud counterparts to the VHE sources, we analysed 12CO (J=1-0) molecular line observations taken by the 4-m mm/sub-mm NANTEN telescope, at Las Campanas Observatory, Chile (Mizuno & Fukui 2004). The NANTEN Galactic Plane Survey data of 1999 to 2003 (see Matsunaga et al. (2001) and references therein for details) were used, and for the W 28 region, the survey grid spacing was 4'.
Figure 2 (upper left panel) shows the 12CO (J=1-0) image integrated over the Local Standard of Rest
range 0 to 10 km s-1, while the right panel shows the image integrated over
=10 to 20 km s-1.
Two prominent 12CO features representing molecular clouds centred at (
spatially correspond with the VHE -ray emission.
As shown in Fig. 2, these molecular clouds span
According to the Galactic rotation model of Brand & Blitz (1993),
ranges formally correspond to kinematic distances of
approximately 0 to 2.5 kpc (overlapping the Sagittarius arm), and 2.5 to 4 kpc (reaching the
Scutum-Crux arm) respectively.
Given the uncertainties in rotation models close to the Galactic centre, such
ranges would cover the distance
estimates for W 28, the most prominent SNR in the region. Much discussion has centred on the
systemic velocity (SV) of W 28 (and hence its distance), and how much W 28 has influenced matter in the region.
(Radhakrishman et al. 1972) and HI absorption features (Lozinskaya et al. 1981) have
suggested SV 18 km s-1. Claussen et al. (1997) have pointed to SV 17 km s-1.
More recent HI studies by Velázquez et al. (2002) suggest SV = +7 km s-1
(which leads to the distance estimate for W 28 at 1.9 kpc). They also suggest a HI shell
may also extend over the
25 to +38 km s-1 range, giving rise to a shock speed of 30 km s-1.
Torres et al. (2003) and Reach et al. (2005) have also studied the large-scale 12CO(J=1-0) emission for
this region using the survey data of Dame et al. (2001), suggesting that the parent molecular cloud under the influence of W 28
is presently centred at
The Galactic longitude-velocity (l-v) diagram (bottom panels of Fig. 2) from our NANTEN data integrated over the
Galactic latitude ranges
shows the distribution of
molecular material in relation to the SV of W 28 from the HI studies of Velázquez. The wider, latter b range shows the effect of
including the cloud component overlapping HESS J1800-240A.
A void or dip in CO emission appears at a similar
range as found
in the HI data, with much of the molecular material appearing to surround the void in positive
with respect to the SV of W 28. A similar longitude-velocity picture was revealed by Torres et al. (2003)
(see their Fig. 22).
|Figure 2: Upper Left: NANTEN 12CO(J=1-0) image of the W 28 region (linear scale in K km s-1) for to 10 km s-1 with VHE -ray significance contours overlaid (green) - levels 4, 5, 6 as in Fig. 1. The radio boundary of W 28, The 68% and 95% location contours of GRO J1801-2320 and the location of the HII region W 28A2 (white stars) are indicated. Upper Right: NANTEN 12CO(J=1-0) image for to 20 km s-1 (linear scale and same maxima as for upper left panel). Bottom panels: distribution of CO emission over the Galactic longitude and plane integrated over Galactic latitude b ranges to ( left) and to ( right). The latter range is used to show the effect of extending the latitude range to encompass component A of HESS J1800-240. The bold circle indicates the suggested systemic velocity (7 km s-1) of W 28 from the HI studies of Velázquez et al. (2002).|
|Open with DEXTER|
|Figure 3: Left: VLA 90 cm radio image from Brogan et al. (2006) in Jy beam-1 (rebinned by a factor 1.2 compared to the original). The VHE significance contours (green) from Fig. 1 are overlaid along with the HII regions (white stars) and the additional SNRs and SNR candidates (with yellow circles indicating their location and approximate dimensions) discussed in the text. Right: ROSAT PSPC image - 0.5 to 2.4 keV (smoothed counts per bin from Rho & Borkowski et al. 2002). Overlaid are contours (cyan - 10 linear levels up to W m-2 sr-1) from the MSX 8.28 µm image. Other contours and objects are as for the left panel. The X-ray Ear representing a peak at the northeastern edge is indicated.|
|Open with DEXTER|
The southern cloud overlaps all components of HESS J1800-240, with a dominant fraction of the cloud overlapping components A and B. The component of this cloud visible in the to 10 km s-1range coincides well with HESS J1800-240B and the HII region W 28A2. The strongest CO temperature peak of this component at ( , ) is within of W 28A2, and is likely the dense material surrounding this HII region. Moreover the peak's velocity at km s-1 (with dispersion of 15 km s-1), suggests a distance (2.4 kpc) similar to that of W 28A2 (2 kpc; Acord et al. 1997), and also W 28. In the km s-1 to 20 km s-1 range, molecular material appears to coincide with all three VHE components of HESS J1800-240. In particular, HESS J1800-240A and C have molecular cloud overlaps only in this latter range.
