Origin of the low-temperature plasma in the Galactic center X-ray emission
Shigeo Yamauchi, Miku Shimizu, Masayoshi Nobukawa, Kumiko K. Nobukawa, Hideki Uchiyama, Katsuji Koyama
aa r X i v : . [ a s t r o - ph . H E ] J un Publ. Astron. Soc. Japan (2014) 00(0), 1–6doi: 10.1093/pasj/xxx000 Origin of the low-temperature plasma in theGalactic center X-ray emission
Shigeo Y
AMAUCHI ∗ , Miku S HIMIZU , Masayoshi N OBUKAWA , Kumiko K.N OBUKAWA , Hideki U CHIYAMA , and Katsuji K OYAMA Department of Physics, Nara Women’s University, Kitauoyanishimachi, Nara 630-8506 Faculty of Education, Nara University of Education, Takabatake-cho, Nara 630-8528 Faculty of Education, Shizuoka University, 836 Ohya, Suruga-ku, Shizuoka 422-8529 Department of Physics, Graduate School of Science, Kyoto University,Kitashirakawa-oiwake-cho, Sakyo-ku, Kyoto 606-8502 ∗ E-mail: [email protected]
Received ; Accepted
Abstract
The Galactic Center X-ray Emission (GCXE) is composed of high temperature ( ∼ ∼ α line map, a pair of horn-like soft diffuse sources areseen at the symmetric positions with respect to Sagittarius A ⋆ . The X-ray spectra of the pair arewell represented by an absorbed thin thermal plasma model of a temperature and N H of 0.6–0.7 keV and 4 × cm − , respectively. The N H values indicate that the pair are located nearat the GC. Then the dynamical time scales of the pair are ∼ yr. The Si and S abundancesand the surface brightnesses in the the S XV He α line band are 0.7–1.2 and 0.6–1.3 solar, and(2.0–2.4) × − erg s − cm − arcmin − , respectively. The temperature, abundances, andsurface brightness are similar to those of the LTP in the GCXE, while the abundances are farlarger than those of known point sources, typically coronal active stars and RS CVn-type activebinaries. Based on these results, possible origin of the LTP is discussed. Key words:
Galaxy: center — X-rays: diffuse background — X-rays: soft diffuse sources
The spectrum of the Galactic Diffuse X-ray Emission (GDXE)has strong K-shell transition lines of highly ionized atoms andneutral iron. The strongest are the He-like iron (Fe XXV He α )and sulfur (S XV He α ) lines, which indicates that the GDXEis composed of a high-temperature plasma of ∼ α line, and a low-temperatureplasma of ∼ α line(Uchiyama et al. 2013). The other component is a power-law with the Fe I K α line. The equivalent widths and scaleheights of the Fe XXV He α and Fe I K α lines are position de-pendent in the GDXE, and hence the GDXE is spatially and spectrally separated into the Galactic Center X-ray Emission(GCXE), the Galactic Ridge X-ray Emission (GRXE), andthe Galactic Bulge X-ray Emission (GBXE) (Yamauchi et al.2016; Nobukawa et al. 2016; Koyama 2018).A long standing question of the GDXE is its origin, whetherit is integrated emission of point sources, diffuse plasma, orelse. Using the deep Chandra observation in the 6.5–7.1 keVband, Revnivtsev et al. (2009) and Hong (2012) made an X-rayLuminosity Function (XLF: the integrated flux of point sourcesas a function of the point source flux) down to the luminosityof ∼ × erg s − , and resolved more than 80 % flux ofthe GBXE into point sources. However, the profiles of the XLF c (cid:13) Publications of the Astronomical Society of Japan , (2014), Vol. 00, No. 0 and integrated spectra of the point sources were largely differ-ent between these authors, which led different prediction of thepoint source composition in the GBXE: RS-CVn type active bi-naries (ABs) and cataclysmic variables (CVs) with the mixingratio of ∼ ∼ ∼
10 times larger than the GBXE andGRXE. We reports properties of a pair of soft diffuse sourcesNE and NW in the LTP map at the northeast and northwestof the Galactic center (GC). The diffuse sources have beennoted by Wang et al. (2002) (Chandra) and Ponti et al. (2015)(XMM-Newton), but the detailed information has not been re-ported. Based on the improved spectral and spatial informa-tion of Suzaku, the origin of NE and NW, and possible inter-pretation of the origin of the LTP in the GCXE are discussed.Throughout this paper, the distance to the GC is 8 kpc (e.g.,Reid 1993; Gillessen et al. 2009), and quoted errors are in the90% confidence limits.
