NASA Technical Reports Server (NTRS) 20140010426: Detection of the Compressed Primary Stellar Wind in eta Carinae*

0.46 MB·English

by  NASA Technical Reports Server (NTRS)

Accepted for publicationin ApJL
PreprinttypesetusingLATEXstyleemulateapjv.04/17/13
DETECTION OF THE COMPRESSED PRIMARY STELLAR WIND IN η CARINAE(cid:2)
M. Teodoro1,2, T. I. Madura1,3, T. R. Gull1, M. F. Corcoran4,5, and K. Hamaguchi4,6
1AstrophysicsScienceDivision,Code667,NASAGoddardSpaceFlightCenter,Greenbelt, MD20771,USA
Accepted for publication in ApJL
ABSTRACT
A series of three HST/STIS spectroscopic mappings, spaced approximately one year apart, reveal
three partial arcs in [Fe II] and [Ni II] emissions moving outward from η Carinae. We identify these
arcswiththeshell-likestructures,seeninthe3Dhydrodynamicalsimulations,formedbycompression
of the primary wind by the secondary wind during periastron passages.
Subject headings: Stars: individual (η Carinae) — stars: massive — binaries: general
1. INTRODUCTION proachingtheobserver. Thecompanionstarspendsmost
η Carinae (hereafter η Car) is now recognized to ofthe orbitnear apastron,so that the undisturbed wind
from η Car A is located on the far side of the system,
be a massive binary system in a highly eccentric or-
bit (e > 0.9). The primary star, η Car A, is in red-shiftedasseenfromEarth,andlargelyunaffectedby
the ionizing radiation from η Car B.
the luminous blue variable (LBV) phase, with a lu-
minosity of 106.7 L(cid:2) and mass of about 120 M(cid:2) 1−Du2riAngUpoefriaηstCroanr pAa,ssleaagdei,nηgCtoarrBapaidppcrhoaanchgeesswiniththine
(Davidson & Humphreys 1997). It is surrounded by
shape of the wind-wind interaction surface and confine-
a massive stellar wind (M˙ ∼ 10−3 M(cid:2)yr−1, v∞ ≈ ment of the far-UV flux from η Car B. Hydrodynamical
420kms−1; Groh et al.2012) thatinteractswith a com- simulations (Madura et al. 2012, T. Madura et al. 2013,
panion,η CarB,thathasnotbeendetecteddirectlyyet. inpreparation)showthat,aftereachperiastronpassage,
The physical parameters of η Car B are, thus, inferred a highly-distorted volume of hot, low density secondary
from X-ray (M˙ ∼ 10−5 M(cid:2)yr−1, v∞ ≈ 3000 kms−1; windpushesoutwardintotheslow,high-densityprimary
Pittard & Corcoran2002)andphotoionizationmodeling wind,leadingtotheformationofathin,highdensitywall
(∼105L(cid:2),34000K<Teff <39000K;Verner et al.2005; surroundingthe lowerdensity, trapped wind of η Car B.
Teodoro et al. 2008; Mehner et al. 2010). This dense wall is accelerated to velocities somewhat
The orbital orientation is such that the orbital plane higher than the terminal velocity of η Car A and ex-
has an inclination of i ≈ 140◦, with the semi-major pands both in the orbital plane and perpendicular to it,
axis pointing along position angle (P.A.7) 314◦ and lon- creating a thin, high-density sheet of trapped primary
gitude of periastron ω ≈ 240◦ (Gull et al. 2009, 2011; wind material.
Madura et al. 2012). In this configuration, η Car B is Using spectral imaging maps from the Space Tele-
in front of η Car A at apastron, and the orbital angular scope Imaging Spectrograph(STIS), on board the Hub-
momentumvectorisparalleltothesymmetryaxisofthe bleSpaceTelescope(HST),weidentify structureswhich
bipolar Homunculus nebula that surrounds the system. are consistent with these walls of compressed primary
Projected on the sky, at apastron, the secondary star wind material. We identify three partial arcs formed by
directly illuminates and photo-ionizes the circumstellar the close passage of η Car B around the primary star
material within 1200 AU of the binary system primar- over the last 3 orbital cycles, and using the STIS maps,
ily to the northwest, where the material is primarily ap- we determine their space velocity and age.