Using the relation between the hydrogen column density and the 12CO(J=1-0) intensity (the X-factor) W(12CO), (Strong et al. 2004), we estimate a total mass for the northeastern cloud from our NANTEN data at for d=2 kpc within an elliptical region of diameter 0.2 ( pc; centred on HESS J1801-233) for the velocity range 0-25 km s-1. An average density (for neutral hydrogen) of cm-3 is also derived. Similarly the total mass of the southern cloud is estimated at for d=2 kpc and combining clouds from a circular area of radius 0.15 (5 pc) for the velocity range 12-20 km s-1, and area 0.3 ( pc) in diameter for the velocity range 0-12 km s-1 (both regions are centred on HESS J1800-240B). The corresponding average density is cm-3. By integrating over the rather broad 0-20 km s-1 and 0-25 km s-1 ranges we assume that the molecular material along this line of sight is physically connected at the same distance (for example kpc) and possibly distrupted or shocked by a local energy source. Systematic effects in the mass estimates arise from the velocity crowding in this part of the Galactic plane, and also the broad velocity range for which X-factor used above may not necessary apply. In the latter case, the X-factor may underestimate the cloud mass since an appreciable fraction of gas may be heated under the assumption of distrupted and/or shock-heated gas. One must allow for 4 kpc distances for some or even all of the km s-1 cloud components, and therefore the conclusion that they are not related to W 28 and other interesting objects at kpc. If the clouds are related, W 28 could play a disrupting role. The level of this disruption is however unclear since several other plausible candidates related to the star formation (discussed later) in this region could also contribute. Some other molecular cloud complexes have also been discussed as possibly disrupted by adjacent SNRs and/or energetic sources (e.g. Yamaguchi et al. 1999; Moriguchi et al. 2000). In Table 2, we present a full summary of cloud masses and densities (for regions centered on the VHE source coordinates as in Table 1) for various combiniations of cloud components and distances of 2 and 4 kpc. Velocity separation of cloud components are based on their apparent distribution in Fig. 2 (bottom panels).
Figure 3 compares the radio (left panel), infrared and X-ray views (right panel) of the W 28 region with the VHE significance contours. The Very Large Array (VLA) 90 cm continuum radio image from Brogan et al. (2006) illustrates the shell-like SNR morphology peaking strongly along the northern and eastern boundaries. HESS J1801-233 can be seen to overlap the northeastern shell of the SNR, coinciding with a strong peak in the 90 cm continuum emission. We note that a thermal component is likely present in this peak, given its spectral index (for ) between 90 and 20 cm (Dubner et al. 2000). Outlines of the SNRs traced by non-thermal radio emission, G6.67-0.42 and G7.06-0.12 (Yusef-Zadeh et al. 2000; Helfand et al. 2006; labelled as G6.51-0.48 and G7.0-0.1 by Brogan et al. 2006) are also indicated. In addition, Brogan et al. notes that the non-thermal radio arc G5.71-0.08, which overlaps well with HESS J1800-240C, could be a partial shell and therefore an SNR candidate. The distances to G6.67-0.42 and G5.71-0.08 are presently unknown. Directly south of W 28, the ultracompact HII region W 28A2 is a prominent radio source, and is positioned within of the centroid of HESS J1800-240B. The other HII regions G6.1-0.6 (Kuchar & Clark 1997) and 6.225-0.569 (Lockman 1989) are also associated with radio emission.
