すざく衛星による Tornado Nebula (G357.7-0.1) の起源の解明

20110921_ASJ2011autum_Tornado_v2.key

鶴 剛，澤田真理，小山勝二 (京都大学)，岡 朋治 (慶應義塾大学)

“Suzaku Discovery of Twin Thermal Plasma from the Tornado Nebula”

Makoto Sawada, Takeshi Go Tsuru, Katsuji Koyama, and Tomoharu Oka

PASJ, to be published (Suzaku SP)

The Tornado Nebula (G357.7-0.1)2

Tornado(G357.7-0.1)

The Tornado Nebula (G357.7-0.1)3

Head

Tail

α ～ -0.6 (Dickel+73)10%偏光(Kundu+74)

Headの拡散Ｘ線スペクトル不明(Gaensler+03)

Ｘ線(Chandra)

OH

OH(1720MHz)-12.4km/s → 11.8kpc(Frail+96)

非熱的電波

Origin Proposed so far4

Head

Tail

中間から両方向に何か出てHead-Tailが出来た

Headから何か出てTailが出来た

電波銀河のone-sideジェット

ヘッドに居る(はずの)パルサーやＸ線連星が作ったnebula

Weiler+80, Miley+80, Shaver+85, Becker+85, Shull+89, Caswell+89, Shaver+85, Manchester+87

決め手無し

電波ローブＸ線連星からのバイポーラジェット

exotic SNの残骸

Suzaku Observation5

1deg

197pointings, 5.26Msec30 refereed papers

6 Doctor Thesis

Observation of the GC region with Suzaku2008.9.19 56.8ks

2009.8.29 125ks

Suzaku XIS (0+1+3)

スペクトル解析

X-ray Image with Suzaku6

Head と Tail と同じ領域から拡散Ｘ線を検出

R=0.5-1.5, G=1.5-3, B=3-7keV

銀河中心の強い拡散Ｘ線放射あり

最重要バックグラウンド

Background - the Key of Analyses7

!

10!4

10!3

0.01

!2

0

2

Inte

nsi

ty (

cou

nts

s"

1 k

eV"

1)

(a) BG 1

!

10!4

10!3

0.01

!2

0

2

Inte

nsi

ty (

cou

nts

s"

1 k

eV"

1)

Energy (keV)

(b) BG 2

0.5 1 2 5 10

Fig. 2. (a) The NXB-subtracted spectra (gray crosses) and the best-fit models (black histograms) of

BG1. The di!erence of e!ective area between the two is corrected and only the results of FIs are shown

for simplicity. The solid, dashed, and dash-dotted curves represent contributions from the GDXE, the

foreground emission, and the CXB, respectively. (b) Same as panel (a), but for BG2.

2.0! 1021 cm!2. Thus for the foreground soft X-ray spectrum, only the normalization was a

free parameter.

The other background component, the Cosmic X-ray background (CXB), is isotropic

and we adopted the empirical form, an absorbed power-law:

Wabs3!C ! (E/keV)!!, (3)

where C is the normalization constant at 1 keV and ! is the photon index of the power-law.

These are fixed to be C = 7.4! 10!7 photons s!1 cm!2 arcmin!2 and ! = 1.486 according

to Kushino et al. (2002). Since the CXB should be absorbed by the interstellar materials

throughout the whole Galaxy, we assumed the hydrogen-equivalent absorption column toward

the CXB to be twice of that for the GDXE (NH).

model, which gives a reasonable fit to the data. Therefore apparent low abundances in this model should

not be taken seriously.

6

wabs1 x (apec1 + apec2 + 6.4keV) + wabs2 x apec3

+ 2 x wabs1 x power

!

10!4

10!3

0.01

!2

0

2

Inte

nsi

ty (

cou

nts

s"

1 k

eV"

1)

(a) BG 1

!

10!4

10!3

0.01

!2

0

2

Inte

nsi

ty (

cou

nts

s"

1 k

eV"

1)

Energy (keV)

(b) BG 2

0.5 1 2 5 10

Fig. 2. (a) The NXB-subtracted spectra (gray crosses) and the best-fit models (black histograms) of

BG1. The di!erence of e!ective area between the two is corrected and only the results of FIs are shown

for simplicity. The solid, dashed, and dash-dotted curves represent contributions from the GDXE, the

foreground emission, and the CXB, respectively. (b) Same as panel (a), but for BG2.

