Imaging and spectroscopy of the faint remnant G 114.3+0.3

6 Pages·0.39 MB·English

by  F. Mavromatakis

Astronomy & Astrophysics manuscript no.
(will be inserted by hand later)
Imaging and spectroscopy of the faint remnant G 114.3+0.3
1 1 2
F. Mavromatakis , P. Boumis , and E. V. Paleologou
2
0 1 University of Crete, Physics Department,P.O. Box 2208, 710 03 Heraklion, Crete, Greece
0 2 Foundation for Research and Technology-Hellas, P.O. Box 1527, 711 10 Heraklion, Crete, Greece
2
Accepted
n
a
J Abstract. WepresentthefirstcalibratedCCDimagesofthefaintsupernovaremnantG114.3+0.3intheemission
3 linesof[Oii],[Oiii],Hα+[Nii]and[Sii].ThedeeplowionizationCCDimagesrevealdiffuseemissioninthesouth
and central areas of the remnant. These are correlated with areas of intense radio emission, while estimates of
1 the[Sii]/Hαratiosuggest thatthedetectedemission originates fromshockheatedgas.Inthemediumionization
v image of [Oiii] we discovered a thin filament in the south matching very well the outer radio contours. This
7 filament is not continuous over its total extent but shows variations in the intensity, mainly in the south–west
2 suggesting inhomogeneous interstellar clouds. Deep long–slit spectra were also taken along the [Oiii] filament
0 clearlyidentifyingtheobservedemissionasemissionfromshockheatedgas.TheHαemissionisafewtimes10−17
1 erg s−1 cm−2 arcsec−2, while the variations seen in the [Oiii] flux suggest shock velocities into the interstellar
0 clouds around or below 100 km s−1. The sulfur line ratio approaches the low density limit implying electron
2
densities less than ∼500 cm−3.
0
/
h Key words.ISM: general – ISM: supernovaremnants – ISM: individualobjects: G 114.3+0.3
p
-
o
r 1. Introduction Fesen et al. (1997) reported the detection of faint fila-
t
s mentarystructuresinHαinthewest,south–westareasof
a
The vast majority of supernova remnants have been dis- the remnant. However, no images were shown due to the
:
v coveredby their non–thermalsynchrotronradioemission. faintnessofthefilaments.Eventhoughthespatialcorrela-
Xi Since the opticalwavelengths suffer significantly more at- tionofthe opticalemissionwith the radiocontoursfavors
tenuation than the radio, the detection of a supernova theirassociation,fluxcalibratedimagesinHαand[Sii]or
r
a remnantinthe opticalbandis adifficult task.The super- long–slitspectra areneeded to establishthe nature of the
novaremnantG114.3+0.3wasinitiallydetectedinthe21 optical flux. In an effort to deepen our knowledge of the
cm continuum survey of Kallas & Reich (1980) and sub- properties of the optically detected remnants, especially
sequently studied in more detail by Reich & Braunsfurth the faintest ones, we performed deep CCD imaging and
(1981). The remnant shows up in the radio as an ellip- spectral observations of G 114.3+0.3. Information about
soidalshelloccupyinganangularextentof∼60′×78′.The the observations and the data reduction is given in Sect.
radio emission is stronger in the south–east and south– 2. In Sect. 3 and 4 we present the results of our imag-
west compared to other areas of the remnant. The sur- ingobservationsandtheresultsfromthelong–slitspectra
face brightness is low, while at the same time the source taken at specific locations of interest. Finally, in Sect. 5
displays a high degree of polarization. More interest was we discuss the physical properties of the remnant.
raised for this system following the proposal of Fu¨rst et
al. (1993) and Kulkarni et al. (1993) that the pulsar PSR
2334+61 is associated with SNR 114.3+00.3.This pulsar 2. Observations
rotates at a period of 495 ms and its spindown rate sug-
gestsanageof∼41000yr.The distance estimatestothis 2.1. Optical images
pulsar–remnant association lie in the range of 2–3 kpc.