2. OBSERVATIONS,DATAREDUCTION,ANDANALYSIS
[email protected]
2CNPq/SciencewithoutBordersFellow. Observations using HST/STIS were accomplished
3NASAPostdoctoral ProgramFellow. throughthree guestobserverprograms,12013(2010Oc-
4CRESST and Xray Astrophysics Laboratory, Code 662,
tober 26), 12508 (2011 November 20), and 12750 (2012
NASA Goddard Space Flight Center, Greenbelt, MD 20771,
USA October18),usingtheG430MandG750Mgratings,cen-
5Universities Space Research Association, 10211 Wincopin tered at 4706˚A and 7283˚A, respectively, in combination
Cir6cDlee,pSaurittmee5n0t0oCfoPluhmysbicias,,MUnDiv2e1r0si4t4y,oUfSMAaryland, Baltimore withthe52(cid:5)(cid:5)×1(cid:5)(cid:5)aperture. Anextendedregion,covering
County, 1000HilltopCircle,Baltimore,MD21250, USA. 1(cid:5)(cid:5) to 2(cid:5)(cid:5)×6.4(cid:5)(cid:5), centeredupon η Car,was mapped using
(cid:2)Based on observations made with the NASA/ESA Hubble 0.05(cid:5)(cid:5) spacing offsets at P.A. constrained by the space-
Space Telescope, obtainedattheSpace TelescopeScienceInsti- craft. Exposures were approximately 30 seconds using
tute,whichisoperatedbytheAssociationofUniversitiesforRe-
CRSPLIT=2. An additional 7 position sub-map, also
search inAstronomy, Inc., under NASAcontract NAS5-26555.
Theseobservationsareassociatedwithprogramnumbers12013, centered on η Car, employed shorter exposures to allow
12508, and 12750. Support for program numbers 12013, 12508, for potential saturation of continuum on the central po-
and 12750 was provided by NASA through a grant from the sitions.
Space TelescopeScience Institute, whichisoperatedbytheAs-
Data were reduced with STIS GTO idl-based8
sociation ofUniversitiesforResearchinAstronomy, Inc., under
NASAcontractNAS5-26555. tools similar to the standard pipeline reduction tools
7 P.A.isdefinedas theangle,intheplaneofthesky,eastward
ofNorth. 8 idlisatrademarkofExelisVisualInformationSolutions,Inc.
2 Teodoro et al.
[Fe II] λ4815 (2012 Oct 18) [Ni II] λ7413 (2011 Nov 20)
(a) +160 km s−1 (b) +200 km s−1 (c) +240 km s−1
−0.5 v = +200 km s−1 N
E
−0.5
0.0 −7.2
A1
0.5
−0.5 (d) +280 km s−1 (e) +320 km s−1 (f) +360 km s−1 sec) A3 A2 −7.4−2−11/2m Å)
δ (arcsec) 0.0 Δδ (arc 0.0 f−11/2 (erg s cλ
Δ
0.5 −7.6
(g) +400 km s−1 (h) +440 km s−1 (i) +480 km s−1 0.5
−0.5 1000 A.U.
0.0 −7.8
0.5 0.0 −0.5
Δα (arcsec)
0.5 1000 A.U.
Fig.2.— Iso-velocity image of [Ni II] λ7413 at +200 kms−1,
obtained in 2011 November 20. The arcs discussed in this work
0.5 0.0 −0.5 0.5 0.0 −0.5 0.5 0.0 −0.5
Δα (arcsec) are indicated by the arrows and were labeled A1, A2, and A3.
The dashed lines show the expected position of three spherical
shellstravelingat aconstant speedof475kms−1 for4.5,10, and
−7.8 −7.6 −7.4 −7.2 15.5 years, for the inner, middle, and outer arcs, respectively. To
fλ1/2 (erg s−1 cm−2 Å−1)1/2 determine the expansion rate, we restricted our measurements to
Fig. 1.—Iso-velocityimagesof[FeII]λ4815fortheobservations theeasterncomponentofA2.
of2012October18. Thevelocitiesareindicatedineachpanel. The
linearstructuresthatappearinpanels(g)and(h)aretheresultof structuresseenatdifferentwavelengths,expandingaway
abadCCDcolumn. Thephysicalscaleisbasedonanadopteddis- from the central source.
