Unveiling Vela - Time Variability of Interstellar Lines in the Direction of the Vela Supernova Remnant II. Na D and Ca II

0.22 MB·

by  N. Kameswara Rao

Mon.Not.R.Astron.Soc.000,1–??(2014) Printed17January2017 (MNLATEXstylefilev2.2)
Unveiling Vela - Time Variability of Interstellar Lines in
the Direction of the Vela Supernova Remnant II. Na D
⋆
and Ca II
7
1
0
2 N. Kameswara Rao1,2, David L. Lambert2, Arumalla B. S. Reddy2, Ranjan Gupta3,

n
a S. Muneer1 & Harinder P. Singh4
J
†,
6
1Indian Institute of Astrophysics, Bangalore 560034, India
1
2The W.J. McDonald Observatory and Department of Astronomy, The University of Texas, Austin, TX 78712-1083, USA

3Inter UniversityCentre for Astronomy and Astrophysics, IUCAA Pune, 411007 India
]
R 4Department of Physics & Astrophysics, Universityof Delhi, Delhi 110007 India
S
.
h
p 17January2017
-
o
r
t ABSTRACT
s
a In a surveyconducted between2011-12ofinterstellarNai D line profilesin the direc-
[ tion of the Vela supernova remnant, a few lines of sight showed dramatic changes in

low velocity absorption components with respect to profiles from 1993-1994reported
1
by Cha & Sembach. Three stars - HD 63578, HD 68217 and HD 76161 showed large
v
decrease in strength over the 1993-2012 interval. HD 68217 and HD 76161 are as-
9
4 sociated with the Vela SNR whereas HD 63578 is associated with γ2 Velorum wind
3 bubble.Here,wepresenthighspectralresolutionobservationsofCaiiKlinesobtained
4 with the Southern African Large Telescope (SALT) towards these three stars along
0 with simultaneous observations of Nai D lines. These new spectra confirm that the
. Na D interstellar absorption weakened drastically between 1993-1994 and 2011-2012
1
0 but show for the first time that the Ca II K line is unchanged between 1993-1994
7 and 2015. This remarkable contrast between the behaviour of Na D and Ca II K line
1 absorption lines is a puzzle concerning gas presumably affected by the outflow from
: the SNR and the wind from γ2 Velorum.
v
i Key words: Star: individual: ISM: variable ISM lines: Supernova Remnants :other
X
r
a
1 INTRODUCTION profiles of Caii K and Nai D lines superimposed on stellar
spectra.
Vela supernova remnant (SNR) is the closest relic of of a
stellar explosion to earth, located at a distance of 287±19 Recently, we completed a study of high resolution Nai
pc, as infered from the VLBI parallax of its pulsar (Dod- DlineprofilestowardssixtyfourOBstarsinthedirectionof
son et al 2003) and represents a supernova explosion that Vela SNR , mainly carried out during 2011-2012, and com-
occurred 11000 years ago (Reichley, Downs & Morris1970). paredthemwithCaiiKandNaiDlineprofilesofthesame
Observationalstudiesofinterstellarlinespresentinthespec- stars earlier obtained by Cha & Sembach (2000) in the pe-
traofstarsinthedirectionofvelaSNRprovideinformation riod1993-1996 at comparablespectral resolution (Raoet al
about the interaction of the remnant with local interstellar 2015-hereafterPaperI).ComparisonoftheprofilesofNaiD
medium(ISM).Suchstudiespredominantlyconcentratedon fromthesetwoepochsrevealedmajordecreasesinlowveloc-
ityabsorptioncomponentstowardsafewstars.WhenPaper
I was written, we did not have Caii K profiles for compar-
ison with K line observations of Cha & Sembach (2000) to
⋆ Based on observations obtained with Southern African Large constrain the cause of this Na D variability. Subsequently,
we could obtain with the high-resolution spectrograph at
Telescope(SALT)andTheVainuBappuTelescope(VBT).
