The XMM-Newton view of the central degrees of the Milky...

The XMM-Newton view of the central degrees of the Milky...
Mon. Not. R. Astron. Soc. 000, 1–?? (0000)
Printed 20 August 2015
(MN LATEX style file v2.2)
The XMM-Newton view of the central degrees of the Milky Way
G. Ponti1? , M. R. Morris2 , R. Terrier3 , F. Haberl1 , R. Sturm1 , M. Clavel3,4 , S. Soldi3,4 ,
A. Goldwurm3,4 , P. Predehl1 , K. Nandra1 , G. Belanger5 , R. S. Warwick6 and V. Tatischeff7
1
Max Planck Institut für Extraterrestrische Physik, 85748, Garching, Germany
Department of Physics and Astronomy, University of California, Los Angeles, CA 90095-1547, USA
3 Unité mixte de recherche Astroparticule et Cosmologie, 10 rue Alice Domon et Léonie Duquet, 75205 Paris, France
4 Service d’Astrophysique (SAp), IRFU/DSM/CEA-Saclay, 91191 Gif-sur-Yvette Cedex, France
5 ESA/ESAC, PO Box 78, 28691 Villanueva de la Cañada, Spain
6 Department of Physics and Astronomy, University of Leicester, University Road, Leicester, LE1 7RH, UK
7 Centre de Sciences Nucléaires et de Sciences de la Matière, IN2P3-CNRS and Univ Paris-Sud, F-91405 Orsay Cedex, France
arXiv:1508.04445v1 [astro-ph.HE] 18 Aug 2015
2
20 August 2015
ABSTRACT
The deepest XMM-Newton mosaic map of the central 1.5◦ of the Galaxy is presented, including a total of about 1.5 Ms of EPIC-pn cleaned exposures in the central 15” and about 200 ks
outside. This compendium presents broad-band X-ray continuum maps, soft X-ray intensity
maps, a decomposition into spectral components and a comparison of the X-ray maps with
emission at other wavelengths. Newly-discovered extended features, such as supernova remnants (SNRs), superbubbles and X-ray filaments are reported. We provide an atlas of extended
features within ±1 degree of Sgr A? . We discover the presence of a coherent X-ray emitting
region peaking around G0.1-0.1 and surrounded by the ring of cold, mid-IR-emitting material
known from previous work as the ”Radio Arc Bubble” and with the addition of the X-ray data
now appears to be a candidate superbubble. Sgr A’s bipolar lobes show sharp edges, suggesting that they could be the remnant, collimated by the circumnuclear disc, of a SN explosion
that created the recently discovered magnetar, SGR J1745-2900. Soft X-ray features, most
probably from SNRs, are observed to fill holes in the dust distribution, and to indicate a direct interaction between SN explosions and Galactic center (GC) molecular clouds. We also
discover warm plasma at high Galactic latitude, showing a sharp edge to its distribution that
correlates with the location of known radio/mid-IR features such as the ”GC Lobe”. These
features might be associated with an inhomogeneous hot ”atmosphere” over the GC, perhaps
fed by continuous or episodic outflows of mass and energy from the GC region.
Key words: Galaxy: centre; nucleus; interstellar medium; ISM: supernova remnants; bubbles; kinematics and dynamics; X-rays: binaries; diffuse background; ISM; plasmas; methods:
data analysis;
1
INTRODUCTION
At a distance of only ∼ 8 kpc, the center of the Milky Way is the
closest Galactic nucleus, allowing us to directly image, with incomparable spatial resolution, the physical processes typical of galactic
nuclei. The central region of the Galaxy is one of the richest laboratories for astrophysics (Genzel et al. 2010; Morris et al. 2012; Ponti
et al. 2013). Within the inner ∼ 200 pc about 3 − 5 × 107 M of
molecular material are concentrated, the so called Central Molecular Zone (CMZ). This corresponds to about 1 % of the molecular
mass of the entire Galaxy and it is concentrated in a region of about
∼ 10−6 of its volume (Morris & Serabyn 1996). In this region
?
[email protected]
c 0000 RAS
many thousands of persistent and transient point-like X-ray sources
are embedded, such as active stars, bright accreting binary systems
(and many more quiescent massive bodies) and cataclysmic variables, which have been beautifully imaged thanks to the superior
spatial resolution of Chandra (Wang et al. 2002; Muno et al. 2003;
2009). One of the best jewels in the GC is Sgr A? , the electromagnetic counterpart of the closest supermassive black hole (BH; Genzel et al. 2010). In addition to this large population of point sources,
extended X-ray sources, such as supernova remnants, non-thermal
filaments, pulsar wind nebulae, and massive star clusters, populate
the GC (Wang et al. 2002). The GC is considered a mini-starburst
environment, giving us the possibility to study the interaction between supernova remnants (SNRs) and molecular clouds and the
impact of massive-and-young star clusters on their surroundings. It
2
G. Ponti et al.
allows us to image, in superb detail, the creation and evolution of
bubbles and superbubbles and the generation of Galactic outflows,
powered by past starbursts and/or accretion events onto Sgr A? , and
their impact on the GC environment.
Warm (kT ∼ 1 keV) and hot (kT ∼ 6.5 keV) thermal plasma
emission plus non-thermal hard X-ray emission associated with Xray reflection nebulae (see Ponti et al. 2013 for a review) pervade
the central region, producing a high background of soft and hard Xray radiation. About 90% (Ebisawa et al. 2001; Wang et al. 2002)
of the soft X-ray emission appears to be due to a diffuse, patchy
and thermal component (Bamba et al. 2002) with a temperature
kT ∼ 1 keV, most probably associated with supernova remnants.
The origin of the hot component is, instead, still highly debated. At
∼ 1.5◦ from the GC, ∼ 80 % of this emission has been resolved
into point sources (e.g., accreting white dwarfs and coronally active stars) by a deep Chandra observation (Revnivtsev et al. 2009).
Although the intensity of the hot plasma emission increases rapidly
towards the GC, point sources continue to make a substantial contribution to the observed hard emission (Muno et al. 2004; Heard
& Warwick 2013a). Additionally, some of the emission may arise
due to scattering of the radiation from bright X-ray binaries by the
dense interstellar medium (Sunyaev et al. 1993; Molaro et al. 2014).
Nevertheless, it is not excluded that a truly diffuse hot-plasma component is also present in the GC (Koyama et al. 2009; Uchiyama et
al. 2013). Such hot plasma would be unbound to the Galaxy and it
would require a huge energy (E ∼ 1055 erg) and energy loss rate
of the mass outflow of ∼ 1043 erg s−1 , corresponding to a rate of 1
supernova/yr, to continuously replenish it (Tanaka 2002). However,
it has recently been proposed that such hot plasma might be trapped
by the GC magnetic field (Nishiyama et al. 2013).
Indeed, the magnetic field is thought to be an important ingredient of the GC environment. The first high-resolution radio
images of the Milky Way center (see bottom panel of Fig. 6), revealed the presence of many straight, long (up to ∼ 20 − 30 pc)
and thin (with width <
∼0.1 pc), linearly polarised vertical filaments
with spectral index consistent with synchrotron radiation (YusefZadeh et al. 1984; 1987a,b; Anantharamaiah et al. 1991; Lang et
al. 1999; LaRosa et al. 2000). These filaments are hypothesized to
be magnetic flux tubes trapping energetic electrons and therefore
tracing the diffuse interstellar GC poloidal magnetic field (Morris
& Yusef-Zadeh 1985; Lang et al. 1999). A staggeringly powerful
poloidal magnetic field pervading the GC, with a field strength of
B >
∼ 50 µG, and very possibly B ∼ 1 mG, has been inferred
(Morris 1990; Crocker et al. 2010; Ferriere et al. 2011). The details of the physical process creating the filaments and energising
the magnetic field are still debated; however, it appears clear that
the magnetic filaments are interacting with the ionised surfaces of
massive molecular clouds.
Recent far-infrared/sub-millimeter polarization studies of
thermal dust emission made it possible to probe the direction of
the interstellar magnetic field inside dense molecular clouds. The
magnetic field threading GC molecular clouds is measured to be
parallel to the Galactic plane (Novak et al. 2003; Chuss et al. 2003;
Nishiyama et al. 2009). Therefore, it appears that the large-scale
GC magnetic field is poloidal in the diffuse interstellar medium
and toroidal in dense regions in the plane. If the strength of the diffuse magnetic field is on the high side (B ∼ 1 mG) a huge amount
of magnetic energy, E ∼ 1055 erg, would be stored in the central
∼ 300 pc. This is comparable to the kinetic energy associated with
the rotation of the gas in the CMZ. Therefore, it is thought to be a
key player for the GC physics and phenomenology.
A large scale structure with a possible magnetic origin and
appearing to be interacting with massive clouds of the CMZ (similar to the non-thermal filaments) is the Galactic center lobe (GCL).
The GCL has a limb brightened shell structure in the 10.5 GHz
map, defined primarily by two spurs (see Fig. 1 and 2 of Law et al.
2009). The eastern one arises from the location of the GC Radio
Arc1 , while the second starts from the Sgr C thread. It was proposed that the GCL is produced by channelling of plasma from the
Galactic plane, induced by energetic GC activity (e.g. episode of
AGN activity, or a large mass outflow due to the high star formation rate, etc.; see Law et al. 2011) or from twisting of poloidal
magnetic field lines by Galactic rotation (Sofue et al. 1984; 1985;
Uchida 1985; 1990; Shibata 1989). Located at the western limb of
the GCL is an interesting feature, AFGL 5376 (Uchida et al. 1994),
an unusually warm, shock heated and extended IR source, thought
to be associated with the GCL.
All major X-ray telescopes devoted a significant fraction of
their time to the study of the GC. Chandra invested several Ms to
monitor both Sgr A? ’s activity (Baganoff et al. 2001; 2003; Neilsen
et al. 2013) as well as diffuse soft and hard X-ray emission (Wang
et al. 2002; Park et al. 2004). Suzaku and Swift also performed large
observational campaigns to scan the Milky Way center (Koyama et
al. 2007; Degenaar & Wijnands 2010) and monitor the transients
in the region (Degenaar et al. 2012). The study of the GC is one
of the key programs of the NuSTAR mission (Harrison et al. 2013;
Barriere et al. 2014; Mori et al. 2013). XMM-Newton completed a
first shallow (∼ 30 ks total cleaned exposure in each point) scan
of the CMZ within a couple of years after launch (see the conference proceedings: Sakano et al. 2003; 2004; Decourchelle et al.
2003). A larger amount of time (more than ∼ 1.5 Ms) has been
invested by XMM-Newton on studying the emission properties of
Sgr A? (Goldwurm et al. 2003; Porquet et al. 2003; 2008; Belanger
et al. 2005; Trap et al. 2011; Mossoux et al. 2014), focussing on the
central ∼ 15 arcmin, only. Using the XMM-Newton observations
from the shallow scan of the CMZ together with a number of the
Sgr A? pointings, Heard & Warwick (2013a,b) have investigated
the distribution of the X-ray emission within the central region of
the Galaxy. With the aim of studying the propagation of echoes of
the past GC activity within the CMZ (Sunyaev et al. 1993; Koyama
et al. 1996; 2008; Revnivtsev et al. 2004; Muno et al. 2007; Inui
et al. 2009; Ponti et al. 2010; 2013; Terrier et al. 2010; Nobukawa
et al. 2011; Capelli et al. 2011; 2012; Clavel et al. 2013; 2014;
Krivonos et al. 2014), recently, a new deep (with ∼ 100 ks exposure at each location) XMM-Newton scan of the CMZ has been
completed (in fall 2012). We present here the combined images of
both the new and old XMM-Newton scans, as well as all the XMMNewton observations within the central degree of the Galaxy.
In §2 we present the data reduction process and the key steps to
produce the GC EPIC mosaic maps. Section 3 introduces the broad
band X-ray images, discussing the (transient) emission from the
brightest point sources, the contribution from the foreground emission, as well as the soft and hard GC diffuse emission. In section 4,
the narrow band images at the energies of the soft X-ray lines are
displayed. Section 5 presents a new technique of spectral-imaging
decomposition of the soft X-ray emission into three physical components. Section 6 presents an atlas of all the new and known diffuse features within the surveyed area. Section 7 presents the comparison with the distribution of column density of intervening mat-
1
This is a well known radio feature (see Fig. 6) composed of an array of
straight, long, thin and linearly polarised vertical filaments, indicating the
importance of the GC magnetic field.
c 0000 RAS, MNRAS 000, 1–??
3
ter. Discussion and conclusions are in § 8 and 9, respectively. Hereinafter, unless otherwise stated, we will state all locations and positions in Galactic coordinates and Galactic cardinal points. Errors
are given at 90 % confidence for one interesting parameter.
2
DATA REDUCTION AND CLEANING
The new XMM-Newton CMZ scan has been performed in 2012
starting on August 30th and ending on October 10th . It comprises
of 16 XMM-Newton observations all performed with all the EPIC
instruments in full-frame CCD readout mode with the medium optical blocking filters applied (we refer to Tab. 8 and 9 for more
details on the instruments set-ups).
This paper is not limited to the use of the 2012 XMM-Newton
scan of the CMZ. Instead it is using all XMM-Newton observations
pointed within 1 degree from Sgr A? . Therefore we combined the
16 observations of the new CMZ scan with the 14 observations of
the previous CMZ scan accumulated between 2000 and 2002. We
also include the 30 observations pointing at Sgr A? and other 49
observations aimed at studies of different sources in the vicinity of
Sgr A? (see Tab. 8 and 9).
We performed the analysis of the EPIC data with the version
13.0.0 of the XMM-Newton Science Analysis System (SAS). Periods of increased particle background have been removed from the
data. To perform this, we first selected the Good Time Intervals
(GTI) starting from the 7-15 keV background light curves, then we
applied a threshold of 8 and 2.5 cts ks−1 arcmin−2 for EPIC pn and
EPIC MOS, respectively (see e.g. Haberl et al. 2012). The chosen
thresholds efficiently cut out all the periods of most extreme activity of soft proton flares. We noted, however, that an enhanced, but
weak, background activity was still present in the data during several observations. Because of the non-uniform distribution of the
GC diffuse emission, lowering the threshold uniformly in all observations, would result in cutting truly good time intervals in observations with higher GC diffuse emission. Thus we decided to visually
inspect the background light curve of all data-sets and select a different threshold for each observation (see Tab. 8 and 9). Such as
in Haberl et al. (2012), when the data from several detectors were
available, we combined the GTIs using only common time intervals, otherwise we included GTIs of the single detector. Most of
the 2012 CMZ scan data were affected by negligible particle flaring activity. On the other hand, many of the previous observations
have been severely affected by soft proton flares (see the reduction
in exposure in Tab. 8 and 9).
To prevent infrared, optical and UV photons from bright
sources in the field of view that would increase the noise and degrade the CCD energy scale, the different XMM-Newton observations have been performed with different filters applied, according
to the optical-UV brightness of the sources in the field of view (see
Tab. 8 and 9). In particular, we used the filter wheel closed observations to remove the internal EPIC background.
2.1
Images and exposure maps
Images and exposure maps, corrected for vignetting, have been produced with an image pixel size of 2” × 2” for each energy band
(for the definition of all bands, see § 2.2). To increase the sky coverage, we selected EPIC-pn events requiring (FLAG & 0xfa0000)
= 0, which also includes events in pixels next to bad pixels or
bad columns. Moreover, we used single to double pixel events.
c 0000 RAS, MNRAS 000, 1–??
EPIC-MOS events were required to have FLAG = 0 and single to
quadruple-pixel events were allowed.
Figure 1 shows the combined EPIC exposure map that covers
the entire CMZ. Such as done in Sturm et al. (2013), EPIC-MOS1
and -MOS2 exposures are weighted by a factor of 0.4 relative to
EPIC-pn, before being added to the latter, to account for the lower
effective area. Therefore, the exposure times obtained correspond
to the equivalent total EPIC-pn exposure time. This allows us to
obtain a better combination of EPIC-pn and EPIC-MOS data for
image display purposes. We note, however, that the fluxes can not
be easily read out directly from these combined images. Therefore,
the line profiles and the measured fluxes/luminosities are computed
from the EPIC-pn and each EPIC-MOS map separately and then
combined (averaged) to obtain a better signal to noise.
The top panel of Fig. 1 shows that more than 1.5 Ms of clean
(after cut of time intervals during increased particle background
activity) exposure time (EPIC-pn equivalent) has been accumulated
on Sgr A? and over ∼ 100 − 200 ks are present in each point of the
CMZ. The few pointings above and below the plane have exposures
between ∼ 15 − 40 ks. Regions with less than 7.2 ks of equivalent
EPIC-pn exposure have been masked out.
To check the impact of the bright transients on the images and
on the physical quantities under investigation, two sets of maps
have been created. The first series keeps all bright transients and
point sources, while the second set removes their emission by excising from the data extended regions including the transients whenever they were in outburst (see section 3.1). The middle and bottom
panels of Fig. 1 show the exposure maps (computed in the same
way) for the observations of new and old CMZ scans, separately.
The maximum exposure times are ∼ 190 ks and ∼ 45 ks during
the new and old scan, respectively.
2.2
Energy bands
We created images in several energy bands (see Tab. 1). Figure
2 shows the EPIC-pn spectra of the extended emission from several regions within the CMZ. In red and black are the spectra from
the G0.11-0.11 and Center Superbubble regions, respectively (see
Fig. 6). Both regions are located within 15 arcmin from Sgr A? ,
thus they have excellent statistics because of the large exposure. In
green and blue are the pn spectra of G0.687-0.146 and Sgr B1 soft,
respectively, both are located further out, thus having lower exposure.
We first selected the standard broad energy bands for the continuum with the softer band being: E = 0.5 − 2 keV; the medium
2 − 4.5 keV and the hard 4.5 − 12 keV (see Tab. 1). We note that,
at high energies, the EPIC-pn camera shows strong instrumental
background emission lines due to Ni Kα (at E ∼ 7.47 keV), Cu
Kα (∼ 8.04 keV) and ZnCu (∼ 8.63 and 8.87 keV) that strongly
contribute to the observed X-ray emission in the hard band (see
Freyberg et al. 2004). To avoid contamination from these strong
internal background lines, we do not consider (for the EPIC-pn images) photons in the 7.2-9.2 keV range (see Fig. 2 and Tab. 1). We
chose these broad energy bands because they are typically used as
input by the standard XMM-Newton point source detection algorithm and for comparison to other similar surveys of nearby galaxies (e.g. M33: Misanovic et al. 2006; Tüllmann et al. 2011; M31:
Henze et al. 2014; Stiele et al. 2011; LMC: Haberl et al. 1999;
SMC: Haberl et al. 2012; Sturm et al. 2013). However we note that,
given the typical GC neutral column density of several 1022 cm−2 ,
the low energy absorption cut-off occurs at the highest energies of
the standard soft band, making standard broad band RGB images
4
G. Ponti et al.
TOTAL CMZ SCAN
0.400
0.200
Galactic latitude
0.000
-0.200
-0.400
-0.600
-0.800
1.500
1.000
0.500
0.000
359.500
359.000
Galactic longitude
0.00e+00
6.48e+03
1.72e+04
3.50e+04
6.41e+04
1.13e+05
1.92e+05
3.23e+05
5.42e+05
9.00e+05
1.49e+06
NEW CMZ SCAN
0.400
0.300
Galactic latitude
0.200
0.100
0.000
-0.100
-0.200
-0.300
-0.400
1.500
1.000
0.500
0.000
359.500
359.000
Galactic longitude
0.00e+00
6.48e+03
0.400
1.72e+04
3.50e+04
6.41e+04
1.13e+05
1.92e+05
3.23e+05
5.42e+05
9.00e+05
1.49e+06
OLD CMZ SCAN
0.300
Galactic latitude
0.200
0.100
0.000
-0.100
-0.200
-0.300
-0.400
1.500
1.000
0.500
0.000
359.500
359.000
Galactic longitude
0.00e+00
6.48e+03
1.72e+04
3.50e+04
6.41e+04
1.13e+05
1.92e+05
3.23e+05
5.42e+05
9.00e+05
1.49e+06
Figure 1. (Top panel) Combined exposure map of all the XMM-Newton EPIC-pn + MOS1 + MOS2 observations within one degree from Sgr A? . Such as
done in Sturm et al. (2013), EPIC-MOS1 and -MOS2 exposure is weighted by a factor of 0.4 relative to EPIC-pn to account for the lower effective area.
The exposure times, thus, correspond to the equivalent total EPIC-pn exposure time. Regions with less than 7.2 ks of equivalent EPIC-pn exposure have been
masked out. The cleaned EPIC-pn equivalent exposure time is reported in seconds. (Medium panel) Similar exposure map for the observations of the new
CMZ scan only. (Bottom panel) Similar exposure map for the old CMZ scan only (regions with less than 7.2 ks of equivalent EPIC-pn exposure are included).
The maximum exposure times are ∼ 1.5 Ms, ∼ 190 ks and ∼ 45 ks during the total, new and old scan, respectively.
poorly sensitive to column density fluctuations. For this reason we
define a second set of broad bands, the ”GC continuum bands” (see
Tab. 1). The first band (E = 0.5 − 1.5 keV) contains mainly emission from foreground sources. The second band (E = 1.5 − 2.6
keV) is selected in order to contain the low-energy GC neutral absorption cut-off, thus making it more sensitive to either column
density or soft gas temperature variations. While the ”GC medium”
(E = 2.6−4.5 keV) and the ”GC hard” (E = 4.5−12 keV) bands
are similar to the standard broad bands.
We also selected images at the energies of the soft emission
lines, such as Si XIII, S XV, Ar XVII and Ca XIX. To perform continuum subtracted line intensity maps as well as line equivalent width
c 0000 RAS, MNRAS 000, 1–??
5
0(%1(((%%
!$%56%%
!$%112%%
0%12%% 3#%12((%%4,%1(1%%
Soft
0.5-2
Standard continuum bands:
Medium
Hard†
2-4.5
4.5-12
!$!#
Fore
0.5-1.5
78%!$5%%
GC continuum bands:
GC Soft
GC Medium GC Hard†
1.5-2.6
2.6-4.5
4.5-12
Soft emission lines:
S XV
Ar XVII
Ca XIX
2.35-2.56
3.03-3.22
3.78-3.99
Continuum subtraction soft emission lines:
Red-Si
Si-S
S-Ar
Ar-Ca
Blue-Ca
1.65-1.77
2.1-2.3
2.7-2.97
3.27-3.73
4.07-4.5
Si XIII
1.80-1.93
!$!%
&'()*+,-./01'2&3404!!05.6!!0*(1),&!"
4";<%!$5%
9":% *$'% -,#'%
!"#$%
0":%%
+,#'%%
&$'()*%%
.$#/%
-,#'%%
&$'()*%-,#'%%
!
"
Fe Kα
6.3-6.5
#
7&.(890:5.6;
CFeK
Figure 2. EPIC-pn spectra of the regions: G0.11-0.11 (red), Center Superbubble (black), G0.687-0.146 (green) and Sgr B1 soft (blue). In dark green
are the energy bands of the broad GC continuum. In orange (bottom right)
is shown the part of the hard energy band excluded in order to avoid contamination by Ni Kα, Cu Kα and ZnCu instrumental background emission
lines. Yellow stripes show the energy bands of the soft and Fe K emission
lines. Blue stripes indicate the regions selected for the determination of the
amount of continuum underlying the soft lines. The dotted lines, from top
to bottom, show the predicted emission of a source with a power-law spectrum (with slope Γ = 1.6) if absorbed by a column density of NH = 3, 5
and 9 × 1022 cm−2 , respectively. The blue and orange labels indicate the
selected broad energy bands for the determination of the continuum underlying the Fe K line emission.
5-6.1
Table 1. Energy bands used for each of the different continuum, and narrow line images. Also shown are the energy bands used to determine the
continuum underlying the line emission. Units are in keV. Several energy
bands, at lower energies compared to the FeK lines, have been computed
to determine the best continuum subtraction for the FeK lines. †To avoid
contribution from the strong internal detector background emission lines,
present in the EPIC-pn camera (such as: Ni Kα, Cu Kα and ZnCu), we do
not consider photons in the 7.2–9.2 keV from this instrument (on the other
hand, we consider such photons detected in the EPIC-MOS cameras).
2.2.1
maps, it is essential to measure the level of the continuum underlying the emission line. Therefore, we created also several images
in energy bands on each side of the soft emission lines2 (selecting,
as far as possible, energy ranges free from line emission; see Fig. 2
and Tab. 1).