The observations presented in this paper were performed
ROSAT PSPC pointed observations showed that the pul-
with the 0.3 m Schmidt Cassegrain telescope at Skinakas
sarisaweakX-raysource,whilenoemissionwasdetected
Observatory,Crete, Greece. The field of G 114.3+0.3was
from the remnant itself (Becker et al. 1996).
observed in July 12 and 14, 1999. The observations were
performed with a 1024 × 1024 Site CCD which had a
Send offprint requests to: F. Mavromatakis,e-mail: pixel size of 25 µm resulting in a 89′ × 89′ field of view
′′
[email protected] and an image scale of 5 per pixel. The number of frames
2 F. Mavromatakis et al.: Optical observations of G 114.3+0.3
Table 1. Log of the exposure times
Hα+[Nii] [Sii] [Oiii] [Oii]
3600a(2)b 3600 (2) 6600(3) 1800(1)
a Total exposuretime in sec
b Numberof individualframes
Table 2. Spectral log
Slit centers No of spectra
α δ (Exp. times)
23h39m46s 61◦21′43′′ 2a (3600)b
23h39m17s 61◦19′06′′ 2a (3600)b
23h36m37s 61◦17′41′′ 2a (3600)b
a Numberof spectra obtained
b Exposure time of individual spectra in sec
taken is given in Table 1 along with the total exposure Fig.1. The field of G 114.3+0.3 in the Hα+[Nii] fil-
times in the specific filters used during the observations. ter. The image has been smoothed to suppress the resid-
ThefiltercharacteristicscanbefoundinMavromatakiset uals from the imperfect continuum subtraction. Shadings
al. (2000). The final images in each filter are the average run linearly from 0 to 20× 10−17 erg s−1 cm−2 arcsec−2.
of the individual frames where appropriate, while during The long rectangles show the positions observed through
theastrometryprocesstheHST Guidestarcataloguewas long–slit spectroscopy. The line segments seen near over-
used. All coordinates quoted in this work refer to epoch exposed stars in this figure and the next figures are due
2000. to the blooming effect.
We employedstandardIRAFandMIDASroutinesfor
thereductionofthedata.Individualframeswerebiassub-
tracted and flat-field corrected using well exposed twi-
light flat-fields. The spectrophotometric standard stars
HR5501,HR7596,HR7950andHR8634wereusedforab-
solute flux calibration.
2.2. Optical spectra
Long–slit spectra were obtained on July 16, 21, and
22, 2001 using the 1.3 m Ritchey–Cretien telescope at
Skinakas Observatory. The data were taken with a 1300
line mm−1 grating and a 800 × 2000 Site CCD covering
the rangeof4750˚A –6815˚A.The slit hada width of7′.′7
and,inallcases,wasorientedinthesouth-northdirection.
The coordinates of the slit centers, the number of avail-
able spectra from each location and the exposure time of
eachspectrumaregiveninTable2.Thespectrophotomet-
ric standardstarsHR5501,HR7596,HR9087,HR718and
HR7950wereobservedinorderto calibratethe spectraof
G 114.3+0.3.
Fig.2. The [Sii] image of the area around G 114.3+0.3.
3. Imaging of G 114.3+0.3
The image has been smoothed to suppress the residuals
3.1. The Hα+[Nii] and [Sii] line images from the imperfect continuum subtraction. Note the very
strong Hii regionLBN 565 in the central east area of the
The low surface brightness of G 114.3+0.3 in the radio, remnant. Shadings run linearly from 0 to 10× 10−17 erg
opticalandX–raybandsseemstobeoneofitsmajorchar- s−1 cm−2 arcsec−2.
acteristics. Especially, in the latter band the emission, if
any, is below the ROSAT detection limit. Our Hα+[Nii]
F. Mavromatakis et al.: Optical observations of G 114.3+0.3 3
Table 3. Typically measured fluxes emissionis strongest,i.e.in the south(see Fig.1 andFig.
4).