† E-mail: [email protected] (NKR); [email protected] Southern African Large Telescope (SALT) Caii K profiles
(DLL);[email protected] (SM); [email protected] (RG); for three of the stars exhibiting dramatic weakening of Na
[email protected](HPS) Dabsorption.Inthepresentpaper,wereportontheCaiiK
2 N. K. Rao et al.
andNaiDlinestowardsthreestarsinwhichDlineabsorp- calibration and corrections for telluric line blending. We
tion at low velocities weakened greatly between 1993-1994 adopted the same local standard of rest (LSR) of Cha &
and 2011-2012. The surprising result is that in sharp con- Sembach(2000)forconvertingheliocentricvelocitiestoLSR
trasttotheNaDlinestheCaiiKlinein2015hasthesame velocities. Decomposition of various components from the
profile in all three cases as it did in 1993-1994. observed line profiles are accomplished by fitting Gaussian
Earlier studies of interstellar absorption lines towards profiles.CentralradialvelocityVLSR,fullwidthathalfmax-
various stars in and around the Vela SNR indicated that imum and equivalent width of the components have thus
high velocity (i.e > 100 km s−1) components, most com- been obtained from these fits. The values of these parame-
monlyseeninCaiiprofiles,manifestshockedgasassociated ters listed in Cha & Sembach (2000) were used as starting
with the SNR (Wallerstein, Silk & Jenkins 1980; Jenkins, values for the gaussian fits and further changes were made
Wallerstein&Silk1984).Sushch,Hnatyk&Neronov(2011) soastomakebothD2 andD1 profilesyieldthesamenum-
pointed out that only stars with distances greater than 500 ber of components, same VLSR and also similar full width
pc show these high-velocity components in their spectrum at half maximum. The combined gaussian fits are made to
(see Paper I figure 18, Cha, Sembach & Danks 1999). For match the observed profiles such that no residuals are left
reference, the pulsar is at about 290 pc and the three stars overthenoise in the surroundingcontinuum.These fits are
of interest here are at 300-500 pc and do not show high- of doubtful validity when the lines are saturated, as is the
velocity interstellar components. The diameter of Vela su- caseforsomeNaDprofiles.Whenmultipleobservationsare
pernova remnant is estimated to be about 8.3 degrees (As- available an average equivalent width of the components is
chenbach 1993; Aschenbach, Egger & Trumper 1995). Two presented in tables. The SALT spectra in the D line region
of the three stars discussed here, HD 68217 and HD 76161, are not corrected for telluric line contamination, since si-
arelocated in theouteredgeof theROSATX-rayimage of multaneoustelluriclinestandardstarobservationswerenot
theSNR.Thethirdone,HD63758, islocated in theγ2 Vel obtained.Asaresultradialvelocityandequivalentwidthof
wind bubbleregion. few D2 components are affected by telluric lines. However,
D1 profilesare fine.TheLSR radial velocities of theNai D
line components obtained from SALTprofiles agree with in
±2km s−1 with theones measured from VBT profiles.
2 OBSERVATIONS
Ourhigh-resolutionspectraofthethreestarsdiscussedhere
wereobtainedattwoobservatories.TheresolvingpowerR= 3 SIGHT LINES TOWARDS WHICH LARGE
λ/dλ of the spectra is comparable to that of the baseline VARIATIONS OF NAI COLUMN DENSITY
spectra by Cha & Sembach (2000) who reported a value of HAVE BEEN OBSERVED
R=λ/dλ≃75000.
Observations of theNaD lines in Paper I LowVLSR velocitycloudsthatarerecognised throughNaD
have been obtained with a 45 meter fiber-fed cross- and Caii absorption components are usually considered to
dispersedechellespectrometeratthe2.3meterVainuBappu be associated with diffuse interstellar medium in the spiral
Telescope (VBT) at the Vainu Bappu Observatory (Rao et arms. However interstellar clouds in the proximity of the
al.2005 ). R = λ/dλ, the spectral resolving power with a Velasupernovaremnantmighthaveexperiencedinteractions
60 µm m slit, was 72000. The spectrum covers the wave- with the remnant through radiation or collisions and thus,
length range of 4000 to 10000˚A with gaps. Beyond about theirphysical conditions and kinematics altered.
