Evolution of supernova remnants in the interstellar medium with a large-scale density gradient. II. The 2-D modelling of the evolution and X-ray emission of supernova remnant RCW86

8 Pages·0.41 MB·English

by  O. Petruk

A&A manuscript no.
(will be inserted by hand later) ASTRONOMY
AND
Your thesaurus codes are:
ASTROPHYSICS
08 (09.19.2 09.09.1 02.08.1 13.25.4 ) 1.2.2008
Evolution of supernova remnants in the interstellar
medium with a large-scale density gradient
II. The 2-D modelling of the evolution and X-ray emission of supernova
remnant RCW86
9 O. Petruk
9 Institutefor Applied Problems in Mechanics and Mathematics, NAS of Ukraine, 3-b NaukovaSt.,Lviv, 290601, Ukraine
9
[email protected]
1

n Received ...; accepted ...
a
J

5 Abstract. The results of the 2D modelling of Supernova 1800 yearsalthough the SNR might be a relatively young
2 remnant (SNR) RCW86 are given. Models of this rem- remnant because it has a radial oriented magnetic field

nant, which for the first time interpret the anisotropy of (Milne 1987) like Tycho, Cas A and Kepler SNRs (Strom
1
surfacebrightnessas aresultofthe evolutionofadiabatic 1994).
v
7 SNRintheinterstellarmediumwithalarge-scalegradient The SN progenitor might be a member of OB-
4 of density, are considered. Estimations on the basic char-
assosiation. In this case we may expect the SN to have
3 acteristics ofRCW86 and the surroundingmedium which
beenoftypeII,butitcouldnotbebornin185A.D.since
1
follow from these models are found. It is shown that the
0 the SNR has much bigger size than the expected one if it
9 observed surface brightness distribution of RCW86 may is situated at the distance d≈2.8 kpc obtained from the
be obtaned in the both proposed up till now models with
9 kinematic survey of ionized H (Rosado et al. 1996).
/ different initial assumptions: one about a Supernova ex-
h plosionin185A.D.andanotheraboutanexplosioninthe From the ratio Hα/Hβ for optical filaments Ruiz
-p OB-asossiation. In order to obtain the observational con- (1981) estimates the distance to the SNR as d≈1 kpc.
o trastofsurfacebrightnessitisnecessarytohaveamedium RCW86, as radio source MSH14-63, was identified by
r
t with the characteristic scale of nonuniformity 11 pc if the Hill (1967). Σ− D relation yields an estimation on the
as age of RCW86 is 1800 years or 20−25 pc when the SNR distance to the SNR d = 2÷3.2 kpc (Milne 1970), but
: is distant from us on 2.8 kpc. The preshock density con- the possible range is, nevertheless,d=1÷10 kpc (Strom
v
i trast between the southwesternand northeasternparts of 1994).
X RCW86 is in range 3.5−4.5.
Both the radio and the X-ray observationsrevealthat
r
a the shape of RCW86 is close to a spherical one with the
Key words: ISM: supernovaremnants – ISM: individual
size 43′ ×40′ (at frequency 843 MHz) and is of a shell-
objects: RCW86 – hydrodynamics – X-rays: interstellar
liketypewiththeapproximatelyaxis-symmetricalsurface
brightness distribution and higher emission in the south-
west(SW)partoftheSNR(Caswelletal.1975,Pisarskiet
al. 1984, Claas et al. 1989, Whiteoak & Green 1996). The
brightestopticalfilamentscoincidewiththemaximumsof
1. Introduction emission in radio and X-ray bands (van den Bergh 1978,
Pisarski et al. 1984).
Supernova remnant (SNR) G315.4-2.3 (RCW86, MSH14-
63)istheresultofasupernovaexplosionwhichisregarded TheimportantresultsininvestigationofRCW86were
asoneofthemostpossiblecandidatesforthe”gueststar” causedbyanalyzingthespectralpropertiesofX-rayemis-
of 185 A.D. according to Chinese manuscripts (Clark & sion from this SNR. The first soft X-ray observations
Stephenson 1977). werereportedandinterpretedbyNaranan(1977).Winkler
Chin & Huang (1994), Schaefer (1995) argue that the (1978),havingusedhardX-raydata,showedthatthe ob-
”gueststar”in185A.D. wasnotaSupernova(SN). Then servedspectrum might be explainedas thermal,with two
it is not right to assume that the age of RCW86 is t = bremsstrahlungcomponents(fromtheforwardandreverse
2 O. Petruk:Evolution of supernovaremnants in theinterstellar medium with a large-scale density gradient
shock waves) with the temperatures T ≈0.22 keV and into account both the NEI plasma emission effects and
low
Thigh >=5 keV. the ISM density gradient. Therefore, previous studies of
The first map of the X-ray surface brightness distri- RCW86 are based on the one-dimentional Sedov (1959)
bution of RCW86 was presented by Pisarski et al. (1984) solutionwhichcannotrestorethe observedmorphologyof
(EINSTEIN observatory). The authors noted that higher theSNR.Inthispaperweshowthattheglobalanisotropy
emission from the SW part might be caused by the in- of the surface brightness distribution of RCW86 may be
teraction of the shock wave with the interstellar medium explainasaconsequenseofthe SNRevolutioninthe ISM
(ISM) with a density 2-3 times higher than the aver- withalarge-scaledensitygradient.Wedonotinvolveinto
age. The one temperature fitting of the spectrum from the analysis the small-scale emission structures which are
the whole remnant (it was accepted that the plasma is seen on the SNR’s maps, because these structures are re-
uniform and in collision ionization equilibrium (CIE)) sponsible for the local rises and do not essentially modify
yielded temperature T = 1.2 keV and column density theglobalbehaviourofthesurfacebrightness.Theplasma
N =2.8·1020cm−2.TheobtainedparametersofRCW86 emission model used here is CIE.