The X-ray morphology as shown (Fig. 3 right panel) in the ROSAT PSPC (0.5 to 2.4 keV) image from Rho & Borkowski (2002) reveals the central concentration of X-ray emission, which is predominantly thermal in nature with characteristic temperatures in the range to 2 keV. An X-ray peak or Ear lies at the northeastern boundary and just outside the 4 significance contour of HESS J1801-233. A non-thermal component to the ear emission ( ) ( erg cm-2 s-1 at 1 keV) with a power-law index has been suggested by Ueno et al. (2003a) based on XMM-Newton observations in the 0.5 to 10 keV energy range. The total kinetic energy of the SNR is estimated at erg, which could be a lower limit due to the possible break-out of the SNR along the southern edge away from the molecular cloud to the north and east (Rho & Borkowski 2002). The HII regions, W 28A2 and G6.1-0.6 are prominent in the 8.28 m image (Fig. 3 right panel) from the Midcourse Space Experiment (MSX), showing that a high concentration of heated dust still surrounds these very young stellar objects.
Our discovery of VHE -ray emission associated with dense ( cm-3) molecular clouds in the W 28 field adds to the list of such associations after the detection of diffuse -ray emission from the Galactic Ridge (Aharonian et al. 2006c), the association of HESS J1834-087 with the old-age SNR W 41 (Lemiére et al. 2005; Albert et al. 2006) and VHE emission discovered from IC 443 (Albert et al. 2007). The VHE/molecular cloud association could indicate a hadronic origin for the parent multi-TeV particles where the -ray emission (multi-GeV to TeV energies) arises from the decay of neutral pions resulting from the interaction of accelerated protons (and higher Z nuclei) with ambient matter of density n. In this case the -ray flux would scale with cloud mass or density, and the total energy in accelerated particles or CRs penetrating the cloud(s). We note that a perfect correlation between the VHE and molecular cloud morphologies is not expected due to complex time and energy-dependent propagation of CR to and within the cloud (see Gabici et al. 2006, for a discussion). Projection effects are also likely to be important for the examples discussed here since the VHE emission could have contributions from clouds at different velocities, not necessarily physically connected to one another. For example the relationship between HESS J1801-233 and the W 28/molecular cloud interaction is not entirely clear due to the overlapping molecular cloud components at km s-1.
One should also consider accelerated electrons as the source of -ray emission, via inverse-Compton (IC) scattering of ambient soft photon fields and/or non-thermal Bremsstrahlung from the interaction of electrons with dense ambient matter. Maximum electron energies however may be considerably lower (factor 10 or more than that of protons) due to synchrotron cooling in magnetic fields and low shock speeds, in the absence of strong electron replenishment. An assessment of the role of accelerated electrons requires consideration of the non-thermal radio and X-ray emission (where a convincing measurement of the latter is so far lacking), and also magnetic fields in this region. Such observations will also provide constraints on synchrotron emission expected from secondary electrons resulting from primary hadron interactions with ambient matter (as discussed above). Relatively high magnetic fields G are inferred in dense molecular clouds (Crutcher et al. 1999). In addition, higher values are indicated from Zeeman splitting measurements in the compact areas (arcsecond scale) surrounding the 1720 MHz OH masers of the northeastern interaction region (Hoffman et al. 2005), coinciding with HESS J1801-233. To the north of W 28, another potential source of particle acceleration is PSR J1801-23, where the VHE emission may arise in an asymmetric pulsar-wind-nebula (PWN) scenario (a primarily leptonic scenario), similar to HESS J1825-137 (Aharonian et al. 2006d). However with a spin-down power of erg s-1 at distance d>9.4 kpc, this pulsar appears unlikely to power any of the -ray sources observed in the region. A PWN scenario would therefore require a so far undetected energetic pulsar.