5600˚A the echelle orders were incompletely recorded on a Our survey of interstellar Na D lines (Paper I) showed
E2V 2048×4096 CCD chip. Although Na D lines are cap- thatthemostofthestarsobservedwithVBTduring2011-12
tured,theCaiiH & K lines occur in theinsensitive region. had their Na D profiles unchanged compared to 1993-1996
The wavelength calibration was done using Th-Ar hollow observationsofCha&Sembach(2000). Most of thelowve-
cathode lamp exposures that were obtained soon after the locity components, generally in the VLSR range of ±25 to 0
exposures on thestar. km s−1 (Cha, Sembach & Danks 1999) of Na D absorption
New high resolution spectra of Caii H & K lines along mighthavetheiroriginsincloudsthatoccurintheinterven-
withNaiDlineshavebeenacquiredwiththehighresolution ing spiral arms. A remarkable result of Paper I was a great
spectrograph (HRS) on the Southern African Large Tele- decrease in strength (almost disappearance) of low-velocity
scope (SALT) on 2015 December 15 (Bramall et al. 2010). absorption components in Na D lines in just three stars –
The HRS blue spectrum with a 1.6 arc sec fiber provides a HD63578, HD68217 andHD76161, in theperiod between
resolvingpowerof66700.SpectrawithBothbluearm(3674- 1993−1996 and 2011−2012. This trio is discussed here in
5490˚A)andredarm(5490-8810˚A)havebeenobtained.The lightoftheSALTspectraofbothNaDandtheCaKlines.
S/NratiooftheredspectrumintheregionoftheNaDlines Twosimplequestionsmaybeaddressedwiththeacquisition
is comparable to that of the VBT spectra ∼80−100) and oftheSALTspectra:(i)Dothe2015SALTCaiiKlinepro-
the spectra obtained by Cha & Sembach (2000). However, filesdifferfromthosethatwereobtainedbyCha&Sembach
the SALT spectra around the Caii K line are of a lower during1993-1994,inasimilarwaytothelargechangesseen
S/N ratio (∼35−50) than the spectra provided by Cha & intheNaDprofilesbetween1993-1994 and2011-2012? and
Sembach (2000). The resolving power of the SALT spectra (ii) Have the Na D profiles evolved further in the interval
is63000 at theDlines asmeasured from thewidthof weak three-four year between the VBT and SALT observations?
telluric lines and comparable at K line. (Cha & Sembach noticed changes in strength and velocity
IRAF routines have been used for spectral reductions in some high velocity absorption components in their ob-
namely flat field corrections, bias subtractions, wavelength serving period of about three years, but our VBT spectra
Unveiling Vela-SALT 3
Absorption Lines of Caii K & Nai towards HD 63578
Table 1.
CaiiK NaiC&S a Nai(VBT)b Nai(SALT)c
C&S SALT D2 D1 D2 D1 D2 D1
VLSR Eq.w Eq.w VLSR Eq.w Eq.w VLSR Eq.w Eq.w Eq.w Eq.w
kms−1 (mA) (mA) kms−1 (mA) (mA) kms−1 (mA) (mA) (mA) (mA)
−14 24 28 −13 31 15 (27) 14
−4 223 178 −6 39 24 ∗ 21
6 34 43 6 232 191 230 191
11 325 270
22 5
a C&S:Cha&Sembach(2000) observationsfrom1993
b VBT:Averageoffournightsfrom2011-2012(seePaperI)
c SALT:Observationfrom2015December15
∗ :Affectedbytelluricline
Absorption Lines of Caii K & Nai towards HD 68217.
Table 2.
CaiiK Nai(C&S)a Nai(VBT)b Nai(SALT)c
C&S SALT D2 D1 D2 D1 D2 D1
VLSR Eq.w Eq.w VLSR Eq.w Eq.w VLSR Eq.w Eq.w Eq.w Eq.w
kms−1 (mA) (mA) kms−1 (mA) (mA) kms−1 (mA) (mA) (mA) (mA)
−18 17 66 −19 9 5 ∗ 3
−8.0 14 15 −8 30 20 −8 10 5 ∗ 3
2 43 12 2 41 22 41 17
9.0 20 25 9 20 16 9 21 10 21 9
a C&S:Cha&Sembach(2000) observationsfrom1994
b VBT:Averageoffournightsfrom2011-2012(seePaperI)
c SALT:Observationfrom2015December15
∗ Affectedbytelluricline
Absorption Lines of Caii K and Nai towards HD 76161
Table 3.