tanceof2350pc(Smith2006). Wenotethattherearenoextended
Comparisonofthesameiso-velocityimagefordifferent
structuresatDopplervelocitieshigherthan+440kms−1.
epochs (2010 October 26, 2011 November 20, and 2012
(Bostroem & Proffitt 2011), but with improved ability October18)revealedthatA1,A2,andA3moveoutward
to spatially align the spectra to sub-pixel accuracy. For from the central source (Figure 3). Because A3 is faint
each emission line of interest, continuum levels were fit- and hard to see in all the images, and A1 is affected by
tedforeachspatialpositionalongthe apertureandare- uncertaintiesinthe continuumsubtractionmethodclose
sampled data cube was generated, with the coordinates tothecentralsource,weconcentrateouranalysisonA2.
of right ascension and declination at 0.05(cid:5)(cid:5) spacing, and We note, however, that all three arcs seem to expand
Doppler velocity at 20 kms−1 intervals. at the same rate, but some extended parts become very
Since we are interested in structures formed in the distorted or diluted, which makes the determination of
primary wind around periastron, here we focus only on accurate positions a difficult task.
the following low-ionization lines: [Fe II] (λ = 4729.39, The position of the eastern portion of A2, relative to
0
4776.05, 4815.88, 7157.13)9 and [Ni II] (λ = 7413.65). the centralsource,wasmeasuredusing a radialflux pro-
0
Observationsfromprogram12013(2010October26)did file cut along its direction (Figure 3). For each dataset
notincludemapsof[FeII]λ7157.13and[NiII]λ7413.65. (2010,2011, and 2012), we extracted the flux profile be-
Visual examinationofiso-velocity imagesin eachdata tween0(cid:5)(cid:5) <r <0.9(cid:5)(cid:5) and100◦ <P.A.<105◦,where the
cuberevealedmultiplearcsinthelow-ionizationspectral emission from A2 is strongest.
lines used in this work. Sample images are presented in We used three different methods for measuring the
Figure 1 for the [Fe II] λ4815 transition. These arcs peakpositionin the radialflux profile curve. The firstis
are visible between P.A. 90◦ and 200◦. We identified a simple gaussianmodel, whichis a goodapproximation
at least three sets of partial arcs, labelled A1, A2, and when the flux profile of the shell is somewhat symmet-
A3 in Figure 2. The two innermost arcs, A1 and A2, ric. The second method relies on the barycenter of the
are conspicuously detected in all of the transitions from flux profile, and the third, developed by Blais & Rioux
[FeII] and [NiII]. The outermost arc, A3, is very faint. (1986), is based on the analysis of the derivative of the
These arcs are not artifacts of the HST/STIS point observedprofile. The position of the peak is the median
spread function (PSF). The HST/STIS PSF diffrac- value of these methods.
tive rings vary slowly with wavelength, are symmetric A comprehensive study comparing several methods
about the central core and have amplitudes substan- to determine peak positions with sub-pixel accuracy,
tially weaker than the partial rings apparent in the im- including the three methods used here, was done by
ages (Krist et al. 2011). As shown in the discussion be- Fisher & Naidu(1996). Thereaderisreferredtothatpa-
low,these featurescorrelatewith multiple forbiddenline per for further details. Based upon the individual mea-
surement errors, we estimate an accuracy of 0.25 pixel
9 All rest wavelengths, λ0, are in ˚A, in vacuum (Zethsonetal. (0.013(cid:5)(cid:5)) for the position of the shell using the median
2012). value of the three methods mentioned before. The to-
The compressed wind of η Car A 3
[Ni II] λ7413 (v=+200 km s−1)
(a) 2011 Nov 20 (b) Orientation and length (c) 2011 Nov 20 (d) 2012 Oct 18 −0.5
2012 Oct 18 of the extraction path
−14
A2 c)
0.0 e
s
−2−1m Å) −15 A2 A1 0.5 Δδ (arc
c 0.5"
−1
s
g
er
st.) ( [Fe II] λ4815 (v=+200 km s−1)
n
co (e) 2010 Oct 26 (f) 2010 Oct 26 (g) 2011 Nov 20 (h) 2012 Oct 18 −0.5
x 2011 Nov 20
fg ( λ10−14 2012 Oct 18 A2 0.0 ec)
o s
l arc
A1 δ (
−15 Δ
A2 0.5
0.8 0.6 0.4 0.2 0.0 0.5 0.0 −0.5 0.5 0.0 −0.5 0.5 0.0 −0.5
Δr (") Δα (arcsec)
Fig.3.— Flux profiles and iso-velocity images of two forbidden lines: (a) and (e) are for [NiII] λ7413 and [FeII] λ4815, respectively,
along the position indicated by the arrow in panel (b), which has a length of approximately 0.9(cid:2)(cid:2) and P.A. = 100◦. The flux profiles
were extracted from the iso-velocity images on the right. To illustrate the detected outward motion, in panels (c), (d), (f), (g), and (h)
the curved solid line marks the position of A2 in 2010 October 26. For better visualization, the flux profiles were shifted in the vertical
direction.