In the Fe K region we selected two energy bands for the Fe Kα
and Fe XXV emission. At energies higher than Fe XXV the presence
of both Fe Kβ, Fe XXVI, and of the Fe K edge can give a significant contribution. At even higher energies (E ∼ 7.5 − 8 keV) the
contribution from internal background emission line (in the EPICpn camera) becomes very important, thus we decided to determine
the continuum emission underlying the Fe K line emission (important to determine the Fe K line intensities and equivalent widths)
through the extrapolation of the continuum red-ward of the Fe K
lines (see Fig. 2 and Tab. 1). The Fe K line emission and its variations will be the prime scientific focus of two future publication
(Ponti et al. in prep.; Soldi et al. in prep.; see also Ponti et al. 2014;
Soldi et al. 2014) and will not be discussed here any further.
All images were exposure corrected and, to remove readout
streaks, the images from EPIC-pn were corrected for out-of-time
events. Noisy CCDs in the MOS data (Kuntz & Snowden 2008)
have been searched with the SAS task emtaglenoise and removed
from the mosaic images.
The energy band red-ward of the Si XIII line extends only down to
1.65 keV, because of the presence of the strong Al Kα background emission
line at E ∼ 1.49 keV (Freyberg et al. 2004).
c 0000 RAS, MNRAS 000, 1–??
Stray-light rejection
Because the GC region is crowded with many bright (transient) Xray sources, several observations, including the new XMM-Newton
scan, are badly affected by stray-light (see Fig. 3). Stray-light is
produced by photons from sources located outside of the XMMNewton EPIC instrument’s fields of view and singly reflected by
the mirror hyperbolas, thus creating concentric arc-like structures
in the detector plane (see XMM-Newton user handbook). The straylight contribution is small (the effective collecting area for straylight is less than ∼ 0.2 % of the effective on-axis area), but a very
bright source can have an important impact up to ∼ 1.4 deg outside
the field of view.
Analogously to the removal of bright transients, we masked
the strongest stray-light artefacts in the images of individual observations. In most cases, affected regions are covered by other unaffected observations, thus leaving no features in the final mosaic
map. To remove a stray-light artefact, we defined a rough region
including the artefact and an individual cut-off value of the surface
brightness. Using this cut-off, we created a mask from an image of
this region in the total energy band that has been smoothed with a
Gaussian kernel with a FWHM of 1000 beforehand. This mask was
multiplied by all images and exposure maps of this observation.
2.2.2
2
Fe K lines:
Fe XXV
6.62-6.8
Continuum subtraction Fe K:
CsFeK
CmFeK
ChFeK
soft
medium
hard
4.0-4.7
4.7-5.4
5.4-6.1
Adaptive smoothing
All images have been smoothed separately using the SAS tool AS MOOTH . To prevent different smoothing patterns from introducing
colour artefacts in RGB images, we adaptively smoothed all energy bands in such images with the same smoothing template. For
the broadband XMM-Newton continuum RGB images (see Fig. 3,
G. Ponti et al.
1.000
1.500
-0.800
-0.600
-0.400
-0.200
0.000
0.200
0.400
TOTAL CMZ SCAN
Red: 0.5-2 keV
Green: 2-4.5 keV
Blue: 4.5-12 keV
0.500
Galactic longitude
0.000
359.500
359.000
6
Galactic latitude
Figure 3. Standard broad energy band (see Tab. 1 and Fig. 2) RGB mosaic image of all XMM-Newton observations within one degree of Sgr A? (see Tab. 8).
This represents the deepest X-ray view of the CMZ region with exposure higher than 0.2 Ms along the Galactic disc and 1.5 Ms in the center (see Fig. 1).
X-ray emission from X-ray binaries, star clusters, supernova remnants, bubbles and superbubbles, HII regions, PWNs, non-thermal filaments, nearby X-ray
active stars, the supermassive BH Sgr A? and many other features are observed (see Fig. 5 and 6). The detector background has been subtracted and adaptive
smoothing applied. Residual features and holes generated by correction of the stray light from GX 3+1 are visible (see also Fig. 1) at Galactic latitudes between
c 0000 RAS, MNRAS 000, 1–??
l ∼ 1.2◦ and l ∼ 1.4◦ and latitudes b ∼ −0.2◦ and b ∼ 0.4◦ .
1.000
1.500
-0.400
-0.300
-0.200
-0.100
0.000
0.100
0.200
0.300
0.400
Red: 0.5-2 keV
Green: 2-4.5 keV
Blue: 4.5-12 keV
0.500
Galactic longitude
0.000
359.500
359.000
7
Galactic latitude
Figure 4. Standard broad energy band RGB image of the XMM-Newton CMZ scan performed in 2012. The CMZ is observed with a uniform exposure (see
Fig. 1).
c 0000 RAS, MNRAS 000, 1–??
8
G. Ponti et al.
Figure 5. Finding chart. The brightest X-ray point sources (all X-ray binaries) are labelled in white (see Tab. 2). In red the positions of some star clusters
are reported, which are placed either in the GC or along the spiral arms of the Galaxy (see Tab. 3). With yellow dashed lines the location of some molecular
complexes are shown. See WWW. MPE . MPG . DE / HEG / GC / for a higher resolution version of this figure.
4, 5, 6, 7, 15 and 17), we required a minimum signal-to-noise ratio
of 6 in the 0.5-12 keV energy band image (i.e., the sum of the three
energy bands composing the RGB image), as well as the standard
minimum and maximum size of the smoothing Gaussian kernel of
1000 (full width half maximum) and 20000 , respectively. The signal
to noise ratio at each pixel is defined as the value at that pixel divided by its standard deviation and the adaptive smoothing that we
applied is making the signal to noise ratio as close at possible to
6, therefore fainter or less exposed areas are more smoothed than
brighter or better exposed regions. For the narrower-band soft-line
images (see Fig. 10, 11, 19 and 20), we use the S xv map as a
template, requiring a minimum signal-to-noise ratio of 4 and the
same standard minimum and maximum of the smoothing kernel.
The same smoothing kernel is then applied to all the other bands of
the RGB images.
2.2.3
Internal particle background subtraction
Unless otherwise specified, internal particle background has been
removed from each broad-band image. Following Haberl et al.
(2012) we first create, for each selected energy band, both the total
emission and the filter wheel closed images. We then re-normalise
and subtract the filter wheel closed images from the total emission
images. The filter wheel closed image re-normalisation factor is
computed by equating, for each instrument, the number of photons in the unexposed corners of the detector to that in the filter
wheel closed images (see Haberl et al. 2012 for more details). This
procedure is reliable and accurate for reasonably long exposures
(t ' 5 − 10 ks). For this reason datasets with total clean EPIC-pn
exposure shorter than 5 ks have not been considered in this analysis.
2.2.4
Continuum subtraction
To subtract the continuum emission from an emission-line image,
we define a narrow band (B) containing the line, typically sandwiched by two nearby but generally wider energy bands (A and
C) that are dominated by continuum emission. Under the assumptions that the emission in the A and C bands is dominated by the
continuum and that the continuum emission can be described by a
simple power-law, we could in principle determine the intensity of
the continuum for each pixel of the band B image. Indeed, using
the fluxes in A and C bands, it is possible to derive the continuum
parameters (spectral index Γ and intensity). However, this requires
the solution of non-linear equations. Therefore, we prefer to implement a different technique based on interpolation. Using Xspec
we simulate, for power-law spectral indices going from Γ = 0.3
to 3.6, the ratio between the observed flux (e.g., number of photons measured) in the continuum in bands A and C, compared to
the simulated continuum flux in band B (e.g. NB /(2 × NA ) and
NB /(2 × NC )). We record these ratios and then plot them as a
function of the hardness ratios (NC − NA )/(NC + NA )), which
is a proxy for the spectral index Γ. We then find the best-fitting
linear relationship between these values, thus determining ConAB
and LinAB that are then allowing us to measure the continuum
emission underlying the line emission in band B (NB ) from the intensity in band A (NA ) and the hardness ratio (NB = 2 × NA ×
[ConAB + LinAB × (NC − NA )/(NC + NA )]. To reduce the uncertainties, we perform the same procedure for band C, determining
c 0000 RAS, MNRAS 000, 1–??
9
Figure 6. Finding charts. (Top panel) Broadband X-ray continuum image. White ellipses show the position and size of known, radio-detected supernova
remnants. Cyan ellipses indicate the position and size of bright diffuse X-ray emission possibly associated with supernova remnants that lack a clear radio
counterpart (or such in the case of G359.12-0.05 that show X-ray emission significantly displaced from the radio emission associated to the radio remnant
G359.07-0.02). The magenta ellipses show the location and dimension of some bright HII regions, while the red ellipses indicate some of the largest nonthermal filaments detected in radio (see Tab. 4). Blue ellipses show some PWN and the yellow dashed ellipses show the regions used to accumulate the spectra
shown in Fig. 2. (Bottom panel) 90-cm radio image of the CMZ region obtained with the VLA (courtesy of LaRosa et al. 2000). For display purposes the radio
supernova remnants are shown with green ellipses. The other regions have the same colour code as the top panel. The image shows the radio flux (Jy beam−1
unit). See WWW. MPE . MPG . DE / HEG / GC / for a higher resolution version of these figures.
c 0000 RAS, MNRAS 000, 1–??
10
G. Ponti et al.
Figure 7. Finding chart. Zoom of the central ∼ 10 arcmin of the Milky Way as seen by XMM-Newton (same energy bands and color scheme as in Fig. 3). The
position of Sgr A? is indicated by the blue cross. The red ellipses show the position and extent of filamentary and diffuse X-ray emission features associated
with, e.g., non-thermal filaments (Tab. 4). The magenta dashed ellipses show the location and extension of the 20 pc bipolar X-ray lobes. The black dashed
ellipses indicate the position and orientation of the CND. See WWW. MPE . MPG . DE / HEG / GC / for a higher resolution version of this figure.
ConCB and LinCB . We then average these values obtaining, for
each pixel: NB = NA × [ConAB + LinAB × (NC − NA )/(NC +
NA )] + NC × [ConCB + LinCB × (NC − NA )/(NC + NA )].
We finally subtract this continuum emission image from the total
emission image B to determine the line intensity map.
3
THE XMM-Newton BROADBAND VIEW OF THE
GALACTIC CENTRE
Figure 3 shows the broad energy band mosaic image of all existing XMM-Newton observations within 1 degree from Sgr A? . Figure 4 shows the Galactic centre image obtained only with data from
the 2012 XMM-Newton campaign. At the GC distance of 7.8 kpc
c 0000 RAS, MNRAS 000, 1–??
11
Figure 8. Finding chart. Chandra RGB image of all the ACIS-I observations pointed at Sgr A? (see Clavel et al. 2013 for data reduction and details on
the image production). The red, green and blue images show the GC soft (1.5-2.6 keV), GC medium (2.6-4.5 keV) and GC hard (4.5-8 keV) energy bands,
respectively. The same regions displayed in Fig. 7 are evidenced here. See WWW. MPE . MPG . DE / HEG / GC / for a higher resolution version of this figure.
(Boehle et al. 2015), 1 arcmin corresponds to 2.3 pc, 10 pc subtends
∼ 4.3’ and ∼ 0.2◦ corresponds to 28 pc. In red, green and blue, the
soft (0.5-2 keV), medium (2-4.5 keV) and hard (4.5-12 keV) continuum bands are shown, respectively. Hundreds of point sources
and strong diffuse emission are clearly observed in the map. These
point sources are characterised by a wide variety of colours, ranging from distinctive red to dark blue.
c 0000 RAS, MNRAS 000, 1–??
3.1
Bright and transient GC point sources during the new
(2012) XMM-Newton CMZ scan
Many X-ray point sources are clearly visible in Fig. 3 and 4. A detailed catalogue of the properties of all the detected point sources is
beyond the scope of this paper. Here we briefly describe the brightest GC sources detected by XMM-Newton and the X-ray transients
12
G. Ponti et al.
G359.969-0.027
-0.010
-0.020
G359.897-0.023
G359.959-0.027
G359.921-0.030
G359.941-0.029
-0.030
G359.969-0.033
G359.933-0.037
Galactic latitude
G359.971-0.038
G359.95-0.04
G359.983-0.040
-0.040
G359.933-0.039
G359.945-0.044
G359.942-0.045
Sgr A*
G359.944-0.052
G359.964-0.053
-0.050
CND
G359.904-0.047
G359.925-0.051
G359.956-0.052
G359.921-0.052
G359.965-0.056
G359.915-0.061
-0.060
G359.962-0.062
G359.899-0.065
-0.070
G359.977-0.076
359.980
359.970
359.960
359.950
359.940
359.930
359.920
359.910
359.900
Galactic longitude
Figure 9. Zoom of the finding chart displayed in Fig. 8.
in the field of the 2012 scan (see Degenaar et al. 2012 for a compendium of previously noted transients).
The brightest GC point source of the 2012 XMM-Newton
CMZ scan is 1E 1743.1-2843 (Porquet et al. 2003; Del Santo et
al. 2006) a persistently accreting neutron star binary detected at an
observed flux level of F2−10keV ∼ 1.1 × 10−10 erg cm−2 s−1 (implying an unabsorbed flux of F2−10keV,unab ∼ 2.6 × 10−10 erg
cm−2 s−1 ; NH ∼ 2 × 1023 cm−2 ; obsid: 0694641201). During the
2012 XMM-Newton campaign we also detected an outburst from a
new, very faint X-ray transient that we name XMMU J174505.3291445. The source has a typical quiescent luminosity at or below LX ∼ 1033 erg s−1 , but on 2012 August 31st (during obsid
0694640201) it was observed to go into outburst for about ∼ 2 hr
and to reach a peak X-ray luminosity of LX ∼ 1035 erg s−1 . The
detailed spectral and multiwavelength analysis of this new transient
will be presented in a separate paper (Soldi et al. in prep.; but see
also Soldi et al. 2014).
Another faint X-ray transient, XMM J174457-2850.3, is detected in two 2012 observations; obsid: 0694641101 - 0694640301.
The observed 2-10 keV flux is 1.1 ± 0.3 and 2.9 ± 0.6 ×
10−13 erg cm−2 s−1 , respectively. A power-law fit to the spectrum
with the photon index fixed to the value reported by Sakano et al.
(2005) yields a column density of NH = (1.4 ± 0.4) × 1023 cm−2 .
This source was discovered in 2001 by Sakano et al. (2005) who re-
ported a tentative detection of an X-ray pulsation of ∼ 5 s, during
the ∼ 25 ks XMM-Newton observation. Both a visual inspection
and timing analysis of the X-ray light curve show no evidence for
bursts and/or dips. However, even considering the 4 times longer
exposure of the new data, we cannot exclude or confirm the ∼ 5 s
periodic modulation because of the lower flux observed. In fact,
during the 2001 XMM-Newton observation (obsid: 0112972101)
XMM J174457-2850.3 had a flux about 10 − 40 times higher
(∼ 45 × 10−13 erg cm−2 s−1 ) than in 2012 (in quiescence XMM
J174457-2850.3 has a typical 2-10 keV flux lower than 0.2×10−13
erg cm−2 s−1 ).
Only upper limits are measured for the other well-known
X-ray transients within the field of view. The two bursters
GRS 1741.9-2853 and AX J1745.6-2901 (Sakano et al. 2002;
Trap et al. 2009; Ponti et al. 2014; 2015) have flux limits of
F2−10keV < 2 × 10−14 erg cm−2 s−1 and F2−10keV < 10−13
erg cm−2 s−1 (obsid: 0694641101, 0694640301), respectively.
Closer to Sgr A? , we find an upper limit on the 2-10 keV flux
of F2−10keV < 5 × 10−12 erg cm−2 s−1 toward three other
sources: CXOGC J174540.0-290031, the low-mass X-ray binary
showing X-ray eclipses (Porquet et al. 2005; Muno et al. 2005),
CXOGC J174540.0-290005 (Koch et al. 2014), and the magnetar discovered on April 25, 2013 (Degenaar et al. 2013; Dwelly
& Ponti 2013; Mori et al. 2013; Rea et al. 2013; Kaspi et al. 2014;
c 0000 RAS, MNRAS 000, 1–??
13
Coti-Zelati et al. 2015), located at distances from Sgr A? of only
∼ 2.9, ∼ 23 and ∼ 2.4 arcsec, respectively.
Finally we observe that both XMMU J174554.4-285456, the
faint transient with a possible pulsation period of about 172 s (Porquet et al. 2005), and SAX J1747.7-2853, the bursting (showing
also superbursts) X-ray transient (Wijnands et al. 2002; Natalucci
et al. 2004; Werner et al. 2004), have flux limits of F2−10keV <
2 × 10−13 erg cm−2 s−1 .
Bright sources outside the 2012 scan
Three very bright sources are outside the field of view during
the 2012 CMZ scan, however they imprint their presence through
bright stray-light arcs. The arc features between and south of the
Sgr A and C complexes (l ∼ 359.6 − 359.9◦ , b ∼ −0.15 − 0.4◦ )
testify that the bright X-ray burster 1A 1742-294 (Belanger et al.
2006; Kuulkers et al. 2007) was active during the 2012 XMMNewton campaign. The very bright arcs east of the Sgr D complex
(obsid: 0694641601) are most probably produced by the very bright
neutron star low-mass X-ray binary GX 3+1 (Piraino et al. 2012) located about 1.18◦ northeast of the arcs3 . On the far west edge of the
2012 scan a brightening is observed. This is due to 1E 1740.7-2942
(Castro et al. 2013; Reynolds & Miller 2010; Natalucci et al. 2014),
a bright and persistent accreting microquasar, at only ∼ 1.5 arcmin
from the edge of the 2012 field of view (see Fig. 3). The lack of
straylight south of the Sgr B region suggests that the BH candidate
IGR J17497-2821 (Soldi et al. 2006; Paizis et al. 2009) was in quiescence during these observations. Two bright X-ray bursters have
been active during the 2003 XMM-Newton observation pointed to
the pulsar wind nebula called The Mouse, i.e., SLX 1744-299 and
SLX 1744-300 (Mori et al. 2005).
3.2
Very soft emission: Foreground emission
Despite the presence of distinctively soft (red) point sources, Fig. 3
shows no strong, diffuse, very soft X-ray emission. This is mainly
due to the very high column density of neutral hydrogen toward the
GC (with typical values in the range NH ∼ 3 − 9 × 1022 cm−2 ;
see also §7). Almost no Galactic centre X-ray radiation reaches us
below E <
∼1.3, 1.7 or 2.3 keV for column density values of NH '
3, 5, or 9 × 1022 cm−2 , respectively (see Fig. 2). The majority
of the ”red” sources present in the 0.5-1.5 keV band are point-like
and are associated with foreground active stars characterised by an
unabsorbed soft X-ray spectrum.
Two clearly extended and soft X-ray emitting sources are
present in Fig. 3. These correspond to Sh2-10 and Sh2-17 (Wang et
al. 2002; Dutra et al. 2003; Law et al. 2004; Fukuoka et al. 2009),
two stellar clusters located in one of the Milky Way spiral arms and
thus characterised by a lower column density of absorbing material,
consequently appearing stronger in the 1-2.5 keV range (visible in
Fig. 3 with orange colours).
3.3
Soft and hard GC emission
Galactic centre radiation with energies above ∼ 2 − 3 keV can
typically reach us and be detected (in green and blue in Fig. 3). GC
sources with a significant continuum component (e.g. power-law or
3
This region is covered only by the observations of the 2012 XMM-Newton
scan, therefore the removal of the stray-light arcs generates regions with null
exposures in the final mosaic maps (e.g., Fig. 3 and 4).
c 0000 RAS, MNRAS 000, 1–??
Bremstrahlung), such as observed from most GC point sources, the
GC stellar clusters (e.g. the Arches, the Quintuplet and the Central cluster) as well as some supernova remnants (such as SNR
G0.9+0.1 and Sgr A East) appear with a bright light blue colour.
Colour gradients confirm the presence of at least two components of the diffuse emission, each having a different spatial distribution (see Figs. 3 and 4). One component dominates the emission
in the soft and medium energy bands, thus appearing with a distinctively green colour. Its distribution appears to be very patchy,
peaking typically at the position of known supernovae remnants.
Another, harder component appears with a dark blue colour (Fig.
3). This harder emission is known to consist of at least two separate
contributions. One, which is associated with intense high-ionisation
Fe K lines, is smoothly distributed and peaks right at the GC; it is
likely produced by faint point sources (Muno et al. 2004; Revnivtsev et al. 2009; Heard & Warwick 2013a). The other, which is associated with neutral Fe K emission lines, has a patchy distribution
peaking at the position of molecular cloud complexes; it is likely
due to an ensemble of X-ray reflection nebulae (see Ponti et al.
2013 for a review). We also note that the Galactic plane emission
is dominated by dark blue colours (Fig. 3), while regions located
◦
◦
at b>
∼0.2 and b<
∼ − 0.35 have a significantly greener colour. We
address this in more detail in § 5, 7 and 8.7.
4
SOFT LINE EMISSION
Figure 2 shows the spectra of the diffuse emission from the regions
marked in magenta in Fig. 6. The ∼ 1.5 to ∼ 5 keV band shows
strong, narrow emission lines, the strongest of which are Si XIII, S
XV , Ar XVII and Ca XIX . This line emission, as well as the underlying continuum and the intra-line emission, are typically well fitted
by a thermal model (e.g. APEC in X SPEC) with temperatures in the
range 0.6 − 1.5 keV (Kaneda et al. 1997; Tanaka et al. 2000; Muno
et al. 2004; Nobukawa et al. 2010; Heard & Warwick 2013b). At
higher energies a power-law component with intense Fe XXV and
Fe XXVI lines is also observed over the entire GC region. Additionally, neutral Fe Kα and Kβ lines are also observed. The neutral
Fe K emission lines are associated with different processes, therefore they will be the focus of separate publications.
4.1
RGB images of soft emission lines
The top and bottom panels of Fig. 10 show the line (continuum
non-subtracted) RGB image and the inter-line continuum RGB image (see caption of Fig. 10 and Tab. 1 for more details). We note
that the soft X-ray line image shows very strong colour gradients
(less dramatic colour variations are observed in the continuum image). In particular, the sources DS1 (the core of Sgr D), the western part of Sgr B1 (i.e., G0.52-0.046, G0.570-0.001), Sgr C, as
well as the Chimney above it, all have a distinctively green-blue
colour, while G359.12-0.05, G359.10-0.5, G359.79-0.26, G359.730.35 and the entire G359.77-0.09 superbubble are characterised by
orange-brown colours. G0.1-0.1, the Radio Arc, the arched filaments (see Fig. of Lang et al. 2002), G0.224-0.032, and G0.40-0.02
are also characterised by red-brown colours, however here a gradation of white and green is also present (please refer to Tab. 3 and
4 and Figs. 5 and 6 for the positions of the regions listed here).
The lobes of Sgr A appear with a whiter colour than the surroundings. In addition, we observe bright red-brown emission along two
broad, linear ridges having relatively sharp edges to the northwest
14
G. Ponti et al.
Red: Si xiii
Green: S xv
Blue: Ar xvii
0.400
Galactic latitude
0.200
0.000
-0.200
-0.400
-0.600
1.500
1.000
0.500
0.000
359.500
359.000
0.000
359.500
359.000
Galactic longitude
Red: Si-S
Green: S-Ar
Blue: Ar-Ca
0.400
Galactic latitude
0.200
0.000
-0.200
-0.400
-0.600
1.500
1.000
0.500
Galactic longitude
Figure 10. (Top panel) Soft lines (continuum unsubtracted) RGB image of the CMZ. The Si XIII line emission is shown in red, S XV in green and Ar XVII+ Ca
XIX in blue. (Bottom panel) RGB image of the energy bands between soft emission lines. The Si-S band emission is shown in red (see Tab. 1 for a definition of
the energy bands), S-Ar in green and Ar-Ca + Blue-Ca in blue. The diffuse emission in this inter-line continuum is very similar to the soft emission line one,
suggesting that the same process is producing both the lines and the majority of the soft X-ray continuum. The colour variations within the map are modulated
primarily by abundance variations (top), temperature of the emitting plasma, continuum shape and absorption.
and northeast of Sgr A? . This latter feature is discussed in detail in
section 8.4.