H
are given in Table 1. Pisarski et al. (1984) also showed
that if this SNR is at the adiabatic stage of its evolution,
it cannot be the result of the type II SN. 2. The 2-D Modelling of RCW86
Nugentetal.(1984)analysedthespectrumofRCW86
(HEAO 1 experiment) with the assumptions: 1) that 2.1. Method and models
plasma emits in the CIE and, 2) for the first time, that
Evolution of the SNR in the medium with a large-scale
the plasma is under nonequilibrium ionisation condition
density distribution is taken as the principal basis for the
(NEI) with T =T . The best fitting of the CIE spectrum
e i
is for T = 5.1 ± 0.14 keV, T = 0.52 ± 0.04 keV, modelspresentedinthispaper.Weacceptherespherically
high low
N = (1.1 ± 0.3) · 1021 cm−2. The NEI model gives symmetrical energy emission during a SN explosion. We
H
N = (4.4±0.3)·1021 cm−2 and the parameters of the assumethatthenonsphericalshapeandtheglobalregular
H
surfacebrightnessanisotropyoftheSNRiscausedonlyby
SNRshowninTable 1.ThedistancetoRCW86issmaller
the large-scale stucture of nonuniform medium.
than to the OB-assosiation but the authors have noted
that the possible deviation from the NEI model may in- Thenewmodelparametersassociatedwithorientation
crease the distance to d=2.5 kpc. of the SNR and his projection on the plan of the sky ap-
Claasetal.(1989)alsousedboththeCIEandtheNEI pearwhen3Dmodellingexecutes.Projectioneffects com-
model for the interpretation of EXOSAT observation of plicate the analysisofthe observations.Ingeneralcase,it
RCW86.Thepossiblecontaminationbythegalacticridge is impossible to reproduce unambiguily the real3Dshape
which increases the flux above 6 keV was subtracted, so for the knowing projection.In case of consideredhere the
in the two-component spectral model the smaller T = axis-symmetrical models, projection effects can easily be
high
3.4±0.2 keV was obtained. The low temperature com- included into model. Simple connections between the pa-
ponent is believed to be a NEI effect. In this paper the rametersofthemodelandtheprojectionexists.Theangle
parameters of RCW86 were also estimated using the sur- δ between the symmetry axis of the 2D SNR and the vi-
facebrightnessdistribution.HigheremissionfromtheSW sualcross-sectionisanewadditionalfreeparemeterofthe
partmustbe causedby a moredense ISM. Therefore,the model.
Sedov (1959) solution is used only for the interpretation Atourpreviouspaper(Paper I)ithavebeendescribed
ofthenortheastern(NE)partoftheSNR.TheNEImodel a new approximate analytical method which allows the
yields T =T /1.3=T /1.3=(3.03±0.2)·107 K. modelling a point explosion in media with a large-scale
s e high
ASCAX-raydataofRCW86areconsideredbyVinket density gradient. We apply this method here to the dis-
al. (1997). The authors reveal a remarkable temperature cribing of an evolution of the SNR RCW86. The possi-
variationoverthe SNR: T =0.8 ÷5 keV. They estimate bilities of the SN explosion in 185 A.D. and in the OB-
e
the preshockdensitycontrastfortheSWandNEpartsof assosiation will be considered separately .
RCW86 as 5÷50 times. We take as the basic initial model’s parameters the
Modelling the SNR emission it is necessary to take three observational results: the visual angular size of the
into considerationthe possibility that the plasma may be SNR Θ ≈ 40′, the temperature of the plasma’s hot com-
under the NEI conditions,since the effects ofthe NEI are ponent T = 3.4 ± 0.2 keV, which corespond to cor-
high
important.Atthesametimethenonuniformityofthesur- rected for the contribution of the galactic ridge emission
rounding ISM is very important, too. Calculations show spectrumfromthe wholeSNR(Claasetal.1989)andthe
that the influence of a nonuniform medium on the evolu- surfacebrightnessdistributionprofilealongthesymmetry
tion and the properties of the emissionof adiabatic SNRs axis of the SNR’s image (Pisarski et al. 1984) (see Fig.4,
may exceed the influence of the NEI in 101 ÷105 times 7). This profile is sensitive to large-scale density distribu-
(Hnatyk & Petruk 1998, hereafter Paper I). At present it tionofISMandallowstoseethepropertiesofthedensity
is practically inpossible to construct models which take gradient.
O. Petruk:Evolution of supernovaremnants in theinterstellar medium with a large-scale density gradient 3
Table 1.ParametersofRCW86derivedfromtheX-rayobservations.RisradiusoftheSNR,E51 istheenergyofSNexplosion
in 1051 erg, noH is hydrogen numberdensity of surroundingmedium, M is swepted up mass, M⊙ is themass of Sun.