2.0! 1021 cm!2. Thus for the foreground soft X-ray spectrum, only the normalization was a

free parameter.

The other background component, the Cosmic X-ray background (CXB), is isotropic

and we adopted the empirical form, an absorbed power-law:

Wabs3!C ! (E/keV)!!, (3)

where C is the normalization constant at 1 keV and ! is the photon index of the power-law.

These are fixed to be C = 7.4! 10!7 photons s!1 cm!2 arcmin!2 and ! = 1.486 according

to Kushino et al. (2002). Since the CXB should be absorbed by the interstellar materials

throughout the whole Galaxy, we assumed the hydrogen-equivalent absorption column toward

the CXB to be twice of that for the GDXE (NH).

model, which gives a reasonable fit to the data. Therefore apparent low abundances in this model should

not be taken seriously.

6

!

10!4

10!3

0.01

!2

0

2In

ten

sity

(co

un

ts s"

1 k

eV"

1)

(a) BG 1

!

10!4

10!3

0.01

!2

0

2

Inte

nsi

ty (

cou

nts

s"

1 k

eV"

1)

Energy (keV)

(b) BG 2

0.5 1 2 5 10

Fig. 2. (a) The NXB-subtracted spectra (gray crosses) and the best-fit models (black histograms) of

BG1. The di!erence of e!ective area between the two is corrected and only the results of FIs are shown

for simplicity. The solid, dashed, and dash-dotted curves represent contributions from the GDXE, the

foreground emission, and the CXB, respectively. (b) Same as panel (a), but for BG2.

2.0! 1021 cm!2. Thus for the foreground soft X-ray spectrum, only the normalization was a

free parameter.

The other background component, the Cosmic X-ray background (CXB), is isotropic

and we adopted the empirical form, an absorbed power-law:

Wabs3!C ! (E/keV)!!, (3)

where C is the normalization constant at 1 keV and ! is the photon index of the power-law.

These are fixed to be C = 7.4! 10!7 photons s!1 cm!2 arcmin!2 and ! = 1.486 according

to Kushino et al. (2002). Since the CXB should be absorbed by the interstellar materials

throughout the whole Galaxy, we assumed the hydrogen-equivalent absorption column toward

the CXB to be twice of that for the GDXE (NH).

model, which gives a reasonable fit to the data. Therefore apparent low abundances in this model should

not be taken seriously.

6

銀河中心拡散Ｘ線foregroundＸ線

nH = (5.2±0.6)x10^22 cm^-2

全てのフリーパラメータがエラーの範囲で一致

銀経/銀緯依存性は低いマージしてBGとして使用

= GCまでの吸収量CXB

Spectral Analyses8

Table 3. Best-fit parameters of the di!use sources.

Parameter NW SE

NH (1022 cm!2) 6.6+1.1!0.8 7.4+1.7

!1.4

kT (keV) 0.73+0.15!0.15 0.59+0.18

!0.15

Z (Z")# 0.89+0.37!0.28 1.7+1.3

!0.7

SX† 2.8+0.2

!0.2 2.4+0.2!0.2

LX‡ 2.4+0.1

!0.1 3.6+0.3!0.3

!2/d.o.f. 85/82 49/41! The chemical abundance of Si, S, and Ar relative to the

solar values. The abundances of the other elements are

fixed at the solar values.

† The average surface brightness in the unit of 10"6 pho-

tons s"1 cm"2 arcmin"2 in the 0.5–7.0 keV band.

‡ The luminosity in the unit of 1035 erg s"1 in the 0.5–

7.0 keV band. The absorption is corrected and the dis-

tances are assumed to be 12.0 kpc (see § 4.1).

(!12 kpc; Frail et al. 1996). In the projected image, the radio emission of the Tornado appears

to be associated with the two X-ray structures at the both ends (see figure 4). Thus we argue

that the two X-ray sources are physically connected to the Tornado nebula.