In spite of the fact that these images show different components (one being dominated by emission lines, the other by continuum emission), they are remarkably similar. No clear diffuse emission component is present in one and absent from the other image.
This indicates that most of the diffuse soft X-ray continuum and
line emission are, indeed, produced by the same process. In addition, the differences in the ratio between photons emitted in the
lines and in the continuum can add valuable information for understanding the radiative mechanism. In fact, such differences could be
due, for example, to different cosmic abundances and/or variations
in the relative contributions of various thermal and nonthermal radiation mechanisms. In order to better highlight these differences, we
map the sum of the interline continua in the same image (see caption of Fig. 11 and Tab. 1). As expected, none of the point sources is
a strong soft line emitter (they in fact appear brighter in the interline
image). We also note that the intense soft X-ray emitting regions in
the Galactic plane, such as Sgr D, Sgr B1, Sgr C, the Chimney and
G359.9-0.125 are characterised by distinctively orange-red colours,
indicating they are strong line emitters.
For an alternative perspective, the bottom panel of Fig. 11
shows the Si XIII + Si-S bands in red, S-Ar + S XV + Ar-Ca in
green, and CFeK in blue. These energy bands are chosen to highc 0000 RAS, MNRAS 000, 1–??
15
Red: Si xiii + S xv + Ar xvii + Ca xiv
Green: Si-S + S-Ar + Ar-Ca + B-Ca
Blue: 4.5-12 keV
0.400
Galactic latitude
0.200
0.000
-0.200
-0.400
-0.600
1.500
1.000
0.500
0.000
359.500
359.000
0.000
359.500
359.000
Galactic longitude
Red: Si xiii + Si-S
Green: S-Ar + Ar xvii + Ar-Ca
Blue: 5-6.1 keV
0.400
Galactic latitude
0.200
0.000
-0.200
-0.400
-0.600
1.500
1.000
0.500
Galactic longitude
Figure 11. (Top panel) RGB image composed of summed, continuum-subtracted line emission (Si XIII + S XV + Ar XVII + Ca XIX) in red, the sum of the
interline continua (Si-S + S-Ar + Ar-Ca + Blue-Ca) in green and the CFeK emission in blue (see Tab. 1). (Bottom panel) RGB image, in red the Si XIII +
Si-S emission, in green the S-Ar + S XV + Ar-Ca and in blue the CFeK emission.
light any energy dependence in the soft X-ray emission that could
be due to column density variations of the obscuring matter or temperature fluctuations of the emitting gas. In fact, the softer energy
bands (Si XIII + Si-S) will be more affected by absorption or low
temperature plasma emission compared to the medium (S-Ar + S
XV + Ar-Ca) or high energy bands. We defer the detailed discussion of the features present in these images to the discussion of the
various physical components presented in § 8 and subsections.
4.2
Continuum subtracted soft emission line maps and
profiles
From top to bottom, the panels of Fig. 12 show the continuumsubtracted Si XIII, S XV, Ar XVII, Ca XIX intensity maps. Although
c 0000 RAS, MNRAS 000, 1–??
the continuum subtraction procedure should naturally remove the
emission from the line-free point sources, small fluctuations in the
continuum subtraction, in the case of the brightest sources, sometimes leave significant residuals. For this reason, we have masked
out the brightest point sources in our computation of these maps.
The different curves of Fig. 13 show the continuum-subtracted
emission profiles (integrated over latitude from the magenta rectangle in Fig. 12) for the individual soft emission lines. The same four
line profiles are compared in Fig. 13 with similar profiles in which
the contribution of specific bright structures has been removed.
16
G. Ponti et al.
0.000
Si xiii
1.000
0.500
0.000
359.500
S xv
Ar xvii
Ca xix
Figure 12. From top to bottom, continuum subtracted Si XIII, S XV, Ar XVII, Ca XIX intensity maps of all the stacked XMM-Newton observations of the CMZ.
5
SPECTRAL DECOMPOSITION
In order to better trace the relative contributions of the diffuse
thermal (soft and hot) and non-thermal components, we have performed a simple component separation using a list of images depicting various energy bands. We use a total of 17 energy bands:
11 for the continuum4 and 6 for the lines (tracing Si XIII, S XV,
Ar XVII, Ca XIX, Fe Kα and Fe XXV, see Tab. 1). This treatment of
the data allows us to be more confident about the spectral decomposition, e.g. compared to single RGB maps, retaining most of the
morphological information on sufficiently large scales (i.e. beyond
few arcmin scales).
The general assumption is that the emission at any position
4
The eleven continuum energy bands used are: 1.0–1.5 keV; 1.5–1.8 keV;
2.0–2.35 keV; 2.55–3.05 keV; 3.25–3.75 keV; 3.95–4.70 keV; 4.70–5.40
keV; 5.40–6.30 keV; 6.50–6.60 keV; 6.80–7.80 keV and 8.20–9.50 keV.
can be represented by the linear sum of three main components,
namely i) a soft plasma with a temperature of 1 keV (Kaneda et al.
1997; Bamba et al. 2002); ii) a hot plasma of temperature 6.5 keV;
and iii) a non-thermal component modeled by an absorbed powerlaw plus a neutral, narrow iron line (with 1 keV equivalent width),
that are subject to an additional absorbing column of NH = 1023
cm−2 . All three components are also absorbed by gas in front of the
GC region and both thermal plasmas are modelled using the APE
model in XS PEC. The resulting model is therefore PHABS ( APEC +
APEC + PHABS ( POWERLAW + G AUSS ))) and has only three free
parameters: the relative normalizations of the three components.
The hot plasma component represents the emission associated with
faint unresolved point sources, whose cumulative spectrum is well
described by a thermal spectrum (Revnivtsev et al. 2009) plus a
possibly truly diffuse hot plasma component (Koyama et al. 2007).
The spectral index of the non-thermal component is assumed to be
Γ = 2, consistent with the values measured through the combined
c 0000 RAS, MNRAS 000, 1–??
17
Figure 13. Longitudinal intensity profiles of the Si XIII (red), S XV (green),
Ar XVII (blue) and Ca XIX (violet) emission lines, integrated over Galactic
latitude within the magenta rectangular region shown in Fig. 12.
spectral fits of XMM-Newton spectra with higher energy data (e.g.,
INTEGRAL and/or NuSTAR; Terrier et al. 2010; Mori et al. 2015;
Zhang et al. 2015).
The strongest assumptions in this approach are that the emission can be represented everywhere with these three components.
This obviously fails on bright point sources or on regions where
the emission is much hotter (e.g. Sgr East). For the soft components the even stronger assumption is that absorption to the GC is
assumed to be uniform over the CMZ at a value of NH = 6 × 1022
cm?2 (Sakano et al. 2002; Ryu et al. 2009), putting aside absorption
in the GC region itself. Clear column density modulations are observed towards different lines of sight (e.g. Ryu et al. 2009; Ryu et
al. 2013). We tested significantly different column densities (up to
NH = 1.5 × 1023 cm−2 characteristic of several GC sources; see
e.g. Baganoff et al. 2003; Rea et al. 2013; Ponti et al. 2015). We
found that if the soft plasma normalization is significantly modified, the overall morphology is consistent. We tested various values
of the other parameters (spectral index or temperatures) and did not
find strong effects on the soft plasma morphology or normalization.
We first produced counts, exposure and background maps
for each observation and each instrument. Background was obtained from cal-closed datasets distributed in the ESAS5 calibration
database. For each observation and instrument, an average RMF is
computed as well as the un-vignetted ARF. For each instrument,
individual observation images were reprojected using the final image astrometry and then combined to compose a mosaic. Average
6
ARF and RMF for each instrument were obtained with the FTOOLS
routines ADDRMF and ADDARF.
For each pixel of the final maps, we fit the measured numbers of counts in all the energy bands and instruments with a model
consisting of the three aforementioned components as well as the
background events number and the Out-of-Time (OoT) events for
the EPIC-pn camera. The free parameters are the normalization of
each individual component. We apply Cash statistics (Cash 1979)
to take into account the low statistics in each pixel. This analysis
5
6
http://xmm2.esac.esa.int/external/xmm sw cal/background/epic esas.shtml
https://heasarc.gsfc.nasa.gov/ftools/ftools menu.html
c 0000 RAS, MNRAS 000, 1–??
NH
(1022 )
F2−4.5 /F 0
F4.5−10 /F 0
0.01
1
3
5
7
10
15
30
50
70
100
150
1
0.776
0.488
0.322
0.223
0.137
0.0698
0.0156
0.0033
0.0008
1.2 × 10−4
6.2 × 10−5
1
0.968
0.912
0.857
0.809
0.737
0.636
0.414
0.244
0.149
0.0754
0.0273
Table 5. Expected ratio of the obscured flux to the un-obscured flux for
different values of the column density of obscuring material. A thermally
emitting gas with temperature of kT = 1 keV (PHABS * APEC model) is assumed in the computations. The predicted observed flux in both the 2 − 4.5
and 4.5 − 10 keV bands (F2−4.5 and F4.5−10 ) is computed and compared
to the respective un-absorbed (F 0) flux.
allows us to perform a rough spectral decomposition, better separating the spectral emission components, although retaining the
maximum spatial resolution.
Figure 14 presents the map of the normalisation (in units of
10−4 times the APEC normalisation) of the soft thermal emission
component. The normalisation of the soft thermal component has a
distribution similar to the one traced by the soft lines and the continuum (Fig. 3, 10, 11 and 12). Enhanced high-latitude soft plasma
emission is observed. The white dashed lines show the position of
two sharp edges in the distribution of this high latitude emission
(see also Fig. 3, 10 and 11). The white solid line shows the edge of
the region having more than 7.2 ks of exposure (see Fig. 1).
6
AN ATLAS OF DIFFUSE X-RAY EMITTING
FEATURES
The patchy and non-uniform distribution of the diffuse emission
makes the recognition of the shape, the border and connection of
the different structures and components difficult. Occasionally, different works report the same X-ray feature with different names
and shapes and, in extreme cases, the same X-ray emitting feature
is associated with different larger scale complexes.
In Tab. 3 and 4 we report all the new X-ray features
discussed in this paper, plus many GC features presented in
previous works. The main purpose of these tables is to provide a first step towards the building of an atlas of diffuse X-ray emitting GC features. The table is available online
at: WWW. MPE . MPG . DE /HEG/GC/ATLAS GC DIFFUSE X- RAY and
will be updated, should the authors be notified of missing extended
features. This exercise is clearly prone to incompleteness and deficiencies, however we believe this might help in providing a clearer
and more systematic picture of the diffuse X-ray emission from
the GC region. The spatial location and size of all these features is
shown in the finding charts in Fig. 5 and 6.
18
G. Ponti et al.
0.500
0.400
Galactic latitude
0.300
0.200
0.100
0.000
-0.100
-0.200
-0.300
-0.400
1.000
0.500
0.000
359.500
359.000
Galactic longitude
0.14
0.145
0.154
0.174
0.212
0.289
0.442
0.747
1.36
2.58
5
Figure 14. Map of the normalisation of the soft thermal gas component (in units of 10−4 times the APEC normalisation). The white lines indicate the extent of
the survey having more than 7.2 ks exposure. The white dashed lines show the position of the two sharp edges in the distribution of the high latitude plasma.
Some bright point sources (i.e., 1E 1743.1-2843, AX J1745.6-2901, 1E 1740.7-2942, GRS 1741.9-2853) have been removed, thereby producing artificial holes
in the maps at their respective locations.
7
THE FOREGROUND COLUMN DENSITY
Given the high column densities of neutral or weakly ionized material absorbing the soft X-ray radiation, it is important to estimate
the effects of X-ray obscuration. For example, a molecular complex
having a column density of NH ∼ 1025 cm−2 , such as the Sgr B2
core, would completely obscure the radiation below about 4 keV, if
placed in front of the GC; see Fig. 2.
To calculate the effects of absorption of the X-ray emission,
we computed the flux generated by a thermally emitting plasma
with temperature of kT = 1 keV (using a PHABS * APEC model), in
both the 2 − 4.5 and 4.5 − 10 keV bands, after being absorbed by a
given column density of neutral material (see also Fig. 2). For each
column density explored, we report in Tab. 5 the ratio of the observed flux (F2−4.5 and F4.5−10 ) over the respective un-absorbed
(F 0) flux. We note that the hard X-ray band starts to be affected
(corresponding to flux reductions up to a factor of 2) for column
densities up to NH ∼ 3 × 1023 cm−2 while it is heavily affected
(flux reduction of a factor of 10 or more) for NH ∼ 1024 cm−2
or higher (see Tab. 5). At lower energies, the obscuration effect is
even more pronounced. Already, for NH ∼ 3 × 1022 cm−2 , the
23
−2
observed flux is less than half and for NH >
∼ 5 × 10 cm it is
less than 0.1 % of its un-obscured flux. This indicates that the softer
band is expected to be heavily affected by absorption.
density estimated from the dust has large uncertainties that can be
mainly ascribed to the uncertainty associated with the dust-to-NH
ratio. In particular, the column densities shown in this map appear
to be systematically larger than what is measured with other methods. For example, the column density of the G0.11-0.11 massive
cloud is estimated to be NH ∼ 5 − 6 × 1023 cm−2 in this map,
while Amo-Baladron et al. (2009) measure NH ∼ 2 × 1022 cm−2 ,
through a detailed modelling of the molecular line emission. The
core of Sgr B2 is estimated by Molinari et al. (2011) to have
NH ∼ 3 × 1025 cm−2 , while modelling of the X-ray emission
(Terrier et al. 2010) suggests NH ∼ 7 × 1023 cm−2 , more than an
order of magnitude lower. Moreover, the average column densities
of G0.40-0.02, G0.52-0.046, G0.57-0.018 and a fourth region (the
magenta ellipse in Fig. 17) are estimated to be NH ∼ 4 × 1023 ,
4 × 1023 , 1.2 × 1024 and 1.5 × 1024 cm−2 , respectively, from
the dust map, while they are measured to be in the range NH ∼
7 − 10 × 1022 cm−2 , from modelling of the X-ray emission. Therefore, the total normalisation of the NH map built from the dust
emission appears to be overestimated. However, the method employed to produce it does not suffer from self-absorption, so it is
presumably giving unbiased relative NH ratios.
7.2
7.1
Column density distribution
The top panel of Fig. 15 shows the neutral Hydrogen column density distribution as derived from dust emission (Molinari et al.
2011)7 . The image shows the NH distribution in logarithmic scale
in the range NH = 4.5×1022 −3.8×1025 cm−2 . This total column
7
We do not show the entire CMZ, because of the limited coverage of the
Herschel dust emission map (Molinari et al. 2011).
X-ray emission modulated by absorption
The bottom panel of Fig. 15 shows the X-ray map with the columndensity contours overlaid for comparison. We observe that, as expected, no soft X-ray emission is observed toward the central part
of the most massive molecular cores. In particular we observe depressed X-ray emission from: i) the Sgr B2 nucleus and its envelope (with NH > 7 × 1023 cm−2 ); ii) the almost perfect coincidence between the hole in soft X-ray emission east of G0.2240.032 (see Fig. 5 and 17) and the shape of the so-called ”Brick”
molecular cloud, M0.25+0.01 (see Fig. 5; Clark et al. 2013); iii)
c 0000 RAS, MNRAS 000, 1–??
19
0.400
0.300
Galactic latitude
0.200
0.100
0.000
-0.100
-0.200
-0.300
-0.400
-0.500
0.800
0.600
0.400
0.200
0.000
359.800
359.600
359.400
Galactic longitude
Figure 15. Top panel: Neutral Hydrogen column density distribution as derived from dust emission (Molinari et al. 2011). The image shows the NH distribution in logarithmic scale from NH = 4.5 × 1022 up to 3.8 × 1025 cm−2 . The green, magenta and white contour levels correspond to NH = 1.5 × 1023 ,
7 × 1023 and 1.5 × 1024 cm−2 , respectively. Bottom panel: X-ray continuum RGB map (Fig. 3) with the column-density contours overlaid.
c 0000 RAS, MNRAS 000, 1–??
20
G. Ponti et al.
BRIGHT AND TRANSIENT POINT SOURCES
Source name
Coordinates‡
Flux†
within the 2012 CMZ scan
1E 1743.1-2843
XMMU J174505.3-291445
XMMU J174457-2850.3
GRS 1741.9-2853
AX J1745.6-2901
CXOGC J174540.0-290031
SGR J1745-2900
XMMU J174554.4-285456
SAX J1747.7-2853
CXOCG J174540.0-290005
0.2608,-0.0287
359.6756,-0.0634
0.0076,-0.1743
359.9528,+0.1202
359.9203,-0.0420
359.9435,-0.0465
359.9441,-0.0468
0.0506,-0.0429
0.2073,-0.2385
359.9497,-0.04269
within the total GC scan
1E 1740.7-2942
1A 1742-294
IGR J17497-2821
GRO J1744-28
XMMU J174654.1-291542
XMMU J174554.4-285456
SLX 1744-299
SLX 1744-300
359.1160,-0.1057
359.5590,-0.3882
0.9532,-0.4528
0.0445,+0.3015
359.8675,-0.4086
359.1268,-0.3143
359.2961,-0.8892
359.2565,-0.9111
110
14
0.3
< 0.02
< 0.1
<5
<5
< 0.2
< 0.2
<5
References
90,92,93
94
90,95
90,59,97
59,90,91,124
98,99
101,102,103,104
98
90,105,106,107
100
111,112,113
90,108,109
90,114,115
90
90
84,85,86
37,59,87,88,89
37,59,87,88,89
Table 2. List of bright and transient point sources during the 2012 XMM-Newton scan as well as bright point sources observed in all scans of the region (see
Fig. 5). To avoid exessive crowding around Sgr A? , CXOCG J174540.0-290005 and SGR J1745-2900 are not shown. †Fluxes are given in units of 10−12 erg
cm−2 s−1 and correspond to the mean flux observed during the 2012 XMM-Newton scan of the CMZ. ‡Coordinates are in Galactic format. References: (1)
Wang et al. 2006a; (2) Yusef-Zadeh et al. 2002; (3) Capelli et al. 2011; (4) Tatischeff et al. 2012; (5) Sakano et al. 2003; (6) Habibi et al. 2013; (7) Habibi et
al. 2014; (8) Krivonos et al. 2014; (9) Clavel et al. 2014; (10) Dutra et al. 2003; (11) Law et al. 2004; (12) Fukuoka et al. 2009; (13) Wang et al. 2002a; (14)
Tsuru et al. 2009; (15) Mori et al. 2008; (16) Mori et al. 2009; (17) Heard & Warwick 2013a; (18) Maeda et al. 2002; (19) Park et al. 2005; (20) Koyama et al.
2007a; (21) Kassim & Frail 1996; (22) Nobukawa et al. 2008; (23) Senda et al. 2002; (24) Renaud et al. 2006; (25) Mereghetti et al. 1998; (26) Gaensler et al.
2001; (27) Porquet et al. 2003a; (28) Aharonian et al. 2005; (29) Dubner et al. 2008; (30) Nobukawa et al. 2009; (31) Sawada et al. 2009; (32) Morris et al.
2003; (33) Morris et al. 2004; (34) Markoff et al. 2010; (35) Zhang et al. 2014; (36) Nynka et al. 2013; (37) Gaensler et al. 2004; (38) Pedlar et al. 1989; (39)
Cotera et al. 1996; (40) Figer et al. 1999; (41) Johnson et al. 2009; (42) Lu et al. 2008; (43) Lu et al. 2003; (44) Yusef-Zadeh et al. 2005; (45) Baganoff et al.
2003; (46) Ho et al. 1985; (47) Bamba et al. 2002; (48) LaRosa et al. 2000; (49) Morris & Yusef-Zadeh 1985; (50) Lang et al. 1999; (51) Anantharamaiah et
al. 1991; (52) Yusef-Zadeh & Morris 1987a; (53) Yusef-Zadeh & Morris 1987b; (54) Yusef-Zadeh & Morris 1987c; (55) Muno et al. 2008; (56) Uchida et al.
1992; (57) Predehl & Kulkarni 1995; (58) Senda et al. 2003; (59) Sakano et al. 2002; (60) Coil et al. 2000; (61) Murakami 2002; (62) Yusef-Zadeh et al. 2007;
(63) Dutra & Bica 2000; (64) Zoglauer et al. 2014; (65) Koyama et al. 2007b; (66) Nakashima et al. 2010; (67) Downes & Maxwell 1966; (68) Tanaka et al.
2009; (69) Tanaka et al. 2007; (70) Wang et al. 2006b; (71) Wang et al. 2002b; (72) Phillips & Marquez-Lugo 2010; (73) Hewitt et al. 2008; (74) Reich &
Fuerst 1984; (75) Gray 1994; (76) Roy & Bhatnagar 2006; (77) Marquez-Lugo & Phillips 2010; (78) Borkowski et al. 2013; (79) Yamauchi et al. 2014; (80)
Inui et al. 2009; (81) Green 2014; (82) Yusef-Zadeh et al. 2004; (83) Nord et al. 2004; (84) Uchiyama et al. 2011; (85) Heinke et al. 2009; (86) Muno et al.
2006; (87) Mori et al. 2005; (88) Skinner et al. 1990; (89) Pavlinski et al. 1994; (90) Degenaar et al. 2012; (91) Ponti et al. 2014; (92) Porquet et al. 2003b;
(93) Del Santo et al. 2006; (94) Soldi et al. 2014; (95) Sakano et al. 2005; (97) Trap et al. 2009; (98) Porquet et al. 2005a; (99) Muno et al. 2005b; (100) Kock
et al. 2014; (101) Degenaar et al. 2013; (102) Dwelly & Ponti 2013; (103) Rea et al. 2013; (104) Kaspi et al. 2014; (105) Wijnands et al. 2002; (106) Natalucci
et al. 2004; (107) Werner et al. 2004; (108) Belanger et al. 2006; (109) Kuulkers et al. 2007; (110) Piraino et al. 2012; (111) Castro et al. 2013; (112) Reynolds
& Miller 2010; (113) Natalucci et al. 2014; (114) Soldi et al. 2006; (115) Paizis et al. 2009 (116) Lu et al. 2013; (117) Do et al. 2013; (118) Yelda et al. 2014;
(119) Bamba et al. 2000; (120) Bamba et al. 2009; (121) Ohnishi et al. 2011; (122) Zhao et al. 2013; (123) Hales et al. 2009; (124) Ponti et al. 2015.
the core of the Sgr C complex8 ; iv) the regions around DB-58 and
at Galactic position l ∼ 0.2, b ∼ −0.48◦ also appear to have darker
colours and, once again, it is possible to find molecular complexes
(M0.018+0.126 and M0.20-0.48) covering roughly the same region
(see Fig. 5 and 15). All these clouds are characterised by very high
23
−2
column densities NH >
∼3 − 7 × 10 cm and they most probably lie in front of Sgr A? and of most of the GC (e.g., according to
the twisted ring model of Molinari et al. 2011). Therefore, they are
absorbing the GC’s extended, soft X-ray emission.
All this evidence suggests that at least the most massive clouds
8
At this location a sharp transition in the soft X-ray emission, with an
arc-like shape, is observed. This is spatially coincident to the edge of a very
dense core of dust, suggesting that the modulation in the soft X-ray emission
is induced by obscuration by the molecular cloud.
located in front of the GC do actually modulate (obscure) the soft
X-ray emission. However, fluctuations in column densities cannot
be the only cause for the observed distribution of soft X-ray emission for two reasons. First, we do observe only weak soft X-ray
emission along some lines of sight having a low column density
of molecular material (such as around the Sgr C and Sgr D complexes and south of the Sgr B1 region). Second, we do detect intense (among the brightest) soft X-ray emission from several regions such as the Sgr A complex and the cores of the Sgr C and
Sgr D complexes, where some of the highest column-density clouds
are found. In particular, in Sgr A very intense soft X-ray emission is
observed along the line of sight toward the 50 km s−1 , the Bridge
(Ponti et al. 2010) and the G0.11-0.11 clouds, some of the highest column density clouds in the CMZ. Although this might be
explained by placing these clouds on the far side of the CMZ, it
c 0000 RAS, MNRAS 000, 1–??