tal error, however, may be larger than that because it (475±29)kms−1. TheageofA2, t(A2),in2010,isthus
is dominated by the dispersion in the measured position obtained from <r (A2)>/V =(9.65±0.55) yr, where
p
for various forbidden lines. < r (A2) >= 0.41(cid:5)(cid:5) is the average projected distance of
p
A2, relative to the central source, at that epoch.
3. RESULTS Assuming that A1 is moving at the same constant
We note that A2 is moving outwards with a proper space velocity as A2, then, for a given epoch, we have
motion μ = (0.05±0.01)(cid:5)(cid:5)yr−1, on the plane of the sky
r (A1) t(A1)
(Figure4(a)), avalue suspiciouslyclosetothe pixelsize. p = , (3)
However,many other regionsacross the field of view are r (A2) t(A2)
p
fixed in position from image to image, giving us con-
fidence that the images are all spatially registered cor- where rp(A1) is the average projected distance of A1
rectly. Furthermore, the motion of portions of A2 de- from the central source, and t(A1) is its age. In 2010
pends on the local Doppler velocity (cf. Figure 4(a)). and 2011, A1 was too close to the central source to be
Hence, the detected motion is not caused by a system- accurately measured. Thus, we used the observations of
atic misalignment of the images. 2012to measure the position and then estimate the age.
The projected radius (rp), at a given Doppler velocity In 2012, we have rp(A1) ≈ 0.22(cid:5)(cid:5), resulting in t(A1) =
(V ), of a shell moving with space velocity V, during a (5.98±0.92) yr. The relation described in equation (3)
D
time t, is given by alsoappliestoA3,resultingint(A3)=(17.13±0.95)yr,
(cid:2) (cid:2) (cid:3)(cid:3) since it was observed at r (A3) ≈ 0.6(cid:5)(cid:5) in 2012. The
V p
r (t)=V t sin arccos D . (1) ratherlargeerrorsaredominatedby uncertaintiesin the
p V position due to the proximity to the central source for
A1, and the weak emission of A3.
Taking the difference of the projected radius of A2, at
Correcting for the time interval between the 2012 and
two epochs separated by a time interval Δt = t −t ,
2 1 2010 observations, and comparing the ages in 2012, we
leads to (cid:2) (cid:2) (cid:3)(cid:3) finally have t(A1) = 5.98, t(A2) = 11.63, and t(A3) =
V 17.13 yr. This means that each shell is created approxi-
rp(t2)−rp(t1)=V Δt sin arccos VD . (2) mately5.6yraftertheearlierone,showingthattheyare
directly tied to the orbital period of the binary system
Settingtheobservationsmadein2010Oct. 26asbase- (P =5.54 yr; Damineli et al. 2008).
line, we have Δt = 1.07 and 1.98 yr, for the two subse- To test for the consistency of our results across the
quent datasets. Thus, we used the measured projected range of Doppler velocities where A2 is observed, we
position for A2 in the range −40 ≤ VD ≤ +40 kms−1 adoptedasimplifiedmodelinwhichweconsiderA2tobe
(this is the small angle regime, where VD/V (cid:5) 1) part of a spherical shell centered on η Car, and allowed
to estimate the space velocity, which resulted in V = it to expand, at a constant space velocity of 475 kms−1,