SEa SWa Centera LBN 565b
Hα+[Nii] 11 10 14 90 3.2. The [Oiii] and [Oii] images
[Sii] 2.4 3.5 2.6 9
[Oiii] 1.0 0.6 – – The morphology of the low ionization line of [Oii] is
[Oii] 2 1.6 – 5 similar to that seen in the Hα + [Nii] and [Sii] im-
ages and is not shown here. Typical fluxes measured in
Fluxesin unitsof 10−17 erg s−1 cm−2 arcsec−2 the [Oii] calibrated image are also given in Table 3.
aMedian valuesover a 112′′× 67′′ box
Contrary to the morphology seen in the low ionization
bMedian values overa 394′′× 394′′ box
images,the[Oiii]morphologyisstrikinglydifferent.Afil-
ament is detected in the south which originates at α ≃
23h40m42s, δ ≃ 61◦25′04′′ and ends at α ≃ 23h36m22s, δ
≃ 61◦20′15′′, while some weak emissioncanbe seenup to
α≃23h35m36s andδ ≃61◦22′06′′(Fig.3).Thefilamentis
not continuous over its full extent but there are maxima
andminimainterchanging inthe[Oiii]flux,mainlyinthe
south–west.This probably reflects inhomogeneities in the
interstellar “clouds” resulting in strong variations of the
shockvelocityuponwhichthe[Oiii]fluxdependscrucially
(Cox & Raymond1985). No [Oiii]emissionis detected in
other areas of the remnant, including those areas where
diffuse Hα, [Nii] or [Sii] emission is observed. The spa-
tial location of the [Oiii] arc like filament nicely matches
the outermost radio contours at 4850 MHz (Fig. 4) sug-
gesting its association with G 114.3+0.3. The [Oiii] arc
probablyliesveryclosetothe leadingedgeofthe forward
shock front. However, the low resolution of the published
radio data (Condon et al. 1994) do not allow us to study
in detail this issue.
4. The long–slit spectra from G 114.3+0.3
Fig.3. G 114.3+0.3 imaged with the medium ionization
The deep low resolution spectra cover the south–east
lineof[Oiii]5007˚A.Theimagehasbeensmoothedtosup-
and south–west parts of the remnant. All spectra clearly
presstheresidualsfromtheimperfectcontinuumsubtrac-
tion. Shadings run linearly from 0 to 7 × 10−17 erg s−1 demonstratethefactthattheobservedemissionmustorig-
cm−2 arcsec−2, while the inlay shows the weak but fila- inate from shock heated gas (Table 4). We note here that
the signalto noiseratiosquotedinTable 4do notinclude
mentary nature of the emission presentonly in the south.
calibration errorswhich are less than 10% for the specific
days of observation. In addition, the ratio of the sulfur
and [Sii] images (Fig. 1 and 2, respectively) reveal weak lines approachesthe low density limit indicating low elec-
diffuse emission in the south–east, south–west and cen- tron densities (e.g. Osterbrock 1989). The spectra taken
tral areas of the remnant. The bright, extended source from positionI (Fig. 1) display strong [Oiii] emission rel-
LBN 565 (Lynds 1965) which is believed to be a back- ativeto Hαsuggestingshockvelocitiesgreaterthan∼100
ground source (Reich & Braunsfurth 1981) is also visi- km s−1 (Cox & Raymond 1985, Hartigan et al. 1987),
ble in the low ionization images. In Table 3, we list typ- while the intensity of the sulfur lines suggests electron
ical fluxes measured in several locations within the field densities of a few tens of cm−3. However,given the errors
of G 114.3+0.3 including the Hii region LBN 565. Since on the individual sulfur fluxes, we estimate that electron
the images are flux calibrated, the Hα+[Nii] and [Sii] densities less than ∼270 cm−3 are allowed.
frames can be used to study the nature of the detected The spectra from position II show roughly the same
emission.These provide evidence that the emissionin the characteristicswiththosefrompositionI,i.e.shockveloc-
south–east and south–west originates from shock heated ity greater than ∼100 km s−1 and low electron densities.
gas ([Sii]/Hα ∼0.6-1.0), while LBN 565 is an Hii region However, due to the lower absolute fluxes the larger er-
([Sii]/Hα ∼0.3-0.4). Towards the center of the remnant rorsonthe sulfur fluxes allow densities up to ∼600cm−3.