Survey observations in the GC region were carried out withthe X-ray Imaging Spectrometer (XIS; Koyama et al. 2007) on-board Suzaku (Mitsuda et al. 2007). This paper utilized theseSuzaku data in the archive. The XISs were composed of 4CCD cameras placed on the focal planes of the thin foil X-rayTelescopes (XRT; Serlemitsos et al. 2007). XIS 1 was a back-side illuminated (BI) CCD, while XIS 0, 2, and 3 were front-side illuminated (FI) CCDs. The field of view (FOV) of theXIS was 17 . ′ × . ′
8. The data from the three sensors (XIS 0, 1,and 3) were used for most of the observations, because XIS 2stopped working in 2006 November. Since the spectral reso-lution of the XIS was degraded due to the radiation of cosmicparticles, the spaced-row charge injection (SCI) technique wasapplied to restore the XIS performance (Uchiyama et al. 2009).After removing hot and flickering pixels, we used the data ofthe ASCA grade 0, 2, 3, 4, and 6.
The XIS pulse-height data are converted to Pulse Invariant (PI)channels using the xispi software in the HEAsoft 6.19 and thecalibration database version 2016-06-07. The data in the SouthAtlantic Anomaly, during the earth occultation, and at the lowelevation angle from the earth rim of < ◦ (night earth) and < ◦ (day earth) are excluded. Figure 1 shows the results of theSuzaku GC survey, covering the full area of the GC by multiplepointings. The non X-ray background (NXB), estimated using xisnxbgen (Tawa et al. 2008), is subtracted. To highlight thecontrast between the HTP and LTP distributions, X-ray imagesof the Fe XXV He α (6.55–6.8 keV) and S XV He α (2.3–2.6keV) bands are separately made. A pair of soft diffuse sources(NE and NW) are detected in two fields of the XIS. The obser-vation logs of the two fields, which include the pair sources NEand NW, and background (BGD), are given in table 1. In the Fe XXV He α line map (HTP distribution), some slightlyenhancement are found near at the giant molecular cloud com-plex (GMC) of Sagittarius (Sgr) A, Sgr C, and Arches cluster(blue dashed lines in figure 1a). The Fe XXV He α line en-hancement is about ∼
10% of the GCXE level. Thus global dis-tribution of the HTP is smooth in the full area of the GC. Inthe LTP distribution, on the other hand, the S XV He α imageshows a largely extended X-ray emission near at ( l , b ) ∼ (0 . ◦ . ◦
4) (GC North, Nakashima et al. 2014). The spectrum of thissoft diffuse source is in collisional ionization equilibrium (CIE)with a temperature of 0.81 keV and solar abundances. Anothersoft diffuse source is a super bubble candidate (G359.79 − − ∼ ∼ ⋆ (hear, XRN complex), firstly reported by Park et al. (2004).This source has different morphology in the Fe XXV He α im-age, slim and faint, which corresponds to the GMC complex(named, Sgr A GMC), or Sgr A XRN. For this source, how-ever, no spectral information has been available. Other softdiffuse sources are found near the Sgr C GMC (the Chimneyand G359.41-0.12), firstly reported by Tsuru et al. (2009). Thetemperatures are ∼ ∼ ∼ ublications of the Astronomical Society of Japan , (2014), Vol. 00, No. 0 Fig. 1.
XIS images in the (a) 6.55–6.8 (Fe XXV He α ) and (b) 2.3–2.6 keV (S XV He α ) bands in the Galactic coordinate. The color bar shows surface brightnessin the logarithmic scale. The unit is 2 × − erg s − cm − arcmin − for (a) and 1 × − erg s − cm − arcmin − for (b). Bright point sources are maskedby the black circles. A stray-light region of the brightest source A1742 −
294 is given by the large black circle. The white dashed squares, the solid lines andthe green horn-like lines indicate the XIS FOVs, the background region (BGD), and a pair of soft diffuse sources (NE and NW), respectively. The black dashedlines in (b) outline other LTP clumps, while the blue dashed lines in (a) are the regions of HTP clumps. In order to figure out the difference between the HTPand LTP structure, the XIS FOVs and horn-like structures in the LTP are also shown in the HTP image.