4 Teodoro et al.
ource (") 00..3300 (a) Sphericvaelx pm = o4d7e5l kfomr As−21 e (") 00..88 ev(ebn)t 9 (1992.41) A3
s c
al ur
m the centr 0000....43430505 2010 Oct 26 e central so 00..66 event 10 (1997.95) A2
e fro (t1exp=9.65 yr) m th 00..44
ed distanc 00..4455 (22t020e1x1p=121 N0O.o7cv2t 21yr08) stance fro 00..22 event 11 (2003.49) A1
Project 00..5500 (t3exp=11.63 yr) Di 00..00 event 12 (2009.03)
−−110000 00 110000 220000 22000099 22001100 22001111 22001122 22001133
Doppler velocity (km s−1) Year
Fig.4.— (a) Observed position of A2 as a function of Doppler velocity for the indicated epochs. The dashed lines show the projected
positionexpectedforasphericalshellexpandingat475kms−1duringt1exp,t2exp,andt3exp. Thereisacleardeparturefromthespherical
symmetry for red-shifted velocities > +100 kms−1. (b) The dashed lines indicate the expected distance from the central source, as a
function of time, of a spherical shell formed during periastron passage, and moving at 475 kms−1. The word ’event’ corresponds to the
periastronpassage at the epoch indicated bythe number inparenthesis. From this figure, A1was formedin2009.03, A2 in2003.49, and
A3in1997.95. In2010.8,thereisnodataforA3becausethisarccanonlybedetectedinthe[NiII]transition,andthistransitionwasnot
includedintheobservations ofthatyear.
during the corresponding age of A2 at eachepoch of ob- this is naturally explained by the primary wind being
servations (t1 , t2 , and t3 in Figure 4(a)). compressed – a factor of 30 – by the secondary wind
exp exp exp
At line of sight velocities near zero, a spherical model during periastronpassage. This is alsosupportedby the
reproduces the observations very well, with standard fact that the average age difference between the arcs is
deviations < 0.015(cid:5)(cid:5). However, it fails to match the approximately 5.6 yr, nicely matching the binary sys-
observed position of A2 at red-shifted velocities > tem’s period.
+100kms−1;A2movesslowerthanpredicted. Thefinal Theprimarymass-lossraterequiredtoproduceashell
standard deviation for the spherical model is 0.05(cid:5)(cid:5), in with the same mass andgeometry as A2, overa time in-
the range from −40 to +220 kms−1. tervalequaltothebinary’speriod,is9.3×10−4M(cid:2)yr−1,
Figure 4(b) shows the comparison between the ob- in agreement with the results from radiative transfer
served and the expected motion of shells formed during spectral modeling (Hillier et al. 2001, 2006; Groh et al.
various periastron passages, and moving at 475 kms−1. 2012). We note, however, that our result must not be
Despite showing an offset between the expected and ob- takenasastrictlowerorupperlimit,sincetheactualge-
servedpositionatagiventime, theirslope arethe same, ometry and thickness of A2 may be different from what
which gives support to our derived space velocity and we assumed.
theassumptionthattheshellsareexpandingatthesame Simulations using 3D smoothed particle hydrodynam-
rate. ics (3D SPH) reveal compressed regions of the primary
wind formed during the periastron passages and main-
4. DISCUSSIONANDCONCLUSIONS tained for multiple cycles by the hot, low density, resid-
ual secondary wind (Figure 5). These simulations pre-
Weusedthecriticaldensityforeachforbiddenline,cal-
culated using the chianti atomic database (Dere et al. dict that these compressed regions accelerate from the
1997; Landi et al. 2013), to infer the typical local den- primary wind terminal velocity (v∞ = 420 kms−1;
sities of the arcs. The outermost arc, A3, seen in the Groh et al.2012)to about460−470kms−1 (T.Madura
[Ni II] transitions but weakly in the [Fe II], must have a etal. 2013,inprep.),whichisconsistentwithourderived
densityintherange0.8−4×106cm−3. Thestrong[FeII] expansion velocity of 475 kms−1 for the shells.