Hydrodynamicsmodela PEMb t, years R, pc d, kpc E51 noH, cm−3 M, M⊙ Ref.d
Free expansion (SNII) CIE 1800 9 1.4 0.1÷0.2 0.13÷0.25 7÷14 [1]
Sedov (SNI) CIE 1800 6.5 1.4 0.1÷0.2 0.3 5 [1]
Free expansion or Sedov NEI 400÷1900 2.5÷9 0.4÷1.6 0.1÷0.2 0.04÷0.2 0.1÷21 [2]
Sedov for NEpart of theSNR NEI 1800 6.7 1.1 0.15 0.11c 4.9 [3]
a All model assumed uniform ISM(density of ISM ρo=const)
b Plasma emission model
c The hydrogen numberdensity of theISM for theSW part is estimated as (noH)SW ∼10(noH)NE
d Reference: [1] – Pisarski et al. (1984); [2] – Nugent et al. (1984); [3] – Claas et al. (1989).
These characteristics are supplemented additionally We will consider the evolution of the SNR in the
with the fourth one depending on the one from two basic nonuniformmediumwiththeflatexponentialdensitydis-
assumptions: the first assumption is that RCW86 is a re- tribution
sultoftheSNexplosionin185A.D.(thatgivesthe ageof
r
the SNR) and the second one is that RCW86 is a result ρo(r,θ)=ρo(0)exp − cosθ , (4)
H
of the SN explosion in the OB-assosiation(this yields the (cid:16) (cid:17)
distance to the SNR). where ρo(0) is the initial density around the place of SN
The temperature T at the shock front in the Sedov explosion,H is the scale ofthe density nonunifirmity,θ is
s
(1959)modeloftheSNRmaybeobtanedfromT :T ≈ an angle between the maximal density decreasing direc-
high s
T /1.3 (Itoh 1977). In Paper I it have been shown that tion and the considered direction. Such exponential dis-
high
the nonsperical SNR may be characterised by the some tribution is the one of the most possible to approximate
”average”characteristictemperatureT ∝(M)−1,which realcontinuousdensitydistributionsininterstellarclouds,
ch
will be likely connected with T : gaseous galactic disc etc.
high
When we fixed T and t we have got from (2) an
T ≈T /1.3=(3.03±0.2)·107 K. (1) ch
ch high estimation on
Wehavealsoshownthatduringtheadiabaticstagethe
E /no(0)=1.5±0.25. (5)
valueofT isclosetoT oftheSedovSNRwiththesame 51 H
ch s
initialmodelparameters(e.g.,theenergyofexplosionE ,
o Fromthisrelationandappropriateexplosionenergyrange
the hydrogen number density in the place of explosion E = 0.1÷ 1 the range for the initial number density
noH(0))withthemaximalerrorabout20÷30%.Therefore, no51(0)=0.06÷0.8 cm−3 follow.
we may write analoguously to the Sedov case that H
Fromthedefinitionofadimentionlesstime(Hnatyk&
Tch ≈Ts =2.08·1011 nEo5(10) 2/5ty−r6/5 K, (2) P(wehtreurke α19A96is)tτhe=setl/f-tsmimwiliathr ctomns=taαnt1A,/2REmo−1is/2tρh1oe/2d(i0s)tRanm5c/2e
(cid:18) H (cid:19) scale) it is obtaned for γ =5/3 that
Tch ≈6.47·109(cid:18)nEoH5(10)(cid:19)(Rpc)−3 K, (3) t=τtm =18.01 nEo5(10) −1/2Rm,pc5/2τ yr, (6)
(cid:18) H (cid:19)
whereE =E ·10−51,t istheageoftheSNRinyears,
51 o yr R is the distanse scale in pc.
R is the average radius of the nonspherical SNR in pc. m,pc
pc
From this relation and (5) a connection between the
Hereafter γ =5/3.
scale high H and the dimentionless time follows:
ForthecalculationoftheX-rayemissionoftheplasma
in the CIE the Raymond & Smith (1977) data were ap- H =R ≃(6.8±0.2)·τ−2/5 pc. (7)
m
proximated.