The two di!use sources are found to exhibit optically thin thermal nature. In fact, the X-

ray emission from the sources are described with CIE plasmas of nearly the same temperatures

(0.6–0.7 keV). The abundances are consistent with the solar values. This indicates that the

X-ray emitting materials mainly consist of interstellar gas. If the distance of the Tornado

is 12 kpc, the X-ray source size of 4! corresponds to 14 pc. Assuming a spherical shape with

uniform density, we derive the physical parameters of the NW and SE plasmas, electron density

(ne), mass (M) and total thermal energy (Eth), as are listed in table 4.

Table 4. Physical parameters of the X-ray emitting plasma.

Parameter NW SE

V EM (1058 cm!3)# 1.0 1.5

ne (cm!3) 0.49 0.59

M (M") 23 29

Eth (1049 erg) 8.3 8.2! These values are taken from the best-fit model in

table 3 at the distance of 12 kpc.

We found no hint of non-thermal emission from the NW and SE sources. Assuming the

photon index of 2, we obtained the 3-! upper limits of the luminosities of non-thermal X-rays

to be 1.1% and 2.0% of those of the thermal plasmas in the 0.5–7.0 keV band for NW and SE,

9

wabs x apec

10^-6 ph/cm2/sec

10!4

10!3

0.01

0.1

!2

0

2

!In

ten

sity

(co

un

ts s"

1 k

eV"

1)

(a) NW

FIs (!5)

BI

Si S Ar

(b) SE

10!4

10!3

0.01

0.1

1 2 5

!2

0

2

!

Energy (keV)

Inte

nsi

ty (

cou

nts

s"

1 k

eV"

1)

FIs (!5)

BI

Si S Ar

Fig. 3. (a) The background-subtracted spectra for NW. The FI and BI spectra are separately given by

blue and red crosses, respectively. For visibility, the FI spectrum is multiplied by 5. The blue and red

histograms are the best-fit model of FI and BI spectra, respectively. (b) Same as panel (a) but for SE.

best-fit parameters are given in table 3.

4. Discussion

4.1. Nature of the X-ray Sources and Association to the Tornado

We find that the absorption column densities (NH) for the NW and SE sources are very

similar to each other (table 3). Since the NH values for the GDXE near the source region are

almost constant at (5.3± 0.4)! 1022 cm!2, the same NH values for NW and SE supports that

the two sources are located at nearly the same distances.

Assuming that the distance of the GDXE is 8 kpc and that the interstellar gas density

along the line-of-sight is constant, the distances of NW and SE can be estimated by the NH

ratio relative to the background emission of (5.3±0.4)!1022 cm!2 as 10.0+1.8!1.4 and 11.2+2.7

!2.3 kpc,

respectively. These distances are almost consistent with the radio distance of the Tornado

8

10!4

10!3

0.01

0.1

!2

0

2

!In

ten

sity

(co

un

ts s"

1 k

eV"

1)

(a) NW

FIs (!5)

BI

Si S Ar

(b) SE

10!4

10!3

0.01

0.1

1 2 5

!2

0

2

!

Energy (keV)

Inte

nsi

ty (

cou

nts

s"

1 k

eV"

1)

FIs (!5)

BI

Si S Ar

Fig. 3. (a) The background-subtracted spectra for NW. The FI and BI spectra are separately given by

blue and red crosses, respectively. For visibility, the FI spectrum is multiplied by 5. The blue and red

histograms are the best-fit model of FI and BI spectra, respectively. (b) Same as panel (a) but for SE.

best-fit parameters are given in table 3.

4. Discussion

4.1. Nature of the X-ray Sources and Association to the Tornado

We find that the absorption column densities (NH) for the NW and SE sources are very

similar to each other (table 3). Since the NH values for the GDXE near the source region are

almost constant at (5.3± 0.4)! 1022 cm!2, the same NH values for NW and SE supports that

the two sources are located at nearly the same distances.