we estimate [Sii]/Hα ∼0.4-0.6 showing that the diffuse The spectra from position III display a different qualita-
emission may be associated with G 114.3+0.3. We note tive behavior compared to those from positions I and II.
here that the areas where the optical flux is due to shock Furthermore, we were able to extract spectra from dif-
heatingcoincidefairlywellwiththeareaswheretheradio ferent apertures along the slit (IIIa, IIIb, IIIc). Spectral
4 F. Mavromatakis et al.: Optical observations of G 114.3+0.3
Table 4. Relative line fluxes
I II IIIa IIIb IIIc
Line(˚A) Fa,b Fa,b Fa,b Fa,b Fa,b
4861 Hβ < 200 < 240 < 140 < 180 187 (3)c
4959 [OIII] 220 (2) – 126 (2) – –
5007 [OIII] 688 (5) 1508 (5) 417 (4) 84 (1) –
6548 [Nii] 242 (3) – – 211 (4) 190 (4)
6563 Hα 1000 (12) 1000 (4) 1000 (13) 1000 (18) 1000 (17)
6584 [Nii] 776 (9) 476 (3) 690 (8) 593 (10) 602 (12)
6716 [Sii] 502 (7) 546 (4) 725 (11) 571 (10) 725 (13)
6731 [Sii] 341 (5) 371 (3) 559 (8) 432 (8) 530 (12)
AbsoluteHα fluxd 2.4 1.5 3.0 4.2 3.3
Hα/Hβ > 5 > 4.2 > 7 > 5.6 5.3 (± 1.8)
[Sii]/Hα 0.8 0.9 1.3 1.0 1.3
F(6716)/F(6731) 1.5 1.5 1.3 1.3 1.4
a Uncorrected for interstellar extinction
b Listed fluxesare a signal to noise weighted average of theindividualfluxes
c Numbersin parentheses represent thesignal to noise ratio of the quoted fluxes
d In units of 10−17 erg s−1 cm−2 arcsec−2
Allfluxesnormalized to F(Hα)=1000
variationsareobservedamongthesespectra.Asignificant
detection of the [Oiii] line is only possible for the spec-
trum at position IIIa suggesting shock velocities greater
than ∼100 km s−1. It is found that electron densities less
than ∼400 cm−3 are allowed for this position. Emission
from the [Oiii]5007˚A line is marginally detected in spec-
trum IIIb suggesting shock velocities less than ∼100 km
s−1. Electrondensities also less than ∼400 cm−3 are esti-
mated from the sulfur line ratio and its 1σ error. Finally,
in position IIIc we do not detect any [Oiii] emission but
for the first time weak Hβ emission is detected.
5. Discussion
ThesupernovaremnantG114.3+0.3isoneofthe faintest
and least observed remnants in optical wavelengths (e.g.
Boumis et al. 2001 on G 17.4-2.3).It is also a weak radio
source, while extended X-ray emission is not detected in
the ROSAT band. The evolved supernova remnant under
studyisnowobservedforthefirsttimeinanumberofop- Fig.4. The correlation between the [Oiii] emission and
ticalemissionlinesapartfromHα.Diffuse emissioninthe
the radio emission at 4850 MHz is shown in this figure.
low ionizationimages is detected in the southand central The thick dashed line outlines the [Oiii] emission, while
areas of the remnant. Its spatial correlation with regions ′
thethindashedlinedefinesacircleof34 radius.The4850
of strong radio emission and the relatively high [Sii]/Hα
MHz radio contours (Condon et al. 1994) scale linearly
ratio (§3.1)lead us to propose that the observedemission from7×10−4Jy/beamto0.05Jy/beam.LBN565isalsoa
is associated with G 114.3+0.3. We have also discovered verystrongradiosource,whileitisnotdetectedin[Oiii].