The possible values of H lie in the range H = 150 pc
2.2. RCW86 as the result of the SN explosion in 185 A.D. fortheGalactic’sdisktofewparsecsforalocalnonunifor-
mityoftheISM.Theinfluenceofthenonuniformmedium
2.2.1. Exponential medium
onthedynamicsoftheSNRbecomemoreapprecieblewith
Let us take as initial SNR characteristicsthe angular size the age of a SNR. The surface brightness distribution of
Θ,the temperatureT ,the surfacebrightnessprofileand a SNR will be close to the Sedov one for small τ. Since
ch
additionally the age of the SNR t=1800 years. the observationaldistribution of the surface brightness in
4 O. Petruk:Evolution of supernovaremnants in theinterstellar medium with a large-scale density gradient
(cid:16)(cid:22)(cid:17)(cid:25) (cid:21)(cid:17)(cid:24)
+ (cid:26)(cid:19)(cid:3)SF
(cid:16)(cid:23)(cid:17)(cid:19) D + (cid:20)(cid:19)(cid:3)SF
(cid:21)(cid:17)(cid:19)
(cid:12)(cid:21)
(cid:16)(cid:23)(cid:17)(cid:23) D[(cid:3)(cid:15) QRQVPRRWKHG
(cid:16)(cid:20)VWHU(cid:12)(cid:3)(cid:3)(cid:16)(cid:23)(cid:17)(cid:27) E (cid:18)6PD[(cid:3)(cid:15)(cid:20)P(cid:20)(cid:17)(cid:24)
(cid:21)(cid:3) 6
(cid:16)P J(cid:11)
(cid:16)(cid:20)VF(cid:16)(cid:24)(cid:17)(cid:21) 6PD[(cid:15)(cid:20) O
J(cid:3) (cid:20)(cid:17)(cid:19) REVHUYHG
U
H
6(cid:15)(cid:3)(cid:16)(cid:24)(cid:17)(cid:25)
(cid:11)
OJ F VPRRWKHG(cid:3)WR(cid:3)(cid:22)(cid:10)
(cid:19)(cid:17)(cid:24)
(cid:16)(cid:25)(cid:17)(cid:19) 6PD[(cid:15)(cid:21)
(cid:24)(cid:17)(cid:19) (cid:20)(cid:19)(cid:17)(cid:19) (cid:20)(cid:24)(cid:17)(cid:19) (cid:21)(cid:19)(cid:17)(cid:19)
+(cid:15)(cid:3)SF
G
(cid:16)(cid:25)(cid:17)(cid:23)
Fig. 2. Ratios of the values of the two peaks in the surface
brightness S distribution versus the high scale H for models
(cid:16)(cid:25)(cid:17)(cid:27) of SNR in the flat exponential media (4). Here t=1800 years
(cid:19)(cid:17)(cid:19) (cid:20)(cid:19)(cid:17)(cid:19) Q (cid:21)(cid:19)(cid:17)(cid:19) (cid:22)(cid:19)(cid:17)(cid:19) (cid:23)(cid:19)(cid:17)(cid:19) and E51/noH(0) = 1.5. In the figure both contrast of surface
(cid:15)(cid:3)DUFPLQ brightness are shown: smoothed to 3′ like observed (Pisarski
et al. 1984) and without smoothing.
Fig. 1. Surface brightness S distribution in the diapason
ε = 0.1÷2 keV along of the median of the SNR in the ISM
withtheexponentialdensitydistribution(4)forthefourcases is the same. Value of the blast energy E (and as conse-
51
oftheinitialmodelparameters.Itissupposehereforallcases quence of (5), the number density no(0)) give the ampli-
H
that E51/noH(0) = 1.5, Hpcτ2/5 = 6.85, t = 1800 years and tude of the profile which may differ with the factor about
Tch = 3.0·107 K. Solid lines show the cases with H = 70 pc, few ten times. It is follow from Fig.1 that the values H
dashed – the cases with H = 10 pc. Energies of explosion: a,
and τ affect on the contrast of the surface brightness.
b: E51 = 1; c, d: E51 = 0.15. Dots represent the cases c, d For finding H, it is possible to accept the contrastbe-
multiplied by thefactor ((no(0)) /(no(0)) )2 =101.66.
H top H bottom tween the brightness in the different points of the SNR’s
image. For example, it may be used the relation between
RCW86isnotsimilarto theSedovone,thedimentionless the two ”near-front” maximums in the distribution. The
time τ have not to be small. most appropriate value of high scale for this choice, as
The observed ratio ξ between the axes of the visual it may be seen from Fig.2, is H = 11.0 pc (and respec-
shape of RCW86 is in range ξ = 1.0÷ 1.2 (Pisarski et tively τ = 0.3). Then calculations show (Fig.4) that the
al. 1984). This corresponds to ranges of τ = 0.01 ÷ 11 most appropriate value of the blast energy is E51 = 0.22
(Fig.5 in Paper I). The scale hight H, as it follow from (no(0)=0.15 cm−3).
H
(7), then must be anywhere between 2.6 and 40 pc. Such FornormalizingthesurfacebrightnessS inphotonen-
wide range of H is a result of very slow decreasing of the ergyrangeε=0.1÷2.0keV,theeffectivevaluesofphoton
visualshape’sanizotropywithtime(thesameFig.).Small energyε=0.1keVandabsorptioncrosssectionσ ascross
scalehight(likeH =2.6pc)itseemstobeimpossiblebe- section at photon energy ε=0.316 keV from Morrison &
cause the surface brightness contrast then must be about McCammon (1983) have been used.