CaiiK Nai(C&S)a Nai(VBT)b Nai(SALT)c
C&S SALT D2 D1 D2 D1 D2 D1
VLSR Eq.w Eq.w VLSR Eq.w Eq.w VLSR Eq.w Eq.w Eq.w Eq.w
kms−1 (mA) (mA) kms−1 (mA) (mA) kms−1 (mA) (mA) (mA) (mA)
−8 16 6 −6 94 32 −9 6 3 ∗ 4
2 38 43 2 60 32 63 28
9 395 365
13 8 4 10 6
27 61 32
a C&S:Cha&Sembach(2000) observationsfrom1993
b VBT:Averageoffournightsfrom2011-2012(seePaperI)
c SALT:Observationfrom2015December15
∗ Affectedbytelluricline
4 N. K. Rao et al.
11..11
11..00
00..99
y
t
si 00..88
n
e
t 00..77
n
I
d 00..66
e
z 00..55
ali
m 00..44
r
o
N 00..33
00..22
00..11
00..00
−1
V (km s )
LSR −200 −−110000 00 110000 200
Figure 1. (Left): Profile of Nai D2 in HD 63578 obtained in 2011 (red line) shown superposed on the profile of Nai D2 observed in
1993 byCha& Sembach (2000)(black line). Theblue-shiftedabsorptioncomponents inthe1993 spectrum (blackline)is absent inthe
2011 spectrum. (Right): Caii K profiles of HD 63578 obtained withSALT in2015 (red line) overplotted on profileobtained by Cha&
Sembach(2000) in1993.
1
0.5
0
-50 0 50 100
Figure 2. The profiles of Nai D in HD 63578 obtained during 2012 (blue line) are superposed with profiles obtained in 2015 with
the SALT (red line). The SALT spectrum was not corrected for (weak) telluric lines. The dotted lines show the velocities of the four
Gaussiancomponents fittedtotheprofiles–seeTable1.
havea limited information on such absorption components. that in 1993. The figure also shows that the D2 profile did
In particular, the three stars observed at SALT have not not change further between 2011 and 2015 (Figure 2) . In
shown high velocity components in any available spectra. striking contrast, the Ca K profiles are unchanged between
High-velocity components appeared more frequently in the 1993 and 2015 (Figure 1-right).
CaiiK line than theNai lines.)
Gaussian components extracted from the spectra are
HD63578:Forthisstarlocatedatadistanceof150pc summarized in Table 1. These show that the Ca K line’s
behind γ2 Velorum and seemingly in the bubble generated two components are unchanged in velocity and equivalent
by γ2 Velorum’s wind ( se Paper I) , Figure 1(left) shows width between 1993 and 2015 except perhapsfor a possible
themuchweakerlow velocityNaD2 linein2011relativeto increase in the redder component in the SALT spectrum.
Unveiling Vela-SALT 5
11..11
11..00
y
t
si
n 00..99
e
t
n
I
00..88
d
e
z
ali 00..77
m
r
o
N 00..66
00..55
00..44
−1
V (km s )
LSR −−220000 −−110000 00 110000 220000
Figure 3.(Leftpanel):ProfileofNaiD2 inthesightlinetoHD68217asobtained in1994byCha&Sembach(blackline)superposed
on the D2 profile obtained on 2012 January 16 with the VBT (red line). The strong absorption component present at VLSR = −9 km
s−1 in 1994 is conspicuously weakened by 2012. (Right panel): The Caii K profile obtained with SALT in 2015 of HD 68217(red line)
is overplotted on the profile obtained in 1994 by Cha & Sembach (2000) (black line). There may be a slight weakening of the sharp
absorptioncomponent at−8kms−1 bythetimetheSALTprofilewasobtained.
1.4
1.2
1
0.8
0.6
0.4
0.2
-50 0 50 100
Figure4. ProfilesofNaiD2andD1obtainedwithVBTin2011(bluelines)arecomparedwithprofilesobtainedon2015December25
withSALT(redline).Theprofilesareverysimilarindicatingnosignificantchangeoccurredbetween2011and2015.TheSALTspectrum
wasnotcorrectedfor(weak)telluriclines.ThedottedlinesshowthevelocitiesoftheGaussiancomponentsfittedtotheNaDprofiles–
seeTable2.
Cha & Sembach note that the two velocity components are at−13and+6kms−1 apparentlycoincidentwiththeCaK
traceable back to 1971 - 1977. components plus another weak Na D component at −6 km
s−1.
