Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.stsci.edu/~wofford/research/talk_12172.pdf Äàòà èçìåíåíèÿ: Sat Nov 17 02:22:58 2012 Äàòà èíäåêñèðîâàíèÿ: Mon Feb 4 04:09:26 2013 Êîäèðîâêà:

WOLF RAYET STARS IN THE EXTRAORDINARY CLUSTER NGC 3125 A1
Aida Wofford Space Telescope Science InsEtute Claus Leitherer, Space Telescope Science Ins5tute, USA

Rupali Chandar, University of Toledo, USA

Aida Wofford

IAU Beijing, August 2012

1


NGC 3125: AN EXTREME WR GALAXY?
-14.0

Log Flux (erg s-1 cm-2 A-1 )

NIII(4640) /CIII(4650)

-14.5

-15.0

-15.5

Figure 2. VLT/FORS1 optical spectra of NGC 3125-A and B. Spectra have been velocity corrected (V r = 865 km s-1 , Lauberts & Valentijn 1989) and WR emission features have been marked.

photometry. We derivea slit-correction factor of 1.3 ± 0.2 for knot A, which dominates the accuracyof our final flux-calibrated spectra. An identical correction factor is found for knot B. We expect our absolute flux calibration to be correct to 15 per cent. Extracted spectra of NGC 3125-A and B are shown in Fig. 2, in which the principal WR emission features are identified, including weak N IV 4058.

2.3 HST imaging Figure 3. 1 â 1-arcsec2 (50 â 50 pc) STIS/Long Pass acquisition (the top A panel) and pre-COSTAR HST /FOC F220W (the bottom panel) images of To complement our FORS1 data set, we have retrieved archival regions A. The two clusters A1 and A2 are marked. North is up and east is HST images, including a STIS/LONG PASS acquisition image to left-hand side. Bt = 40 s, = 7200 å) from programme GO 9036 (P.I. C. Leitherer) ( c of region A and a FOC/F220W1 image (t = 500 s, c = 2280 å) 500 7000 from programme GO 4800 (P.I. P. Conti) of the central starburst region. Spectra have STIS/LONG PASS acquisition image resolves region A The ijn 1989) and two clusters, separated by 0.5 or 25 pc at the distance of into NGC 3125 (upper panel of Fig. 3). We designate these two components A1 and A2, with A1 corresponding to the brighter of the pair to the 0.2 for knot east. The components have FWHM of 0.17 and 0.13 arcsec, and ted spectra.a flux ratio of F A1 / F A2 1.9 at 7200 å, although an absolute expectflux calibration was not possible. our In t. Extracted the HST /FOC F220W image A1 is bright whereas A2 is barely n whichdetected, with F A1 (F220W)/ F A2 (F220W) 10 (lower panel of the g weak Fig. 3). Assuming that A1 and A2 possess comparable intrinsic enN IV ergy distributions ­ which is reasonable since both clusters host WR stars (Section 2.1) ­ A2 must suffer from significantly higher extinction. Aperture photometry reveals F A1 (F220W) = (1.1 ± 0.2) â Figur Figure 1 F A2 (F220W) 3 pc) STIS/Long Pass acquisition (the top e 4. 1 â 1-arcsec2 (50 â 50 pc) pre-COSTAR HST /FOC F220W 10-15 erg s-1 cm-13.andâ 1-arcsec2 (50 â 50â 10-16 erg s-1 cm-1 . ( = panel) and pre-COSTAR HST /FOC F220W (the bottom panel) imagesc of 2280 å) image of region B. North is up and east is to left-hand ed archivThe HST /FOC image also resolves region B into two compoal side. regions1 and B2, o clusters A1 and A2 are marked.B1 to the and east is A. The tw with the brighter component, North is up nents 4) ­ B ition image (Fig. to left-hand side. south-east with F B1 (F220W)/ F B2 (F220W) 1.2. It is not clear . Leitherer) which 2.4 HST spectroscopy = 2280 å) cluster within region B hosts the WR population since these are spatially unresolved in our VLT/FORS1 imaging. High spatial ral starburst Archival HST /STIS spectroscopyofregion NGC 3125-A1 (Chandar resolution optical imaging is not yet available. It was not possiet al. 2004) was obtained in programme GO 9036 (P.I. C. ble to s region A perform aperture photometry on B1 and B2 individually in

W Blue compact dwarf galaxolf­Rayet clusters in NGC 3125 1825 y RedshiS (z~0.004), distance~11.5 Mpc 12+log O/H~8.3 (LMC) 2 super star clusters: A & B Archival HST/FOC F220W (=2040 2520 å) images show that A & B each hosts 2 clusters ­ A1 has the strongest He II 1640 emission from WR stars observed in a local starburst Wolf­Rayet clusters in NGC 3125 1825 (Chandar et al. 2004) ­ A2 is heavily reddened ­ B also exhibits He II 1640 emission but it is unclear if both B1 and B2 host WR stars · Most importantly, regarding A1, it is unclear if the number raEo of stars is extreme, i.e., WR/O~0.8.