emission from A1 and A2, indicates that their densities The departure from spherical symmetry for high red-
must be >4×106 cm−3. shifted Doppler velocities suggeststhat the shells do not
AssumingthatA2ishalfasphericalshell(cf. Figure5) expand with the same velocity in all directions. This
with averageradius of 0.4(cid:5)(cid:5) (projected distance from the is explained by the fact that 3D SPH simulations show
central source, at zero Doppler velocity, in 2010) and that material located in the orbital plane moves faster
thickness < 0.05(cid:5)(cid:5) (since we cannot spatially resolve the than material located off the plane. Thus, the observed
arcs), and using the critical density calculated before, Doppler velocity is a combination of velocities from dif-
we estimate a total mass of 2.6 × 10−3 M(cid:2), for solar ferent emitting regions along a specific viewing angle,
which,inturn,dependsupontheorientationoftheshells
metallicity.
onthe sky, ultimately defined by the orbitalorientation.
Models for the primary star predict that its wind den-
sity should drop below 106 cm−3 for projected distances During periastronpassages,SPH3D simulations show
greater than 0.2(cid:5)(cid:5) (Hillier et al. 2001; Groh et al. 2012). that the secondary star creates a cavity deep within the
Since we detected arcs exceeding 106 cm−3 out to 0.6(cid:5)(cid:5), primary wind, pushing a shell outwards. Since the low
The compressed wind of η Car A 5
Fig. 5.—3DSPH simulationsshowingtheformationandgeometry ofhighdensityregionsduetothecompressionoftheprimarywind
bythesecondarywind,intwodifferentregimesfortheprimarymass-lossrate: 8.5×10−4 M(cid:4)yr−1 (left)and2.5×10−4 M(cid:4)yr−1 (right).
Upper panels: slices in the orbital plane (xy). Lower panels: slices perpendicular to the orbital plane (yz). In the high mass-loss rate
regime,thereisagapinthepolarregionthatisnotobservedinthelowermass-lossrateregime. Thelowerandupperlimitsofthedensity
scalecorrespondtothecriticaldensityvaluesfor[NiII]and[FeII],respectively.
density secondarywind is moving at ≈3000kms−1, the Assuming fixed parameters for the secondary wind, a
interactionwiththehighdensity,relativelyslowprimary large primary mass loss rate decreases the bow shock
windresultsintheformationofahighdensity,thinshell opening angle θ∞, so that the compressedprimary wind
acceleratedtovelocitiesslightlyhigherthantheterminal material is trapped near the orbital plane (Figure5(c));
velocityoftheprimarywind. Hence,thegeometryofthe a low mass-loss rate increases θ∞, and the compressed
shell is strongly affected by the physical parameters of primary wind can extend well off the orbital plane (Fig-
the binary system, and the projected appearance on the ure5(d)). This means that the geometry of the circum-
sky is defined by the inclination of the orbital plane. stellararcsisanexcellentdiagnosticoftheprimarymass-
Thelatitudinalextentofthe shellsiscontrolledbythe loss rate: a lower mass loss rates produces a nearly con-
asymptoticopeningangle,θ∞,ofthewind-windcollision tinuous arc seen in projection around the star, while a
region, given by (Usov 1992) higher mass loss rate produces a shorter, broken arc.
(cid:2) (cid:3) The existence of a discontinuity in the arcs that we
θ∞ = 23π 1− η24/5 η1/3, (4) oPb.Ase.r≈ve1,3m5◦o,reyiperldosnoaulnocweedrilnimAi1t taostahleacmkaosfs-elomsisssriaotneaotf
theprimarystar. SPHsimulationsin3Dshowthatsuch
where η = M˙Bv∞B/M˙Av∞A is the ratio of the sec- broken rings are only apparent if the primary mass-loss
ondary to the primary wind momentum, M˙ and v∞ are rateisabove2.5×10−4M(cid:2)yr−1,whichsuggeststhat,at
least for the last event, the primary mass-loss rate must
themass-lossrateandterminalvelocityofthewind,and
have exceeded this value.
thesubscriptAandBrefertotheprimaryandsecondary
Ourresultsalsoconstrainthe orbitalorientation,with
star, respectively.
6 Teodoro et al.
periastron occuring between 200◦ < ω < 270◦, i.e., dur- 5.6 yr between them remains the same, validating their
ing periastron, the primary star is between the observer connection with the orbital period.
and the secondary star. The opposite configuration,
where the secondary, at periastron, is between the ob-
We are grateful to an anonymous referee for com-
server and the primary star, would produce blue-shifted
ments and suggestions that helped to improve the qual-
shells seen to the northwest of the central source, rather
ity and presentation of this work. M.T. is supported
than the red-shifted features to the southeast that we
byCNPq/MCT-Brazilthroughgrant201978/2012-1. T.
observe.