Table 2: Details for the molecular clouds towards the VHE sources in the W 28 field, assuming a distance d.In the case of a hadronic origin and following Eq. (10) of Aharonian (1991), we can estimate the CR density enhancement factor in units of the local CR density required to explain the VHE emission, given an estimate for the cloud masses and assumptions on distance. Converting the VHE energy spectra in Table 1 to an integral value for E>1 TeV, assuming distances of 2 and 4 kpc for the various cloud components, and that all the VHE emission in each source is associated with the cloud component under consideration, we arrive at values for in the range 13 to 32 (Table 2).
Overall, these levels of CR enhancement factor would be expected in the neighbourhood of CR accelerators such as SNRs. If the clouds were all at 2 kpc, an obvious candidate for such particle acceleration is the SNR W 28, the most prominent SNR in the region. Despite its old age, multi-TeV particle acceleration may still occur in W 28 (Yamazaki et al. 2006), with protons reaching energies of several 10's of TeV depending on various SNR shock parameters such as speed, size and ambient matter density. In addition, CRs produced at earlier epochs have likely escaped and diffused throughout the region, a situation discussed at length in Aharonian & Atoyan (1996). Aharonian & Atoyan show for slow diffusion (diffusion coefficient at 10 GeV cm2 s-1 as might be expected in dense environments) CR enhancement factors in the required range could be found in the vicinity (within 30 pc - note that if at 2 kpc distance, HESS J1800-240 would lie 10 pc from the southern circular boundary of W 28) of a canonical SNR as an impulsive accelerator up to 105 yr after the SN explosion (see their Fig. 1). In this sense, W 28 as a source of CRs in the region could be a plausible scenario.
The W 28 field however is a rich star formation region, and several additional/alternative sources of CR acceleration may be active. The SNR G6.67-0.42 is positioned directly to the southeast of HESS J1801-233 (Fig. 3 left panel) while the SNR G7.06-0.12 is situated north of HESS J1801-233 and on the west side of the HII region M 20. M 20 itself may also be an energy source for the molecular clouds in this region. The SNR candidate G5.71-0.08 (Brogan et al. 2006) may also be responsible in some way for HESS J1800-24C given the good positional overlap between the two. These radio SNR/SNR candidates are without a distance estimate making it unclear as to how they relate to the molecular clouds in the region. The morphology of HESS J1800-240 displays several peaks, perhaps resulting from changes in cloud density and/or the presence of additional particle accelerators and local conditions. For HESS J1800-24B, a potential energy source is the unusual ultra-compact HII region W 28A2 (G5.89-0.39), representing a massive star in a very young phase of evolution. W 28A2 exhibits very energetic bipolar molecular outflows (Harvey & Forveille 1988; Acord et al. 1997; Sollins et al. 2004) which may arise from the accretion of matter by the progenitor star. The outflow ages are estimated at between 103 to 104 yr. Recent observations (Klaassen et al. 2006) suggest both outflows extend over a combined distance of (or 1.2 pc at d=2 kpc), with total kinetic energy of erg. Surrounding the outflows is a very dense (>104 cm-3) molecular envelope of diameter 0.5 to 1. Despite the lack of any model to explain multi-TeV particle acceleration in such HII regions, its kinetic energy budget and its spatial overlap with a VHE source makes W 28A2 a tempting candidate for such acceleration. Already, there are two examples of VHE emission possibly related to the environments of hot, young stars - TeV J2032+4130 (Aharonian et al. 2005b) and HESS J1023-575 (Aharonian et al. 2007c). In this context, the HII regions G6.1-0.6 and 6.225-0.569 may also play a similar role in HESS J1800-24A. Among the prominent open clusters in the area, NGC 6523 and NGC 6530 southeast of HESS J1800-240, and NGC 6514 associated with M 20 north of HESS J1801-233 may also provide energy for CR production. Finally, if the VHE emission is associated with truly distant cloud components approaching the Scutum-Crux arm at 4 kpc, undetected background particle accelerators would then play a role.