104.7 times. Possiblepresence in the shape anizotropythe Obtaned SNR’s luminocity L0.1−2 = 3.5·1034 erg/s
x
residual component related to anizotropy which the SNR is consistent with esteblished by Pisarski et al. (1984)
might haveonthe free expansionstage (due to ashperical L0.2−4 =2·1034d2 erg/s.
x kpc
explosion) is the second reason which complicates using InFig.5itisshownthemapsofthesurfacebrightness
the small anizotropy of the shape for estimating the H distribution for this model in comparison with the distri-
and τ. butionofthe spectralindexα=−∂lnF /∂lnε,whereF
ε ε
Physically the most correct limitation on H and τ we is the continuum flux at photon energy ε.
may get from the profiles of the surface brightness. In
Fig.1 are shown the profiles of the surface brightness for
2.2.2. Power law medium
the four cases of the initial model peremeters. Relations
(5) and (7) have place for each set of the parameters. We Itisusefultonote,thatconcreeteappearanceofsurround-
may see that for fixed H and τ the shape of the profile ing density distribution is not strongly essential then size
O. Petruk:Evolution of supernovaremnants in theinterstellar medium with a large-scale density gradient 5
0.03
(cid:21)(cid:17)(cid:24)
(cid:20)
(cid:21)
(cid:21)(cid:17)(cid:19) nonsmoothed (cid:21)(cid:16)(cid:20)(cid:3)V0.02 (cid:22)
6(cid:12)k (cid:16)PLQ (cid:23)
(cid:18)PD[(cid:15)(cid:20)(cid:3)(cid:20)(cid:17)(cid:24) V(cid:3)DUF (cid:24)
(cid:11)6 QW
J X
O observed FR0.01
6(cid:15)(cid:3)
(cid:20)(cid:17)(cid:19)
smoothed to 3'
0.00
(cid:19)(cid:17)(cid:24)
(cid:24)(cid:17)(cid:19) (cid:20)(cid:19)(cid:17)H(cid:19), pc (cid:20)(cid:24)(cid:17)(cid:19) (cid:21)(cid:19)(cid:17)(cid:19) (cid:19)(cid:17)(cid:19) (cid:20)(cid:19)(cid:17)(cid:19) Q (cid:21)(cid:19)(cid:17)(cid:19) (cid:22)(cid:19)(cid:17)(cid:19) (cid:23)(cid:19)(cid:17)(cid:19)
(cid:15)(cid:3)DUFPLQ
Fig. 3.Thesame asin Fig. 2 fortheratios between themax- Fig. 4. Observed (line 1) distribution of the surface bright-
imum in the surface brightness distribution Smax,1 and the ness S along of the symmetry axis of RCW86 (Pisarski et al.
value of the brightness in the visual geometric center of the 1984) and the distributions in the models of SNRs with the
SNR S . age t = 1800 years: line 2 – the exponential medium (4) with
c
H = 11 pc and E51 = 0.22, noH(0) = 0.15 cm−3; line 3 – the
power law medium (9) with ro = 21.8 pc and the same E51,
ofSNRdoesnotexceedafewscalehights.Namely,similar no(0);line4–theexponentialmedium(4)withH =5pcand
H
parameters of model may be obtained under assumption E51 =0.17, noH(0)=0.11 cm−3; line 5 – the uniform medium,
that the SNR evolve in other type of media, if contrastof E51 = 0.17, noH = 0.11 cm−3. All distributions are smoothed
density along the surface of the SNR will be similar. Let to 3′.
us consider, for example, a medium with the spherically-
symmetrical power law density distribution created by
steller wind
higher density (e.g., interstellar cloud) where density is
distributed according to the exponential law (4). In this
ρo(r˜)=ρ (r˜/R )ω, (8)
o m case one part of the SNR (the NE part of RCW86) will
evolveintheuniformmediumwhereastheevolutionofthe
thentheSNexplosionpositionisdisplacedonthedistance
anotherpart(theSWpartofRCW86)willbedetermined
r from the center of symmetry r˜=0, therefore, the den-
o by the variation of the density in the dense region. In
sity distribution as a function of the distance r from the
this compositemodelthe surfacebrightnessinthe central
explosion point r =0 is:
part of the SNR will be close to the brightness which the
ω Sedov model gives, but the maximal brightness will be
r2+r2+2rr cosθ
ρo(r,θ)=ρo(0) o o . (9) determined by the concrete distribution of the density in
r
p o ! the nonuniform region.
We take ω =−2 and calculate a number of models in When we take as initial parameter the ratio of the
ordertogetanappropriater usingthesurfacebrightness maximum of the brightness distribution to the value of
o
contrastbetweenthetwomaximumsinthebrightnessdis- thebrightnessinthevisualgeometricalcentreoftheSNR,
tribution. we obtaine another model parameters for the case of the
Observedsurfacebrightnesscontrastisobtanedinthis exponential law density distribution, namely, H = 5.0 pc
model with ro = 21.8 pc. Characteristic scale Hch for (Fig.3) and E51 = 0.17 (Fig.4). We may calculate now
medium (9) with such r may be estimated. It equals the profile of the surface brightness for the Sedov model
o
H = 11 pc. Other parameters of the model are really wich discribes the NE part of RCW86 in the assumption
ch
close to the same parameters of the SNR in exponential of the composite medium (Fig.4). The initial parame-
medium (4) and shown in Tables 2 and Fig.4. ters for it are the same as for the SW part, E51 = 0.17,
no(0) = 0.11 cm−3. We may see that the composite
H
medium model (exponential plus constant density dis-
2.2.3. Exponential plus uniform medium
tributions) more better describes the observational sur-
RCW86 may evolve in the ISM with the contact between face brightness distribution than the exponential medium
the uniform medium ρo = const and the region with the alone.