a filament in the medium ionization line of [Oiii] in the
south which is very well correlated with the outermost
radio emission at 4850 MHz. Its projected thickness is e.g.CTB1(Fesenetal.1997),CTB 80(Mavromatakiset
∼20′′ anditisverylikelythatthisfilamentroughlydelin- al. 2001) and G 17.4-2.3 (Boumis et al. 2001). The strik-
eates the outer edge of the expanding shell. Both the cal- inglydifferentmorphologiesbetweenthe lowandmedium
ibratedimages and the long–slitspectra suggestthat it is ionizationlinespointtothepresenceofsignificantinhomo-
theresultofemissionfromshockheatedgas.G114.3+0.3 geneities and density variations in the preshock medium.
belongs to the group of remnants which display strong It is known that the Hα or [Sii] lines are produced in
[Oiii] emission in only certain sections of the remnant, cooler areas behind the shock front, while the [Oiii] line
F. Mavromatakis et al.: Optical observations of G 114.3+0.3 5
is emitted from regions much closer to the front. Since possible that the remnant has passed the adiabatic phase
in the former case higher column densities are sampled ofitsevolutionandhasenteredthe snow–plowphase.For
compared to the latter, the presence of inhomogeneities all these cases, the forward shock velocity at the onset of
and density variations would mainly affect the recombi- the radiative snow–plow phase lies in the range of 250 –
nation zone where the low ionization lines are produced 350 km s−1.
(see Hester 1987). Such effects may also explain the vari- It is proposed that now after ∼ 41000 yrs (Kulkarni
ationsseeninthe [Oiii]/Hαratiointhe long–slitspectra. et al. 1993) following the supernova explosion which gave
Emission from the Hβ line is below our detection birth to PSR 2334+61, the remnant is probably in the
threshold, while the low significance of the Hβ emission radiative phase of its evolution. Its current radius is 25
detected at position IIIc results in an extinction c of 0.75 D pcandtheshockfrontispropagatinginainhomo-
2.5kpc
(±0.45)oranA of1.6(±0.9).UsingthecodeofHakkila geneous interstellar medium as the optical data indicate.
V
et al. (1997) to estimate the interstellar absorptionin the Soft X–ray emission from G 114.3+0.3 is not detected in
directionofG114.3+0.3,we findalogarithmicextinction the ROSAT observation of PSR 2334+61 (Becker et al.
of 1.5 (±0.3) for a distance of 2 kpc and 2.0 (±0.3) for 1996) and we obtain an upper limit on the unabsorbed
a distance of 3 kpc. Our measurement is compatible with soft X–ray luminosity of ∼6 × 1035 erg s−1 (0.1 – 2.0
bothestimateswithinthe2σ rangedue tothe largeerror. keV).DeepX–rayobservationswouldhelptoactuallyde-
Anextinctionintherangeof1.0–1.5couldbemorefavor- terminetheX–raypropertiesoftheremnantorplacenew,
able since it would imply Hβ fluxes below our detection more stringent upper limits.
limit, although other values cannot be excluded.
In order to obtain quantitative estimates of basic su-
6. Conclusions
pernova remnant parameters we will use the relations
ThesupernovaremnantG114.3+0.3wasobservedinma-
n[Sii] ≃ 45 nc×Vs2,100, (1) jor optical lines. New diffuse structures were detected in
the south–eastandcentralareasofthe remnant.The flux
and
calibrated images imply that the diffuse emission is asso-
E =2×10−5β−1 n V2 r 3, (2) ciated with G 114.3+0.3. In addition, we discovered an
51 c s,100 s,pc [Oiii] filament in the south which is very well correlated
givenbyFesen&Kirshner(1980)andMcKee&Cowie with the radio contours at 4850 MHz. Long–slit spectra
(1975), respectively. The factor β is of the order of 1–2, along this filament suggest that the emission arises from
n[Sii] is the electron density derived from the sulfur line shock heated gas, while the sulfur line ratios indicate low
ratio,n isthepreshockclouddensity,E istheexplosion electron densities.
c 51
energyinunitsof1051erg,V istheshockvelocityinto
s,100
the clouds in units of 100 km s−1, and r the radius of Acknowledgements.