21
ATLAS OF DIFFUSE X-RAY EMITTING FEATURES
Name
STAR CLUSTERS:
Central cluster
Quintuplet
Arches
Sh2-10[†
Sh2-17[†
DB-05[†
Other name or
associated features
Coordinates
(l, b)
Size
arcmin
References
G0.12+0.02
DB-6
DB-58
G0.33-0.18
359.9442, -0.046
0.1604, -0.0591
0.1217, 0.0188
0.3072,-0.2000
0.0013, 0.1588
0.31 -0.19
0.33
0.5
0.7
1.92
1.65
0.4
45,116,117,118
1,63,11
1,2,3,4,5,6,7,8,9,39,40,11
10,11,12,63,11
13,63,11
22,63,11
359.03,-0.96
359.07,-0.02
359.12,-0.05
359.10,-0.51
359.41,-0.12
359.46,+0.04
359.73,-0.35
359.77,-0.09
359.84,-0.14
359.79,-0.26
0.00,-0.16
359.94, -0.04
359.93,-0.09
359.963, -0.053
0.109,-0.108
0.13,-0.12
0.224,-0.032
0.34,+0.045
26 × 20
22 × 10
24 × 16
22 × 22
3.5 × 5.0
6.8 × 2.3
4
20 × 16
15 × 3
8 × 5.2
5.88
1
3.2 × 2.5
13.6 × 11
3×3
2.3 × 4.6
14 × 8.8
X-R 48,51,75,76,81,119,120
R 14,48,51,66
X 66
X-R 37,48,51,56,74,75,81,120,121
X 14
X 14
X 58
X 15,16,17,58
X 15,16,17,58
X 15,16,17,58
X This work
X-R 32,33,34,17
R 35,38,43,47,58,60,61
X-R 5,18,19,20,48,75,81
X This work
X 17
X This work
R 21,48,51,81,82
0.40,-0.02
4.7 × 7.4
X 22
0.519,-0.046
0.57,-0.001
0.570,-0.018
0.61,+0.01
0.867,+0.073
1.17,+0.00
1.02,-0.17
2.4 × 5.1
1.5 × 2.9
0.2
2.2 × 4.8
7.6 × 7.2
3.4 × 6.9
10 × 8.0
This work
This work
X 23,24,58,59,68,80
X 22,65,79
R 25,26,27,28,29,48,75,81,82
X 31
R 30,31,48,51,75,77,81,82
10 × 10
R 73,81,82
SNR and Super-bubbles candidates:
G359.0-0.9†††
G358.5-0.9 - G359.1-0.9
G359.07-0.02
G359.0-0.0
G359.12-0.05
G359.10-0.5†††
G359.41-0.12
Chimney[†
G359.73-0.35†‡
G359.77-0.09
Superbubble
G359.9-0.125
G359.79-0.26\
G0.0-0.16†\
20 pc lobes
G359.92-0.09‡
Parachute - G359.93-0.07
Sgr A East
G0.0+0.0
G0.1-0.1
Arc Bubble
G0.13,-0.11[
G0.224-0.032
G0.30+0.04
G0.3+0.0
G0.34+0.05
G0.33+0.04
G0.42-0.04
Suzaku J1746.4-2835.4
G0.40-0.02
G0.52-0.046
G0.57-0.001
G0.57-0.018†
CXO J174702.6-282733
G0.61+0.01†
Suzaku J1747.0-2824.5
G0.9+01♥
SNR 0.9+0.1
DS1
G1.2-0.0
Sgr D SNR
G1.02-0.18
G1.05-0.15
G1.05-0.1
G1.0-0.1
G1.4-0.1
1.4,-0.10
Table 3. Atlas of diffuse X-ray emitting features. The first two columns in the table indicate the name primarily used in this work to refer to the feature as well
as the other names used in the previous literature. The third and fourth columns show the coordinates of each feature as well as its approximate projected size.
Finally, the fifth column provides references to selected works discussing the feature. For convenience, we report in Tab. 2 all the references ordered according
to the numbering used in this table. The other names column shows the different designations used in previous literature. In the case of bubbles, these features
are not necessarily referring to the same structure but to features forming the bubble candidate. †Possibly due to a thermal filament. ‡The interpretation as a
SNR is probably obsolete. †‡ Most probably a foreground feature. \ This feature appears to be part of the superbubble G359.77-0.09. †\ New extended X-ray
feature, possibly part of the superbubble G359.77-0.09. [ This feature appears to be part of the Arc bubble. Possibly connected to G0.61+0.01. ♥ X-ray
emission primarily non-thermal, therefore it appears also in the next table. †† New extended X-ray feature, possibly part of the superbubble G359.77-0.09. [†
The low X-ray absorption towards these star clusters indicate that they are located in front of the GC region. ††† Because of the low X-ray absorption column
density (NH ∼ 2 × 1022 cm−2 ) this is most probably a foreground source (Bamba et al. 2000; 2009). [† The Chimney is most probably either part of a large
scale structure (see §8.7) or an outflow from G359.41-0.12 (Tsuru et al. 2009), therefore most probably it is not a separate SNR.
appears that these regions are characterised by truly enhanced soft
X-ray emission (see e.g. §8.4, 8.5 and 8.6). We defer the detailed
disentangling of these effects to an elaborate spectral study of these
regions.
8
DISCUSSION
In the process of systematically analysing all XMM-Newton observations of the central degrees of the Galaxy, we have discovered several new extended features and have produced an atlas of
known, extended soft X-ray features. Here we discuss their general
properties and investigate the origin/existence of several specific
features.
c 0000 RAS, MNRAS 000, 1–??
8.1
General properties
To compute the total observed (absorbed) flux from the CMZ
we first mask out the emission from the brightest binaries, by
excluding a circle with a radius of: 10 around SAX J1747.72853; 1.50 for AX J1745.6-2901 and GRS1741.9-2853; 20 for
XMMU J174445.5-295044; 2.50 for 1E 1743.1-2843; 3.50 for
IGR J17497-2821 and 1E 1740.7-2942 (see white circles in Fig.
5). We then compute the total observed count rate from two boxes,
one with a size of 1.5◦ × 0.35◦ (l × b) and centered on Sgr A? and
one with a bigger size of 2.08◦ × 0.413◦ centered on l = 0.232◦ ,
b = 0.080◦ . We then measure the total count rate within these re-
22
G. Ponti et al.
ATLAS OF DIFFUSE X-RAY EMITTING FEATURES
Name
Other name or
associate features
Radio and X-ray filaments and PWN candidates:
Snake
G359.15-0.2
G539.40-0.08
G359.43-0.14
Sgr C Thread
Ripple filament
G359.54+0.18
G359.55+0.16
X-ray thread
Suzaku J174400-2913
Crescent
G359.79+0.17
Curved filament
Pelican
G359.85+0.47
Cane
G359.87+0.44
G359.85+0.39
Sgr A-E
G359.889-0.081- wisp
XMM J174540-2904.5
G359.89-0.08
G359.897-0.023
G359.899-0.065
Sgr A-F
G359.90-0.06
G359.904-0.047
G359.915-0.061
G359.91-1.03
G359.921-0.030
F7
G359.921-0.052
The Mouse
G359.23-0.82
G359.925-0.051
G359.933-0.037
F2
G359.933-0.039
F1
G359.941-0.029
G359.942-0.045
G359.944-0.052
G359.945-0.044
G359.95-0.04
G359.956-0.052
G359.959-0.027
F5
Southern thread
G359.96+0.09
359.96+0.09
G359.962-0.062
G359.964-0.053
F3
G359.965-0.056
F4
G359.969-0.033
G359.970-0.009
F8
G359.971-0.038
F6
G359.974-0.000
F9
G359.977-0.076
Cannonball
J174545.5-285829
G359.983-0.040
G359.98-0.11
G0.007-0.014
G0.008-0.015
G0.014-0.054
G0.017-0.044
MC2
G0.02+0.04
G0.021-0.051
G0.029-0.08
G0.032-0.056
G0.029-0.06 - F10
G0.03-0.06
G0.039-0.077
G0.062+0.010
G0.06+0.06
Northern thread
G0.09+0.17
G0.08+0.15
G0.097-0.131
Radio Arc
GCRA
G0.16-0.15
G0.116-0.111
G0.13-0.11
G0.15-0.07
Steep spectrum of Radio Arc
XMM J0.173-0.413
G0.17-0.42
S5
G0.223-0.012
G0.57-0.018†
CXO J174702.6-282733
G0.61+0.01†
G0.9+0.1
Coordinates
(l, b)
Size
arcsec
References
359.15,-0.17
359.40,-0.08
359.43,-0.14
359.45,-0.01
359.548,+0.177
359.55,+0.16
312 × 54
27.5 × 5.1
21.4 × 3.9
500 × 42
320 × 55
56.1 × 8.0
R 48
X 41
X 41
R 48,51
R 43,44,48,51
X 13,41,42,43,79
359.791,+0.16
300 × 74
R 63,50,51
359.859,+0.426
359.87,+0.44
300 × 54
420 × 50
R 48,50,51
R 48
359.889,-0.081
20 × 5
R X 5,35,41,42,43,44,50,55
359.897,-0.023
359.899,-0.065
6.4 × 4
6.5 × 2.5
X 55
X 42,44,55
359.904,-0.047
359.915,-0.061
359.919,-1.033
359.921,-0.030
359.921,-0.052
359.30-0.82
359.925,-0.051
359.934,-0.0372
359.933,-0.039
359.941,-0.029
359.942,-0.045
359.944,-0.052
359.945,-0.044
359.950,-0.043
359.956,-0.052
359.959,-0.027
359.96,+0.11
6.5 × 3
7×2
138 × 36
7.5 × 3
5.5 × 2
156 × 108
8 × 2.2
12 × 3
5×2
6×2
5×3
9 × 1.5
6 × 2.5
10 × 4
4 × 2.5
9×3
500 × 40
X 55
X 55
R 48
X 42,55
X 55
PWN F 37,48,57,123
X 55
X 41,42,55
X 42,55
X 41,55 Stellar wind
X 55
X 41,55
X 41,1,42,55 PWN
X 55,70 PWN
X 55
X 41,42,55
R 48,50,51
359.962,-0.062
359.964,-0.053
359.965,-0.056
359.969,-0.033
359.970,-0.009
359.971,-0.038
359.974,-0.000
359.977,-0.076
359.983,-0.0459
359.983,-0.040
359.979,-0.110
0.008,-0.015
0.014,-0.054
0.017,-0.044
0.0219,+0.044
0.021,-0.051
0.029,-0.08
0.0324,-0.0554
5.5 × 3.5
16 × 3.5
9×3
5×2
10 × 2.5
16 × 8
7×2
6×4
30 × 15
6.5 × 4.5
Streak
11 × 3.5
18 × 14
15 × 4
Streak
15 × 12
29 × 18
35 × 6
X 55
X 41,42,45,55 PWN
X 42,55
X 55
X 41,42,55 PWN
X 41,42,55 PWN
X 42
X 55
X-R PWN 36,122
X 42,55
R 50
X 41,55
X 55
X 41,42 FeKa
R 50
X 55
X 55
FeKa 41,42,55 PWN
0.039,-0.077
0.062,+0.010
0.09,+0.17
22 × 15
40 × 25
714 × 48
X 55
R 50,55
R 48,49,50,51
0.097,-0.131
0.167,-0.07
70 × 50
1690 × 145
X 55
R 38,48,49,50,51
0.116,-0.111
0.13,-0.11
0.138,-0.077
0.173,-0.413
0.17,-0.42
0.223,-0.012
0.57,-0.0180
0.61,+0.01
0.9,+0.1
50 × 40
55 × 12
X 55
71,17,41,42 PWN
R 50
X This work
180 × 18
912 R 82
50 × 20
0.33
132 × 288
X 41
X 23,24,58,59,68,79,80
X 22,65,79
PWN
Table 4. Atlas of diffuse X-ray emitting features. This table has the same structure as Tab. 3. Because of possible misplacements between the peak emission
of the radio and X-ray counterparts of filaments, SNR, PWN and other diffuse structures (generally related to the different ages of the population of electrons
traced at radio and X-ray bands), when available we give the best X-ray position (following the preference: Chandra, XMM-Newton, Suzaku), otherwise we
state the radio position. We cite the literature results separating thes between the X-ray (X) from the radio (R) detections. †Possibly part of a young supernova
remnant. For convenience, we report in Tab. 2 all the references ordered according to the numbering used in this table.
c 0000 RAS, MNRAS 000, 1–??
23
al. 1989; 2004; see figures 7a, 16b,c, and 17a,b,c of the latter reference, where the filament is labelled ”S5”), but because it is not
continuous or exactly parallel with the filaments of the Arc, it is
not completely evident that it is an extension of the Arc in three dimensions. The X-ray filament has a hard X-ray colour and does not
appear in the soft line images, indicating a non-thermal emission
spectrum.
Three other nonthermal radio filaments have been found
to have X-ray emission along some portion of their lengths:
G359.54+0.18, G359.89-0.08, and G359.90-0.06 (Lu, Wang &
Lang 2003; Sakano et al. 2003; Lu et al. 2008; Johnson et al. 2009;
Morris, Zhao & Goss 2014; Zhang et al. 2014). XMM 0.173-0.413
is the only one of the four known cases where the X-ray emission is
not at or near a location where the radio filament shows unusually
strong curvature.
8.3
Figure 16. XMM-Newton image in the 2-12 keV band showing the new
X-ray filament located south of the Radio Arc. The filament is located at
l ' 0.173◦ and b ' −0.413◦ and appears as a thin (< 0.15 arcmin) and
long (∼ 2.7 arcmin) filament running along the north-south direction, as
the Radio Arc. The brightest source in the image is SAX J1747.7-2853.
gions and convert it into a flux9 assuming that the soft X-ray diffuse
emission is dominated by a thermally emitting plasma with a temperature of kT = 1 keV (Kaneda et al. 1997; Bamba et al. 2002).
In particular, we assumed an APEC emission component with temperature kT = 1.08 keV, absorbed by a column density of neutral
material of NH = 6 × 1022 and with Solar abundance.
The fluxes of the integrated continuum and line intensities
(line plus continuum) within the big and small boxes are reported
in Tab. 6. This corresponds to an observed 2 − 12 keV luminosity
of L2−12 = 3.4 × 1036 erg s−1 and L2−12 = 2.6 × 1036 erg s−1
for the big and small boxes, respectively, assuming a distance of
8 kpc to the Galactic center (Reid 1993; Reid et al. 2009).
The top and bottom panels of Fig. 17 show the X-ray (2.5–4.5 keV)
and 850 µm (Pierce-Price et al. 2000) maps of the Sgr B1 region.
Four enhancements of X-ray emission are clearly present (shown
by the green dashed ellipses in Fig. 17). In particular, G0.5700.001, G0.52-0.046 and G0.40-0.02 correspond to holes in the dust
distribution derived from the 850 µm radiation (see bottom panel
of Fig. 17). To reinforce this evidence, we observe that their X-ray
edges can be traced in the dust distribution all around the X-ray enhancements, suggesting a tight connection between the two. Such
phenomenology is typical of SNe exploding within or near molecular clouds and interacting with them, creating bubbles in the matter
distribution (Ferreira & de Jager 2008; Lakicevic et al. 2014). Indeed, in this case, the SN ejecta might have cleared the entire region
that is not filled with hot, X-ray emitting plasma, pushing away the
ambient molecular material. However, this is not the only possibility. The apparent X-ray enhancements could have resulted instead
from a higher obscuration surrounding the submm holes, leading to
higher X-ray extinction at the edges. We note that none of these regions has a known radio SNR counterpart (see Fig. 17). However,
the radio emission might be confused within the very high radio
background of diffuse emission created by G0.30+0.04, the several
HII regions present in this region (see Fig. 6), and the bright, extended synchrotron background of the GC. Enhanced X-ray emission has already been reported towards G0.40-0.02 (Nobukawa et
al. 2008) and close to G0.570-0.001 (see cyan ellipses in Fig. 17).
8.3.1
8.2
A new X-ray filament, XMM 0.173-0.413
We observe a new X-ray filament extending ∼ 2.2 arcmin perpendicular to the Galactic plane and situated at l = 0.173◦ ,
b = −0.413◦ , which is almost directly toward negative longitudes
from the GC Radio Arc. It coincides with the brightest segment
of a much longer radio filament that extends toward the southernmost extensions of the filaments of the Radio Arc (Yusef-Zadeh et
9
To perform this task we use WEBPIMMS: https:heasarc.gsfc.nasa.govcgibinToolsw3pimmsw3pimms.pl. As explained in § 2.1, the combined count
rate is the sum of the EPIC-pn plus the EPIC-MOS count rates after scaling
the latter exposure maps by 0.4. We then use the EPIC-pn, medium-filter,
rate-to-flux conversion computed with WEBPIMMS.
c 0000 RAS, MNRAS 000, 1–??
SNR excavated bubbles within the CMZ?
Spectral analysis
To further investigate the origin of these structures, we extracted a
spectrum from each of these features (in either obsid 0694641301
or 0694641201). We fitted each spectrum with a model composed
of a SNR emission component (fitted with a PSHOCK model; a constant temperature, plane-parallel plasma shock model, meant to reproduce the X-ray emission from a supernova remnant in the Sedov
phase) plus the emission components typical of the GC environment such as a hot thermal plasma (with temperature in the range:
kT = 6.5 − 10 keV), and an Fe Kα emission line, all absorbed
by foreground neutral material (PHABS *( PSHOCK + APEC + GAUS )
in XS PEC). We assume that all these components have Solar abundances. Possible confusion effects produce uncertainties associated
with the determination of the correct sizes of these candidate SNRs,
so some of the results presented here, such as the dynamical timescales, could thereby be affected.
24
G. Ponti et al.
G0.30+0.04
2-4.5 keV emission
0.100
G0.224-0.032
0.050
G0.570-0.001
G0.40-0.02
Galactic latitude
G0.61+0.01
0.000
G0.57-0.018
-0.050
G0.52-0.046
Tail of stars from Quintuplet
-0.100
-0.150
0.600
0.550
0.500
0.450
0.400
0.350
0.300
0.250
0.200
0.150
Galactic longitude
Scuba 850 micron
0.100
G0.224-0.032
Galactic latitude
0.050
G0.40-0.02
G0.570-0.001
0.000
-0.050
-0.100
G0.52-0.046
-0.150
0.600
0.550
0.500
0.450
0.400
0.350
0.300
0.250
0.200
0.150
Galactic longitude
Figure 17. (Top panel) 2-4.5 keV map. The dashed white ellipse shows the location of the radio SNR G0.30+0.04. The dashed and solid cyan ellipses show
the positions of known X-ray SNRs and superbubbles, respectively. The green dashed ellipses indicate enhancements of soft X-ray emitting gas (G0.40-0.02
was already observed in X-rays by Nobukawa et al. 2008), the magenta ellipse is used in the spectral analysis as a background region (Back in Tab. 7). The
dashed yellow ellipse shows the location of the massive stars that could be in the tidal tail of the Quintuplet cluster (Habibi et al. 2013; 2014). (Bottom panel)
850 µm map of the GC obtained with the SCUBA bolometer (Pierce-Price et al. 2000). The four dashed ellipses indicated by the green ellipses and showing
enhanced X-ray emission correspond to holes in the 2mm gas distribution.
c 0000 RAS, MNRAS 000, 1–??
25
Big box
Small box
Flux
Surf. Bright.
Flux
Surf. Bright.
F1−2 keV = 19.0
F2−4.5 keV = 155.0
F4.5−12 keV = 290.7
FSi xiii = 4.4
FS xv = 15.1
FAr xvii = 15.1
FCa xix = 18.4
f1−2 keV = 6.2
f2−4.5 keV = 50.7
f4.5−12 keV = 95.1
fSi xiii = 1.5
fS xv = 4.9
fAr xvii = 4.9
fCa xix = 9.9
F1−2 keV = 12.9
F2−4.5 keV = 119.3
F4.5−12 keV = 219.0
FSi xiii = 3.4
FS xv = 12.0
FAr xvii = 11.9
FCa xix = 14.3
f1−2 keV = 6.9
f2−4.5 keV = 64.3
f4.5−12 keV = 118.0
fSi xiii = 1.8
fS xv = 6.4
fAr xvii = 6.4
fCa xix = 7.7
Table 6. Fluxes (F) and surface brightnesses (f) of the continuum and the line intensities (line plus continuum) integrated over the small and big boxes described
in § 8.1. The fluxes are given in units of 10−12 erg cm−2 s−1 , while the surface brightnesses are given in 10−15 erg cm−2 s−1 arcmin−2 .
G0.40-0.02 shows a best-fit temperature and normalisation of
the warm plasma associated with the SNR component of kT =
−2
0.55 ± 0.1 keV and Apsho = 2.4+5
, while the column
−1 × 10
density is observed to be NH = 7.7 ± 0.8 × 1022 cm−2 (see
also Nobukawa et al. 2008). To test whether the observed X-ray
enhancement is due to a real variation of the intensity of the soft
X-ray emission or whether it is the product of lower extinction,
we also extracted a spectrum from a nearby background comparison region (magenta in Fig. 17 and ”Back” in Tab. 7). This second region has the same size as G0.40-0.02 and it is located in
a fainter region in X-rays, characterised by higher NH , as suggested by the 850 µm map (see bottom panel of Fig. 17). This
background region shows a slightly higher absorption column density, NH ∼ 8.8 ± 1 × 1022 cm−2 , and no significant warm plasma
component. If we impose the presence of a warm plasma component having the same temperature and τ (the ionisation timescale
of the shock plasma model) as observed in G0.40-0.02 we obtain
an upper limit to its normalisation of Aphsho < 6×10−3 . This suggests that the enhanced X-ray emission towards G0.40-0.02 is due
to a real excess of X-ray emission and is not a simple byproduct of
lower extinction. The thermal energy, the dynamical timescale and
the size of G0.40-0.02 are Eth ∼ 1.9 × 1050 erg, tdy ∼ 3700 yr
and 8.6 × 5.5 pc2 , respectively, as expected for a young SNR in the
Sedov-Taylor phase (derived from the equations shown in Maggi et
al. 2012).
found neither radio nor nucleosynthetic decay products (such as
44
Ti), questioned such an interpretation. Further investigations are
required to understand the link, if any, between these features.
The morphology of G0.224-0.032 appears more complex,
compared to the other SNR candidates. The edge of the X-ray emission is well defined only towards the Brick molecular cloud (designated M0.25+0.01 in Fig. 5) that, with its very high column density, can obscure the soft X-ray emission there. In any case, the
fit of its X-ray spectrum shows parameters typical of a SNR. In
particular, we derive a thermal energy and a dynamical time of
Eth ∼ 2.6 × 1050 erg and tdy ∼ 1800 yr, respectively. Therefore, G0.224-0.032 might be a new SNR partly obscured by the
brick molecular cloud. In such a case, the true size and the energy
estimate are likely larger.
Overall, we remark that, if these SNR candidates are real, their
dynamical timescales are extremely short, which would imply an
extremely high supernova rate. The SN rate in the CMZ has been
estimated to he as high as 0.4 SN per millenium (Crocker et al.
2011). However, we caution that our dynamical time-scales could
be off either because of a higher ambient density than we have assumed, or because absorption or confusion effects do not allow us
to distinguish the proper border of the SNRs, or to detect any colder
and more extended portions that might be present.
Similar parameters characterise G0.52-0.046 (kT = 0.77 ±
−3
0.3 keV, Apsho = 5+7
, NH = 7.9 ± 1.1 × 1022 cm−2 ).
−3 × 10
Therefore, this feature also appears to be consistent with an SNR
origin. However, although we derive a dynamical age of the same
order (tdy ∼ 1700 yr), the energy inferred for the SN explosion is
substantially lower, Eth ∼ 5 × 1049 erg.
8.3.2
The spectrum of G0.57-0.001 is characterised by significantly
lower statistics and a higher column density of absorbing material. The best fit prefers a low temperature plasma, at the limit
of detection. We fix its temperature to a relatively low value of
−3
kT = 0.6 keV and find Apsho = 4.3+11
and NH =
−4 × 10
22
−2
9.5 ± 2 × 10 cm . The derived thermal energy and dynamical
times are Eth ∼ 2.6 × 1049 erg and tdy ∼ 1600 yr, respectively.
The soft X-ray excess of this feature lies spatially very close to a
region of X-ray excess that has been traced by Fe XXV line emission (Nobukawa et al. 2008). We also note that the region defining
G0.57-0.001 almost completely contains a diffuse X-ray source detected both by the Chandra and ASCA satellites (Senda et al. 2002).
The Chandra image shows a very compact (∼ 10” radius) and hot
shell, G0.57-0.018, possibly the youngest SNR in the Galaxy, less
than about 100 yr old. However, Renaud et al. (2006), because they
c 0000 RAS, MNRAS 000, 1–??