Figure 4 also compares the EGRET and VHE spectra. Given the degree-scale EGRET PSF, GRO J1801-2320 remains
unresolved at scales of the VHE sources. Although the peak of the EGRET emission coincides with HESS J1801-233, we therefore cannot rule
out unresolved MeV/GeV components from HESS J1800-240.
Observations with GLAST will be required to determine the MeV/GeV components of the VHE sources.
|Figure 4: Energy fluxes of HESS J1801-233 and HESS J1800-240 (for regions defined in Table 1) compared to the E>100 MeV counterpart GRO J1801-2320. The power law fits and data points (summarised in Table 1) are also indicated: HESS J1801-233 (solid blue line and points); HESS J1800-240 (open red points and solid line); GRO J1801-232 (solid black points and grey 1 confidence band).|
|Open with DEXTER|
The VHE/molecular cloud association could indicate a hadronic origin for the VHE sources in the W 28 field. Under assumptions of connected cloud components at a common distance of 2 kpc, or, alternatively, separate cloud components at 2 and 4 kpc, a hadronic origin for the VHE emission implies cosmic-ray densities 10 to 30 times the local value. W 28 could provide such densities in the case of slow diffusion. Additional and/or alternative particle accelerators such as HII regions representing very young stars, other SNRs/SNR candidates and/or several open clusters in the region may also be contributors. Alternatively, if cloud components at km s-1 are at distances kpc, as-yet undetected particle accelerators in the Scutum-Crux arm may be responsible. Detailed modeling (beyond the scope of this paper), and further multiwavelength observations of this region are highly recommended to assess further the relationship between the molecular gas and potential particle accelerators in this complex region, as well as the nature of the acclerated particles. In particular, further sub-mm observations (e.g. at high CO transitions) will provide more accurate cloud mass estimates, and allow to search for disrupted/shocked gas towards the southern VHE sources. Such studies will be valuable in determining whether or not W 28 and other energetic sources have disrupted molecular material at line velocities >10 km s-1.
The support of the Namibian authorities and of the University of Namibia in facilitating the construction and operation of HESS is gratefully acknowledged, as is the support by the German Ministry for Education and Research (BMBF), the Max Planck Society, the French Ministry for Research, the CNRS-IN2P3 and the Astroparticle Interdisciplinary Programme of the CNRS, the UK Particle Physics and Astronomy Research Council (PPARC), the IPNP of the Charles University, the Polish Ministry of Science and Higher Education, the South African Department of Science and Technology and National Research Foundation, and by the University of Namibia. We appreciate the excellent work of the technical support staff in Berlin, Durham, Hamburg, Heidelberg, Palaiseau, Paris, Saclay, and in Namibia in the construction and operation of the equipment. The NANTEN project is financially supported from JSPS (Japan Society for the Promotion of Science) Core-to-Core Program, MEXT Grant-in-Aid for Scientific Research on Priority Areas, and SORST-JST (Solution Oriented Research for Science and Technology: Japan Science and Technology Agency). We also thank Crystal Brogan for the VLA 90 cm image and the referee for valuable comments.
Table 3: Numerical summary of differential energy spectra for HESS J1801-233 and HESS J1800-240 (Fluxes F and errors ). Results are derived using both standard and hard cuts. In cases where two flux points are extracted the weighted average flux and error (with are 99% correlated) is quoted. Statistical significances S (from Li & Ma 1983, ApJ 272, 317) from both hard-cuts and standard cuts are given when available. A minimum significance is required for a flux point otherwise an upper limit is quoted.