6 O. Petruk:Evolution of supernovaremnants in theinterstellar medium with a large-scale density gradient
Table 2. Parameters of RCW86 obtaned from the different
models of the SNR. Plasma emission model is CIE. Lx is in
thephotonenergyrangeε=0.1÷2.0keV.R(0)andR(π)are
(cid:24)(cid:17)(cid:19) (cid:24)(cid:17)(cid:19)
the radii of shock in directions θ = 0 and θ = π respectively,
D is theshock velocity.
SF Parameters Initial density distribution
5(cid:17)(cid:3)(cid:19)(cid:17)(cid:19) (cid:19)(cid:17)(cid:19) of model Ea PLb EUc Ea Ea
Observed
Θ, arcmin 40 40 40 40 40
Tch, 107 K 3.0 3.0 3.0 3.0 3.0
Additionally supposed
(cid:16)(cid:24)(cid:17)(cid:19) (cid:16)(cid:24)(cid:17)(cid:19)
t, years 1800 1800 1800 — —
d, kpc — — — 2.8 2.8
D
H,pc 11 ∼11d 5.0 26 22
δ, degree 0 0 0 0 30
Obtained
E t, years 1800 1800 1800 4300 4300
Eo, 1051 erg 0.22 0.22 0.14 2.0 2.0
(cid:24)(cid:17)(cid:19) (cid:24)(cid:17)(cid:19) no(0), cm−3 0.15 0.15 0.10 0.1 0.1
H
R(0), pc 7.50 7.42 6.84 17.96 18.25
R(π),pc 6.27 6.20 5.78 14.96 14.78
F D(0), km/s 1800 1740 1490 1810 1870
S
5(cid:17)(cid:3)(cid:19)(cid:17)(cid:19) (cid:19)(cid:17)(cid:19) TD((0π)),,1k0m7/Ks 142.4650 142.1280 130.0730 142.5500 142.8320
T(π),107 K 2.17 2.03 1.58 2.16 2.07
no(R,0), cm−3 0.08 0.08 0.10 0.18 0.19
H
no(R,π), cm−3 0.27 0.29 0.31 0.05 0.04
H
(cid:16)(cid:24)(cid:17)(cid:19) (cid:16)(cid:24)(cid:17)(cid:19) d, kpc 1.18 1.17 1.08 2.83 2.83
M, M⊙ 6.9 6.8 — 62 62
lg(Lx,erg/s) 34.54 34.53 — 35.33 35.35
α(5 keV) 1.90 1.88 — 1.91 1.92
(cid:16)(cid:24)(cid:17)(cid:19) (cid:19)(cid:17)(cid:19) (cid:24)(cid:17)(cid:19) a E is theflat exponential medium (4)
5(cid:15)(cid:3)SF b PL is thepower law medium (9) with ω=−2 and
r =21.8 pc
o
Fig.5.aThesurfacebrightnessS (inerg s−1 cm−2 st−1)dis- c EU is theexponential medium (4) plus uniform
tribution in the photon energy range ε = 0.1−2 keV. b The d Characteristic scale hight Hch
surfacedistributionofthespectralindexαatε=5keVforthe
SNR model in the exponential medium (4) with H =11.0 pc.
Parameters of the model are t = 1800 years, E51 = 0.22,
no(0)=0.15 cm−3 and δ=0o. estimated(e.g.,Westerlund1969).Inthissectionthenext
H
question is the main: what scale hight has the medium
aroundtheRCW86iftheSNR’sprogenitorhaveexploded
It is nesessary to note, when RCW86 is the SNR cre- in this OB-assosiation? Therefore the nonuniform media
atedbytheexplosionin185A.D.,weseeitnearthemax- with exponential density distribution (4) is used.
imum disclosing projection (δ ≈0o). Really, another pro- We take as the initial SNR characteristics the obser-
jections hide the real contrasts (e.g., of axes size, surface vational angular size Θ, the temperature Tch, the surface
brightness; see Paper I). If we see the contrast of surface brightness profile and, additionally, the distance to the
brightness decreased already, then the real one must be remnant d≈2.8±0.4 kpc (Rosado et al. 1996).
greater and therefore H must be smaller. ItispossiblenowtoestimatetheaverageSNR’sradius
ObtanedparametersofRCW86aresummarisedinTa- from Θ and d: Rpc = 0.5dkpcΘ′/3.438 ≈ 16.3±2.3 pc.
ble 2 (the first three columns). Therefore we have got from (3) an estimation on
E /no(0)≃23±10. (10)
51 H
2.3. RCW86 as the result of the SN explosion in the OB-
This means that the explosion energy must be high and
association
the initial number density have to be low.