Expanding molecular shells
The observed temperatures, ages and sizes of these SNR candidates
are consistent with a Sedov-Taylor framework expanding into an
average ambient density between 1 − 10 cm−3 (higher density environments would result in older and cooler SNRs; Ostriker & McKee 1988).
If the X-ray enhancements described above truly arise from
SNRs interacting with and carving bubbles inside or near the surfaces of molecular clouds (preceded, perhaps, by the wind of the
massive progenitor), we should observe clear traces of such events
also in the kinematics of the surrounding molecular matter. Such
a complex and delicate investigation is beyond the scope of the
present paper. Nevertheless, we note that Tanaka et al. (2009) discovered an expanding SiO shell (SiO0.56-0.01) centered at l ∼
0.56◦ , b ∼ −0.01◦ and having a size of ∼ 3.0 × 3.4 pc2 . The
center and size of the expanding SiO shell closely match the peak
and size of the X-ray emission of G0.57-0.001 and suggest an association between the two. In particular, high-velocity clumps have
been found consistent with the idea that the SiO shell consists of
swept-up material. Tanaka et al. (2009) calculated a kinetic energy
of Ekin ∼ 1050.4 erg for SiO0.56-0.01. This strongly suggests that
26
G. Ponti et al.
parameter
kT
Apsho
τ †‡
Size
NH
χ2 /dof
ne
tdy
Eth
unit
keV
s cm−3
pc
1022 cm−2
cm−3
yr
erg
G0.40-0.02
G0.52-0.046
G0.570-0.001
G0.224-0.032
Back
0.55 ± 0.1
−2
2.4+5
−1 × 10
> 1.7 × 1011
8.6 × 5.5
7.7 ± 0.8
807/777
0.77+0.7
−0.2
+9
5−3 × 10−3
> 7.5 × 1010
5.9 × 2.7
7.9 ± 1.1
342/364
0.6‡
−3
4.3+11
−4 × 10
> 1.4 × 1010
4.2 × 2.1
9.5 ± 2
141/134
0.54 ± 0.1
−2
3+6
−2 × 10
> 3.5 × 1011
5.4 × 2.6
7.4 ± 1
369/338
0.55†
< 6 × 10−3
7.5 × 1011 †
8.6 × 5.5
8.8 ± 1
834/741
1.4
3.7 × 103
1.9 × 1050
1.4
1.7 × 103
5.0 × 1049
2.1
1.6 × 103
2.6 × 1049
11
1.8 × 103
2.6 × 1050
Table 7. Best-fit and derived parameters of the SNR candidates described in § 8.3. †Parameter unconstrained. ‡Value weakly constrained by the high column
density of neutral material, therefore fixed for the corresponding fit. †‡ Ionisation time-scale of the shock plasma model.
G0.57-0.001 is indeed a SNR caught in the process of carving its
bubble. Further studies of the gas kinematics around the other Xray enhancements are required to establish their real nature.
We note that the thermal energy estimated for these SNRs is
observed to be systematically lower than the theoretical value for
the remnant of a standard type II SN expanding into the interstellar
medium. This might result from a relatively higher ambient density
in the GC, leading to greater energy dissipation, or from a significant fraction of the energy budget going into the inflation of the
bubbles and the production of cosmic rays.
We also note that G0.570-0.001, G0.52-0.046 and G0.40-0.02
are located within the trail of massive stars that have been hypothesised to have tidally escaped from the Quintuplet cluster (see Fig.
17 and Habibi et al. 2013; 2014). This raises the possibility that
some of these SNRs might be associated with SN explosions from
stars originating in this massive, young stellar cluster.
8.4
Origin of the Sgr A X-ray lobes
All the soft X-ray maps (see Figs. 3, 10, 11, 7, 12 and 14) show
the presence of two extended features, with a size of roughly 5 −
10 pc, located to the Galactic north and south of Sgr A? , the socalled ”bipolar Sgr A lobes” (Morris et al. 2003; 2004; Markoff et
al. 2010; Heard & Warwick 2013).
8.4.1
Lobe morphology
The lobes appear to have roughly oval shapes with co-aligned major axes oriented perpendicular to the Galactic plane. They appear joined at the position of Sgr A? suggesting the latter is their
point of origin (see Fig. 8). The top panel of Fig. 11 and Fig. 7
show that the lobes’ emission is characterised by a smaller ratio of
soft X-ray lines to continuum (therefore characterised by a greener
colour) compared to the surrounding regions (appearing with a
redder colour) such as the superbubble, G0.1-0.1 and the Radio
Arc (Fig. 11, 7 and 8). This suggests that the lobes, although they
show thermal emission lines (see §4.2), have either a stronger nonthermal component or significantly hotter thermal emission than
the surrounding regions10 . We also note that the eastern portion of
what appears to be part of the southern lobe has a colour as red as
10
A hot plasma, with temperatures of ∼ 2 − 4 keV, produces intense Xray emission but weaker soft X-ray lines, compared to a plasma having a
temperature around 1 keV.
G0.1-0.1 and the superbubble regions (see Fig. 10). Therefore, this
emission might not be associated with the lobes, but rather with
G0.1-0.1 or the edge of the superbubble (however, a gap such as
might be produced by a foreground dust lane, appears to separate
the lobes’ emission from G0.1-0.1). Figures 7 and 8 also show that
the surface brightness of the northern lobe decreases with distance
from Sgr A? (Heard & Warwick 2013).
The bottom panel of Fig. 11 shows that the lobes have an orange colour, indicating harder soft X-ray emission compared to the
surrounding regions. In particular, a brighter and harder (yellowgreen) linear structure outlines the northern lobe, shaping it to have
a well-defined and symmetric spade structure, with a sharp transition at the border. The sharpness of the transition suggests the presence of a limb-brightened shock, indicating that the lobe is a bubble
enclosed by a thin shell of hot, compressed material. This claim is
strengthened by our analysis of the Chandra data (see Fig. 8 and 9,
Baganoff et al. 2003; Lu et al. 2008; Muno et al. 2008). The superior Chandra spatial resolution, in fact, allows us to note that these
projected linear features are running right along the lobes’ edges,
indeed confirming the presence of a shock (Fig. 8 and 9). We also
note that the emission from the northern half of the lobe seems to
be mainly due to three harder filaments converging at the top in a
cusp, having radio continuum counterparts (Zhao et al. 2015) and
associated Paschen-α emission, indicating that these are thermal
features (see Fig. 18 and 19).
In the southern lobe, two bright knots are observed in the center and at the tip. Interestingly these appear to be located approximately at the same distance and in the opposite direction, compared
to Sgr A? , as two enhancements present in the northern lobe. Moreover, the two bright knots have a green-yellow colour (upper panel
of Fig. 11) similar to their apparent counterparts in the northern
lobe. This suggests both: i) a similar physical origin for these features in both the north and south lobes and; ii) that the process that
created the lobes is symmetric about the Galactic plane and its engine is (or is located close to) Sgr A? . The obvious interpretation
of this morphology is that energetic events simultaneously ejected
diametrically opposed blobs of hot gas.
However, upon closer inspection, the northern and southern
lobes do not appear completely symmetric. For example, compared
to the northern lobe, the western side of the southern lobe and the
region close to Sgr A? appear suppressed (see Fig. 7, 8 and discussion following). On the other hand, the eastern side of the southern lobe appears to extend further (e.g. further east compared to
G359.977-0.076) than the corresponding boundary of the northern
lobe (located close to e.g. G359.974-0.000). As described before,
c 0000 RAS, MNRAS 000, 1–??
27
the emission around and east of G359.977-0.076 has a different
colour and might therefore be associated with either the superbubble, with G0.1-0.1, or it could be a feature that is independent of
either of these and of the lobes.
The bottom panel of Fig. 15 shows that the region with depressed soft X-ray emission south of Sgr A? and on the western
side of the southern lobe spatially coincides with the presence of
the 20 km s−1 molecular cloud, which is thought to be located in
front of Sgr A? (Coil et al. 2000; Ferrière 2009) and to have a large
column density. Soft X-ray emission could be produced there but
be completely obscured to us by this intervening cloud (see Fig. 2).
To reinforce this idea, we note that, in fact, at this position, hard
X-ray radiation (between 4.5 and 12 keV) is observed by Chandra to have a non-thermal spectrum and to be extended (Morris
et al. 2003). Therefore, this hard radiation, which is able to penetrate the cloud, could be produced by strong shocks at the bubble’s
border. Similar hard non-thermal filaments are observed in several
places at the border of the northern lobe (Morris et al. 2003). This
strongly suggests that the actual border of the southern lobe is located further west than the images reveal, and that the lobe’s soft
X-ray emission is obscured there. Therefore, once the effect of absorption by molecular clouds (e.g. the 20 km s−1 cloud) is considered, the symmetry between the northern and southern lobes
appears more clearly. The linear or filamentary structures (such
as G359.974-0.000, G359.970-0.009, G359.959-0.027, G359.9450.044, G359.942-0.045, 359.933-0.039, see Fig. 8) observed in the
northern lobe might be present in the southern one as well, but be
suppressed by the intervening absorption.
Figure 18 shows the comparison between the Chandra and
Paα emission. The Paα map clearly show the presence of tendrils
of foreground absorbing material, running north-south along the
extension of the lobe. Interestingly, the same tendrils are also evident as absorption lanes in the X-ray image. We also note that two
Paα emitting luminous stars are contributing to, if not dominating, the ionisation of the gas surrounding them. However, there is
a close correspondence between the Paα emission and the soft Xray emission at the northern tip of the lobe, so strong shocks might
be contributing both to the ionization and to the production of very
hot post-shock gas that can emit X-rays. Indeed, we hypothesize
that some portion of the hot wind that we see in the softest X-ray
bands, presumably emanating from near Sgr A? , is undergoing a
shock where it encounters ambient interstellar material at b ∼ 0.02
degrees, and that it is thereby blocked from continuing to higher
latitudes. That shock is manifested both as a horizontal feature in
the Paα image, and as a diminution in the brightness of the X-ray
emission proceeding north from that latitude. There is still some
weaker, extended X-ray emission at higher latitudes, indicating that
not all of the outflowing wind is completely blocked. In addition,
the portion of the X-ray emitting plasma lying behind the shock
front appears to be absorbed along the shock front, creating a horizontal shadow in the extended X-ray emission, and suggesting that
the shock at b ∼ 0.02 degrees has created a thick, compressed layer
that absorbs the X-rays coming from behind it.
8.4.2
Lobes collimated by the Circumnuclear Disk (CND)
Morris et al. (2003) observed that the CND has a size and orientation that are consistent with it being the agent that collimates an
isotropic outflow from the Sgr A? region, thereby creating the bipolar lobes of hot plasma. Those authors further suggested that the
sequence of enhancements along the axis of the lobes might have
resulted from a series of energetic mass ejections from the immec 0000 RAS, MNRAS 000, 1–??
diate environment of Sgr A? . A similar scenario has recently been
discussed by Heard & Warwick (2013b). At an outflow velocity of
103 km s−1 , it would have taken 2 − 6 × 103 yr to inflate the lobes.
Assuming a thermal emission model, an electron density ne in the
range 1 − 10 cm−3 can be inferred (Morris et al. 2003; Heard &
Warwick 2013b) and, assuming that the line-of-sight depth is equal
to the projected width, the total hot plasma mass involved in the
X-ray lobes is only about 1 − 3 Solar masses.
8.4.3
Energetics of the Lobes
Except for the broad intensity enhancements along the axis of the
lobes, the surface brightness is relatively smoothly distributed. Assuming a continuous and constant outflow, we examine the energy
budget of the lobes. In particular, integrating the measured energy
density over a cylinder of 5 pc radius and 12 pc height (the approximate sizes of the lobes), a thermal energy of Eth ∼ 9 × 1049 erg
is estimated.
It has been estimated that the massive stars in the central parsec collectively lose ∼ 5 × 10−3 M yr−1 in stellar winds, with
velocities ranging from ∼ 300 to 1000 km s−1 (Geballe et al. 1987;
Najarro et al. 1997; Paumard et al. 2001). The total kinetic energy
thermalised by shocks is then E ∼ 5×1038 erg s−1 for such a mass
outflow rate and an outflow velocity of 1000 km s−1 (Quataert &
Loeb 2005). Therefore the energy released within the time needed
to inflate the lobes (∼ 4×103 yr) is equivalent to E ∼ 5×1049 erg,
therefore giving an important contribution to the generation of the
lobes.
The lobes might also be traceable to the accretion flow onto
Sgr A? . We note that, as estimated by Wang et al. (2013), only
∼ 1 % of the matter initially captured at the Bondi radius presently
reaches the innermost regions around Sgr A? . The rest of the accretion power, estimated to be ∼ 1039 erg s−1 (Wang et al. 2013), is
probably converted to kinetic energy and used to drive an outflow
that carries away the bulk of the inflow, sculpting the environment
with its ram pressure. If so, within the lobe inflation time, a total
energy of ∼ 1050 erg would have been deposited.
Within the CMZ, X-ray reflection nebulae indicate that a few
hundred years ago Sgr A? was more active, being ∼ 106 times
brighter than at present for approximately 5 − 10 % of the time in
the past millennium (experiencing LX ∼ 1039 erg s−1 ; see Ponti
et al. 2013 for a review). Could the lobes have been created by similar events that occurred over the past 104 yr? The light crossing
time of the CMZ limits our capability to trace Sgr A? ’s past activity beyond about 103 yr ago, so it is difficult to directly trace the
echoes of possible energetic events on such time scales. However,
if the process has been active over the past (5 − 10) × 103 yr at
roughly the same rate, (therefore active at L ∼ 2 × 1039 erg s−1
for ∼ 103 yr in the past 104 yr), a total integrated energy equal to
∼ 5×1049 erg should have been generated. If Sgr A? ’s past activity
was characterised by outbursts with associated outflows having kinetic luminosity comparable to the radiated power (therefore much
higher than in soft state stellar mass black holes; Ponti et al. 2012),
then these events could be the primary source (or at least contribute)
to form the lobes.
All these processes appear similarly likely to have an impact
on the formation of the lobes, from an energetic point of view.
However, we note that the first two mechanisms are powered by
a quasi-continuous outflow from Sgr A? or from the central stellar cluster. In such scenarios, therefore, the sharpness of the edges
at the extremities of the lobes remains rather puzzling, favouring
explosive-outbursting scenarios.
28
G. Ponti et al.
Figure 18. (Left panel) Chandra RGB image of the northern lobe. See Figure 8 for more details. (Right panel) Pa-α image of the northern lobe, tracing
thermal, ionized gas (from Wang et al. 2010). Continuum emission from stars has been removed (Dong et al. 2011), so the only stars that appear are those that
have strong Pa-α emission lines. Two Paα-emitting luminous stars located at l ∼ 359.935◦ , b ∼ 0.21◦ and l ∼ 359.925◦ , b ∼ 0.45◦ (Mauerhan et al. 2010;
Dong et al. 2012) are probably responsible for at least part of the ionisation indicated by the Pa-α. See WWW. MPE . MPG . DE / HEG / GC / for a higher resolution
version of these figures.
8.4.4
Are Sgr A’s lobes the SNR of SGR J1745-2900?
The recent discovery of a young magnetar, SGR J1745-2900 (Degenaar et al. 2013; Dwelly & Ponti 2013; Mori et al. 2013), most
probably in orbit around Sgr A? (Rea et al. 2013), raises the quest
for finding its young SNR. SGR J1745-2900 is estimated to be only
about 9 × 103 yr old and to be located at <
∼0.07 − 2 pc from
Sgr A? (Rea et al. 2013). Therefore, the supernova that generated
SGR J1745-2900 should have exploded near the centroid of the
lobes and likely inside the inner radius of the CND. If the shock
propagated at the sound speed (vs ∼ 750 km s−1 ), then a present
size of ∼ 6.5 pc would be expected. This is slightly smaller than
the observed size of the lobes (∼ 12 pc), but is consistent with
the observed size within the uncertainties in the age and the sound
speed. We also note that the thermal energy content in the lobes
(Eth ∼ 9 × 1049 erg) is lower, but of comparable order of magnitude, to the energy released by typical supernova explosions. It
is therefore plausible that the lobes are indeed the SNR associated
with SGR J1745-2900.
Another viable possibility is that the lobes have been generated by the supernova that created the PWN candidate G359.9450.044, located only ∼ 8” from Sgr A? , with an estimated age of
few thousand years (Wang et al. 2006). It is therefore possible that
at least one SN exploded close to Sgr A? within the last ∼ 10 kyr. If
so, its blast wave likely propagated into a pre-existing hot, low density cavity created by Sgr A? ’s outflows and the collective winds of
the central stellar cluster. Given the density and the temperature in
the lobes (kT ∼ 2 keV; Morris et al. 2003), the shock is estimated
to reach 15 pc in 9 × 103 yr (Wang et al. 2005), a value consistent
with the observed size of the lobes.
As mentioned above, the presence of the sharp edges to the
lobes seems to favour an explosive mechanism for their creation.
A supernova exploding inside the CND would expand into the preexisting, stationary outflow from the center and be collimated in
the same way. Hydrodynamical simulations typically show that SN
shock fronts are reflected away when encountering the walls of
a dense molecular cloud, such as the CND (Ferreira & de Jager
2008). In this scenario the sharp edge of the lobes would be due to
the SN shock front. If the supernova recurrence time is longer than
or comparable to the lobe expansion time (a few thousand years)
then this would not appear as a stationary process. Assuming a recurrence time between 1−10×104 yr (similar to the SN recurrence
time of the central young cluster, in which ∼ 100 massive stars presumably becoming SN over a ∼ 107 yr time interval), we estimate
a time-averaged kinetic power release of 3−30×1038 erg s−1 . This
indeed suggests that: i) SN explosions of the central star cluster can
c 0000 RAS, MNRAS 000, 1–??
29
contribute to powering the lobes; ii) the lobes are quasi-stationary
features; and iii) it is not unlikely that we observe such features
created by a rare event such as a SN explosion.
Finally, we note that, although the characteristics of the Xray emission from the lobes appear consistent with being the X-ray
remnant of a recent supernova that exploded within a few parsecs
of Sgr A? , the lack of associated nonthermal radio emission from
such a young SNR is problematical for this hypothesis.
Mori et al. (2009) and Heard & Warwick (2013) estimate a
total thermal energy contained within the superbubble of Eth ∼
1051 f 1/2 erg (where f is the volume filling factor of the emitting
plasma). Such a large energy content does, indeed, require multiple
supernova events. Those authors also estimate for G359.77-0.09
and G359.79-0.26 an ionisation time-scale of tion ∼ 3 × 104 yr
(assuming f ∼ 1).
8.5.2
8.5
G359.77-0.09 and G359.79-026: a ring from a hot
superbubble southwest of Sgr A?
A series of diffuse, soft features appear to the southwest of Sgr A?
(see Fig. 3), namely G359.77-0.09, G359.79-0.26 and a newly
recognised extended feature, G0.0-0.16. If these are physically connected, they form, in projection, a roughly elliptical shape whose
major axis has an inclination of about 60◦ with respect to the Galactic plane (see Fig. 3, 6, 7, 10, 11, 12 and 14). These features show
similar colours and strong, soft line emissions, indicating a similar thermal origin (see Fig. 3, 10 and 11). This elliptical structure
appears brightest at the softest energies, however it is not observed
below 1.5 keV, suggesting a location near the Galactic centre and a
low temperature for its plasma, compared to the surrounding emission. We note that this feature is characterised by a very bright edge
with strong Si XIII emission on the outside of the ellipse, suggesting a lower temperature of the edge compared to the interior.
The ellipse center is located at l ∼ 359.9◦ , b ∼ −0.125◦ and
it has minor and major axes of about 7.8 and ∼ 12 arcmin, respectively (corresponding to 18 and 28 pc). Both Mori et al. (2009) and
Heard & Warwick (2013) considered that these structures/group
of structures were physically connected and form a superbubble
candidate. We note that the recognition of such an elliptical ring
critically depends on the presence of the dark lane running from
l ∼ 0.02◦ , b ∼ −0.22◦ to l ∼ 0.05◦ , b ∼ −0.07◦ (see Fig. 3,
10, 11, 7 and 14), which separates G0.0-0.16 from the emission
of G0.1-0.1; this dark lane helps define the quasi-continuous elliptical morphology of the ring. However, the lane might simply be
due to absorption by foreground material, in which case G0.0-0.16,
forming the eastern part of the ring, could simply be connected to
G0.1-0.1.
8.5.1
S XV emission filling the superbubble
The S XV emission provides a key piece of information to better understand the superbubble. We note, in fact, that the S XV emission
completely fills the region inside the ring with a roughly uniform
brightness (several times brighter than in the surrounding region)
and sharply drops at the ring’s edge. This indicates that the superbubble is a shell of hot gas that we see projected onto the plane
of the sky (see middle top panel of Fig. 12). To further corroborate this, we note that a sharp emissivity drop appears to be located
just outside of the ring, running all around the ring’s external edge.
Such X-ray depression might be produced by a high column density of cold gas pushed away by the superbubble’s shock front and
accumulated in large quantities just outside the shock. If so, the
observed depression could be indicating that the superbubble is located in front of G0.1-0.1. Fig. 7 also shows a small depression in
the top part of the northern lobe that could easily be explained by
absorption associated with the superbubble if it is located in front
of the lobes. This situation would then be somewhat analogous to
the colour variation in the south lobe.
c 0000 RAS, MNRAS 000, 1–??
Origin
The origin of such a superbubble is not clear. We note that many
of the massive stars that are suggested to have escaped from the
Quintuplet cluster (Habibi et al. 2013; 2014) are projected inside
the superbubble. It is therefore possible that explosions of stars lost
by the Quintuplet cluster have contributed to energising the superbubble.
A more speculative point is that the estimated age of the superbubble is of the same order of magnitude as the recurrence time
of tidal disruption events by Sgr A? : tT DE ∼ 1 − 3 × 104 yr
(Alexander & Hopman 2003). While Sgr A? appears, in projection, to be located inside the superbubble, it is ∼ 6.7 arcmin off
from the superbubble’s center. Khokhlov & Melia (1996) suggested
that an explosion associated with a tidal disruption event would
liberate a large amount of energy on the order of E ∼ 1052 erg
that would propagate as a powerful shock wave into the local interstellar medium. As with the remnant of the SGR J1745-2900,
we expect that the shockwaves of a tidal disruption event would
interact with the CND. However, the unbound part of the tidally
disrupted star would be ejected into a limited solid angle, producing a strongly elongated and asymmetric remnant (Khokhlov et al.
1996; Ayal et al. 2000). Such a remnant would then appear as a
very energetic shell of hot gas and remain visible for a time comparable to the age of a typical SNR. Assuming a shock survival time
of ∼ 1 − 10 × 104 yr, we could potentially observe a few remnants
resulting from tidal disruptions. We suggest that the superbubble
G359.77-0.09 has properties that make it a possible candidate. No
other feature with properties obviously related to a tidal disruption
event appear to be observed close to Sgr A? .
8.6
The arc bubble: a second superbubble in the GC
Highly enhanced soft X-ray emission is observed east of Sgr A?
from the region called G0.1-0.1. This feature appears in Figures 10
and 11 as a slightly elliptical feature of enhanced emission with
center at l ∼ 0.09◦ , b ∼ −0.09◦ and with radius of ∼ 5 arcmin
(corresponding to ∼ 10 pc). The top panel of Fig. 11 shows that
G0.1-0.1 and the Radio Arc regions both show distinct red emission. This indicates large equivalent widths of the emission lines
from this plasma and therefore a thermal origin. However, the top
panel of Fig. 10 and the bottom of Fig. 11 show strong colour gradients within these regions, indicating that they might have different
contributions from distinct components.
The PWN candidate G0.13-0.11 (l = 0.131◦ , b = −0.111◦ ;
Wang et al. 2002; Heard & Warwick 2013) stands out from the general thermal emission in G0.1-0.1, appearing as a distinct point-like
source characterised by a light blue/white colour11 , indicating its
non-thermal origin (dominated by intense soft and hard continuum
11
Note that in the top panel of Fig. 10 despite an enhancement of diffuse
emission from the region surrounding the core of G0.13-0.11, no point like
source is detected in the soft X-ray line image.
30
G. Ponti et al.
emission with no soft line emission, see the top panel of Fig. 11)12 .