Publications of the Astronomical Society of Japan , (2014), Vol. 00, No. 0
Table 1.
List of data used for spectral analyses.
Observation ID Pointing Position Observation time (UT) Exposure time Region ∗ ( l , b ) Start – End (ks)503007010 (0 . ◦ + . ◦ . ◦ + . ◦ ∗ See figure 1.
The X-ray spectra of NE and NW after the subtraction of theNXB (see section 2) are made from the source regions. TheBGD region is selected from a nearby blank sky of similarGalactic latitude to those of NE and NW. The X-ray spectrumof BGD is made with the same process as NE and NW. In orderto make the spectra of NE, NW, and BGD with good statistics,we utilize the two XIS fields, which include a large fraction ofNE, NW, and BGD with long exposure times (table 1). Theflux of the NXB are ∼
9% and ∼
14% of the total counts of thebackground region (BGD) in the 1–8 keV band for FI and BI,respectively, and the statistical error is less then 1%. Thus, theuncertainty caused by the NXB subtraction is not significant inthe following spectral analysis. The flux of the BGD spectrumis fine-tuned, taking account of the longitude and latitude dif-ferences between BGD and the pair sources (NE and NW) andusing the e-folding longitude and latitude scales of the GCXE of0 . ◦
63 and 0 . ◦
26, respectively (Uchiyama et al. 2013; Yamauchiet al. 2016; Koyama 2018). The fine-tuning factor is 1.3 forNE and 1.1 for NW. Then, the BGD spectrum multiplied bythis fine-tuning factor is subtracted from the NE and NW spec-tra. The resultant NE and NW spectra are shown in figure 2,in which many emission line features are found. In order toincrease photon statistics, the spectra with the FI sensors (XIS0 and 3) are co-added, but the XIS 1 spectrum is treated sepa-rately, because the response functions of the FIs and BI are dif-ferent. Response files, Redistribution Matrix Files (RMFs) andAncillary Response Files (ARFs) are made using xisrmfgen and xissimarfgen (Ishisaki et al. 2007), respectively. Theabundance tables, and the atomic data of the lines and continuaof the thin thermal plasma are taken from Anders & Grevesse(1989) and ATOMDB 3.0.9, respectively.The NE and NW spectra are fitted with a CIE plasma modelof solar abundances, vapec in XSPEC version 12.9.0u. Thismodel is rejected ( χ /d.o.f. of 163/119 and 177/119, respec-tively) with residuals at the Si, S, and Ar lines. Therefore, theNE and NW spectra are re-fitted by the same CIE model, but theabundances of Si, S, and Ar (=Ca) are treated as free parame-ters. Then an improved fit is obtained with χ /d.o.f. of 147/116and 158/116, respectively . The best-fit model is shown in fig- Exactly speaking, this model is not accepted from the statistical point of
Table 2.
The best-fit parameters of NE and NW.
Parameter ValueNE NW N H (cm − ) (4.4 +0 . − . ) × (4.2 +0 . − . ) × kT e (keV) 0.64 +0 . − . +0 . − . Si ∗ (Solar) 0.7 +0 . − . ± ∗ (Solar) 0.6 +0 . − . ± ∗ (Solar) 2.3 ± +1 . − . Others ∗ (Solar) 1 (fixed) 1 (fixed)Normalization † (4.0 +2 . − . ) × − (1.7 +1 . − . ) × − χ /d.o.f. 147/116 158/116 ∗ Abundance relative to the solar value (Anders & Grevesse1989). † Defined as 10 − × R n H n e dV / 4 πD Ω (cm − arcmin − ),where n H , n e , D and Ω are hydrogen density (cm − ), electrondensity (cm − ), distance (cm) and solid angle (arcmin ) of thesource, respectively.ure 2, while the best-fit parameters are given in table 2. Theflux in the S XV He α line band (2.3–2.6 keV) are 2.4 × − erg s − cm − arcmin − for NE and 2.0 × − erg s − cm − arcmin − for NW. Chandra found many point sources in the GC region (e.g., Munoet al. 2009). The flux of the integrated point sources in the re-gions of NE and NW is only ∼ N H values of the pair sources,NE and NW, are ∼ × cm − (table 2), roughly consistentwith those of the point sources located at the GC region (Sakano2000; Sakano et al. 2002). Therefore, here and after, we assumethat NE and NW are located near the GC region at the distanceof 8 kpc.The pair of soft diffuse sources, NE and NW, have horn-likestructures standing above the Galactic plane (see, figure 1b). view, which may be due to systematic errors caused in the BGD subtractionprocess. Taking account of the possible systematic errors, we regard themodel is a good approximation of the NE and NW spectra. ublications of the Astronomical Society of Japan , (2014), Vol. 00, No. 0 − . C oun t s s − k e V − NE1 2 − − ( da t a − m ode l ) / e rr o r Energy (keV) − . C oun t s s − k e V − NW1 2 − − ( da t a − m ode l ) / e rr o r Energy (keV)
Fig. 2.