At Na D, Cha & Sembach fitted their 1993 spectrum
with strong components (Table 1) at −4 and +11 km s−1 Jenkins, Wallerstein & Silk (1984) obtained an ultra-
which differ by several km s−1 from their velocities of −14 violet spectrum of the star with the IUE on 1979 June 15.
and +6 km s−1 obtained from the much weaker Ca K line. They measured for both Ci, Sii, and Siii–Feii lines LSR
For the VBT and SALT Na D lines, there are components radialvelocitiesof+9,+7and+2kms−1,respectively.The
6 N. K. Rao et al.
11..11
11..00
00..99
y
t
si 00..88
n
e
00..77
t
n
I
00..66
d
e
z 00..55
i
l
a
m 00..44
r
o 00..33
N
00..22
00..11
00..00
−1
V (km s )
LSR −−220000 −−110000 00 110000 220000
Figure 5. (Left panel): Profiles of Nai D2 in the sight line towards HD 76161 obtained with VBT on 2011 December 25 (red line) is
superposedonthe1993profileobtainedbyCha&Sembach(2000)(blackline).Strongabsorptioncomponents atVLSR =9and27km
s−1 havealmostdisappearedbythetimeofVBTobservationsin2011.(Rightpanel):CaiiKprofilesofHD76161obtainedwithSALT
in2015(redline)overplottedon1993profileobtainedearlierbyCha&Sembach(2000).
1.4
1.2
1
0.8
0.6
0.4
0.2
-50 0 50 100
Figure 6. Nai D2 and D1 profiles obtained in 2011 with (blue lines)VBT are compared with profiles obtained on 2015 Dec 15 with
SALT(redline).Theprofilesareverysimilarindicatingnosignificantchangeoccurredbetween2012and2015.TheSALTspectrumwas
notcorrectedfortelluriclines.Thedotted linesshowthevelocitiesoftheGaussiancomponents fittedtotheNaDlines–seeTable3.
+9 km s−1 velocity is coincident with the +9 km s−1 com- butremainedwithoutadditionalchangein2015(Figure4).
ponent of Ca K. The −4 km s−1 component that has been Table 2 shows that the Gaussian components for the Ca K
present in 1993 spectrum in Na D lines was not seen in ei- line have close counterparts in the Na D profiles and, in
therCiorinsinglyionized linesofCa, S,SiandFe.Thus, particular, thesharp feature at −8km s−1 which weakened
thiscomponentprobablyreferstoacoldgascloudseenonly greatly at Na D may also have weakened at Ca K (Figure
in NaD. 3-right). The four Na D components identified by Cha &
Sembach are present in the VBT and SALT spectra with
HD 68217:ThisstarliesneartheedgeoftheROSAT the −8 km s−1 feature weaker in the latter spectra. A Na
imageoftheVelaSNR.Figure3(left)showsthatasharpNa
D component at −8 km s−1 almost vanished by 2011-2012 D spectrum obtained in 1989 by Franco (2012) with the
Unveiling Vela-SALT 7
sameinstrumentusedbyCha&Sembachshowsthe−8km the Vela nebula. A few earlier reports for stars behind the
s−1 featureweakerthanitwas in1994 butstronger thanin Vela SNR have noted modest variations in the Na D lines
VBT and SALT spectra which suggests that the feature’s atlow velocity-see,for example,observations ofthevisual
evolution is not a simple one. binary HD 72127 (Hobbs, Wallerstein & Hu 1982; Hobbs
HD 76161: This star which like HD 68217 lies near et al. 1991; Welty. Simon & Hobbs 2008). The magnitude
the edge of the ROSAT image of the Vela SNR (paper I) of the weakening reported by us earlier is unprecedentedin
provides the third and the extreme example in which low published studies of interstellar Na D lines (see, for exam-
velocitystrongNaDabsorptionspresentin1993hadalmost ple,McEvoyetal.2015) andisplausiblyattributabletoin-
disappeared by 2011 - see Figure 5(left) : the saturated ab- teractions between ambient diffuse interstellar gas and the
sorptioncomponentatVLSR +9kms−1 hasvanishedalmost high-velocity gas in and around the Vela nebula from the
completely by 2011. A redward weaker absorption compo- Vela supernova and the winds from massive stars. It seems
nent at +27 km s−1 present in 1993 also weakened dras- highlypertinenttonotethat outof forty onestarswith Na
tically by 2012. However a weak absorption component at D lines observed previously by Cha & Sembach and again
about −9 km s−1 seems unchanged over the period of 1993 attheVBTnoneshowedacomparablestrengtheningofthe
to 2012 baseline. The Na D profile remained unchanged in NaDlinesbetween1993–1994and2011–2012.(Asnotedfor
theperiod 2011 to2015 (Figure 6). HD 68217, A low-velocity component Na D did strengthen
The 1993 Caii K profiles of 1993 and 2015 (Figure 5- between 1998 and 1994. Also, HD 73882, a star behind the
right)appearwithcomponentsat−8and+2kms−1 (Table Vela SNR showed a strengthening of the Cai 4226 ˚A line
3). The Caii K profile obtained between 1971-77 period by between 2006 and 2012 (Galazutdinov et al. 2013) and pre-
Wallerstein, Silk & Jenkins (1980) appears to be similar to sumablytheNaDlinesalsostrengthenedoverthisinterval.)