The point-like head of G0.13-0.11 appears to be accompanied by a
tail extending to the south for 4.5 − 5 pc; in Fig. 19, it appears with
a white-violet colour.
Heard & Warwick (2013) suggest that the SN that generated
G0.13-0.11 might be the source of the soft X-ray emission from this
region. Those authors present a spectral study of the X-ray emission
from G0.1-0.1 and find a gas temperature of kT = 1.1 ± 0.1 keV,
and a column density of NH = 5.6 ± 0.5 × 1022 cm−2 , indicating a GC location of this emission, and abundances that are about
1.8 times Solar. Assuming that the plasma volume is only 3.5 pc3 ,
corresponding to only 1.5 arcmin radius around the PWN (see the
red circle in Fig. 20), the authors estimated a thermal energy of
Eth = 3.1 × 1049 erg (and a plasma ionisation time-scale of at
least t = 1.8 × 104 yr), thus consistent with being produced by a
single supernova explosion.
8.6.1
Energetics of the arc bubble
We note from Figs. 19, 3, 10, 11, 7, 12 and 14 that G0.1-0.1 extends
further from G0.13-0.11 than the 1.5 arcmin region size considered
by Heard & Warwick (2013), with no clear boundary at 1.5 arcmin
(see Fig. 20). To illustrate this, Fig. 19 shows the soft emission lines
RGB image with colour scales chosen to highlight intensity variations present within this region. The left panel of Fig. 20 shows the
contours of the S XV emission overlaid on the 20 µm MSX image
(Price et al. 2001). Figure 20 clearly shows that the empty mid-IR
bubble (the so called arc bubble; Levine et al. 1999; RodriguezFernandez et al. 2001; Simpson et al. 2007) is completely filled
with warm X-ray emitting plasma and that the soft X-ray emission is not confined to within ∼ 1.5 arcmin of G0.13-0.11, but it
extends much further, for about 7 arcmin. Assuming a uniform surface brightness, if the bubble is 4.5 times larger, we would expect
a thermal energy of Eth ∼ 1.5 × 1051 erg, thus most probably
requiring multiple supernova events, and supporting the notion that
G0.1-0.1 is a second superbubble candidate in the GC.
Rodriguez Fernandez et al. (2001) have noticed that the radio
Arc bubble is filled with continuum X-ray emission seen by ASCA
which they ascribed to X-ray sources inside the bubble. Here we
find that the bubble is in fact filled with diffuse thermal X-rays,
most likely originating from SN explosions of massive stars associated with the Quintuplet cluster.
We also note that the soft line emission is at least as extended
as the arc bubble and is highly inhomogeneous (see Figs. 19 and
20). Three depressions having roughly circular shapes can be discerned in G0.1-0.1 (Fig. 20). Two cavities are located at about the
same latitude, with centers close to l = 0.057◦ , b = −0.067◦
and to l = 0.116◦ , b = −0.071◦ and with radii of ∼ 1.6 arcmin
(corresponding to ∼ 3.7 pc) and ∼ 1 arcmin, respectively. These
cavities appear to be surrounded by a thin rim of brighter emission.
A third depression is centered at l = 0.083◦ , b = −0.123◦ , with
∼ 1.8 arcmin radius. This cavity also seems to be confined by a
thin shell of brighter material, except for its southern edge, where
it appears open (see Fig. 19), possibly because of the presence of a
dark absorbing lane.
12 Please note that an unrelated point source at l = 0.142◦ , b = −0.109◦
is located at the same latitude, but ∼ 1.5 pc to the Galactic east of the PWN
candidate G0.13-0.11.
8.6.2
Association with the Quintuplet cluster?
Despite its offset position from the centre of the mid-IR arc bubble, the Quintuplet cluster is often considered responsible for creating and maintaining the bubble with a combination of supernovae
and strong stellar winds (e.g., Simpson et al. 2007). Johnson et al.
(2007) invoke a possible non-uniformity of the ambient medium as
a possible origin of this asymmetry. This might also apply to the Xray emission. However, we note that the Quintuplet cluster is moving supersonically within the CMZ (Stolte et al. 2014). Given its
projected velocity, the cluster would have taken ∼ 100 kyr to cross
the width of the IR arc bubble, in which case the bubble would have
been inflated on a time scale much smaller than typical superbubble
formation times (Castor et al. 1975; Weaver et al. 1977; Mc Low
& McCray 1988). Several SN explosions in that amount of time
would therefore have been required. Furthermore, the Quintuplet
cluster would have been located in the middle of the two northern
cavities about 4 × 104 and 9 × 104 yr ago, respectively (the right
panel of Fig. 20 illustrates the direction of the cluster’s motion and
the position of the cavities). The relatively small cavities observed
in the northern part of G0.1-0.1 are unlikely to have been generated
by multiple cluster stars. In fact, a supernova exploding in the hot
plasma of G0.1-0.1 is expected to undergo a significantly different evolution than a typical SNR. In particular, the sound velocity
is significantly larger than in a typical low pressure medium. Tang
& Wang (2005) have shown that the shock velocity follows a Sedov solution but quickly deviates from it when it becomes mildly
supersonic. This translates into a much faster evolution and much
larger cavities would be expected if the SN exploded 4 × 104 and
9×104 yr ago, when the quintuplet cluster was at that location. It is
more likely, therefore, that the two cavities to the Galactic west of
the Quintuplet might have been generated by supernova explosions
of massive stars either stripped from the Quintuplet cluster (Habibi
et al. 2013) or having no association with it.
The large thermal energy filling the arc superbubble could
have been produced by some combination of winds from the young
stars and by multiple supernova explosions, including the supernova explosion associated with the PWN G0.13-0.1113 .
8.7
8.7.1
A hot atmosphere around the GC - A link to the GC
lobe? And to the Fermi bubbles?
General morphology
As observed in all soft X-ray maps (Figs. 3, 10 and 11) and confirmed by the soft plasma intensity map (obtained through spectralimages decomposition, Fig. 14), the regions at higher Galactic latitudes are significantly brighter in soft X-rays than the regions closer
to the disc, presumably in part because of the smaller extinction at
the higher latitudes.
The western border of this enhanced emission is defined by
a relatively sharp edge between l = 359.63◦ , b = 0.06◦ and
l = 359.55◦ , b = 0.46◦ (Fig. 14, see also Figs. 10 and 11). The
soft X-ray emission peaks above the location of the GC Radio Arc,
appearing as a continuation of the Radio Arc itself. Further west the
soft X-ray emission appears to fade with increasing Galactic longitude. In particular, the spectral decomposition provides hints for the
presence of an eastern edge, fainter but similar to the western edge,
of this hot GC atmosphere. However the presence of an edge to
13
We note that the PWN G0.13-0.11 is located right in the middle of the
mid-IR arc bubble.
c 0000 RAS, MNRAS 000, 1–??
31
0.150
0.100
0.050
Galactic latitude
0.000
-0.050
-0.100
-0.150
-0.200
-0.250
0.300
0.200
0.100
0.000
359.900
359.800
359.700
Galactic longitude
Figure 19. RGB image with colour scale chosen to highlight enhancements and depressions in the diffuse emission east of Sgr A? . In red the sum of the Si-S
and S XV bands is shown. In green the sum of S-Ar plus the S XV and Ar XVII bands is shown. The blue image shows the sum of the Ar-Ca plus Blue-Ca and
Ca XIX bands (see Tab. 1 for the definition of the energy bands).
the west is less obvious because of the soft X-ray extinction likely
caused by a series of molecular clouds (e.g., the Brick) present at
that location and because of the partial high-latitude coverage of
this region.
8.7.2
Radiative process
The presence of intense, soft X-ray emission lines (see Figs. 10 and
11) and, in particular, the good fit of a spectral decomposition based
on a thermally emitting gas (see Fig. 14) indicate that most of the
high-latitude emission is generated by a thermal radiative process
in a warm plasma. To demonstrate this, we accumulate the EPICpn spectrum from a circular region of 8.28 arcmin radius centered
at l = 0.181◦ , b = 0.359◦ . The resulting spectrum is well fitted
with an absorbed thermal emission component (APEC) with kT =
0.96 ± 0.1 keV, NH = (2.3 ± 0.2) × 1022 cm−2 and Aapec =
(1.5 ± 0.4) × 10−2 cm−5 .
might have two contributions, one associated with the G0.1-0.1
superbubble (filling the mid-IR arc bubble; see § 8.6), while the
second is associated with enhanced soft X-ray emission due to the
presence of the Radio Arc and its polarized radio plumes at higher
latitudes (Seiradakis et al. 1985; Tsuboi et al. 1986; Yusef-Zadeh
& Morris 1988). If that is indeed the case and if the two structures
have different X-ray colours (e.g. the superbubble produces lower
temperature thermal X-ray lines, while the Radio Arc has a larger
continuum to lines ratio), then we should observe variations in the
X-ray colour distribution. In particular, we would expect a whiter
colour and a green-yellow colour (similar to the one characterising the lobes of Sgr A) at the location of the Radio Arc region
compared to G0.1-0.1, in the top panel of Fig. 10 and in Fig. 11,
respectively. This idea is, indeed, in agreement with the colour variations and the evolution of the line intensities observed between the
G0.1-0.1 and Radio Arc complexes (Fig. 10 and 11).
8.7.4
8.7.3
Eastern edge
The eastern edge of the high-latitude emission rises from the position of the Radio Arc. This raises the interesting question of
whether the soft X-ray emission at the location of the Radio Arc
c 0000 RAS, MNRAS 000, 1–??
Western edge, the Chimney and AFGL5376
Running almost parallel to the western edge of the high-latitude
plasma is another region of enhanced soft X-ray emission, located
near the Galactic plane, the so called Chimney (l = 359.45◦ ; Tsuru
et al. 2009). The Chimney appears as a column of soft X-ray emitting plasma extending all the way between the core of Sgr C and
32
G. Ponti et al.
0.100
0.100
North Lobe
0.050
0.050
Arches cluster
0.000
Galactic latitude
Galactic latitude
0.000
-0.050
-0.100
PWN G0.13-0.11
-0.150
Quintuplet cluster
-0.050
167 km/s
Tail of massive stars
-0.100
PWN G0.13-0.11
-0.150
South Lobe
-0.200
-0.200
0.250
0.200
0.150
0.100
0.050
0.000
359.950
359.900
-0.250
0.250
0.200
0.150
0.100
0.050
0.000
359.950
359.900
-0.250
Galactic longitude
Galactic longitude
Figure 20. (Left panel) 20 µm MSX map of the GC. The contours indicate the intensity of S XV emission. Soft X-ray emission fills the arc bubble observed in
the mid-IR. The green solid circle and the white dashed ellipses indicate the position of the PWN G0.13-0.11 and three structures in the soft X-ray emission
map (see right panel). (Right panel) Soft X-ray map of the GC (the same energy bands used in Fig. 19 are displayed). The position of the PWN G0.13-0.11
is indicated by a red circle with 1.5 arcmin radius. At least three sub-structures, appearing like holes, are observed within G0.1-0.1 (here indicated with white
dashed ellipses). The positions of the Arches and Quintuplet star clusters are indicated by yellow dashed circles. The direction of the supersonic motion of
the Quintuplet cluster is indicated and its past location is indicated by the yellow dashed line. The inferred positions of the Quintuplet 4 × 104 and 9 × 104
years ago are indicated with yellow crosses. The cyan dashed ellipse indicates the region in which many massive stars that might have been expelled by the
Quintuplet cluster are located.
the northern limit of the XMM-Newton scan (b ∼ 0.15◦ ; see Fig.
10, 11 and 15). Tsuru et al. (2009) suggested that the Chimney
is an outflow emanating from the supernova remnant candidate
G359.41-0.12. They estimated a thermal energy and dynamical
time for G359.41-0.12 and the Chimney of Eth = 5.9 × 1049 erg,
tdy = 2.4 × 104 yr and Eth = 7.6 × 1049 erg, tdy = 4 × 104 yr,
respectively. The energetics and time-scales are consistent with typical GC supernova remnants and Tsuru et al. suggested that the very
peculiar morphology of the outflow producing the Chimney might
be due to a peculiar distribution of molecular clouds that block the
plasma expansion in the other directions (Tsuru et al. 2009). We
note that the morphology of the Chimney resembles that of the
Radio Arc. It originates near the Galactic plane (where dense and
massive molecular clouds are located) and extends almost perpendicular to the Galactic plane. Within the gap between the Chimney
and the western edge (see Fig. 6), a bright non-thermal radio filament with an X-ray counterpart is observed: G359.54+0.18 – the
Ripple filament, with a radio length of 0.08◦ (Lu et al. 2003; YusefZadeh et al. 2005). It is oriented parallel to the edge of the soft Xray plasma distribution (Bally et al. 1989; Yusef-Zadeh et al. 1997;
2004; Staguhn et al. 1998; see also Yamauchi et al. 2014). Similar to the Radio Arc, other X-ray and non-thermal radio filaments
are observed at the base of the Chimney. The high concentration of
non-thermal filaments indicates the importance of magnetic structures in this region (e.g., Morris 2014).
The bright IR source AFGL5376 is located further to the
northwest, along the continuation of the sharp X-ray edge and the
Chimney (unfortunately just off the XMM-Newton map; see Uchida
et al. 1990; 1994). It is associated with high-velocity CO emission and defines the most prominent portion of a strong large scale
(∼ 90 pc) shock front that extends all the way down to Sgr C
(Uchida et al. 1994). Because the Chimney appears to be associated
with a shock, and because it is spatially coincident with magnetic
filaments along its length (Yusef-Zadeh et al. 2004), we suggest
that it is not a simple supernova remnant, but is a phenomenon associated with a footpoint of a larger scale structure, the GCL.
8.7.5
Is the outflow confined inside the GCL? Inside the Fermi
bubbles?
The GC is considered a mini-starburst environment, producing intense outflows (Crocker 2012; Yoast-Hull et al. 2014). The warm
plasma detected at high-latitudes is therefore, most probably, associated with intense star formation and it can be a pervasive atmosphere above the entire CMZ. Even in the absence of another
confining force, the gravitational potential (e.g., that of Breitschwerdt et al. 1991; Launhardt et al. 2002) would bind the ∼ 1 keV
plasma (having a sound speed of ∼ 500 km s−1 ; Muno et al. 2004)
to the Galaxy, but in hydrostatic equilibrium, would allow it to extend to heights of several hundred parsecs. If so, it would require a
large average star formation rate and concomitant energy input to
generate and maintain it.
However, the detection of edges in its distribution suggests
that such plasma might be confined within known structures. As
noted by Blanton (2008), the locations of the eastern and western
footpoints of the GCL (Law 2011) correspond to the positions of
the Radio Arc and the Chimney, respectively. Indeed, the GCL and
its possible magnetic nature might confine the warm plasma observed in soft X-rays. This opens the exciting possibility that the
observed high-latitude enhanced X-ray emission from the GC ”atmosphere” is indeed the warm plasma filling the GCL. Based on
the spectral fit, we deduce a density of ne = 0.06 cm−3 inside
the GCL. Assuming uniform physical conditions inside the GCL (a
cylinder of 45 pc radius and 160 pc height) and extrapolating over
the entire GCL, we estimate a mass of ∼ 4 × 103 M filling the
GCL with a total thermal energy of ∼ 1052 erg. This value is of the
c 0000 RAS, MNRAS 000, 1–??
33
same order of magnitude as the energy required to inflate the GCL
as estimated by Law (2011).
Just after the first detection of the GCL, Uchida et al. (1985)
noted its similarity with the lobes in nearby radio galaxies (although smaller in size and strength). The authors interpreted the
lobes as created by a magneto-dynamic acceleration mechanism
where the magnetic twist is produced by the rotation of a contracting disc of gas in the Galactic plane. Under such conditions, the
plasma is accelerated into a conical cylinder with a helical velocity
field (Uchida et al. 1985). Alternatively, Bland-Hawthorn & Cohen (2003) suggested that the GCL could be produced by a largescale bipolar Galactic wind, that would be the result of a powerful (E = 1054−55 erg) nuclear starburst that took place a few
106 yr ago. These authors show that dust is associated with the
entire GCL structure and they suggest that the GC (and the centers of many Galaxies) would drive large-scale winds into the halo
with a recurrence time of about 10 Myr (Bland-Hawthorn & Cohen
2003). Other, alternative, scenarios for the origin of the GCL involve outflows associated with enhanced activity of Sgr A? (Ponti
et al. 2013) or intense star formation (Crocker et al. 2011; 2012).
It is not excluded that the GCL could be simply one part of
an even larger scale feature extending over a physical scale of several kiloparsecs above and below the Galaxy, the so called Fermi
bubbles (Su et al. 2010). These gamma-ray bubbles, detected with
Fermi, are interpreted as produced by highly relativistic particles
emitting brightly at GeV energies and beyond and they appear
to contain and confine soft X-ray emitting plasma traced by the
ROSAT all sky survey, from very large scales down to the Milky
Way’s center (Su et al. 2010). However, close to the Galactic plane,
the bubbles’ edges start to become confused. Whatever their origin
might be, the Fermi bubbles appear to originate (and be collimated)
from the CMZ, within the region that XMM-Newton scanned here.
Additional XMM-Newton observations at high Galactic latitudes, in particular, inside and at the border of the GCL, covering the AFGL 5376 source and the edges of the base of the Fermi
bubbles will be needed to measure the extent of this high-latitude
emission and to help disentangle the hypotheses for its origin. Furthermore, higher spatial resolution observations (such as provided
by Chandra) at high latitudes would allow one to pin down what
fraction of the extended, high-latitude X-ray emission is associated
with faint point sources that are relatively less subject to extinction
than sources near the plane.
8.8
Soft X-ray emission from the Sgr D and Sgr E regions
Intense, soft, diffuse X-ray emission is observed from G359.120.05, the region around 1E1740.7-2942. A radio SNR (G359.070.02) is observed at about the same position (LaRosa et al.
2000). G359.12-0.05 has an emission spectrum typical of an SNR
(Nakashima et al. 2010). In particular, the high extinction suggests it is located at the GC. Nakashima et al. (2010) suggest
that G359.12-0.05 might be associated with the great annihilator
and therefore be the second system (such as SS433 and the radio
SNR W 50) where a BH is associated with its SNR.
The core of the Sgr D complex is also observed to show enhanced medium energy emission (see Fig. 3). In radio, a SNR
southwest of Sgr D’s core and H II regions are clearly observed
(Fig. 6; LaRosa et al. 2000). Sawada et al. (2009) analysed X-ray
data from the XMM-Newton and Suzaku satellites and observed soft
X-ray emission from two diffuse X-ray sources, DS1 and one associated with the core of Sgr D. They suggest that DS1 is a new SNR
at the GC.
c 0000 RAS, MNRAS 000, 1–??
8.9
Star formation estimate from counts of supernova
remnants
We observe a total of ∼ 10 − 12 supernova remnant candidates
in the CMZ (plus ∼ 5 independent radio SNR; see Tab. 7) plus
two superbubbles, each likely created by many (3-10) supernova
events. These remnants have typical estimated ages of a few tens
of thousands of years and temperatures of kT ∼ 0.4 − 1.5 keV.
Due to the presence of the two superbubbles requiring multiple
SNe and the high absorption towards the GC (such as in the star
forming region Sgr B) that hampers us from observing a potentially larger population of remnants characterised by lower temperatures, the number of SNR observed in the GC is most probably under-estimated. However, assuming lifetimes of 10-40 kyr,
the observed number of SNR yields a rate, averaged over the past
several thousands of years, of rSN ∼ 3.5 − 15 × 10−4 yr−1 ,
consistent with other estimates (Crocker 2012). This implies a kinetic energy input higher than 1.1 × 1040 erg s−1 . To estimate the
star formation rate, we assume that all stars with masses greater
than 8 M produce supernovae and that all SNR are observable.
Therefore, we multiply the supernova rate by the integral of the
initial mass function (IMF) over all masses divided by the integral
of the IMF above 8 M . To reflect the GC IMF, we assume the
Kroupa (2002) formulation. The star formation rate then results to
be: rSFR ∼ 0.035 − 0.15 M yr−1 . If the IMF in the CMZ is
top-heavy, as some have argued, then a smaller star formation rate
is implied.
As noted also by Mori et al. (2008; 2009) and Heard &
Warwick (2013), we observe that the two superbubbles have far
hotter temperatures (higher density and smaller size) than all the
ones observed in the Galactic plane or in the Large Magellanic
Cloud (typically with temperatures kT ∼ 0.1 − 0.3 keV, densities ne ∼ 0.01 − 0.03 cm−2 , sizes l ∼ 140 − 450 pc; but see
also Sasaki et al. 2011; Kavanagh et al. 2012). This could simply
be the consequence of the high extinction towards the GC, hiding
a population of normal and very soft superbubbles, or it could be a
characteristic feature of GC superbubbles, inducing a different evolution because of the interaction with the peculiar GC environment.
Further investigation is required to solve this problem.
9
CONCLUSIONS
We have systematically analysed more than 100 XMM-Newton observations pointed within one degree of Sgr A? and have created
the deepest, few arcsec resolution, X-ray images of the CMZ. This
includes a total of about 1.5 Ms of EPIC-pn cleaned exposure in the
central 15 arcmin and about 200 ks at all other points of the Central Molecular Zone (CMZ). We present here, for the first time, not
only broad-band X-ray continuum maps, but also mosaicked maps
of both soft line intensities and inter-line emission from the entire
CMZ region.
• The remarkably similar distributions of both the soft line emitting plasma (Si XIII, S XV, Ar XVII and Ca XIX) and the soft continuum (intra-line bands) indicate that most of the diffuse soft X-ray
emission arises from a thermal process generating both continuum
and lines.
• Starting from the mosaic maps of the different narrow energy bands and assuming the GC emission is produced by three
different components, we fit the maps at different energies and
derive the integrated intensity map of the thermal plasma emission. Integrating over the entire CMZ, the total observed (un-
34
G. Ponti et al.
absorbed) flux is: F2−4.5 keV = 4.2 × 10−11 erg cm−2 s−1 ,
F4.5−12 keV = 1.2 × 10−10 erg cm−2 s−1 , corresponding to a
luminosity of L2−12 = 3.4 × 1036 erg s−1 at an assumed 8 kpc
distance.
• Counting the number of supernova remnants in the CMZ, we
estimate a supernova rate between rSN ∼ 3.5 − 15 × 10−4 yr−1 ,
consistent with other estimates (Crocker 2012), that corresponds
to a star formation rate of rSFR ∼ 0.035 − 0.15 M yr−1 over
the past several thousand years. This implies a kinetic energy input
greater than 1.5 × 1040 erg s−1 .
• We report the discovery of a new X-ray filament
XMM J0.173-0.413 perpendicular to the Galactic plane and
south of the GC Radio Arc spatially corresponding to a nonthermal radio filament. XMM J0.173-0.413 is the first of the four
cases known where the X-ray emission is not at or near a location
where the radio filaments show unusually strong curvature.
• The soft GC X-ray emission is absorbed not only by highcolumn-density foreground clouds located in the Galactic disk, but
also by some clouds located on the near side of the CMZ, such as
the core and envelope of Sgr B2, M0.25+0.01 (the ”Brick”), and
even a few clouds at higher Galactic latitudes, M0.18+0.126 and
M0.20-0.48. However, the majority of the observed variations in
the soft X-ray emission are true emissivity modulations and not a
product of absorption.
• Several SNR candidates are identified by their soft X-ray
emission that appears to fill holes in the column density distribution of gas-and-dust derived from submillimeter maps.
• Our data shed new light on two quasi-symmetric lobes situated
to Galactic north and south of Sgr A? . The Northern lobe shows a
bright and sharp transition at its edge, suggesting the presence of a
shock. Such features are possibly the remnant of the SN that generated SGR J1745-2900 or the PWN candidate G359.945-0.044.
Alternatively, the lobes might constitute a long-lived bipolar structure produced by an isotropic outflow produced by either 1) the
cumulative winds from the young stars of the central cluster, 2) a
wind associated with the accretion flow onto Sgr A? , or 3) the same
process that generated the X-ray reflection nebulae (if such activity
has been recurrent over the past millennia).
• The uniform X-ray colour of the superbubble G359.9-0.125,
its sharp external edge and its being filled with S XV emitting plasma suggest that the soft X-ray features southwest of
Sgr A? form a unique shell-like structure with total energy Eth ∼
1051 erg, therefore making it a superbubble candidate in the GC
(high absorption indicates that G359.9-0.125 is located at the GC).