Assuming that the distance of the GDXE is 8 kpc and that the interstellar gas density

along the line-of-sight is constant, the distances of NW and SE can be estimated by the NH

ratio relative to the background emission of (5.3±0.4)!1022 cm!2 as 10.0+1.8!1.4 and 11.2+2.7

!2.3 kpc,

respectively. These distances are almost consistent with the radio distance of the Tornado

8

10!4

10!3

0.01

0.1

!2

0

2!

Inte

nsi

ty (

cou

nts

s"

1 k

eV"

1)

(a) NW

FIs (!5)

BI

Si S Ar

(b) SE

10!4

10!3

0.01

0.1

1 2 5

!2

0

2

!

Energy (keV)

Inte

nsi

ty (

cou

nts

s"

1 k

eV"

1)

FIs (!5)

BI

Si S Ar

Fig. 3. (a) The background-subtracted spectra for NW. The FI and BI spectra are separately given by

blue and red crosses, respectively. For visibility, the FI spectrum is multiplied by 5. The blue and red

histograms are the best-fit model of FI and BI spectra, respectively. (b) Same as panel (a) but for SE.

best-fit parameters are given in table 3.

4. Discussion

4.1. Nature of the X-ray Sources and Association to the Tornado

We find that the absorption column densities (NH) for the NW and SE sources are very

similar to each other (table 3). Since the NH values for the GDXE near the source region are

almost constant at (5.3± 0.4)! 1022 cm!2, the same NH values for NW and SE supports that

the two sources are located at nearly the same distances.

Assuming that the distance of the GDXE is 8 kpc and that the interstellar gas density

along the line-of-sight is constant, the distances of NW and SE can be estimated by the NH

ratio relative to the background emission of (5.3±0.4)!1022 cm!2 as 10.0+1.8!1.4 and 11.2+2.7

!2.3 kpc,

respectively. These distances are almost consistent with the radio distance of the Tornado

8

非熱的成分 < 1-2% (3σ)

•まるで双子（一卵性）•nH>nH(GC)→D=10-11kpc　電波(11.8kpc)とほぼ一致．

(Lx=3e+35ergs/s, Eth=8e+49ergs) x2

Ｘ線はイメージ，スペクトル共に対称なバイポーラー構造を持つ

拡散Ｘ線は電波Tornadoと物理的にassociate

Bipolar Structure9

•分子雲 (H2, OHメーザー)

はHeadからのみ検出

•Ｘ線がバイポーラ構造なら...分子雲だって...

•野辺山45m CO観測

357.85 357.80 357.75 357.70 357.65 357.60Galactic longitude (degree)

0.00

0.05

0.10

0.15

0.20

0.25

Gal

actic

latit

ude

(deg

ree)

分子雲も対称なバイポーラー構造

•OHメーザーと同じ速度帯で，Head，Tailの両方から分子雲発見．

Bipolar Jet from an X-ray Binary10

357.85 357.80 357.75 357.70 357.65 357.60Galactic longitude (degree)

0.00

0.05

0.10

0.15

0.20

0.25

Gal

actic

latit

ude

(deg

ree)

•Ｘ線連星からのジェットが分子雲に衝突→Ｘ線，電波

•中心天体 < 1-2e+33 ergs/s (3σ)

•過去に活動→現在は静か．

•非熱的Ｘ線は弱い•Ｘ線シンクロトロンを放射する電子は冷えた

•冷却時間

•電波シンクロトロンを放射する電子の冷却時間 ～10^5 yr

•Ｘ線熱放射をする電子の冷却時間 ～10^6 yr

活動＆ジェット放出時期～50yearより以前

Summary

•電波トルネード方向から「すざく」で2つの拡散Ｘ線を検出．高い精度のスペクトルの取得に成功→光学的に薄い熱的プラズマ．2つは極めて良く似ている．

•Ｘ線吸収から求めた距離→GCより少し遠い，銀河系内天体．電波トルネードと同じ距離．電波トルネードと2つの拡散X線は物理的にassociate．

•拡散熱的Ｘ線はHeadとTailの中間に対して対称形である．

• HeadとTailの両者からCO分子雲を検出

• 50年以上前にX線連星が活動しバイポーラジェットを放出，分子雲に衝突，熱化してX線プラズマが生成された．X線シンクロトロンは既に冷却して見えない．

11

Thank you.