Let us suppose now that RCW86 is a result of the SN The age of the SNR as it follows from (2) is any-
explosionin the OB-assosiationwhich distance from us is where from t = 3500 years for T = 2.8 · 107 K and
ch
O. Petruk:Evolution of supernovaremnants in theinterstellar medium with a large-scale density gradient 7
(cid:20)(cid:17)(cid:24) (cid:19)(cid:17)(cid:19)(cid:22)
(cid:20)
(cid:21)
(cid:20)
(cid:12)PD[(cid:3)(cid:15)(cid:21) (cid:16)(cid:21)(cid:16)PLQ(cid:3)V(cid:19)(cid:17)(cid:19)(cid:21) (cid:22)
6 F
(cid:18)D[(cid:3)(cid:15)(cid:20)(cid:20)(cid:17)(cid:19) V(cid:3)DU
P QW
6 X
OJ(cid:11) REVHUYHG d (cid:19)R 6(cid:15)(cid:3)FR(cid:19)(cid:17)(cid:19)(cid:20)
d
R
(cid:22)(cid:19)
(cid:19)(cid:17)(cid:24) (cid:19)(cid:17)(cid:19)(cid:19)
(cid:20)(cid:24)(cid:17)(cid:19) (cid:21)(cid:19)(cid:17)(cid:19) (cid:21)(cid:24)(cid:17)(cid:19) (cid:22)(cid:19)(cid:17)(cid:19) (cid:22)(cid:24)(cid:17)(cid:19) (cid:19)(cid:17)(cid:19) (cid:20)(cid:19)(cid:17)(cid:19) Q (cid:21)(cid:19)(cid:17)(cid:19) (cid:22)(cid:19)(cid:17)(cid:19) (cid:23)(cid:19)(cid:17)(cid:19)
U(cid:19)(cid:3)(cid:15)(cid:3)SF (cid:15)(cid:3)DUFPLQ
Fig. 6. Ratios of the values of the two peaks in the X-ray Fig.7.DistributionofthesurfacebrightnessS alongthesym-
surfacebrightnessS distributionversusH fortheSNRsinthe metryaxisoftheSNRRCW86:line1–observational(Pisarski
exponentialmedia(4),whent=4300years,E51/noH(0)=20.3 et al. 1984); line 2 – model in exponential medium (4) with
andδ=0o.DashedlineisthesamerelationforthesameSNR H =25.7pcandt=4300years,E51 =2.0,noH(0)=0.1cm−3,
models, except δ=30o. All ratios are smoothed to 3′. δ = 0o; line 3 – the same model, except H = 22.3 pc and
δ=30o. All distributions are smoothed to 3′.
E /no(0) = 13, up to t = 5350 years in case T =
51 H ch
3.2·107 K and E /no(0)=32. We may see, with agree- bution of the spectral index are also very close to shown
51 H in Fig.8 case of δ =0o.
mentwith Rosadoetal.(1996),if RCW86is the remnant
oftheSNexplodedinOB-assosiationitcannotbeborned
in 185 A.D. 2.4. Contribution of the ejecta
For the distance d = 2.8 kpc and T = 3.0·107 K
ch
the age t = 4300 years and ratio E /no(0) = 20.3 have Nugent et al. (1984) have made the conclusion that the
51 H observeddataonRCW86canbefittedwithouttakinginto
been obtained. We calculate the evolution of RCW86 in
considerationthe emissionfromthe ejecta.Theseauthors
theexponentialmedium(4)withtheseinitialparameters.
have also suggested that the value of β = M /E (M
It may be seen from Fig.6 that the scale hight H = ej 51 ej
25.7pcisthemostappropriateforthecaseofthemaximal is the mass of ejecta in M⊙) lies much probably in the
range 1 to 10. In our models of the SNR the blast energy
disclosure of the remnant (δ =0o).
are E = 0.22 in the case of SN explosion in A.D.185
InordertoestimateanexplosionenergyE andanum- 51
o and E = 2.0 in the case of SN explosion at a distance
berdensityno(0)itwascalculatedanumberofthemodels 51
H d=2.8kpc.Terefore,themassofejectamaybeestimated
with such H and different E . It may be seen from Fig.7
that the surface brightnessdoistribution alongthe symme- as Mej = 0.22÷2.2M⊙ or Mej = 2÷20M⊙. Swepted up
try axis for E = 2.0, no(0) = 0.1 cm−3 is near of the massesarerespectively7and62M⊙ andexceedsMej in3
51 H to 31times.We expect thatunderour assumptionofCIE
observedone.Thewholemapsofthe surfacebrightnessS
conditions and obtained no ∼ 0.1 cm−3, such swepted
and the spectral index α distribution are shownin Fig.8. H
up masses are enough for RCW86 to be in Sedov phase
Fig.7 demonstrates also the surface brightness distri-
of his evolution. So, we may assume that in both cases
bution along the symmetry axis for the case of the dif-
the ejecta do not considerably modify the emission from
ferent projection of the SNR, when δ = 30o. We see that
RCW86. However, if to take into account the NEI effects
distribution, like observational, may be obtaned also un-
and consider the value of the β up to the possible upper
der different projection of the 2D SNR on the plan of the limit which have been put as β ∼ 40, the contribution of
skywiththesmallerH =22.3pc.Nowwecannotseparate
the ejecta into the emissionmight be essential(Nugent et
one case of projection from another but we see also that
al. 1984). On the other hand, if RCW86 would evolve in
the model with δ = 30o does not essentially change the verydense ISM(no ∼103),it wouldentry inSedovstage
SNR’s parameters which are summarised for these differ- after swepting upHM ∼ 19M (Dohm-Palmer & Jones
ej
ent models of RCW86 in the last two columns of Table2.
1996).