X-ray spectra of NE and NW (upper panels) and the residuals from the best-fit model (lower panels). Only the spectra obtained with the FI sensors inthe 1–4 keV band are displayed for brevity.
Assuming a cone shape geometry with a diameter of the base of ∼ ′ (35 pc) and a height of ∼ ′ , the volume ( V ) is estimatedto be V = 3.3 × cm . Using the best-fit volume emissionmeasure, and assuming a filling factor = 1 and n e = 1.2 n H ,where n e and n H are the electron and hydrogen densities, re-spectively, we obtain the mean hydrogen density ( n H ), the ther-mal energy ( E th ), gas mass ( M gas ), and sound velocity ( c s )for NE to be n H =0.3 cm − , E th =3 × erg, M gas =120 M ⊙ ,and c s =4 × cm s − . For NW, we obtained n H =0.2 cm − , E th =2 × erg, M gas =80 M ⊙ , and c s =4 × cm s − . Thedynamical time scale ( t dyn ) is estimated to be t dyn ≃ × yr.Most of these physical parameters of NE and NW are con-sistent with those of middle-aged SNRs. However, the plasmasize is larger than typical middle-aged SNRs, and the horn-likemorphology is largely different from that of a single SNR. Thephysical parameters and the dynamical time scales of NE andNW are similar to each other, and the pair positions are sym-metric with respect to Sgr A ⋆ . These suggest that the pair, NEand NW, originated from the same event, possibly a past activityin the GC region.Remarkable features in the LTP map are the presence ofmany bright soft diffuse sources (Mori et al. 2008; Mori et al.2009; Tsuru et al. 2009; Heard & Warwick 2013; Nakashimaet al. 2014; Ponti et al. 2015), including NE and NW, near theGC. These soft diffuse sources have similar temperature andSi–S abundances to those of the LTP in the GCXE (Uchiyamaet al. 2013). In the scenario of the point source origin forthe LTP, a candidate source in the luminosity range of > erg s − has been regarded to be ABs with a thermal spec-trum of temperature > ∼ ∼ –10 erg s − . In this luminosity range,the candidate source may not be only ABs, but includes coronalactive stars (CAs) with the temperature of < ∼ ∼ –10 , because the totalLTP luminosity of the GCXE is ∼ erg s − (Uchiyama etal. 2013). Since this huge number of point sources would leada uniform LTP distribution, the presence of many bright softdiffuse sources disfavors the point source origin.Most of the soft diffuse sources have dynamical time scalesof ∼ yr, which corresponds to the last epoch of the highstar formation activity of ∼ –10 yr ago (Yusef-Zadeh et al.2009). The abundances of Si and S of NE and NW and mostof other soft diffuse sources are larger than CAs (typically ∼ ∼ yr ago (Ponti et al. 2015), or the past flares ofSgr A ⋆ (e.g., Koyama 2018). The origin of a power-law com-ponent with the Fe I K α line (6.4 keV) would also be the sameactivities near the GC. Acknowledgement
The authors are grateful to all members of the Suzaku team.KKN is supported by Research Fellowships of JSPS for YoungScientists. This work was supported by the Japan Society forthe Promotion of Science (JSPS) KAKENHI Grant NumbersJP24540232 (SY) and JP16J00548 (KKN).
Publications of the Astronomical Society of Japan , (2014), Vol. 00, No. 0