theprofile observed by Cha & Sembach (2000) in 1993. The better established signature of these interactions
Jenkins, Wallerstein & Silk (1984) measured LSR ra- between high-velocity gas and the ambient interstellar
dial velocity of Sii and Siii+ Feii as −3 km s−1 and −7 medium is the presence of high-velocity (|VLSR ∼ 100| km
km s−1 respectively from their IUE spectrum obtained in s−1) absorption components seen in stars with distances
1979 September 26, very similar values to that of Caii K greater than about 500 pc. High-velocity gas was first re-
components measured bu Cha & Sembach (2000) in 1993. portedbyWallerstein &Silk(1971) andThackeray &War-
Later, Nichols& Slavin(2004) detected Cilines at the ren(1972) from observationsoftheCaiiKlinewith exten-
LSR velocity of 8 km s−1 in the same IUE spectrum ob- sive follow-up at the Caii K line and Na D lines by Cha &
tainedin1979.ThisvelocitycorrespondstothestrongNai Sembach (2000). (High-velocity Na D gas associated with
component that disappeared after 1993. It appears that a the SNR Monoceros Loop has been shown to be variable
strong neutral cloud was present at +9km s−1 for at least (Dirks & Meyer 2016).) High-velocity gas associated with
a period of 14 years prior to 1993 and then vanished some Vela has been studied in the ultraviolet (Jenkins, Waller-
time in the next 18 years. stein&Silk1976;Wallerstein, Silk&Jenkins1980; Jenkins
Nichols&Slavin(2004)investigatedCilinesin54stars et al.1981; Jenkins,Wallerstein & Silk 1984). With respect
inVelaSNRregionincludingHD76161, andnotedthered- to the Na D and Ca K line, a characteristic of the high-
shift of Ci lines with respect to ionized lines and suggested velocity gas is the several 100-fold increase with respect to
thatCilinesfromtheground-stateoftheatomareacombi- thevalueforlow velocity gas ofthecolumn densityratioof
nationoftwoabsorptioncomponents:onenearzerovelocity Caii to Nai which is attributed to thedestruction of inter-
which might haveits origins in theclouds in thespiral arm stellargrains andrelease ofsubstantialamountsofCa (and
alongthesightlinestothestarandtheothercomponentata otherelements)butnotNawhichislittledepletedbygrains
highervelocityarisingfromcloudthatmighthaveinteracted (Spitzer 1978; Danks & Sembach 1995; Sembach & Danks
withtheVelaSNR.Theyalsosuggestthatredshiftobserved 1994).
inCiground-statelines might haveitsorigins inaHishell Surely, the outstanding characteristic of the low-
onthebacksideoftheremnant.Dubneretal.(1998)discov- velocity interstellar lines of the three sightlines where the
eredashellofHithrough21cmemissionaroundVelaSNR Na D lines are strikingly weakened between 1993-1994 and
that follows closely the ROSAT’s outer X-ray bright shell. 2012-2015 isthattheCaKlineshowsnodetectablechange
This 30 km s−1 expanding The Hi shell is thought to be a between 1993-1994 and 2015 in two cases (HD 63578 and
result of recombined postshocked gas behind the advancing HD 76161) and a possible mild weakening in the third case
shock front. There is the possibility that the cold neutral (HD68217) wherethecentraldepthofthecomponentcoin-
cloud that disappeared in the direction of HD 76161 might cidentinvelocitywiththevaryingNaDfeaturedropsfrom
havebeen a shocked cloud. 77%in1994to88%in2015,adroponlyslightlylargerthan
the the amplitude of the noise in the SALT spectrum. In
short, the Ca K line is effectively unaffected by the factors
responsible for the large Na D changes and, in particular,
4 CONCLUDING REMARKS
theCa K’svelocity componentcoincident with thevariable
Thefocusofthepaperhasbeenthreestarsin thedirection Na D component are unaffected. This contrast between Na
oftheVelaSNRinwhichasevereweakeningoflow-velocity DandCaKconstrainsthefactors drivingtheweakeningof
absorption components of interstellar Na D lines occurred theNa D line.