· · · · ·



NIV(4058)

CIV(5808)

A B

4000

4500

HeII(4686)

5000 5500 Wavelength (A)

6000

6500

7000

A B

HST ACS HRC F330W, F555W, F814W
2

Aida Wofford

IAU Beijing, August 2012


HST SPECTROSCOPY
We observed ·A & B with STIS from 1150 10270 å ·A with COS from 1150 1450 å at higher spectral resoluEon. STIS (Space Telescope Imaging Spectrograph) ·52"x2" long slit, projected onto either ·25"x25" FUV or NUV MAMA detector or ·52"x52" CCD detector · GraEngs - G140L (1150 1730 å, nominal =0.9 å) - G230L (1570 3180 å, nominal = 2.4 å) - G430L (2900 5700 å, nominal =4.1 å) - G750L (5240 10270 å, nominal =7.4 å) COS (Cosmic Origins Spectrograph) ·Circular aperture of 2.5" in diameter ·GraEng ­ G130M (1150 1450 å, nominal =0.06 å; HST ACS HRC F330W (3100 3650 å) image with COS & effecEve =0.3 å or 75 km/s)
STIS apertures overlaid (before centering).
Aida Wofford IAU Beijing, August 2012 3


STIS 2D SPECTRA

· G140L 1150 1730 å · Showing full range · VerEcal band is geocoronal Ly emission filling slit · Bright spectral traces are from A1, B1, and B2 · The bright spot near the right end of the spectral trace of A1 is He II 1640.

· · · ·



G230L 1570 3180 å · G430L 2900 5700 å · G750L 5240 10270 å · Showing parEal Showing full range · Showing parEal range (H+[N II]) Featureless range (H+[O III]) · Bright spectral Bright spectral · Bright spectral traces from A1, A2, traces are from A1, traces are from A1, and B B1, and B2 A2, B1, and B2



Aida Wofford

IAU Beijing, August 2012

4


UV WR STAR SIGNATURES IN STIS SPECTRA
A1 · N IV] 1486, · He II 1640, · N IV 1718 detecEons · Strongest He II 1640 A2 · Too faint B1 · Weak N IV] and He II 1640 B2 · N IV] 1488, · He II 1640, detecEons · N IV 1720 at edge of graEng and suspicious



Ai

da Wofford

IAU Beijing, August 2012

5


Figure shows 1150 9000 å range OpEcal WR lines shown in bopom two panels The WR lines are: ·N IV] 1486 ·He II 1640 ·N IV 1718 ·N III 4640 ·C III 4650 ·He II 4686 ·C IV 5808
Aida Wofford

FULL STIS SPECTRUM OF A1



IAU Beijing, August 2012

6


FULL STIS SPECTRUM OF STIS SPECTRUM OF B
· The top two panels correspond to B1 · The opEcal spectral traces of B1 & B2 are faint in the opEcal · The opEcal WR lines are too weak to determine reliable fluxes


Aida Wofford

IAU Beijing, August 2012

7


A1: AGE=3 MYR
Showing comparison between recEfied STIS spectrum of A1 (black) and a Starburst99 model (brown; (Leitherer et al. 1999 and ref. therein; SSP, 3 Myr LMC metallicity, Kroupa IMF from 0.1 100 M
¤, empirical stellar library of SMC/LMC stars). The library stops at 1600 å. In gray we show the residual.




Aida Wofford



IAU Beijing, August 2012

8


A1: REDDENING
· Our determinaEon of c(H) is uncertain because our slit is wide and the HII gas is extended, so the spectral resoluEon of the Balmer lines is poor. From c(H) we obtained
­ E(B V)HII=0.22 which is between ­ E(B V)HII=0.27 of C04 & ­ E(B V)HII=0.16 of HC6.