Alternatively, it might be the remnant of a very energetic event at
the GC, such as a tidal disruption event.
• We discover new evidence for the GC superbubble G0.1-0.1,
also known as the arc-bubble from mid-IR observations: its soft Xray (e.g. S XV) emission completely fills the mid-IR bubble, and
indicates a thermal energy as large as Eth ∼ 1.5 × 1051 erg. At
present the Quintuplet cluster, which is moving at very high speed
through the CMZ, is located at the border of the superbubble. However, it was more centrally located a few 104 yr ago and it could
have, at least in part, energised it. We do not observe similar soft Xray emission trailing the Arches cluster, but this might be ascribed
to its younger age.
• We suggest that the Galactic Center Lobe might be a magnetic
structure filled with warm, soft, X-ray-emitting plasma. In fact, we
observe: i) enhanced soft X-ray emission at high Galactic latitudes;
ii) enhanced soft X-ray emission at, and between, the longitudes
of the Radio Arc and the Chimney associated with Sgr C, corresponding to the east and west foot-points of the GCL; iii) a sharp
edge (at l = 359.63◦ , b = 0.06◦ and l = 359.55◦ , b = 0.46◦ ),
running parallel to the nonthermal ripple filament (G359.54+0.18)
and Sgr C thread, defining the western border of the enhanced soft
X-ray emission. The GCL could be the relatively small base of an
even larger structure, the so-called Fermi Bubbles. Additional observations will be needed to clarify this.
• A new very faint X-ray transient, XMMU J17450.3-291445,
has been discovered during the new XMM-Newton campaign to
reach a peak luminosity of LX ∼ 1035 erg s−1 for ∼ 2 hr (Soldi et
al. 2014).
ACKNOWLEDGMENTS
This research has made use both of data obtained with XMMNewton, an ESA science mission with instruments and contributions directly funded by ESA Member States and NASA, and data
obtained from the Chandra Data Archive. We kindly acknowledge
Sergio Molinari for providing the Herschel map, Casey Law for the
GBT images and Namir Kassim for the VLA 90-cm map. GP acknowledges Roland Crocker, Barbara De Marco and Pierre Maggi
for useful discussions. GP also acknowledges Frederick Baganoff
and Nanda Rea for discussions about the origin of the lobes and
the association with the SNR of SGR J1745-2900. We thank the
referee for a careful reading of the paper. GP acknowledges support via an EU Marie Curie Intra-European Fellowship under contract no. FP7-PEOPLE-2012-IEF-331095. The GC XMM-Newton
monitoring project is partially supported by the Bundesministerium
für Wirtschaft und Technologie/Deutsches Zentrum für Luft- und
Raumfahrt (BMWI/DLR, FKZ 50 OR 1408) and the Max Planck
Society. Partial support through the COST action MP0905 Black
Holes in a Violent Universe is acknowledged. The authors thank
the ISSI in Bern.
REFERENCES
Aharonian, F., Akhperjanian, A. G., Aye, K.-M., et al. 2005,
A&A, 432, L25
Alexander, T., & Hopman, C. 2003, ApJ, 590, L29
Amo-Baladrón, M. A., Martı́n-Pintado, J., Morris, M. R., Muno,
M. P., & Rodrı́guez-Fernández, N. J. 2009, ApJ, 694, 943
Anantharamaiah, K. R., Pedlar, A., Ekers, R. D., & Goss, W. M.
1991, MNRAS, 249, 262
Ayal, S., Livio, M., & Piran, T. 2000, ApJ, 545, 772
Baganoff, F. K., Bautz, M. W., Brandt, W. N., et al. 2001, Nature,
413, 45
Baganoff, F. K., Maeda, Y., Morris, M., et al. 2003, ApJ, 591, 891
Bally, J., & Yusef-Zadeh, F. 1989, ApJ, 336, 173
Bamba, A., Yokogawa, J., Sakano, M., & Koyama, K. 2000, PASJ,
52, 259
Bamba, A., Murakami, H., Senda, A., et al. 2002, arXiv:astroph/0202010
Bamba, A., Yamazaki, R., Kohri, K., et al. 2009, ApJ, 691, 1854
Barrière, N. M., Tomsick, J. A., Baganoff, F. K., et al. 2014, ApJ,
786, 46
Bélanger, G., Goldwurm, A., Melia, F., et al. 2005, ApJ, 635, 1095
Bélanger, G., Goldwurm, A., Renaud, M., et al. 2006, ApJ, 636,
275
Blanton, M. C. 2008, Ph.D. Thesis,
Bland-Hawthorn, J., & Cohen, M. 2003, ApJ, 582, 246
c 0000 RAS, MNRAS 000, 1–??
35
EPIC-pn
Exp
Exp
Mod
ID
EPIC-MOS1
Exp
Exp
Mod
ID
EPIC-MOS2
Exp
Exp
Mod
ID
EPIC-pn
mode
filter
EPIC-MOS1
mode
filter
EPIC-MOS2
mode
filter
rev
0694640101
0694640201
0694640301
0694640401
0694640501
0694640601
0694640701
0694640801
0694640901
0694641001
0694641101
0694641201
0694641301
0694641401
0694641501
0694641601
2335
2335
2335
2335
2335
2335
2335
2335
2335
2335
2335
2335
2335
2335
2335
2335
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
003
003
003
003
003
003
003
003
003
003
003
003
003
003
003
003
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
NEW CMZ XMM-Newton scan
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
6.0
6.0
6.0
8.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
2.0
2.0
2.0
2.5
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.5
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
0112970101
0112970201
0112970401
0112970501
0112970701
0112970801
0112971001
0112971301
0112971501
0112971601
0112971701
0112971801
0112971901
0112972101
0145
0145
0143
0144
0139
0144
0145
0143
0240
0240
0240
0240
0240
0318
U
S
S
S
S
S
S
S
S
S
S
S
S
S
002
003
003
003
003
003
003
003
003
003
003
003
003
003
U
S
S
S
S
S
S
S
S
S
S
S
S
S
OLD CMZ XMM-Newton scan
002
U
002
FF
001
S
002
eFF
001
S
002
eFF
001
S
002
eFF
001
S
002
eFF
001
S
002
eFF
001
S
002
FF
001
S
408
SW
001
S
002
eFF
001
S
002
eFF
011
S
010
SW
001
S
002
eFF
001
S
002
eFF
001
S
002
eFF
med
med
med
med
med
med
tck
med
med
med
med
med
med
med
FF
FF
FF
FF
FF
FF
FF
TU
FF
FF
TU
FF
FF
FF
med
med
med
med
med
med
med
med
med
med
med
med
med
med
FF
FF
FF
FF
FF
FF
FF
RF
FF
FF
SW
FF
FF
FF
med
med
med
med
med
med
med
med
med
med
med
med
med
med
6.0
6.0
6.0
6.0
6.0
6.0
5.0
6.0
6.0
6.0
6.0
6.0
5.0
6.0
2.0
2.0
2.0
2.0
2.0
1.5
1.5
1.5
2.0
2.0
2.0
2.0
1.5
1.5
2.0
2.0
2.0
2.0
2.0
1.5
1.5
1.5
2.0
2.0
2.0
2.0
1.5
1.5
FF
FF
tck
tck
FF
FF
med
med
FF
FF
med
med
6.0
6.0
2.0
2.0
2.0
2.0
eF
eF
FF
FF
F med
F med
med
med
FF
FF
FF
FF
med
med
med
med
FF
FF
FF
FF
med
med
med
med
6.0
6.0
6.0
6.0
2.0
2.0
2.0
1.5
2.0
2.0
2.0
1.5
FF
FF
med
med
FF
FF
med
med
FF
FF
med
med
6.0
7.0
2.0
2.0
2.0
2.0
FF
FF
FF
FF
med
med
med
med
FF
FF
FF
FF
med
med
med
med
FF
FF
FF
FF
med
med
med
med
8.0
8.0
8.0
8.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
FF
FF
FF
thn
thn
med
FF
FF
FF
thn
thn
med
FF
FF
FF
thn
thn
med
7.0
7.0
8.0
2.0
2.5
2.0
2.0
2.5
2.0
FF
FF
FF
med
med
med
FF
FF
FF
med
med
med
FF
FF
FF
med
med
med
8.0
8.0
8.0
2.5
2.5
2.5
2.5
2.5
2.5
FF
FF
FF
FF
FF
FF
FF
med
med
med
med
med
med
med
FF
FF
FF
FF
FF
FF
FF
med
med
med
med
med
med
med
FF
FF
FF
FF
FF
FF
FF
med
med
med
med
med
med
med
7.0
7.0
8.0
7.0
7.0
8.0
7.0
2.0
2.0
2.0
2.0
2.0
2.5
1.8
2.0
2.0
2.0
2.0
2.0
2.5
1.8
FF
FF
FF
FF
FF
med
med
med
med
med
FF
FF
FF
FF
FF
med
med
med
med
med
FF
FF
FF
FF
FF
med
med
med
med
med
7.0
7.0
6.0
7.0
6.0
2.0
2.0
1.8
1.8
2.0
2.0
2.0
1.8
1.8
2.0
0111350101
0111350301
0406
0516
U
S
002
001
S
S
0202670501
0202670601
0202670701
0202670801
0788
0789
0866
0867
U
S
S
S
002
003
003
003
U
S
S
S
0302882601
0302884001
1139
1236
S
S
003
003
S
S
0402430301
0402430401
0402430701
0504940201
1339
1340
1338
1418
S
U
S
S
001
002
001
003
S
U
S
S
0511000301
0511000401
0505670101
1508
1610
1518
S
S
U
003
003
002
S
U
U
0554750401
0554750501
0554750601
1705
1706
1707
S
S
U
003
003
002
S
S
S
0604300601
0604300701
0604300801
0604300901
0604301001
0658600101
0658600201
2069
2070
2071
2072
2073
2148
2148
S
U
U
S
S
S
S
003
002
002
003
003
001
001
S
S
U
S
S
S
S
0674600601
0674600701
0674600801
0674601001
0674601101
2245
2246
2248
2249
2247
S
S
S
S
S
003
003
003
003
003
S
S
S
S
U
Pointing toward Sgr A?
2002
006
S
005
006
S
005
2004
003
U
003
001
S
002
001
S
002
001
S
002
2006
001
S
002
001
S
002
2007
002
S
003
002
U
002
002
S
003
001
S
002
2008
001
S
002
002
U
002
002
U
002
2009
001
S
002
001
S
002
001
S
002
2011
001
S
002
001
S
002
002
U
002
001
S
002
001
S
002
002
S
003
002
S
003
2012
001
S
002
001
S
002
001
S
002
001
S
002
002
U
002
pn
c/s
Threshold
M1
M2
c/s
c/s
OBSID
Table 8. List of all XMM-Newton observations considered in this work. Exposure Mode: U, S stand for unscheduled and scheduled, respectively. Filters: Med,
thn, tck stand for medium, thin and thick filters, respectively. FF, eFF, SW, Ti, TU stand for full frame, extended full frame, small window, timing and time
uncompressed, respectively.
Borkowski, K. J., Reynolds, S. P., Hwang, U., et al. 2013, ApJ,
771, LL9
Breitschwerdt, D., McKenzie, J. F., & Voelk, H. J. 1991, A&A,
245, 79
Capelli, R., Warwick, R. S., Porquet, D., Gillessen, S., & Predehl,
P. 2011, A&A, 530, AA38
Capelli, R., Warwick, R. S., Porquet, D., Gillessen, S., & Predehl,
P. 2012, A&A, 545, A35
Cash, W. 1979, ApJ, 228, 939
Castor, J., McCray, R., & Weaver, R. 1975, ApJ, 200, L107
Castro, M., Maiolino, T., D’Amico, F., et al. 2013,
arXiv:1302.6213
Chuss, D. T., Davidson, J. A., Dotson, J. L., et al. 2003, ApJ, 599,
c 0000 RAS, MNRAS 000, 1–??
1116
Clark, P. C., Glover, S. C. O., Ragan, S. E., Shetty, R., & Klessen,
R. S. 2013, ApJ, 768, L34
Clavel, M., Terrier, R., Goldwurm, A., et al. 2013, A&A, 558, A32
Clavel, M., Soldi, S., Terrier, R., et al. 2014, MNRAS, 443, L129
Coil, A. L., & Ho, P. T. P. 2000, ApJ, 533, 245
Cotera, A. S., Erickson, E. F., Colgan, S. W. J., et al. 1996, ApJ,
461, 750
Coti Zelati, F., Rea, N., Papitto, A., et al. 2015, MNRAS, 449,
2685
Crocker, R. M., Jones, D. I., Melia, F., Ott, J., & Protheroe, R. J.
2010, Nature, 463, 65
Crocker, R. M., & Aharonian, F. 2011, Physical Review Letters,
36
G. Ponti et al.
OBSID
rev
0030540101
0144220101
0152920101
0144630101
0203930101
0205240101
0304220301
0304220101
0303210201
0302882501
0302882701
0302882801
0302882901
0302883001
0302883101
0302883201
0305830701
0302883901
0302884101
0302884201
0302884301
0302884401
0302884501
0406580201
0410580401
0410580501
0400340101
0506291201
0504940101
0504940401
0504940501
0504940601
0504940701
0511010701
0511000101
0511000501
0511000701
0511000901
0511001101
0511001301
0511000201
0511000601
0511000801
0511001001
0511001201
0511001401
0505870301
0603850201
0655670101
0504
0596
0607
0688
0868
0956
1048
1063
1065
1139
1139
1139
1139
1139
1139
1139
1157
1236
1236
1236
1236
1236
1236
1241
1243
1245
1244
1322
1418
1418
1418
1418
1418
1505
1508
1508
1508
1508
1508
1508
1510
1510
1512
1512
1512
1512
1511
1891
2065
EPIC-pn
Exp
Exp
Mod
ID
S
U
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
N
N
S
N
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
N
003
002
003
003
003
003
004
003
003
003
003
003
003
003
003
003
003
003
003
003
003
003
003
003
000
000
003
000
003
003
003
003
003
003
003
003
003
003
003
003
003
003
003
003
003
003
003
003
000
EPIC-MOS1
Exp
Exp
Mod
ID
S
U
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
EPIC-MOS2
Exp
Exp
Mod
ID
EPIC-pn
mode
filter
Other observations of the CMZ
001
S
002
SW
002
U
002
SW
001
S
002
FF
001
S
002
SW
001
S
002
eF
001
S
002
FF
002
S
003
SW
001
S
002
SW
001
S
002
SW
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
002
S
003
Ti
002
S
003
Ti
001
S
002
FF
001
S
002
Ti
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
FF
001
S
002
Ti
tck
med
tck
med
F med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
tck
tck
med
med
med
med
med
med
med
med
thn
thn
thn
thn
thn
thn
thn
thn
thn
thn
thn
thn
med
med
med
EPIC-MOS1
mode
filter
SW
FF
FF
TU
FF
FF
FF
FF
TU
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
tck
med
tck
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
tck
tck
med
med
med
med
med
med
med
med
thn
thn
thn
thn
thn
thn
thn
thn
thn
thn
thn
thn
med
med
med
EPIC-MOS2
mode
filter
SW
FF
FF
SW
FF
FF
FF
FF
TU
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
TU
TU
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
FF
tck
med
tck
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
med
tck
tck
med
med
med
med
med
med
med
med
thn
thn
thn
thn
thn
thn
thn
thn
thn
thn
thn
thn
med
med
med
pn
c/s
Threshold
M1
M2
c/s
c/s
6.0
6.0
6.0
2.0
1.5
1.5
2.0
1.5
1.5
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
7.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
8.0
6.0
6.0
6.0
6.0
1.5
1.5
1.5
1.5
1.5
2.0
2.0
2.0
2.0
2.0
2.0
2.0
1.5
2.0
2.0
2.0
2.5
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.5
2.0
2.0
2.0
2.5
2.5
2.0
2.5
2.5
2.5
2.5
2.5
2.5
2.0
2.0
2.0
1.5
1.5
1.5
1.5
1.5
2.0
2.0
2.0
2.0
2.0
2.0
2.0
1.5
2.0
2.0
2.0
2.5
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.5
2.0
2.0
2.0
2.5
2.5
2.0
2.5
2.5
2.5
2.5
2.5
2.5
2.0
2.0
2.0
Table 9. List of all XMM-Newton observations considered in this work. Exposure Mode: U, S stand for unscheduled and scheduled, respectively. Filters: Med,
thn, tck stand for medium, thin and thick filters, respectively. FF, eFF, SW, Ti, TU stand for full frame, extended full frame, small window, timing and time
uncompressed, respectively.
106, 101102
Crocker, R. M. 2012, MNRAS, 423, 3512
Decourchelle, A. 2003, XMM-Newton Proposal, 210
Degenaar, N., & Wijnands, R. 2010, A&A, 524, A69
Degenaar, N., Wijnands, R., Cackett, E. M., et al. 2012, A&A,
545, A49
Degenaar, N., Reynolds, M. T., Miller, J. M., Kennea, J. A., &
Wijnands, R. 2013, The Astronomer’s Telegram, 5006, 1
Del Santo, M., Sidoli, L., Bazzano, A., et al. 2006, A&A, 456,
1105
Do, T., Lu, J. R., Ghez, A. M., et al. 2013, ApJ, 764, 154
Dong, H., Wang, Q. D., Cotera, A., et al. 2011, MNRAS, 417, 114
Dong, H., Wang, Q. D., & Morris, M. R. 2012, MNRAS, 425, 884
Downes, D., & Maxwell, A. 1966, ApJ, 146, 653
Dubner, G., Giacani, E., & Decourchelle, A. 2008, A&A, 487,
1033
Dutra, C. M., & Bica, E. 2000, A&A, 359, L9
Dutra, C. M., Ortolani, S., Bica, E., et al. 2003, A&A, 408, 127
Dwelly, T., & Ponti, G. 2013, TheAstronomer’s Telegram, 5008,
1
Ebisawa, K., Maeda, Y., Kaneda, H., & Yamauchi, S. 2001, Science, 293, 1633
Ferreira, S. E. S., & de Jager, O. C. 2008, A&A, 478, 17
Ferrière, K. 2009, A&A, 505, 1183
Ferrière, K. 2011, The Galactic Center: a Window to the Nuclear
Environment of Disk Galaxies, 439, 39
Figer, D. F., Kim, S. S., Morris, M., et al. 1999, ApJ, 525, 750
Freyberg, M. J., Briel, U. G., Dennerl, K., et al. 2004, Proc. SPIE,
5165, 112
Fukuoka, R., Koyama, K., Ryu, S. G., & Tsuru, T. G. 2009, PASJ,
61, 593
Gaensler, B. M., Pivovaroff, M. J., & Garmire, G. P. 2001, ApJ,
556, L107
Gaensler, B. M., van der Swaluw, E., Camilo, F., et al. 2004, ApJ,
616, 383
Geballe, T. R., Wade, R., Krisciunas, K., Gatley, I., & Bird, M. C.
1987, ApJ, 320, 562
Genzel, R., Eisenhauer, F., & Gillessen, S. 2010, Reviews of Modern Physics, 82, 3121
Goldwurm, A., Brion, E., Goldoni, P., et al. 2003, ApJ, 584, 751
Gray, A. D. 1994, MNRAS, 270, 835
Green, D. A. 2014, arXiv:1409.0637
Haberl, F., & Pietsch, W. 1999, A&AS, 139, 277
Haberl, F., Sturm, R., Ballet, J., et al. 2012, A&A, 545, A128
Habibi, M., Stolte, A., Brandner, W., Hußmann, B., & Motohara,
K. 2013, A&A, 556, AA26
Habibi, M., Stolte, A., & Harfst, S. 2014, A&A, 566, AA6
Hales, C. A., Gaensler, B. M., Chatterjee, S., van der Swaluw, E.,
& Camilo, F. 2009, ApJ, 706, 1316
Harrison, F. A., Craig, W. W., Christensen, F. E., et al. 2013, ApJ,
c 0000 RAS, MNRAS 000, 1–??
37
OBSID
obs date
0694640101
0694640201
0694640301
0694640401
0694640501
0694640601
0694640701
0694640801
0694640901
0694641001
0694641101
0694641201
0694641301
0694641401
0694641501
0694641601
2012-09-07
2012-08-30
2012-08-31
2012-09-02
2012-09-05
2012-09-06
2012-10-02
2012-10-06
2012-09-12
2012-09-23
2012-09-24
2012-09-26
2012-09-26
2012-09-30
2012-10-06
2012-10-08
0112970101
0112970201
0112970401
0112970501
0112970701
0112970801
0112971001
0112971301
0112971501
0112971601
0112971701
0112971801
0112971901
0112972101
2000-09-23
2000-09-23
2000-09-19
2000-09-21
2000-09-11
2000-09-21
2000-09-24
2000-09-19
2001-04-01
2001-03-31
2001-03-31
2001-04-01
2001-04-01
2001-09-04
0111350101
0111350301
2002-02-26
2002-10-03
0202670501
0202670501
0202670501
0202670601
0202670601
0202670701
0202670801
2004-03-28
2004-03-28
2004-03-30
2004-03-30
2004-03-30
2004-08-31
2004-09-02
0302882601
0302884001
2006-02-27
2006-09-08
0402430301
0402430401
0402430701
0504940201
2007-04-01
2007-04-03
2007-03-30
2007-09-06
0511000301
0511000401
0505670101
2008-03-03
2008-09-23
2008-03-23
0554750401
0554750501
0554750601
2009-04-01
2009-04-03
2009-04-05
0604300601
0604300701
0604300801
0604300901
0604301001
0658600101
0658600201
2011-03-28
2011-03-30
2011-04-01
2011-04-03
2011-04-05
2011-08-31
2011-09-01
0674600601
0674600701
0674600801
0674601001
0674601101
2012-03-13
2012-03-15
2012-03-19
2012-03-21
2012-03-17
Exp pn
Exp M1
Exp M2
Exp pn
Exp M1
Exp M2
NEW CMZ XMM-Newton scan
41978
43452
43605
45035
46616
46619
40041
41616
41619
52954
51442
51460
44976
46606
46621
40042
41614
41621
42539
44099
44120
40041
41616
41619
43031
44617
44604
46021
47607
47620
40041
41616
41619
40008
41559
41577
53842
56260
56348
45816
46751
46920
49746
51483
51486
40005
41585
41585
38739
45038
40041
38736
32935
40042
42539
40041
42202
46041
40041
40008
46667
32466
39167
27250
38980
46616
41616
40073
33180
41614
44117
41616
43784
47614
41616
41588
48012
33767
40518
27795
38985
46619
41619
40075
33185
41621
44120
41619
43786
47620
41619
41598
48018
33770
40507
27803
OLD CMZ XMM-Newton scan
12870
15806
15611
13499
17394
17392
25411
29365
29391
21119
24914
24911
19518
23419
23413
19969
23892
23892
12599
16492
16482
12800
0
13091
20293
25020
25017
0
3996
3949
11000
0
11799
9927
14513
14542
4698
9191
9191
21687
26039
26055
12252
12999
21880
10289
19383
13462
8774
0
6752
0
0
1900
4147
20130
14679
16894
23849
14084
23221
17198
12529
0
7017
0
0
2069
8379
23515 2
14637
16892
23847
14081
23218
17198
12529
0
7017
0
0
2069
8379
3517
40030
8261
52118
9877
52120
9880
45847
0
0
56926
0
78857
91795
0
0
0
0
0
78921
93131
0
0
0
0
0
78915
93126
1700
4787
3160
6365
3163
6370
50962
36886
21240
7392
50955
36892
22820
8949
50958
36876
22825
8960
3305
5058
64200
4863
4358
65143
4868
4342
65153
31934
38634
31485
33358
40216
37464
33363
40218
37466
28768
32872
33771
19941
32571
47653
39634
30121
39149
36149
21140
33917
49169
41109
30119
39156
36129
21143
33914
49177
41115
8594
6802
16784
19841
8956
9296
8209
18358
21416
8173
9301
8212
18358
21419
8178
Pointing toward Sgr A?