between 1993-1994 and 2011-2012 (Rao et al. 2016). Such Estimation of the Nai column density for two of the
low-velocity componentsareassumedtobelongtothelocal threeexamples is an uncertain procedure for the1993-1994
interstellar medium and not tobean immediate product of spectra because the dominant components are saturated.
the Vela SNR or other energetic sources in the vicinity of FortheVBT andSALTspectra,thevariable componentof
8 N. K. Rao et al.
the Na D lines is little saturated; the equivalent widths of TheprincipalnextchallengesuggestedbyourVBTand
theD2andD1linesapproachthe2:1weaklinelimit.ForHD SALT spectroscopy of Na D and Ca K lines is to catch
68217, the −8 km s−1 component appears to provide both and follow large scale changes occurring along a line of
the Na D and Ca K line and assuming that the lines are sight. Panoramic optical spectroscopy will be able to ob-
unsaturated,thecolumndensityratioN(Caii)/N(Nai)was serve the behaviour of many interstellar lines (see, for ex-
about 3 in 1994 and increased to 17 in 2015. For the other ample,Pakhomov,Chugai&Iyudin2012). Sincethereisas
two stars, theratio N(Caii)/N(Nai)for thehighly-variable yetnopredictorofwhich starisabout toundergoachange
component was between about two and four in 2015. The in its interstellar lines, detection of an onset will require
1993-1994ratioswereobviouslymuchsmallerbecauseofthe routine high-resolution monitoring of a set of stars in and
larger Nai column densities and theseratios appeartypical around the Vela SNR but this presents a severe scheduling
of diffuse clouds (Danks & Sembach 1995). The ratio of 17 challengeandinvolvesaconsiderableconsumptionofobserv-
for HD 68217 in 2017 is typical of values for high-velocity ingtime.Althoughobservation oftransient phenomenaisa
clouds in the Vela SNR (Sembach & Danks 1994) but the ‘hot’ contemporary topic, its remit does not yet to extend
change between 1993-1994 and 2015 seems unlikely to be to exploration of the questionsraised by ourpaper.
due to release of Ca from grains because the Ca K line’s
equivalentwidth isunchangedoverthisintervalandthein-
creaseintheratioarisesfromlossofneutralNapresumably
5 ACKNOWLEDGMENTS
to ionization. High-velocity clouds moving tangentially to
theline of sight may appear as low velocity cloudsand this We thank SALT astronomer, Brent Miszalski, for his con-
likelihoodisincreasedforstarssuchasourtriolocatednear siderable help in conducting the HRS observations for us.
theedge of theSNR. Wealso thankDr.EdJenkins,thereferee,forhisinsightful
Interactions between the SNR and an ambient diffuse commentsabouthowshocksmayaccountforourresultsand
diffusecloudmaybreakthecloudintosmallerparts.Pakho- othercorrections. Wealso would liketothankBabaVargh-
mov,Chugal&Iyudon(2012–seealsoKlein,McKee&Col- ese for providing the figures of Caii K superpositions. This
lela 1994) discuss how a fast-moving shock interacts with researchhasmadeuseoftheSIMBADdatabase,operatedat
a diffuse cloud. Three stages of interaction are envisaged. CDS,Strasbourg, France.Wealso would like tothank staff
First, the shock propagates through the cloud. In the sec- of theVBO at Kavalurfor their help with observations.
ondstage,thecloudisaccelerated.Finally,thecloudisfrag-
mented. It seems that the interaction can destroy a small
(diameter of a few au) cloud in a few years. Although de-
REFERENCES
tailed calculations remain to be done to show that almost
completedisappearanceofNaDlinesmaybeachievedwith Aschenbach, B.,1993,Adv.Space. Res.,14,45
almost no change of the Ca K line, it seems, as the referee Aschenbach, B.,Egger,R.,Trumper,J.,1995, Nature,373,587
haspointedout,unlikelythatlarge-scalefragmentation can Bramall,D. G., Sharples, R.,Tyas, L., et al. 2010, inSociety of
lead to a large Na D reduction without a change in Ca K Photo-OpticalInstrumentationEngineers(SPIE)Conference
and, moreover, changes in radial velocity are likely to ac- Series,Vol.7735,4R
Cha,A.N.,Sembach, K.R.,2000,ApJS,126,399
company column density changes.