· From UV conEnuum slope we obtained
­ E(B V)*=0.48 which is in agreement with ­ E(B V)*=0.52 of C04.

· Using E(B V)HII=0.16 and the SMC exEncEon curve of Gordon et al. 2003 matches best the dust free model.

Aida Wofford

IAU Beijing, August 2012

9


WR/O
· We determine the mass of A1 by comparing the luminosity at 1500 å of the SMC exEncEon corrected STIS spectrum and a 3 Myr model. · N(O) was determined assuming a Kroupa IMF (0.1 100 M
¤) from our cluster mass determinaEon. · N(WNL) was determined from L(WNL)=1.2e37 erg/s (Schaerer et al. 1999). · N(WC) was determined from L(WC)=1.8e36 erg/s (Hadfield & Crowther 2006) · Our number raEo, WR/O, is intermediate between Chandar et al. 2004 (0.84) and Hadfield & Crowther (0.2). Note that, HC6 used L(WN5 6)=1.8e37 erg/s instead of L(WL).
Aida Wofford IAU Beijing, August 2012 10


WR STAR POPULATION OF A1
A1 Age Mass EW(1640) FWHM(1640) L(1640) N(WNL) L(4686) N(WL) L(5801) N(WC4) N(O) N(WR)/N(O) B2 EW(1640) FWHM(1640) L(1640) N(WN5 6) 7.2 6.7 4.26E+38 24 3 Myr Mass=1.7x105 M
¤ 7.9 å 7.6 å 2.39E+39 erg/s 200 1.69E+38 erg/s 106 6.95E+37 erg/s 21 468 ~0.42 THIS WORK THIS WORK

Aida Wofford

IAU Beijing, August 2012

11


SUPERIOR SPECTRAL RESOLUTION OF COS IN COMPARISON TO STIS
Showing COS vs. STIS spectrum of A1 ·COS spectrum degraded to effecEve resoluEon of 0.3 å ·STIS spectrum has nominal resoluEon of 0.9 å ·COS conEnuum level is twice the value of that of STIS so here it is scaled down. This is because in STIS we did a background subtracEon, but this is not possible with COS data.


Aida Wofford IAU Beijing, August 2012 12


LINES IDENTIFIED IN THE COS SPECTRUM OF A1
Foreground
· Milky Way (H I, C II, N I, Si II, Si III, Si IV, S II, Ni II) · Geocoronal (H I, O I, clipped)

Intrinsic

· OB stellar wind (C III, N V, Si IV, O V)

· Photosphere (Fe V) ISM · (H I, C II, N I, O I, Si II, Si III, Si IV, S II, Ni II)
Aida Wofford

IAU Beijing, August 2012

13


A1: O V 1371 å
We compare the COS spectrum of A1 (black) with a Starburst99 model (brown; SSP, 3 Myr, Kroupa IMF, LMC metallicity, theoreEcal stellar library). The broad absorpEon around 1371 å due to O V from very massive stars and photospheric Fe V presents a wavelength offset.
Aida Wofford IAU Beijing, August 2012 14


·

·

Our analysis of the UV STIS observaEons of NGC 3125 A1 show that... ­ EW(1640) is 3 Emes the mean of local starburst galaxies (Chandar et al. 2004) and 3 Emes the value of the strongest LBG (Erb et al. 2010), which suggests that A1 must have a large fracEon of WR stars relaEve to the number of O stars, WR/O. ­ Given the expected diluEon of the He II 1640 emission line by the nearby conEnuum produced by massive B stars, and given the expected value of EW (1640) for a stellar populaEon, which is maximum 3 å, the number of WR/O would be expected to be extreme. ­ We speculate that this could be due to 1) a top heavy IMF, 2) a few massive stars with very strong winds. ­ The nebular analysis, which is ongoing, will provide more clues about the stellar populaEon properEes. The analysis of the COS observaEons of NGC 3125 A1 show... ­ The strongest O V 1371 å absorpEon from a starburst observed in the local universe ­ Since this line originates in the most massive stars (Walborn et al. 1985), this emphasizes how extreme NGC 3125 A1 is. ­ The O V line is blueshiSed. ­ Currently, stellar populaEon synthesis models are unable to reproduce the line, which is heavily affected by clumping.

TAKE AWAY POINTS

Aida Wofford

IAU Beijing, August 2012

15


THANK YOU.

Aida Wofford

IAU Beijing, August 2012

16