I. M. was supported by an appointment to the NASA
Finally, we note that our results assume a constant
PostdoctoralProgramattheGoddardSpaceFlightCen-
shell space velocity. However, as can be seen in Fig-
ure4(b), there is anoffset of0.07(cid:5)(cid:5) betweenthe expected ter, administered by Oak Ridge Associated Universities
through a contract with NASA. We thank Don Lindler
and the observed distance from the central source as a
forhisconsistenthelpwiththe idlproceduresnecessary
functionoftime. Weknow,fromthecomparisonbetween
thederivedspacevelocity(475kms−1)andtheterminal for production of the data cubes. M.T. would like to
thank David Fanning for making his idl Coyote library
velocity of the primary wind (420 kms−1), that this off-
publicly available. CHIANTI is a collaborative project
setisduetotheaccelerationthattheseshellsexperienced
involving George Mason University, the University of
at some point during the early stagesof their formation.
Michigan(USA) andthe UniversityofCambridge(UK).
Thus, the ages derived in this letter are overestimated,
This research has made extensive use of NASA’s Astro-
i.e the shells areyounger. Nevertheless,the difference of
physicsDataSystemandidl Astronomy User’s Library.
Facilities: HST (STIS).
REFERENCES
Blais,F.,&Rioux,M.1986,SignalProcessing,11,145 Hillier,D.J.,Gull,T.,Nielsen,K.,etal.2006, Astrophysical
Bostroem,K.A.,&Proffitt,C.2011,STISDataHandbook, HST Journal,642,1098
DataHandbooks,-1 Krist,J.E.,Hook,R.N.,&Stoehr,F.2011,inSPIEOptical
Damineli,A.,Hillier,D.J.,Corcoran,M.F.,etal.2008, Engineering+Applications,ed.M.A.Kahan(SPIE),
AstronomicalJournal,384,1649 81270J–81270J–16
Davidson,K.,&Humphreys,R.M.1997,AnnualReviewof Landi,E.,Young,P.R.,Dere,K.P.,DelZanna,G.,&Mason,
AstronandAstrophys,35,1 H.E.2013, AstrophysicalJournal,763,86
Dere,K.P.,Landi,E.,Mason,H.E.,MonsignoriFossi,B.C.,& Madura,T.I.,Gull,T.R.,Owocki,S.P.,etal.2012,
Young,P.R.1997,AstronomyandAstrophysicsSupplement AstronomicalJournal,420,2064
Series,125,149 Mehner,A.,Davidson,K.,Ferland,G.J.,&Humphreys,R.M.
Fisher,R.B.,&Naidu,D.K.1996,inlink.springer.com(Berlin, 2010,AstronomicalJournal,710,729
Heidelberg:SpringerBerlinHeidelberg),385–404 Pittard,J.M.,&Corcoran,M.F.2002,Astronomyand
Groh,J.H.,Hillier,D.J.,Madura,T.I.,&Weigelt, G.2012, Astrophysics,383,636
AstronomicalJournal,423,1623 Smith,N.2006,AstrophysicalJournal,644,1151
Gull,T.R.,Madura,T.I.,Groh,J.H.,&Corcoran,M.F.2011, Teodoro,M.,Damineli,A.,Sharp,R.G.,Groh,J.H.,&Barbosa,
Publications oftheAstron.Soc.ofAustralia,743,L3 C.L.2008, AstronomicalJournal,387,564
Gull,T.R.,Nielsen,K.E.,Corcoran,M.F.,etal.2009, Monthly Usov,V.V.1992,AstronomicalJournal,389,635
Notices oftheRAS,396,1308 Verner,E.,Bruhweiler,F.,&Gull,T.2005,Astrophysical
Hillier,D.J.,Davidson,K.,Ishibashi,K.,&Gull,T.2001, Journal,624,973
AstrophysicalJournal,553,837 Zethson,T.,Johansson,S.,Hartman,H.,&Gull,T.R.2012,
AstronomyandAstrophysics,540,A133