Cha,A.N.,Sembach, K.R.,Danks,A.C.,1999,ApJ,515,L25
Nonetheless, the fact that these remarkable changes in
Danks,A.C.,Sembach, K.R.,1995,AJ,109,2627
Na D lines unaccompanied by Ca K line changes occur on
Dirks,C.,Meyer,D.M.,2016,ApJ,819,45
sightlinestoSNRorthrougharegioncrossedbyavigorous Dodson, R., Legge, D., Reynolds, J. E, McCulloch, P. M., 2003,
stellar wind and have not been seen sight lines through the ApJ,596,1137
ambientdiffusestellarmediumsuggeststhatshocksmaybe Franco,G.A.P.,2012,A&A,543,A39
the key to understanding our remarkable discovery. Shock Galazutdinov,G.,Kre lowski,J.,Beletsky,Y.,Valyavin,G.,2013,
fronts even at low velocities can emit Ly-alpha photons PASP,125,1329
(Shull&McKee1979). Lymanαat10.2 eVcanionizeneu- Hobbs,L.M.,Ferlet,R.,Welty,D.E.,Wallerstein,G.,1991,ApJ,
tralNawithitsionization potentialof5.14eVbutnotCa+ 378,586
Hobbs,L.M.,Wallerstein,G.,Hu,E.M.,1982,ApJ.,252,L17
with its ionization potential of 11.9 eV. As the front ap-
Jenkins, E.B., Silk, J., Leep, E.M., Wallerstein, G. , 1981, ApJ,
proaches the cloud, the increase in Lyman α flux could be
248,977
ratherabrupt;thezonecontainingLymanαphotonsahead
Jenkins,E.B.,Wallerstein,G.,Silk,J.1976,ApJS,32,681
ofthefrontcanbenarrowbecauseLymanαphotonsscatter
Jenkins,E.B.,Wallerstein,G.,Silk,J.,1984,278,649
manytimesinaneutralhydrogenmedium.(Thisappealing Klein,R.I.,McKee,C.F.,Colella,P.,1994,ApJ,420,213
argument we owe to thereferee.) McEvoy,C.M.,Smoker,J.V.,Dufton,P.L.,etal.2015,MNRAS,
A strictly geometrical explanation but perhaps over- 451,1396
looking the SNR and stellar wind association may be pos- Nichols,J.S.,Slavin,J.D.,2004,ApJ,610,285
sible. The column density ratio of Na D/Ca K for diffuse Pakhomov, Yu. V., Chugai, N.N., Iyudin, A.F., 2012, MNRAS,
424,314
clouds spans the range from 300 to 0.02 (Welty, Morton &
Rao, N.K.,Sriram,S., Jayakumar, K., Gabriel, F., 2005, JApA,
Hobbs1996). IfathincloudofhighNaD/CaKmovedout
26,331
of the line of sight to our three stars, the Na D line would
Rao,N.K.,Muneer,S.,Lambert,D.L.,Verghese,B.A.,2016,MN-
weakensharplybut,perhaps,theCaKlinewouldbelargely
RAS,455,2529(paperI)
unchanged. Observations of sight lines through the general Reichley,P.E.,Downs,G.S.,Morris,G.A.,1970,ApJL,159,L35
interstellar medium do not support this idea as a common Sembach,K.R.,Danks,A.C.,1994,A&A,289,539
an eventas we findtowards theVela SNR. Shull,J.M.,McKee,C.F.,1979,ApJ,227,131
Unveiling Vela-SALT 9
Spitzer, L., 1978 ‘Physical processes inthe interstellar medium’,
Wiley-Interscience
Sushch,I.,Hnatyk, B.,Neronov,A.,2011,A&A,525,A154
Thackeray, A.D.,Warren,P.R.,1972,MNRAS,160,23
Wallerstein,G.,Silk,J.,1971, ApJ,170,289
Wallerstein,G.,Silk,J.,Jenkins,E.B.,1980, ApJ,240,834
Welty,D.E.,Morton,D.C.,Hobbs,L.M.,1996,ApJS,106,537
Welty,D.E.,Simon,T.,Hobbs,L.M.,2008,MNRAS,388,323