s,pc
the remnant in pc. The long–slit spectra show that an The authors would like to thank R. Fesen for his help-
upper limit on the electron density is ∼500 cm−3 and we ful comments. We would also like to thank the referee
use it in Eq. 1 to obtainnc × V2s,100 < 11. Using this last for his remarks and suggestions. Skinakas Observatory
result in Eq. 2, we find E51 < 1.6 D32.5kpc, a rather coarse is a collaborative project of the University of Crete, the
upper limit. The distance to the remnant in units of 2.5 Foundation for Research and Technology-Hellas and the
kpc is denoted by D2.5kpc (Kulkarni et al. 1993). Max-Planck-Institut fu¨r Extraterrestrische Physik. This
Sinceadirectmeasurementoftheinterstellarhydrogen work has been supported by a P.EN.E.D. program of the
densityisnotavailable,adensityof0.1cm−3 (Fu¨rstetal. GeneralSecretariatofResearchandTechnologyofGreece.
1993,Frailet al.1994)andanE51 of1 and0.1 willbe as- This research has made use of data obtained through
sumedinthefollowingcalculations.AccordingtoCioffiet the High Energy Astrophysics Science Archive Research
al. (1988) and the above range of parameters the onset of Center Online Service, provided by the NASA/Goddard
the pressure–driven snow-plow phase (PDS) would occur Space Flight Center.
at 38 pc and 19 pc, respectively. These radii when com-
paredwiththecurrentradiusof∼25pcat2.5kpcsuggest
References
that in the former case the onset of the PDS phase will
occur in another ∼40000 yrs, while the PDS phase has Becker W., Brazier K.T.S., Tru¨mper J. 1996, A&A 306, 464
already started in the latter case. We note here that an BoumisP.,MavromatakisF.,PaleologouE.V.2001,A&A,sub-
interstellar density of e.g. 0.5 cm−3 would lower the lim- mitted
iting radii to 14 pc and 10 pc, respectively, implying that Cioffi D.F.,McKeeC.F.,BertschingerE.1988, ApJ334, 252
Condon J. J., Broderick J. J., Seielstad G. A., Douglas K.,
theremnantiswellwithintheradiativephaseofitsevolu-
Gregory P. C. 1994, AJ107, 1829
tion. Since neither the distance, nor the ambient medium
Cox D. P., RaymondJ. C. 1985, ApJ298, 651
density are accurately known, we cannot confidently de-
Cox D. P.1972, ApJ178, 159
termine the current stage of evolution of G 114.3+0.3.
Fesen R.A., Kirshner R.P. 1980 ApJ242, 1023
However, since the majority of the cases we have exam- Fesen R.A.,WinklerP.F., RathoreY.,DownesR.A.,Wallace
inedresultsinradiilowerthanthecurrentradiusitmaybe D., Tweedy R.W.1997, AJ113, 767
6 F. Mavromatakis et al.: Optical observations of G 114.3+0.3
FrailD.A.,GossW.M.,WhiteoakJ.B.Z.1994,ApJ437,781
Fu¨rst E., Reich W., Seiradakis J. 1993, A&A276, 470
HartiganP.,RaymondJ.andHartmannL.1987,ApJ316,323
Hakkila J., Myers J. M., Stidham B. J. and Hartmann D. H.,
1997, AJ 114, 2043
Hester J. 1987, ApJ314, 187
Kallas E. and Reich W. 1980, A&AS42, 227
Kulkarni S.R., Predehl P., Hasinger G., Aschenbach B. 1993,
Nature 362, 135
LyndsB. T. 1965, ApJS 12, 163
Mavromatakis F., Papamastorakis J., Paleologou E. V., and
VenturaJ. 2000, A&A353, 371
Mavromatakis F., Ventura J., Paleologou E. V., and
Papamastorakis J. 2001, A&A 371, 300
McKee C. F., Cowie L. 1975, ApJ195, 715
Osterbrock D. E. 1989, Astrophysics of gaseous nebulae, W.
H. Freeman & Company
Reich W. and Braunsfurth E. 1981, A&A99, 17