Modellingrevealalsothatformodelwithδ =30o both
the surface brightness distribution and the surface distri-
8 O. Petruk:Evolution of supernovaremnants in theinterstellar medium with a large-scale density gradient
It is shown that observed surface brightness distribu-
tion of RCW86 may be obtaned in the models with the
twodifferentinitialassumptions:oneabouttheSupernova
explosion in 185 A.D. and another about the explosionin
(cid:20)(cid:19)(cid:17)(cid:19) (cid:20)(cid:19)(cid:17)(cid:19) the OB-asossiation. Data we posses do not allow us to
diside which of those models is true. The parameters ob-
taned for RCW86, basing on these two assumption, are
F
5(cid:15)(cid:3)S(cid:19)(cid:17)(cid:19) (cid:19)(cid:17)(cid:19) summarasedinTable 2.Themodelsgivetheobservational
contrasts of the surface brightness when to consider the
ISM with the scales of nonuniformity 11 pc (if the age of
RCW86is1800years)or20−25pc(whentheSNRisdis-
(cid:16)(cid:20)(cid:19)(cid:17)(cid:19) (cid:16)(cid:20)(cid:19)(cid:17)(cid:19) tant from us on 2.8 kpc). The preshock density contrasts
between the southwestern and the northeastern parts of
RCW86inrange3.5−4.5areobtainedforbothconsidered
D
assumptions.
Acknowledgements. I am very grateful to Dr.Bohdan Hnatyk
E
for the permanent attention to my work as well as for always
usfull advices.
(cid:20)(cid:19)(cid:17)(cid:19) (cid:20)(cid:19)(cid:17)(cid:19)
References
Caswell, J., Clark, D., Crawford, D. 1975, Austr. Phys. Ap.
F
5(cid:15)(cid:3)S(cid:19)(cid:17)(cid:19) (cid:19)(cid:17)(cid:19) Suppl.,37, 46
Chin, Y.-N., Huang,Y.-L. 1994, Nature, 371, 398
Claas, J., Smith,A., Kaastra, J., et. al. 1989, ApJ, 337, 399
Clark, D., Stephenson, F. 1977, Historical Supernovae, Perga-
mon, Oxford
(cid:16)(cid:20)(cid:19)(cid:17)(cid:19) (cid:16)(cid:20)(cid:19)(cid:17)(cid:19)
Dohm-Palmer, R.,Jones, T. 1996, ApJ, 471, 279
Itoh, H.1977, Pub.Astr. Soc. Japan, 29, 813
Hill, E. 1967, Austr.J. Phys., 20, 297
Hnatyk,B.,Petruk,O.1996,KinematicsandPhysicsofCeles-
tial Bodies, 12, 35
(cid:16)(cid:20)(cid:19)(cid:17)(cid:19) (cid:19)(cid:17)(cid:19) (cid:20)(cid:19)(cid:17)(cid:19)
Hnatyk,B., Petruk, O.1998, A&A,in press (Paper I)
5(cid:15)(cid:3)SF
Klimishin,I.1984,Shock WavesintheStar Envelopes,Nauka,
Fig.8.aThesurfacebrightnessS (inerg s−1 cm−2 st−1)dis- Moscow
tribution in the range 0.1−2 keV and b the surface distribu- Milne, D. 1970, Austr. J. Phys., 23, 425
tionofthespectralindexαatε=5keVfortheSNRmodelin Milne, D. 1987, Austr. J. Phys., 40, 771
Morrison, R., McCammon, D. 1983, ApJ, 270, 119
the nonuniform exponential medium (4) with H = 25.7 pc.
Naranan,S.,Shulman,S.,Yentis,D.etal.1977,ApJ,213,L53
Parameters of the model are t = 4300 years, E51 = 2.0,
no(0)=0.1 cm−3 and δ=0o. Nugent,J.,Pravdo,S.,Garmire,G.etal.,1984,ApJ,284,612
H
Pisarski, R., Helfland,D., Kahn,S. 1984, ApJ, 277, 710
Raymond,J., Smith, B. 1977, ApJS, 35, 419
3. Conclusions Rosado, M., Ambrosio-Cruz, P., Le Coarer, E., et al., 1996,
A&A,315, 234
Observations reveal complicated RCW86 morphology. Ruiz, M. 1981, ApJ,243, 814
ClosetosphericalshapeoftheSNRcoexistswiththesur- Schaefer, B. 1995, AJ, 110, 1793
face brightness distribution which is far from the Sedov Sedov, L. 1959, Similarity and Dimensional Methods in Me-
one. When we take into account the nonuniform ISM we chanics, Academic, New York
may explain such morphology. Strom, R. 1994, MNRAS,268, L5
van den Bergh, S. 1978, ApJS,38, 120
Weconsideredherethe2-DmodelsofRCW86inafew
Vink,J., Kaastra, J., Bleeker, J. 1997, A&A,328, 628
types of nonuniform media. It is shown that the features
Westerlund, B. 1969, AJ, 74, 879
of the surface brightness distribution allow us to restore
Winkler, P. 1978, ApJ,221, 220
the SNR characteristics. As it is described, the contrasts
Whiteoak, J.B., Green, A. 1996, A&AS,118, 329
in the surface brightness depend on the gradient of the
density of the ISM and the age of the SNR, thereas the
amplitude of the surface brightness distribution depends
on the energy of the Supernova explosion and the initial ThisarticlewasprocessedbytheauthorusingSpringer-Verlag
a
density around the place of explosion. LTEX A&Astyle fileL-AA version 3.