52105
52120
16960
16996
2004
110170
5733
6087
0
107784
108572
0
650
848
112204
585
538
0
120863
122251
127470
132469
132503
130951
132997
133036
2006
4937
6563
6568
4987
6563
6570
2007
101319
93947
94022
93594
97566
96461
32338
33912
33917
11092
12649
12652
2008
5057
6615
6620
5058
4358
4342
96601
97787
97787
2009
38034
39614
39619
42434
44016
44018
32837
38816
38818
2011
45306
48467
48491
42305
48579
48584
37321
38642
38494
36568
37589
37573
48210
47757
47646
47585
49169
49159
51324
52903
52908
2012
19594
21167
21172
14040
15616
15618
21041
22615
22618
22034
23616
23619
25682
24638
24628
40030
15377
Table 10. List of all XMM-Newton observations considered in this work. Total and cleaned exposure time (in seconds) for each camera, respectively.
770, 103
Heard, V., & Warwick, R. S. 2013a, MNRAS, 428, 3462
Heard, V., & Warwick, R. S. 2013b, MNRAS, 434, 1339
Heinke, C. O., Tomsick, J. A., Yusef-Zadeh, F., & Grindlay, J. E.
2009, ApJ, 701, 1627
Henze, M., Pietsch, W., Haberl, F., et al. 2014, A&A, 563, A2
Hewitt, J. W., Yusef-Zadeh, F., & Wardle, M. 2008, ApJ, 683, 189
Ho, P. T. P., Jackson, J. M., Barrett, A. H., & Armstrong, J. T.
1985, ApJ, 288, 575
Inui, T., Koyama, K., Matsumoto, H., & Tsuru, T. G. 2009, PASJ,
61, 241
Johnson, J. L., Greif, T. H., Bromm, V., Klessen, R. S., & Ippolito,
J. 2009, MNRAS, 399, 37
Kaneda, H., Makishima, K., Yamauchi, S., et al. 1997, ApJ, 491,
638
c 0000 RAS, MNRAS 000, 1–??
Kaspi, V. M., Archibald, R. F., Bhalerao, V., et al. 2014, ApJ, 786,
84
Kassim, N. E., & Frail, D. A. 1996, MNRAS, 283, L51
Kavanagh, P. J., Sasaki, M., & Points, S. D. 2012, A&A, 547, A19
Khokhlov, A., & Melia, F. 1996, ApJ, 457, L61
Koch, E. W., Bahramian, A., Heinke, C. O., et al. 2014, MNRAS,
442, 372
Koyama, K., Hyodo, Y., Inui, T., et al. 2007, PASJ, 59, 245
Koyama, K., Inui, T., Hyodo, Y., et al. 2007, PASJ, 59, 221
Koyama, K., Maeda, Y., Sonobe, T., et al. 1996, PASJ, 48, 249
Koyama, K., Hyodo, Y., Inui, T., et al. 2007, PASJ, 59, 245
Koyama, K., Inui, T., Matsumoto, H., & Tsuru, T. G. 2008, PASJ,
60, 201
Koyama, K., Takikawa, Y., Hyodo, Y., et al. 2009, PASJ, 61, 255
Krivonos, R. A., Tomsick, J. A., Bauer, F. E., et al. 2014, ApJ,
38
G. Ponti et al.
OBSID
obs date
0030540101
0144220101
0152920101
0144630101
0203930101
0205240101
0304220301
0304220101
0303210201
0302882501
0302882701
0302882801
0302882901
0302883001
0302883101
0302883201
0305830701
0302883901
0302884101
0302884201
0302884301
0302884401
0302884501
0406580201
0410580401
0410580501
0400340101
0506291201
0504940101
0504940401
0504940501
0504940601
0504940701
0511010701
0511000101
0511000501
0511000701
0511000901
0511001101
0511001301
0511000201
0511000601
0511000801
0511001001
0511001201
0511001401
0505870301
0603850201
0655670101
2002-09-09
2003-03-12
2003-04-02
2003-09-11
2004-09-04
2005-02-26
2005-08-29
2005-09-29
2005-10-02
2006-02-27
2006-02-27
2006-02-27
2006-02-27
2006-02-27
2006-02-27
2006-03-29
2006-04-04
2006-09-08
2006-09-08
2006-09-08
2006-09-09
2006-09-09
2006-09-09
2006-09-18
2006-09-22
2006-09-26
2006-09-24
2007-02-27
2007-09-06
2007-09-06
2007-09-06
2007-09-06
2007-09-06
2008-02-27
2008-03-03
2008-03-04
2008-03-04
2008-03-04
2008-03-04
2008-03-04
2008-09-23
2008-09-23
2008-09-27
2008-09-27
2008-09-27
2008-09-27
2008-03-10
2010-04-07
2011-03-19
Exp pn
Exp M1
Exp M2
Other observations of the CMZ
27689
27842
27844
46746
49905
49843
50182
51639
51774
8469
0
8661
46544
50438
50446
46919
50625
50604
20031
20213
20226
8051
8237
8250
23472
0
0
7561
9176
9178
5237
6851
6869
5937
7558
7571
5936
7566
7569
5937
7540
7558
9814
11432
11448
4896
6518
6526
6399
11266
11256
4987
6565
6568
4987
6565
6570
4987
6565
6570
4987
6565
6568
4036
5616
5621
6787
8364
8370
28034
29607
29609
0
32558
0
0
32116
0
40001
41575
41580
0
38616
38621
5058
6615
6620
5058
6615
6620
5057
6615
6620
5058
6615
6620
5058
6615
6620
7455
9004
9004
6943
8500
8500
5058
6615
6620
5058
6615
6620
5058
6615
6620
5057
6615
6620
5058
6615
6620
5058
6615
6620
5058
6615
6620
5035
6602
6620
5034
6615
6620
5034
6615
6620
5034
6615
6620
29885
31614
31494
22503
21643
21663
0
103934
103954
Exp pn
Exp M1
Exp M2
27339
28820
48486
0
39078
14946
6417
5621
23
6364
2937
5537
4437
3137
8614
3898
0
4787
4000
4987
4987
4037
6787
13896
0
0
16312
0
4958
5058
5006
1720
4558
5803
546
4658
4506
5058
5057
3800
5058
5058
5035
5034
5034
5034
7249
16919
0
27495
31550
50082
316
43003
15251
6615
5816
314
7999
4564
7164
6066
4766
10248
5547
1028
6365
5578
6565
6565
5616
8364
14809
32367
30108
17481
30937
6515
6615
6563
3175
6115
7362
796
6215
6063
6514
6615
5132
6615
6615
6615
6615
6615
6615
7250
18271
80716
27494
31461
50097
311
43013
15243
6620
5821
315
8002
4569
7171
6069
4771
10258
5539
1028
6368
5583
6570
6568
5621
8369
14814
32326
30096
17486
30937
6520
6620
6568
3181
6120
7368
800
6220
6068
6518
6620
5137
6620
6620
6620
6620
6620
6620
7255
18266
80729
Table 11. List of all XMM-Newton observations considered in this work. Total and cleaned exposure time (in seconds) for each camera, respectively.
781, 107
Kroupa, P. 2002, Science, 295, 82
Kuntz, K. D., & Snowden, S. L. 2008, A&A, 478, 575
Kuulkers, E., Shaw, S. E., Paizis, A., et al. 2007, A&A, 466, 595
Lakićević, M., van Loon, J. T., Meixner, M., et al. 2014,
arXiv:1410.5709
Lang, C. C., Morris, M., & Echevarria, L. 1999, ApJ, 526, 727
Lang, C. C., Goss, W. M., & Morris, M. 2002, AJ, 124, 2677
LaRosa, T. N., Kassim, N. E., Lazio, T. J. W., & Hyman, S. D.
2000, AJ, 119, 207
Launhardt, R., Zylka, R., & Mezger, P. G. 2002, A&A, 384, 112
Law, C., & Yusef-Zadeh, F. 2004, ApJ, 611, 858
Law, C. J., Backer, D., Yusef-Zadeh, F., & Maddalena, R. 2009,
ApJ, 695, 1070
Law, C. J., Brentjens, M. A., & Novak, G. 2011, ApJ, 731, 36
Levine, D., Morris, M., & Figer, D. 1999, The Universe as Seen
by ISO, 427, 699
Lu, F. J., Wang, Q. D., & Lang, C. C. 2003, AJ, 126, 319
Lu, F. J., Yuan, T. T., & Lou, Y.-Q. 2008, ApJ, 673, 915
Lu, J. R., Do, T., Ghez, A. M., et al. 2013, ApJ, 764, 155
Maeda, Y., Baganoff, F. K., Feigelson, E. D., et al. 2002, ApJ, 570,
671
Markoff, S. 2010, Proceedings of the National Academy of Science, 107, 7196
Marquez-Lugo, R. A., & Phillips, J. P. 2010, MNRAS, 407, 94
Mauerhan, J. C., Muno, M. P., Morris, M. R., Stolovy, S. R., &
Cotera, A. 2010, ApJ, 710, 706
Mac Low, M.-M., & McCray, R. 1988, ApJ, 324, 776
Mereghetti, S., Sidoli, L., & Israel, G. L. 1998, A&A, 331, L77
Misanovic, Z., Pietsch, W., Haberl, F., et al. 2006, A&A, 448,
1247
Molaro, M., Khatri, R., & Sunyaev, R. A. 2014, A&A, 564,
AA107
Molinari, S., Bally, J., Noriega-Crespo, A., et al. 2011, ApJ, 735,
L33
Mori, H., Maeda, Y., Pavlov, G. G., Sakano, M., & Tsuboi, Y.
2005, Advances in Space Research, 35, 1137
Mori, H., Tsuru, T. G., Hyodo, Y., Koyama, K., & Senda, A. 2008,
PASJ, 60, 183
Mori, H., Hyodo, Y., Tsuru, T. G., Nobukawa, M., & Koyama, K.
2009, PASJ, 61, 687
Mori, K., Gotthelf, E. V., Zhang, S., et al. 2013, ApJ, 770, LL23
Morris, M., & Yusef-Zadeh, F. 1985, AJ, 90, 2511
Morris, M., & Yusef-Zadeh, F. 1989, ApJ, 343, 703
Morris, M. 1990, Galactic and Intergalactic Magnetic Fields, 140,
361
Morris, M., & Serabyn, E. 1996, ARA&A, 34, 645
Morris, M., Baganoff, F., Muno, M., et al. 2003, Astronomische
Nachrichten Supplement, 324, 167
Morris, M., Howard, C., Muno, M., et al. 2004, The Dense Interstellar Medium in Galaxies, 281
Morris, M. R., Meyer, L., & Ghez, A. M. 2012, Research in Astronomy and Astrophysics, 12, 995
Morris, M. R., Zhao, J.-H., & Goss, W. M. 2014a, IAU Symposium, 303, 369
Morris, M. R. 2014b, arXiv:1406.7859
c 0000 RAS, MNRAS 000, 1–??
39
Mossoux, E., Grosso, N., Vincent, F. H., & Porquet, D. 2015,
A&A, 573, A46
Muno, M. P., Baganoff, F. K., Bautz, M. W., et al. 2003, ApJ, 589,
225
Muno, M. P., Baganoff, F. K., Bautz, M. W., et al. 2004, ApJ, 613,
326
Muno, M. P., Pfahl, E., Baganoff, F. K., et al. 2005a, ApJ, 622,
L113
Muno, M. P., Lu, J. R., Baganoff, F. K., et al. 2005b, ApJ, 633,
228
Muno, M. P., Bauer, F. E., Bandyopadhyay, R. M., & Wang, Q. D.
2006, ApJS, 165, 173
Muno, M. P., Baganoff, F. K., Brandt, W. N., Park, S., & Morris,
M. R. 2007, ApJ, 656, L69
Muno, M. P., Baganoff, F. K., Brandt, W. N., Morris, M. R., &
Starck, J.-L. 2008, ApJ, 673, 251
Muno, M. P., Bauer, F. E., Baganoff, F. K., et al. 2009, ApJS, 181,
110
Nakashima, S., Nobukawa, M., Tsuru, T. G., Koyama, K., &
Uchiyama, H. 2010, PASJ, 62, 971
Natalucci, L., Bazzano, A., Cocchi, M., et al. 2004, A&A, 416,
699
Natalucci, L., Tomsick, J. A., Bazzano, A., et al. 2014, ApJ, 780,
63
Najarro, F., Krabbe, A., Genzel, R., et al. 1997, A&A, 325, 700
Neilsen, J., Nowak, M. A., Gammie, C., et al. 2013, ApJ, 774, 42
Nishiyama, S., Tamura, M., Hatano, H., et al. 2009, ApJ, 690,
1648
Nishiyama, S., Yasui, K., Nagata, T., et al. 2013, ApJ, 769, L28
Nobukawa, M., Tsuru, T. G., Takikawa, Y., et al. 2008, PASJ, 60,
191
Nobukawa, M., Sawada, M., Matsumoto, H., Tsuru, T. G., &
Koyama, K. 2009, Advances in Space Research, 43, 1045
Nobukawa, M., Koyama, K., Tsuru, T. G., Ryu, S. G., & Tatischeff, V. 2010, PASJ, 62, 423
Nobukawa, M., Ryu, S. G., Tsuru, T. G., & Koyama, K. 2011,
ApJ, 739, L52
Nord, M. E., Lazio, T. J. W., Kassim, N. E., et al. 2004, AJ, 128,
1646
Novak, G., Chuss, D. T., Renbarger, T., et al. 2003, ApJ, 583, L83
Nynka, M., Hailey, C. J., Reynolds, S. P., et al. 2014, ApJ, 789, 72
Ohnishi, T., Koyama, K., Tsuru, T. G., et al. 2011, PASJ, 63, 527
Ostriker, J. P., & McKee, C. F. 1988, Reviews of Modern Physics,
60, 1
Paizis, A., Ebisawa, K., Takahashi, H., et al. 2009, PASJ, 61, 107
Park, S., Muno, M. P., Baganoff, F. K., et al. 2004, ApJ, 603, 548
Park, S., Muno, M. P., Baganoff, F. K., et al. 2005, ApJ, 631, 964
Paumard, T., Maillard, J. P., Morris, M., & Rigaut, F. 2001, A&A,
366, 466
Pavlinsky, M. N., Grebenev, S. A., & Sunyaev, R. A. 1994, ApJ,
425, 110
Pedlar, A., Anantharamaiah, K. R., Ekers, R. D., et al. 1989, ApJ,
342, 769
Piraino, S., Santangelo, A., Kaaret, P., et al. 2012, A&A, 542, L27
Phillips, J. P., & Marquez-Lugo, R. A. 2010, MNRAS, 409, 701
Ponti, G., Terrier, R., Goldwurm, A., Belanger, G., & Trap, G.
2010, ApJ, 714, 732
Ponti, G., Fender, R. P., Begelman, M. C., et al. 2012, MNRAS,
422, L11
Ponti, G., Morris, M. R., Terrier, R., & Goldwurm, A. 2013, Cosmic Rays in Star-Forming Environments, 34, 331
Ponti, G., Morris, M. R., Clavel, M., et al. 2014, IAU Symposium,
c 0000 RAS, MNRAS 000, 1–??
303, 333
Ponti, G., Bianchi, S., Muñoz-Darias, T., et al. 2015, MNRAS,
446, 1536
Porquet, D., Rodriguez, J., Corbel, S., et al. 2003a, A&A, 406,
299
Porquet, D., Decourchelle, A., & Warwick, R. S. 2003b, A&A,
401, 197
Porquet, D., Grosso, N., Burwitz, V., et al. 2005a, A&A, 430, L9
Porquet, D., Grosso, N., Bélanger, G., et al. 2005, A&A, 443, 571
Porquet, D., Grosso, N., Predehl, P., et al. 2008, A&A, 488, 549
Predehl, P., & Kulkarni, S. R. 1995, A&A, 294, L29
Price, S. D., Egan, M. P., Carey, S. J., Mizuno, D. R., & Kuchar,
T. A. 2001, AJ, 121, 2819
Pierce-Price, D., Richer, J. S., Greaves, J. S., et al. 2000, ApJ, 545,
L121
Quataert, E., & Loeb, A. 2005, ApJ, 635, L45
Rea, N., Esposito, P., Pons, J. A., et al. 2013, ApJ, 775, L34
Reich, W., & Fuerst, E. 1984, A&AS, 57, 165
Reid, M. J. 1993, ARA&A, 31, 345
Reid, M. J., Menten, K. M., Zheng, X. W., Brunthaler, A., & Xu,
Y. 2009, ApJ, 705, 1548
Renaud, M., Paron, S., Terrier, R., et al. 2006, ApJ, 638, 220
Revnivtsev, M. G., Churazov, E. M., Sazonov, S. Y., et al. 2004,
A&A, 425, L49
Revnivtsev, M., Sazonov, S., Gilfanov, M., Churazov, E., & Sunyaev, R. 2006, A&A, 452, 169
Revnivtsev, M., Sazonov, S., Churazov, E., et al. 2009, Nature,
458, 1142
Reynolds, M. T., & Miller, J. M. 2010, ApJ, 716, 1431
Rodrı́guez-Fernández, N. J., Martı́n-Pintado, J., & de Vicente, P.
2001, A&A, 377, 631
Roy, S., & Bhatnagar, S. 2006, Journal of Physics Conference Series, 54, 152
Sakano, M., Koyama, K., Murakami, H., Maeda, Y., & Yamauchi,
S. 2002, ApJS, 138, 19
Sakano, M., Warwick, R. S., Decourchelle, A., & Predehl, P.
2003a, MNRAS, 340, 747
Sakano, M., Warwick, R. S., & Decourchelle, A. 2003b, Workshop on Galaxies and Clusters of Galaxies, 9
Sakano, M., Warwick, R. S., Hands, A., & Decourchelle, A. 2004,
Memorie della Societa Astronomica Italiana, 75, 498
Sakano, M., Warwick, R. S., Decourchelle, A., & Wang, Q. D.
2005, MNRAS, 357, 1211
Sasaki, M., Breitschwerdt, D.,Baumgartner, V., & Haberl, F. 2011,
A&A, 528, A136
Sawada, M., Tsujimoto, M., Koyama, K., et al. 2009, PASJ, 61,
209
Seiradakis, J. H., Lasenby, A. N., Yusef-Zadeh, F., Wielebinski,
R., & Klein, U. 1985, Nature, 317, 697
Senda, A., Murakami, H., & Koyama, K. 2002, ApJ, 565, 1017
Senda, A., Murakami, H., & Koyama, K. 2003, Astronomische
Nachrichten Supplement, 324, 151
Shibata, K., Tajima, T., Steinolfson, R. S., & Matsumoto, R. 1989,
ApJ, 345, 584
Simpson, J. P., Colgan, S. W. J., Cotera, A. S., et al. 2007, ApJ,
670, 1115
Skinner, G. K., Foster, A. J., Willmore, A. P., & Eyles, C. J. 1990,
MNRAS, 243, 72
Sofue, Y., & Handa, T. 1984, Nature, 310, 568
Sofue, Y. 1985, PASJ, 37, 697
Soldi, S., Walter, R., Eckert, D., et al. 2006, The Astronomer’s
Telegram, 885, 1
40
G. Ponti et al.
Soldi, S., Clavel, M., Goldwurm, A., et al. 2014, IAU Symposium,
303, 126
Staguhn, J., Stutzki, J., Uchida, K. I., & Yusef-Zadeh, F. 1998,
A&A, 336, 290
Stiele, H., Pietsch, W., Haberl, F., et al. 2011, A&A, 534, A55
Stolte, A., Hußmann, B., Morris, M. R., et al. 2014, ApJ, 789, 115
Sturm, R., Haberl, F., Pietsch, W., et al. 2013, A&A, 558, A3
Su, M., Slatyer, T. R., & Finkbeiner, D. P. 2010, ApJ, 724, 1044
Sunyaev, R. A., Markevitch, M., & Pavlinsky, M. 1993, ApJ, 407,
606
Tanaka, Y., Koyama, K., Maeda, Y., & Sonobe, T. 2000, PASJ, 52,
L25
Tanaka, Y. 2002, A&A, 382, 1052
Tanaka, K., Kamegai, K., Nagai, M., & Oka, T. 2007, PASJ, 59,
323
Tanaka, K., Oka, T., Nagai, M., & Kamegai, K. 2009, PASJ, 61,
461
Tang, S., & Wang, Q. D. 2005, ApJ, 628, 205
Tatischeff, V., Decourchelle, A., & Maurin, G. 2012, A&A, 546,
AA88
Terrier, R., Ponti, G., Bélanger, G., et al. 2010, ApJ, 719, 143
Trap, G., Falanga, M., Goldwurm, A., et al. 2009, A&A, 504, 501
Trap, G., Goldwurm, A., Dodds-Eden, K., et al. 2011, A&A, 528,
A140
Tsuboi, M., Inoue, M., Handa, T., et al. 1986, AJ, 92, 818
Tsuru, T. G., Nobukawa, M., Nakajima, H., et al. 2009, PASJ, 61,
219
Tüllmann, R., Gaetz, T. J., Plucinsky, P. P., et al. 2011, ApJS, 193,
31
Uchida, Y., Sofue, Y., & Shibata, K. 1985, Nature, 317, 699
Uchida, K., Morris, M., & Serabyn, E. 1990, ApJ, 351, 443
Uchida, K., Morris, M., & Yusef-Zadeh, F. 1992, AJ, 104, 1533
Uchida, K. I., Morris, M. R., Serabyn, E., & Bally, J. 1994, ApJ,
421, 505
Uchiyama, H., Nobukawa, M., Tsuru, T., Koyama, K., & Matsumoto, H. 2011, PASJ, 63, 903
Uchiyama, H., Nobukawa, M., Tsuru, T. G., & Koyama, K. 2013,
PASJ, 65, 19
Wang, Q. D., Lu, F., & Lang, C. C. 2002, ApJ, 581, 1148
Wang, D. X., Xiao, K., & Lei, W. H. 2002, MNRAS, 335, 655
Wang, Q. D., Gotthelf, E. V., & Lang, C. C. 2002, Nature, 415,
148
Wang, W., Jiang, Z. J., & Cheng, K. S. 2005, MNRAS, 358, 263
Wang, Q. D., Dong, H., & Lang, C. 2006a, MNRAS, 371, 38
Wang, Q. D., Lu, F. J., & Gotthelf, E. V. 2006b, MNRAS, 367,
937
Wang, Q. D., Dong, H., Cotera, A., et al. 2010, MNRAS, 402, 895
Wang, Q. D., Nowak, M. A., Markoff, S. B., et al. 2013, Science,
341, 981
Weaver, R., McCray, R., Castor, J., Shapiro, P., & Moore, R. 1977,
ApJ, 218, 377
Werner, N., in’t Zand, J. J. M., Natalucci, L., et al. 2004, A&A,
416, 311
Wijnands, R., Miller, J. M., & Wang, Q. D. 2002, ApJ, 579, 422
Wijnands, R., in’t Zand, J. J. M., Rupen, M., et al. 2006, A&A,
449, 1117
Yamauchi, S., Shimizu, M., Nakashima, S., et al. 2014,
arXiv:1409.4520
Yelda, S., Ghez, A. M., Lu, J. R., et al. 2014, ApJ, 783, 131
Yoast-Hull, T. M., Gallagher, J. S., III, & Zweibel, E. G. 2014,
ApJ, 790, 86
Yusef-Zadeh, F., Morris, M., & Chance, D. 1984, Nature, 310, 557
Yusef-Zadeh, F., & Morris, M. 1987a, ApJ, 320, 545
Yusef-Zadeh, F., & Morris, M. 1987b, AJ, 94, 1178
Yusef-Zadeh, F., & Morris, M. 1987c, ApJ, 322, 721
Yusef-Zadeh, F., & Morris, M. 1988, ApJ, 329, 729
Yusef-Zadeh, F., Wardle, M., & Parastaran, P. 1997, ApJ, 475,
L119
Yusef-Zadeh, F., Law, C., Wardle, M., et al. 2002, ApJ, 570, 665
Yusef-Zadeh, F., Hewitt, J. W., & Cotton, W. 2004, ApJS, 155,
421
Yusef-Zadeh, F., Wardle, M., Muno, M., Law, C., & Pound, M.
2005, Advances in Space Research, 35, 1074
Yusef-Zadeh, F., Muno, M., Wardle, M., & Lis, D. C. 2007, ApJ,
656, 847
Wang, Q. D., Lu, F., & Lang, C. C. 2002, ApJ, 581, 1148
Zhang, Z.-Y., Gao, Y., Henkel, C., et al. 2014, ApJ, 784, LL31
Zoglauer, A., Reynolds, S. P., An, H., et al. 2015, ApJ, 798, 98
c 0000 RAS, MNRAS 000, 1–??
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement