ngc-346.org

+OPTIONS: ^:{}

MUSE cube of N66/NGC 346 in the SMC

• [2023-06-01 Thu] This file was getting far too big. I have moved specific sub-projects to their own files
• The Deep Red Lines paper
  • file:ngc-346-drl-spectra.org for description of data analysis steps
  • file:ngc-346-drl-discuss.org for …
    • … searching for candidate identifications …
    • … and comparison with other regions
  • cloudy-h2.org for Cloudy PDR models
• The Bow Shock paper
  • cloudy-bow-shock.org for the Cloudy models
  • file:ngc-346-bow.org for everything else (for now)

Compact sources of interest

• The Giant Raman Wing Star
  • 0:59:05.1826 -72:10:35.340
  • This is mYSO-C in Rubio:2018a
    • A_V = 2.5 +/- 1
    • log T_eff / K = 4.5 +0.1 -0.2
    • log L / L_sun = 5.1 ± 0.25
    • Mass = 26 ± 5 M_sun
    • Log Age / yr = 4.25 ± 1.0
    • Class I (massive protostellar disk)
    • 20 times brighter than any other source in O I 8446
• The central globule star
  • That is, center of the MUSE field
  • SSN 17
  • 0:59:02.9347 -72:10:34.976
  • It has a mini star cluster around it (Sc 2)
  • This has no low ionization line emission
  • The star must be in front of the globule, possibly far from it
  • Evidence for some [Fe III] emission, possibly jet source
• The Sc 1 globule star
  • 0:59:05.9596 -72:10:30.145
  • Star is 1 arcsec south of SSN 62 (O9V), which is probably unrelated foreground
  • Globule is just 3 arcsec north of W 2 (O8.5III), which is brightest star in central cluster
  • Bright O I 8446 centered on star, which is at head of globule
    • Also cometary tail-like extension along tail of globule
  • Based on Ha brightness around the edge, globule seems ionized principally from South (by W 2) but also to a lesser extent from the West (similar to the fainter globules to south of Sc 1)
• The East globule star
  • The globule that is just north of the bow shock
    • Along the chain of filaments
  • 0:58:59.7724 -72:10:16.596
  • Very bright in O I 8446, second only to mYSO-C
• The Sc 5 globule
  • 0:59:00.4173 -72:10:04.941
  • This has the highest density ionized gas apart from the globule around mYSO-C
  • Is it ionized by a nearby O star?
  • Or is it just at a similar z-position to one of the bright ones, such as W2 or W1?
    • Seems to be pointing more closely at W2
  • Double source in O I
• The Sc 4 globule
  • 0:59:02.5807 -72:10:07.095
  • Tiny globule with star in its tip
  • Bright unresolved source in O I 8446
• South East Foreground globule star A
  • Two stellar sources with very different line emission spectra
    • The one to E is more central on the globule (but may just be projected)
    • The one to the W is resolved in O I 8446 and is strong in [S II] but not [O I]
• South East Foreground globule star B
  • Just visible on bottom edge of MUSE field
  • 0:59:08.5152 -72:11:03.105
  • Projected on or inside a high extinction globule
  • Bright in O I 8446
• South foreground ridge star
  • 0:59:02.0775 -72:11:02.425

Details of Raman profiles

• Most of this is in the globule around mYSO-C
  • All the other emission (for example, from the filaments) is very weak
• [2022-10-26 Wed] I have done new continuum subtractions with a broader window, which work better for the Ha wings

Extent of flat top to profile

• This should tell us the wavelength where the profile becomes optically thin

Absorption lines

The O I lines

• These are seen only in the globule spectrum with any certainty
  • Note that the csub-101 spectrum loses the Raman wings, since they are more than 100 pixels wide
• In other regions, they are too easily confused with the two Deep Neutral UILs on either side of them

The Si II line

• This is at 6480.811 Å rest STP
• Which is 6484.48 with systemic redshift
• We do see absorption at that wavelength in the YSO spectrum
  • Although it is less deep than the O I lines

The He II lines

• In the Ha profile, we see absorption at 6545, which may well be the Raman scattered counterpart of He II 1025.272, which is just to the blue of H I 1025.722
  • 6547.4 observed wavelength in YSO means 6543.6 rest optical
    • Difference in wave numbers from Ha is 45.2 /cm
    • Which gives UV wavelength of 1 / (97492.283 + 45.2) = 1025.25, which is close enough
• This line is seen in emission in the PNe NGC 6888 and NGC 6881
  • Choi:2020c

The new broad absorptions [2022-10-24 Mon]

• There are many more features in the near Raman wings, which can be seen most clearly around the GLOB-E region, but generally in the neutral gas
• These are the approximate observed wavelengths (rest wavelengths)
  • Red wing 6600 (6596)
    • This is the strongest, with a parabolic shape in the near red wing
    • In the csub-101 cube, the trough has negative values
    • The spatial distribution seems to track the Raman wings
    • But it might just be tracking the shadow of the Ha core emission (the depression halo around the strong line due to subtracting the median-filtered continuum)
    • We should try a 1001 pixel median filter to see if it works better
    • Note that there is maybe even a hint that we are seeing a very shallow version of this in Orion
  • Red wing 6612 (6608)
    • Just a bit further out in the red wing, and might be a bit broader
    • The gap between this one and the last one is classified as a candidate emission line #1603 in the spreadsheet, but if I am right about the two absorption lines, then it is not really an emission line but will just be showing a little bit of the Raman wing.
  • Blue wing XXX
    • There are some genuine emission lines, probably Deep Neutral, which complicate things on this side: #1524 (6505 observed) and #1547 (6534 observed)
    • There may be some absorption here, but it is hard to tell

General points about the globules and filaments

• The background globules show bright heads and less bright flanks
  • This is evidence for the local anisotropy of the radiation field
  • And that the tail axis is aligned with the radiation
  • Suggesting that the globules have been sculpted to some extent (unless it is simply chance)
• The foreground ridge, on the other hand, does not look like this
  • It seems more to be illuminated from the side

Lines of interest

• Still to do:
  • [O I] 6300, 6363
  • [S III] 6312, 9069
  • O I 8446
  • [Cl IV] 8045
  • [C I] 8727
  • [Fe III]

General thoughts

• Properties of N66 from Lopez:2014a
  • Q(H) = 4e50 /s = 40 x Orion
  • n = 100 pcc = 0.03 x Orion
    • In both cases there are large variations, however
  • R = 64 pc = 30 x Orion
  • therefore (n R) and (Q n) are both similar to Orion, so ionization parameter is similar too

Comparison of size scales and luminosities between Orion and SMC

• SMC distance: 61.7 kpc
  • Distance modulus: 5 log10(D/10 pc) = 18.95
  • The brightest star in the vicinity of source C is V = 15
    • => M_V = 15 - 18.95 = -3.95
    • This is only 0.5 mag brighter than th1C, so probably O6V
  • Sun has M_V = +4.81 => V = 23.76
  • 10 pc = 33.43 arcsec, 50 pc = 167.2 arcsec
• Orion
  • Distance: 410 pc => 150 times closer
    • Distance modulus: 5 log10(D/10 pc) = 8.06
  • Size of Orion S region: R = 30 arcsec = 0.06 pc
    • 0.2 arcsec @ SMC = 1 MUSE pixel !!!
  • Separation between Trapezium stars: about 5 arcsec = 0.01 pc
    • 0.03 arcsec @ SMC
  • Separation of Orion S from theta 1 C: R = 55 arcsec = 0.1 pc
    • 0.4 arcsec @ SMC = 2 pixels
  • Separation of Orion-BNKL from theta 1 C = 75 arcsec = 0.15 pc
    • 0.5 arcsec @ SMS = 2.5 pixels
    • BN and Src I are high-mass protostars in the BNKL region
  • Separation of th2A from th1C: R = 140 arcsec = 0.28 pc
    • 0.93 arcsec @ SMC = 5 pixels
  • Entire region studied in Orion paper: 360 x 270 arcsec
    • 2.4 x 1.8 arcsec @ SMC
• Orion star magnitudes
  • th1C:
    • V = 5.13
      • A_V = 1.5 approx
      • M_V = 5.13 - 1.5 - 8.06 = -4.43
    • I = 4.73
    • K = 4.57
      • A_K = 0.15 approx
      • M_K = 4.57 - 0.15 - 8.06 = -3.64
• NGC 346 star magnitudes:

The photospheric profiles around Lyβ

• I need to look at some more POWR models
• Observationally it is impossible to tell because of all the ISM H_2 lines that get in the way. These completely destroy the continuum near Lyβ

POWR models

FUV atlas

• https://archive.stsci.edu/prepds/fuvostars/datalist.html
• Example spectra are at ../fuv-spectra/fuvostars/
  • I got the spectrum of the O2 III star W3

Observations of the filaments and star forming regions in N66

• HST observations of pre-MS population
  • These find several sub-clusters of pre-MS stars
  • Gouliermis:2006h (Paper I)
  • Nota:2006x
  • Sabbi:2007h
    • This has the sub-clusters that I am currently plotting in the figures
    • But it might be better to use the ones from Schmeja instead
    • Or even the ones from IR studies
  • Hennekemper:2008u (Paper III) - not sure what Paper II is
    • Detailed analysis of
      1. OB stars
      2. Upper main sequence (A-F)
      3. Pre-main ssequence
    • Find that OB stars are less reddened than other upper MS stars
  • Gouliermis:2014a
    • Detailed analysis of clustering properties, using autocorrelation function
      • Find a broken power-law behavior, which they model as a two-component system
        1. Central cluster with core radius of 2.5 pc
        2. Extended fractal component
• near-IR observations
  • Rubio:2000k present 2.12 H_2 and 2.14 micron continuum
  • They find” several embedded sources”
  • Numbered 1-5 and associated with mid-IR peaks as follows:
    1. Tip of plume
    2. I
    3. D
    4. E
    5. C
• mid-IR observations
  • Contursi:2000f
    • Find several clumps of PAH emission:
      • labelled A-H
  • Simon:2007r
    • Spitzer
  • Whelan:2013d
    • More Spitzer
  • Ruffle:2015h
    • color-color classification of point sources in the entire SMC
    • overlap between YSOs and H II regions
  • Sewio:2013f
    • Spitzer photometry of entire SMC
    • Their Fig 12 compares ACS H alpha with Spitzer 3.6 micron for all their sources in NGC 346
    • Most of them are inside globules
      • We need to decide what term to use for the globules:
        • PIGs
        • Globules
        • EIYSOs
        • “Dusty columns” is what Sewio:2013f calls them
        • EGGs (Evaporating gaseous globules)
• CO/C I/C II observations
  • Rubio:2000k
    • CO
  • Requena-Torres:2016g
    • APEX and Sofia observations
    • Mainly [C II] 158 micron but also [C I] 609 micron and CO (various transitions and isotopes)
    • Extra comments [2022-11-05 Sat]
      • They find that the [C II] is well correlated with the CO on large scales (> 10 pc) but is anti-correlated on scales of a few parsecs
      • They find that [C I] tends to follow CO, particularly 13CO
      • It looks to me like the CO and [C I] come from more or less the same place as our optical deep neutral lines
      • but the [C II] emission is very different
        • it has a peak to the NE of W2, in a spot that is completely dark in all optical emission
        • the mid-IR dust emission is also weak there
        • On the other hand, there is weaker [C II] emission that does seem to trace the filaments and globules seen in the optical and infrared
• Analysis of clustering/sub-clustering
  • Schmeja:2009q
    • Quote from abstract

      Several PMS clusters are aligned along filaments of higher stellar density pointing away from the central part of the region.
                  

      • This is exactly what I had spotted in the star distribution
• Optical UBV observations
  • See below Papers on the O stars

Ionizing stars in N66

Contribution of the WR+WR binary HD 5980

• Considerable variation over last 20 years
• Conclusions
  • For the low state, we have Q_0 = 1e50 approx
  • And a very low Q_2 (so no ionizing to He II)
• High state (1994):
  • log L = 6.6
  • T_* = 47 kK
  • Mdot / sqrt(f) = 111e-5
  • Vinf = 750 km/s
  • R_10 = 28 Rsun
• Low state (2014):
  • log L = 6.23
  • T_* = 48 kK
  • Mdot / sqrt(f) = 14e-5
  • Vinf = 2100 km/s
  • R_10 = 19 Rsun
• Transformation for POWR models
  • All their models are for nominal log L = 5.3
  • They use a transformed radius: R_t = R_* (V_inf / 2500 km/s)^2/3 (Mdot sqrt(D) / 1e-4)^-2/3
    • where D = 1/f is clumping factor
  • For a different luminosity, we must rescale R_* as L^1/2
  • And Mdot scales as L^3/4
  • So we have:
    • High state:
      • R_* = 28 sqrt(10**(5.3 - 6.6)) = 6.268 Rsun
      • R_t = R_* ((750/2500)/ (11.1))**(2/3) = 0.09 6.268 = 0.56412 Rsun => -0.248 on log10 scale
        • Unfortunately, that is off the bottom of the grid!
    • Low state:
      • Luminosity is 10**(6.23 - 5.3) = 8.5 times higher than fiducial
      • R_* = 19 sqrt(10**(5.3 - 6.23)) = 6.513 Rsun
      • R_t = R_* ((2100/2500)/ (1.4))**(2/3) = 0.7114 6.513 = 4.633 Rsun => 0.6659 on log 10 scale
• So for the low state, the closest model is 06-14: T = 44.7 kK, log Rt = 0.7
  • This comes with R_* = 7.5 => rescaled: 7.5 sqrt(8.5) = 21.9 Rsun
    • (close enough to 19)
  • Q_0, Q_1, Q_2, Q_3 = 49.01, 47.69, 0.00, 45.77 - so very soft
    • Rescaled back to real luminosity: Q_0 = 49.01 + 6.23 - 5.3 = 49.94
  • Mdot = -4.23 => rescaled: (8.5)**(3/4) 10**(-4.23) = 2.93e-4
    • this is not entirely consistent with the original 1.4e-4
  • V = 1600 km/s
• We should also look at the closest hotter model 07-13: T = 50000 K, log Rt = 0.8
  • R_* = 5.9 => rescaled: 5.9 sqrt(8.5) = 17 Rsun
    • So this model and the previous one nicely bracket 19 Rsun
  • Q_0, Q_1, Q_2, Q_3 = 49.09, 48.47, 36.46, 47.42 - so not so soft, but still almost no He+ ionizing radiation
    • Rescaled ionizing luminosity: 49.09 + 6.23 - 5.3 = 50.02
  • Mdot = -4.53 => rescaled: (8.5)**(3/4) 10**(-4.53) = 1.47e-4
    • Very close to original

Notable stars by luminosity

• SSN 7, MPG 435, DEHLS 1001, Walborn 1
  • O4If+O5-6
  • r = 11.6 arcsec = 3.48 pc
  • V = 12.6
• Evans 001, SSN 5, Sk 80, AzV 232, MPG 789
  • O7 Iaf+
  • V = 12.31
  • Why does this have V-I = +0.154?
    • It has B-V = -0.19
  • 00 59 31.94 -72 10 46.05
    • Way off to the East, off my maps
    • r ≈ 120 arcsec
  • UV spectrum studied in detail in Bouret:2021h
• SSN 4, MPG 185
• HD 5980, MPG 755 (not in SSN)
  • WR+WR+O
    • Eclipsing binary plus at least one more O star
    • Brighter WR used to be an LBV
    • See Hillier:2019j
  • V = 11.31
  • M = 60 Msun + 60 Msun
  • 00 59 26.5847969465 -72 09 53.927725716
  • r = 105.7 arcsec
• Av 229, SSN 1?, MPG 755
• Av 214
• SSN 11, MPG 342, Dufton 1010
  • O5-6 V((f))
  • V = 13.6
  • Just to W of red box
  • r = 24.85
  • could be important, marginally
• SSN 15, MPG 368, Dufton 1012
  • O6: V((f))
  • V = 14.1
  • Quite close, in Sc 2
  • r = 17.15 arcsec
  • could be important, marginally
• Av 234
  • B3 Iab
  • V = 12.80
  • Way off to north
  • r = 360 arcsec
  • Too far away
• Av 210, Sk73
  • B1.5 Ia
  • V = 12.8
  • Way off to the West
  • Too far away
• Av 226
  • O7 IIIn((f))
  • V = 14.37
  • Way to South
  • 00 59 20.70 -72 17 10.52
  • r = 600 arcsec - too far!

Notable stars by proximity to SSN 152

• SSN 168, MPG 454
• SSN 9, MPG 355, W 3
  • V = 13.45
  • O2III(f)
  • r = 22.68 arcsec = 6.8 pc
    • 7.55 times farther than SSN 14 => 57 times smaller flux per Q(H)
  • This was first discussed by Walborn & Blades (Walborn:1986y)
    • They classified it as

      spectrum of type O3 III(f*), identical to that of HDE 269810 in the Large Magellanic Cloud
                  

  • Current classification from Heydari-Malayeri:2010i
    • Or is it?
• SSN 14
  • 00 59 05.972 -72 10 33.85
  • V = 14.1
  • r = 3 arcsec = 0.9 pc
  • New ID for this star
    • I think that this is MPG 470
      • 00 59 05.984 -72 10 34.05 from 2MASS
      • This is within 0.2 arcsec
      • O8.5 III
    • This is the same as Walborn 2
• SSN 17, MPG 396
  • V = 14.383
  • O7V
• SSN 18, MPG 487
  • V = 14.47
  • r = 8.37 arcsec
• SSN 22, MPG 476
  • V = 14.562
  • r = 4.06 arcsec
  • No spectral classification!
  • Just to side of SSN 9/W2
• SSN 40, MPG 451
  • V = 15.1
  • B0V
  • r = 3.17 arcsec but to south
• SSN 43, MPG 481
  • V = 15.08
  • B - V = -0.23
  • r = 5 arcsec

[#A] Reminder on distances: 1 arcsec = 0.3 pc

Other interesting stars and objects

• SSN 50, MPG 508
  • This is the one that shows the broad Ha line
  • It is in Sc 8
  • There is no spectral classificaction
• N66A Compact H II region
  • Studied in depth by Heydari-Malayeri:2010i
  • First line of their discussion:

    N66A is clearly the most compact Hii region of the N66 complex in the optical.
            

    So that is clearly untrue, if we count the region around IR Source C, which is far more compact

  • Exciting star is N66-A
    • Spectral type: O8 V

The twins of SSN 14

• Since SSN 14 does not seem to have any spectroscopy, we need to compare it with stars that do in order to estimate its spectral type.
• I found all the ones within 0.3 mag of V and V-I
  • SSN 13, MPG 324
    • O4V((f+)
    • V = 14.13
    • V-I = -0.15
    • B-V = -0.31
    • U-B = -1.14
    • r = 37 arcsec
  • SSN 15, MPG 368
    • Already measured: see above
    • B-V = -0.23
    • O6: V((f) or O5.5V((f+))
  • SSN 16, MPG 482
    • B0.5V (surprising that it is so late)
    • V = 14.34
    • V-I = -0.05
    • B-V = -0.07
    • U-B = -1.19
    • r = 40 arcsec, located to right of base of plume

Papers on the O stars

• Walborn:1978k
  • Brightest star is NGC 346 No. 1, O4 III (n)(f)
    • This is SSN 7 and it is now classified as O4If+O5-6 (Dufton:2019n)
    • Also MPG 435
• Evans:2006z
  • FLAMES/GIRAFFE spectroscopy
  • They use their own numbering system:
    • For instance, NGC 346-001 is Sk 80 and MPG 789
  • 19 O stars and 84 early B (not complete)
• Dufton:2019n
  • Extend the Evans:2006z by adding new targets
  • 47 O-type and 287 B-type spectra (still not complete?)
  • Their NGC 346-1001 is SSN 7/MPG 435 - the big one
• Niemela:1986k
  • Spectroscopy of the brighter stars

[#A] Ionizing luminosities and fluxes

Which are the most important stars for mYSO C?

Name	SSN	MPG	Sp. Type	T_eff, kK	V	BC	M_bol	L/Lsun	R/Rsun	log q_0	Q_0	d (“)	F_bol	L_glob/L_sun	Q_0(glob)
W 1	7	435	O4If+O5-6	43	12.6	-3.98	-10.63	1.4e6	21.29	24.46	8.0e49	11.56	3.72	2.6e3	1.5e47
W 2	14*	470	O8.5 III	36	14.1	-3.45	-8.60	2.1e5	11.76	24.04	9.2e48	3.04	8.06	5.7e3	2.5e47
W 3	9	355	O2 III(f)	52.5	13.45	-4.58	-10.38	1.1e6	12.66	24.9	7.7e49	22.68	0.76	5.3e2	3.7e46
SSN 168	168		B0?	30	16.7 - 1	-2.90	-6.45	3.0e4	6.40	23.2	4.0e47	0.2	266.05	0	0
Sk 80	5	789	O7 Iaf+	36	12.31	-3.45	-10.39	1.1e6	26.92	24.07	5.2e49	122.4	0.03	1.8e1	8.7e44
HD 5980		755	WR+WR+O	43	11.6 + 0.75	-3.98	-10.88	1.8e6	24.14	X	1e50	105.7	0.06	4.0e1	2.2e45
W 4	11	342	O5-6 V((f))	43	13.6	-3.98	-9.63	5.5e5	13.34	24.47	3.2e49	24.85	0.32	2.2e2	1.3e46
SSN 13	13	324	O4 V((f+))	48	14.13	-4.31	-9.43	4.6e5	9.79	24.8	3.7e49	36.99	0.12	8.4e1	6.8e45
SSN 15	15	368	O6: V((f))	42.5	14.1	-3.95	-9.10	3.4e5	10.74	24.4	1.8e49	17.15	0.41	2.9e2	1.5e46
SSN 17	17	396	O7V	38	14.383	-3.61	-8.48	1.9e5	10.04	24.15	8.7e48	11.53	0.51	3.6e2	1.6e46
SSN 18	18	487	O8V	36	14.47	-3.45	-8.23	1.5e5	9.94	23.94	5.2e48	8.37	0.76	5.4e2	1.9e46
SSN 46	46	500	O6V:	42.5	15.267	-3.95	-7.93	1.2e5	6.38	24.4	6.2e48	16.39	0.16	1.1e2	5.8e45
SSN 22	22	476	O7?V?	38	14.562	-3.61	-8.30	1.6e5	9.22	24.15	7.3e48	4.06	3.44	2.4e3	1.1e47
SSN 26	26	655	OC5-6Vz	43	14.90	-3.98	-8.33	1.7e5	7.42	24.47	9.9e48	60	0.02	1.2e1	6.9e44
SSN 37	37	593	O5.5V	43	15.07	-3.98	-8.16	1.4e5	6.73	24.47	8.1e48	40	0.03	2.2e1	1.3e45
						X	X	7.6e6	X	X	4.3e50	X	X	1.3e4	6.3e47

• M_bol = V + BC - AV - DM
  • DM = 18.95
  • AV = 0.3 approx (assuming E(B-V) = 0.1)
  • I have added an extra 1 mag extinction by hand to SSN 168
• L/Lsun = 10**(0.4 (M_sun - M_bol))
  • M_sun = 4.73
• R = sqrt(L / 4 π σ T^4)
• So in terms of bolometric flux, the nearby B star totally dominates
• But we should work out the external luminosity incident on the globule
  • And not include SSN 168 since that is on the inside
  • Sum over stars of luminosity x area covering fraction: ∑_i L_i Ω_i / 4π
  • Ω_i / 4π = A_glob / 4π d^2 = π r_glob^2 / 4π d^2 = r_glob^2 / 4 d^2
  • Take r_glob = 1 arcsec
• The result is a total of 1.3e4 Lsun, of which half comes from W 2
• So this is less than the inner luminosity from SSN 168 (3e4), if that was all trapped
  • Also, the external luminosities are upper limits, assuming that everything is in the plane of the sky
• I have done the same for the ionizing luminosity
  • I get a total external incident luminosity of Q_ext = 6.3e47 s^-1
  • As opposed to an internal luminosity from the B0V star of 4e47 s^-1
  • The external incident luminosity is split up as follows
    • 40% from W 2
    • 24% from W 1
    • 17% from SSN 22
    • 6% from W 3
    • 13% from the rest (each 3% or less)
• [ ] Now, we need to compare this with the Ha luminosity of the globule and of the inner sources

Relative contributions to the global ionization balance

• The list above has all the stars with Q_0 > 5e48 I think
  • Except for some more that a long way to S or N
• Total luminosity is 4.3e50
  • Could be 5.3e50 if we add in both WR stars as equals
  • Lopez:2014a give 50.6 -> 4e50
    • This is based on a simple conversion of the Hα luminosity of log L(Ha) = 38.6
    • [ ] We should compare that with the flux measured in our MUSE field
• Recombination time:
  • Dense gas:
    • n_e = 100 pcc
    • α ≈ 2e-13
    • 1 / α n = 1600 years
  • He II gas
    • n_e = 11 pcc
    • α = 1.5e-12
    • 1 / α n = 1920 years
  • So almost the same! The lower density is compensated by a higher recombination rate
  • So recomb timescale of a few thousand years is small compared with stellar evolution timescale, even for WR stars

Conversion of spectral type to ionizing luminosity

• Galactic case is done by Martins:2005a
  • Application to Carina in Smith:2006a
  • But what about the SMC abundances?

Spectral type to effective temperature

• Massey:2005z show the S. Type - T_eff relations for MW vs SMC file:ngc-346_att/screenshot-20210527-134032.png
  • They find that SMC stars a bit hotter (green line) for both supergiants and dwarfs
  • I have inserted these temperatures as a column in the above table
• [ ] But what about bespoke stellar parameter determinations?
  • Massey:2000l have
    • Q = (U - B) - 0.72 (B - V)
    • This is supposedly a reddening-free index
    • Then
      • For Q < -0.6 and either (B - V)_0 < 0.0 or (U - B)_0 < -0.6,
      • log Teff = 4.2622 + 0.64525 Q + 1.09174 Q^2, (V)
      • log Teff = 5.2618 + 3.42004 Q + 2.93489 Q^2, (III)
      • log Teff = -0.9894 - 22.767380 Q - 33.09637 Q^2 16.19307 Q^3 (I)
  • However, I can’t be bothered to look up all the UBV photometry necessary to do this

Luminosities, radii

• We need the gravity to get the q_1 ionizing flux from the stellar atmosphere
  • function of T_eff, g, Z
• But then we need the R to convert to luminosity Q_1
• But we also have L_bol = 4 π R^2 T_eff^4
  • And we can find L_bol from M_V and bolometric correction
  • BC is known as a function of T_eff
    • E.g., BC = -6.90 log Teff + 27.99
    • Massey:2005z, equation (2)
    • From this, we can calculate the radii from M_V
      • If we know A_V
• This is done in the table

Masses, gravity

• [ ] The gravity should have a calibration with luminosity class

Ionizing fluxes

• If the fluxes q_0 are not a strong function of g, then we don’t need to find the gravity
• We can use the same q_0 as in Martins:2005a, but using our T_eff, rather than their Galactic calibration
• Apparently, the direct effect of metallicity on q_0 is small, but q_1 should be higher at lower metallicities
  • Quote regarding this from Martins:2005a section 6:

    From the modelling side, Kudritzki (2002) and Mokiem et al. (2004)
    have investigated the effect of a change of the metal content on the
    spectral energy distribution of O dwarfs using CMFGEN models. They
    found that H ionising fluxes are essentially unchanged when Z is
    varied between twice and one tenth the solar content. They argue that
    the redistribution of the flux blocked by metals at short wavelengths
    takes place within the Lyman continuum, which explains the observed
    behaviour. However, they show that the SEDs are strongly modified
    below ~450 \AA, spectra being softer at higher metallicity (see also
    Sect. 5.3.1). Morisset et al. (2004) have computed various WM-BASIC
    models at different metallicities and showed how Z affected the
    strength of mid-IR nebular lines emitted in compact H I regions. The
    softening of the SEDs when metallicity increases is crucial to
    understand the behaviour of observed excitation sequences.
            

Helium-plus ionizing luminosities

• This is important for the bow shock ionization
• We are also interested in the ionization of Ar III -> Ar IV
  • This requires 40.735 eV, compared with 35.121 eV for O III and 54.418 eV for He II
  • Ar IV can persist in the He II zone since Ar IV -> V requires 59.58 eV

Consistency check on the ionization fraction and thickness of the He II zone

• Based on the notebook ~/Dropbox/muse-hii-regions/notebooks/01-02-yet-more-line-ratios.md
• Updated [2024-04-14 Sun] with new concordance solutions for He++ ionization
• There I find that Q_2 = (6.8 +/- 3.4)e45 /s
  • Previous value of 2.8e45 /s was not taking into account extinction and assuming a larger covering fraction
  • Even more previous value of Q_2 = 4.7e45 /s was not using the right temperature
• And I also found that n_H = (35 +/- 10) cm^-3 is the hydrogen nucleon density
  • The He++ fraction is measured to be x_He++ = 0.1
    • This comes from the value of about \langle x_He++ ⟩ = 0.016 for the entire line of sight, but then taking into account the He++ shell being half of the H+ shell, which is in turn 30% of the total emission measure
  • The electron density is n_e = x_e n_H where x_e = 1 + y (1 + x_He++) assuming no neutral helium
    • so n_e = 1.09 n_H
• The distance of the inner edge from the star is 4 arcsec = 4 9.23e+17 cm = 1.2 pc
• So the ionizing flux is F_2 = Q_2 / 4 π R^2 = 3.95e07 phot/s/cm^2
• This should be equal to n_e n_i α_B h = y_He x_e x_He++ n^2 α_B h
  • Assuming y_He = 0.082 and α_B = 1.2e-12 cm^3/s
  • y_He x_e x_He++ = 0.082 1.09 0.1 = 0.009
  • h = F / y_He x_e x_He++ n^2 α_B = 3.95e07 / 0.009 35**2 1.2e-12 = 3.0e18 cm = 0.97 pc
    • 1 arcsec = 0.3 pc
    • So this is 3 arcsec - marginally OK

Estimating He++ ionization fraction

• Now that we have the F_2 flux determined we can estimate the He++ x from local ionization balance:
  • In principle, this will be consistent with the value that we already determined
  • (1 - x) y n σ F_2 = x y x_e n^2 α
  • x / (1 - x) = σ F_2 / 1.09 α n
  • We have all these ingredients except for σ, which we can easily find since it is hydrogenic
  • Turns out to be about 1.6e-18 cm^2 at threshold
  • Therefore, we can find
    • σ F_2 / α n = 1.6e-18 3.95e+07 / 1.2e-12 35 = 1.5
    • x / (1 - x) = 1.5 => x = 0.63
  • This is not consistent with the estimate from the He II / He I ratio
  • Is this consistent with the Cloudy models?
    • Yes, more or less - but it requires the density to fall with radius

Luminosities from POWR models

from pathlib import Path

datapath = Path("../stars/powr/colors")
datafiles = datapath.glob("*_colors.txt")

rslt = [
    ["Grid", "Model", "Q0", "Q1", "Q2", "Q3"],
    None,
]

for datafile in datafiles:
    with open(datafile) as f:
        lines = f.readlines()

    _, grid_id, _, model_id = lines[0].split()
    # print(grid_id, model_id)
    _, _, _, Q0 = lines[6].split()
    _, _, _, Q1 = lines[7].split()
    _, _, _, Q2 = lines[8].split()
    _, _, _, Q3 = lines[9].split()

    rslt.append([grid_id, model_id, Q0, Q1, Q2, Q3])

print(rslt)

Grid	Model	Q0	Q1	Q2	Q3
OB-I	54-42	50.00	49.52	45.79	48.85
OB-I	50-42	49.66	49.08	45.49	48.37
SMC-OB-II	49-44	49.25	48.72	45.02	48.12
SMC-OB-I	50-44	49.32	48.82	44.80	48.24
SMC-OB-III	50-42	49.59	49.05	44.63	48.34
OB-I	53-42	49.92	49.41	45.53	48.74
OB-I	50-40	50.09	49.52	45.37	48.80
SMC-OB-III	49-42	49.52	48.96	44.68	48.24
SMC-OB-II	50-44	49.33	48.82	45.28	48.23
SMC-OB-I	49-44	49.24	48.73	44.36	48.14
OB-I	49-42	49.57	48.97	44.79	48.24
SMC-OB-III	49-44	49.26	48.70	44.65	48.01
SMC-OB-I	36-38	48.82	47.81	42.06	46.54
SMC-OB-I	49-42	49.51	48.99	44.65	48.38
SMC-OB-II	50-42	49.58	49.07	45.52	48.46
SMC-OB-II	42-38	49.57	48.85	41.43	48.09
SMC-OB-I	50-42	49.58	49.07	45.19	48.47
SMC-OB-II	49-42	49.51	48.98	45.32	48.36
OB-I	52-42	49.81	49.34	45.93	48.72
OB-I	56-42	50.08	49.67	46.99	49.10
OB-I	51-42	49.75	49.20	45.23	48.50
SMC-OB-III	50-44	49.33	48.79	44.75	48.10

Rearrange and rescale the models

• The Q0’ (H0) and Q2’ (He+) columns are ionizing luminosities rescaled so that the model has a bolometric luminosity of log L = 5.98, as implied by the cooler Rivero-Gonzalez solution
  • [X] We should also rescale the Mdot values - now done in the final column
  • All these models now give Q_0 = 49.80 +/- 0.01
• The observationally derived Q2 is 45.83 +/- 0.22 ((6.8 +/- 3.4)e45)
  • [2024-04-12 Fri] Updated to slightly higher value
  • [2024-04-14 Sun] Updated to significantly higher value with realistic uncertainties
  • For T = 50,000 K only the strong wind model SMC-OB-II (strong wind) 50-42 model is consistent with the observed Q2 value
    • log Mdot = -5.66=> Mdot = 2e-6
    • The weak wind model has 10x too small a value of Q2, and intermediate wind model has 2.5x too small a value
  • The higher gravity model (50-44) is similiar in that only the stronger wind model works
    • log Mdot = -5.66 => Mdot = 2e-6
  • The lower temperature model (49-42) only marginally consistent with observed Q2 for the strongest wind (SMC-OB-II: Q2 = 45.60), which implies Mdot = 2.34e-6
• In summary, the He++ ionizing luminosity is a strong function of Teff and Mdot, but very weak gravity dependence
  • Acceptable solutions require Mdot >= 2e-6 and tendo to favor 50 kK over 49 kK
  • Is there any observational constraint on the mass loss rate?
  • [X] But we need to have error bars on our value for Q2. This is now done.

Grid	Model	log L	V_inf	log M_dot	Q0	Q1	Q2	Q3	Q1/Q0	1e4 Q2/Q0	Q3/Q0	Q0’	Q2’	Mdot’
OB-I	56-42	6.230	4136	-7	50.08	49.67	46.99	49.10	0.389	8.128	0.105	49.83	46.74	-7.44
OB-I	54-42	6.150	4077	-7	50.00	49.52	45.79	48.85	0.331	0.617	0.071	49.83	45.62	-7.37
OB-I	53-42	6.070	4001	-7	49.92	49.41	45.53	48.74	0.309	0.407	0.066	49.83	45.44	-7.30
OB-I	52-42	5.980	3898	-7	49.81	49.34	45.93	48.72	0.339	1.318	0.081	49.81	45.93	-7.22
OB-I	51-42	5.910	3835	-7	49.75	49.20	45.23	48.50	0.282	0.302	0.056	49.82	45.3	-7.15
SMC-OB-II	50-44	5.520	3311	-5.891	49.33	48.82	45.28	48.23	0.309	0.891	0.079	49.79	45.74	-5.66
SMC-OB-I	50-44	5.520	3311	-6.891	49.32	48.82	44.80	48.24	0.316	0.302	0.083	49.78	45.26	-6.66
SMC-OB-III	50-44	5.52	3311	-7.891	49.33	48.79	44.75	48.10	0.288	0.263	0.059	49.79	45.21	-7.66
SMC-OB-II	50-42	5.760	3020	-5.771	49.58	49.07	45.52	48.46	0.309	0.871	0.076	49.8	45.74	-5.66
OB-I	50-42	5.830	3800	-7	49.66	49.08	45.49	48.37	0.263	0.676	0.051	49.81	45.64	-7.09
SMC-OB-I	50-42	5.76	3020	-6.771	49.58	49.07	45.19	48.47	0.309	0.407	0.078	49.8	45.41	-6.66
SMC-OB-III	50-42	5.76	3020	-7.771	49.59	49.05	44.63	48.34	0.288	0.110	0.056	49.81	44.85	-7.66
OB-I	50-40	6.240	3578	-7	50.09	49.52	45.37	48.80	0.269	0.191	0.051	49.83	45.11	-7.35
SMC-OB-II	49-44	5.450	3234	-5.933	49.25	48.72	45.02	48.12	0.295	0.589	0.074	49.78	45.55	-5.63
SMC-OB-I	49-44	5.450	3234	-6.933	49.24	48.73	44.36	48.14	0.309	0.132	0.079	49.77	44.89	-6.63
SMC-OB-III	49-44	5.450	3234	-7.933	49.26	48.70	44.65	48.01	0.275	0.245	0.056	49.79	45.18	-7.63
SMC-OB-II	49-42	5.700	2978	-5.799	49.51	48.98	45.32	48.36	0.295	0.646	0.071	49.79	45.6	-5.63
OB-I	49-42	5.750	3695	-7	49.57	48.97	44.79	48.24	0.251	0.166	0.047	49.8	45.02	-7.01
SMC-OB-I	49-42	5.700	2978	-6.799	49.51	48.99	44.65	48.38	0.302	0.138	0.074	49.79	44.93	-6.63
SMC-OB-III	49-42	5.700	2978	-7.799	49.52	48.96	44.68	48.24	0.275	0.145	0.052	49.8	44.96	-7.63
SMC-OB-I	36-38	5.260	1979	-6.993	48.82	47.81	42.06	46.54	0.098	0.002	0.005	49.54	42.78	-6.23
SMC-OB-II	42-38	5.800	2316	-5.687	49.57	48.85	41.43	48.09	0.191	7.24e-5	0.033	49.75	41.61	-5.43

Correction for radius

• The POWR models are a bit sub-luminous compared with W3
  • SMC-OB-I 50-42 has log L = 5.76
  • W3 has log L = 5.98 (cooler) to 6.08 (hotter)
  • Take the cooler one to start with, for consistency with the POWR Teff
  • So to rescale all the luminosities, we would add 0.22 in log space
  • Each model has a different nominal luminosity, so I am going to have to add in
• Scaling of Mdot when luminosity is adjusted
  • Q = Mdot D^1/2 (Rstar Vinf)^-3/2
    • where D is the clumping factor (D = 1/f) at 10 R_*
      • clumping increases gradually from sonic point to 10 R_*
    • all models assume D = 10
  • Models are with log Q = -13.0, -12.0, -14.0 for SMC-OB-I, II, III, respectively
    • Although in some places it says -13.5, -12.5, -14.5
    • But that is using a different definition of Q that does not include the sqrt(D) factor
  • Work through for SMC-OB-II 50-42
    • R_STAR = 10.137 R_SUN
    • V = 3020 km/s
    • log M_dot = -5.771
    • D = 10
    • => log Q = log10(sqrt(10) 10**-5.771 (10.137 3020)**(-3/2)) = -12.0
  • Therefore, we should have Mdot scaling as Q R^3/2 or Q L^3/4
    • This is now implemented in the final column of the above table

Correction for wind velocity

• From the previous section, we also have Mdot scaling as Q Vinf^3/2
• We have a measured Vinf for W 3: 2800 km/s
• So we should do an additional rescaling of Mdot by (2800 / Vinf)^3/2, where Vinf is the value from the POWR model
• This is now also implemented in the final column

What is this “excitation parameter” that the early papers mention?

• It has units of pc cm^-2, which is length

New paper [2022-07-24 Sun] on NGC 346 O star winds

• Rickard:2022z

First few rows of their results table

SSN	MPG	V sin i	T.	log g	log L	E(B-V)	§min	log M	Voo	D	B	Comment
		(km s-1)	(kK)	(cgs)	(Lo)	(mag)	(kms-1)	(Moyr-‘)	(km s-)
9	355	130(2)	51.7	4.0	6.12	0.13	35	-6.65	2800	200	0.8	MSB
13	324	≥113(2)	42.0	3.8	5.60	0.11	15(1)	-7.4	≥2300	20	1.0	SB1(2),MSB
14	470	145	37.0	4.4	5.37	0.16	5	-9.5	≥600	20	1.0	MSB
15	368	58(2)	39.0	4.0	5.45	0.12	15(1)	-7.6	≥2100	20	1.0	SB1(2),MSB
17	396	196	37.0	4.0	5.30	0.11	10	-8.7	≥1000	1	0.8	MSB

Conclusions for Walborn 3

• log L = 6.12
• T = 51.7 kK
• Mdot = -6.65, so 2.239e-07 Msun/yr, which is very low for a supergiant
• This is more consistent with the hotter solution, whereas I was using the cooler one

What to call the stars?

• My current plan is just to use the Sabbi ID numbers, which are in order of V brightness
  • SIMBAD says:

    dic: Table 2: <Cl* NGC 346 SSN NNNNNN> N=79960 among (Nos 1-132733). Table 3: <[SSN2007] Sc NN> (Nos 1-16).
            

  • So I can just abbreviate it as SSN 152, SSN 168, etc
• Alternative lists:
  • MPG gets used a lot

Different components of mYSO C

• There are two continuum sources, presumably stars
  • Separated by about 0.3 arcsec
  • Sabbi ID 152 and 168
    • MPG 454 corresponds to both of them
  • The NE source
    • Sabbi ID 152
• Other correspondences
  • SSTS3MC 14.7725-72.1766 from Simon:2007r
    • They estimate 3.4e4 Lsun and 15 Msun
    • Their source includes both stars
  • Contursi:2000f give spectrum (their Peak C in Fig 6)
    • They estimate G_0 = 8.7e5
    • [ ] We can use this to get a distance from the star maybe
  • Whelan:2013d source PS 9
    • Spectrum is different from other sources
    • Prominent Silicate emission
    • Strangely low PAH 7.7/11.3 ratio although normal 6.2/11.3 ratio
  • Sewio:2013f
    • Source Y535
  • Rubio:2018f have the most comprehensive study
    • They have near-IR imaging and spectra
    • And they compare with the literature measurements

SED modelling of the YSOs

• Robitaille:2006f is what everyone uses
• Although Zhang:2018k claim their models are more physically consistent

Massive YSOs in other regions

• van-Gelder:2020z do X-shooter spectroscopy of sources in the LMC
  • Their Fig 13 shows Brγ luminosity vs K-band magnitude for a whole bunch of sources from MW, SMC, and LMC:
  • The ones that are way above the trend line might be good candidates for looking for Raman scattering
  • The SMC sources are in this category: relatively bright in Brγ
    • For Source C, observed K_S = 12.05, meaning M_K = 12.05 - 18.95 = -6.9
    • This means it must be the rightmost orange square, so L(Brγ) = 30 L_o, which corresponds to Q(H) = 8.554e+48 /s (see conversion factor from below)
    • So that would be like an O7V star if 100% covering factor
    • [0/2] But some questions remain?
      • [ ] How is Brγ flux apportioned between different components (two compact sources plus surrounding globule)?
        • This could be addressed by using the Hα fluxes from HST and/or MUSE
      • [ ] How feasible is external, rather than internal illumination?
        • For the globule it must be external because of the ionization gradient (seen in [O III] compared with Hα for instance)
  • Conversion of Brγ to other measures:
    • Brγ is 7 → 4 transition
    • Equivalent Hβ
      • Table 4.2 of Osterbrock book has j(Brγ) / j(Hβ) = 0.0281
        • For T = 1e4 K in low-density limit
      • So 1 L_o of Brγ means 35.6 L_o of Hβ
    • Equivalent Hα
      • j(Hα) / j(Hβ) = 2.87
      • => 1 L_o of Brγ = 102.1 L_o of Hα
      • That is worth remembering: Hα / Brγ = 100 more or less
    • Equivalent Q(H)
      • L(Hβ) = e(Hβ) α(Hβ) VEM
      • Q(H) = α_B VEM
      • => Q(H) = [α_B / e(Hβ) α(Hβ)] L(Hβ)
      • α_B = 2.6e-13 cm^3/s
      • e(Hβ) α(Hβ) = 1.24e-25 erg cm^3/s
        • (This is 4π j(Hβ) / n_e n_p in the Osterbrock table)
      • => Q(H) = (2.6e-13 / 1.24e-25) 35.6 3.82e33 [L(Brγ) / L_o]
      • => Q(H) = 2.85e+47 [L(Brγ) / L_o]
      • Remember, this is assuming 100% area covering fraction of ionization-bounded matter (no escape) and no EUV absorbed by dust.
        • So it is a lower limit on Q(H)
    • So th1C would be way above the line, at M_K, L(Brγ) = -3.6, 30
      • but that is assuming all ionizing photons trapped nearby
      • In reality, Orion S has much lower flux than this would imply
      • [ ] need to calculate Hα flux from Orion S

Jets in the core of NGC 346

• Possible jets seen in H alpha ACS image
  • Note: I used the standard org-attach method to attach this screenshot, which I made with CleanShot because I wanted to annotate it
  • This has the disadvantage that the link is not a normal image link
• [ ] Look at evidence from the MUSE spectra for high-velocity Hα
  • I think there was some indication to the N of the embedded sources
• [ ] Do a cross-reference with the V and I-band images to check they aren’t just stars
• These are smaller than the jet in McLeod:2018a from LMC N180
  • That has total length of 11 pc (46 arcsec)
  • Ours have lengths of about 0.3 arcsec if they are real

Comparison of globules with those in other regions

• [2023-05-16 Tue] Return to this issue to get some facts to mention in the DRL paper
• Compare sizes, densities, etc. with other samples such as
  • NGC 346 globules
    • Small one around source C
      • R ≈ 0.3 pc
    • Larger ones have R up to about 0.5 pc, but there are many smaller ones too
    • How close are they to the core of the cluster?
      • MUSE field has a radius of about 15 pc
      • Closest “normal” globules to cluster center have a projected distance of about 1 pc from center of cluster, which has W 2 (O 8.5)
      • Globule surrounding Source C has separation of about 0.5 pc
      • W 1 (O4) is projected right on top of the upper filament, but it must be a way in the foreground since the globules do not point towards it
      • Brightest rims are seen in ones off to E (in Sc 2) with a separation of about 3 pc, so those are probably the physically closest (apart from MYSO C globule)
      • Conclusion is distance of 3 to 15 pc for most, but 0.5 pc for the one around C
      • [ ] Estimate incident flux from density and size
    • Densities in nebula are about 100 pcc from the [S II] data
  • Orion
    • Orion proplyds
    • Orion globules
      • (the far out ones)
      • and the ones in the Dark Bay
    • Orion bar - not really the same,
    • Orion S considered as a globule
      • SW minibar has R = 0.03 pc taking the long axis (much smaller if we take the short axis)
  • Eagle Nebula EGGs
    • McCaughrean:2002a
    • Hester:1996w
      • R = 300 to 1000 AU = 1.5 to 5 mpc
      • Much smaller than in other regions
  • Carina
    • Sahai:2012b - coins frEGG term
    • Menon:2021j
    • Hartigan:2015a
      • Southern pillars typical size = 0.1 pc
      • Some larger globules, for instance Figs 14, 15: Western Wall, with R = 0.4 pc
      • Some much smaller ones, especially in Tr 14 region
    • McLeod:2016a
      • MUSE maps of many globules in multiple emission lines
    • Reiter:2019a
      • Tadpole - small: 0.01 pc
  • frEGGS in Cygnus OB2
    • R = 0.05 to 0.1 pc

Analysis of Raman wings of SMC N66 mYSO C

• Optically thick out past 100 Å
  • Must have column of > 1e22 H^0 / cm^2
  • Suggests low dust cross section

What we need for a paper

• Spectrum of Raman wings
  • Obviously Ha
  • But also look at Hb too
• Deal with the sky problems
  • Nebular over-subtraction
    • Puts the emission lines in absorption in the faint regions
  • Telluric over-division
    • Puts telluric absorption bands in emission in spectra of bright stars
• Spatial profiles in wings, continuum, Ha, [S II], [O III]
  • When is it spatially resolved and when not?
• Look at more lines
  • [S III] and [Ar III] to see if they look more like Ha (central peak) or like [O III] (hole in middle)
  • Extinction from Ha/Hb - does the core of the line have nebular extinction (very small) or stellar extinction (A_V = 2.5)?
• Density/temperature diagnostics
  • We have an [S II] density, but in the presence of gradients/inhomogeneities that will be biased towards low densities
  • Can we get any [Fe III] densities?
    • Mabel used 4986 / 4658
• Kinematics
  • Local maximum in sigma(Ha) - but is that the line core, or the effect of the Raman wings?
• Analysis
  • Optical depth of Raman scattering
    • Does it require dust destruction?
  • General thoughts on the young cluster
    • many of the stars seem to be aligned in chains
    • there is a 3-pronged “bird foot” pattern coming out of the brightest compact sub-cluster
    • the chains are approximately aligned with the filaments seen in the gas

Going through the lines we need

• I am currently doing these for the sharper cube
  • ADP.2016-10-12T05_43_23.882.fits

Stars of moderate interest

This is a list of moderately interesting stars that are showing up in my line analysis

SSN 258

• This is the slightly fainter star below SSN 62 in the globule just N of our main source
• It is not in SIMBAD but has V = 17.26 (M_V = -1.69) and V-I = 0.167
  • That puts it around B4 spectral type according to Fig 5 of Martayan:2010j
• It has strong [N I] 5200 line emission
  • Possible spatially extended, but not much
• Also shows Si II 5056

SSN 552

• This one is just outside the cluster box in the star figures
• In a globule
• It is quite faint and red
  • V = 18.217, V-I = 0.616
• Shows the Si II 5056, 5041 line but not much else

SSN 73

• Simbad link shows it as “emission line star”, B1
  • Probably a Be star, although unclear whether pre-MS or post-MS
  • See Martayan-2010j
  • Also Paul:2017t
• Has lots of lines that show up in the pass band around 6300 Å
• The above spectrum shows the globule close to our main source (SSN 152+168) in orange and SSN 73 in yellow
• In addition to the usual [O I], [S III], Si II, we have lines emission lines at 6247, 6317, 6383
  • Some of these are also seen weekly in our main compact source

[O I] 6300, 6363 (+ [S III], Si II)

• This is done in 03-03-oi-lines-sharp-cube.md
• As well as the [O I] lines, we have [S III] 6312 and Si II 6347 and 6371
  • The Si II lines come from compact sources only

[Ar IV] 4711, 4740

• The 4711 is blended with He I 4713, but we can see the effect of both
• the 4740 line is great - it is very diffuse, but has an inner hole the same as [O III]
  • This could correspond to the inner wind bubble

He II 4686

• Another great line - it shows a beautiful emission arc just inside the [Ar IV] emission
• We can calculate the flux and then convert to a > 4 Ryd luminosity and see if this is consistent with expected stellar EUV
• The arc shows no extinction by the globule filaments, and so must be foreground to them
• I need to do this for the summed cube too to increase signal-noise

Larger field images from MCELS

Other things in the FORS1 spectra

High velocity knots

• There are several of these
• Most are redshifted by 100 to 150 km/s
• But some are blueshifted too
• Mostly visible in low-ionization lines such as [S II]
  • In some cases the high-velocity component is as bright as the nebula
• Probably associated SNR, which may or may not be related to WR binary HD 5980
• This has already been well-observed
  • Danforth:2003m is the most complete study
  • Strangely, though they claim there is no low-ionization emission
  • Whereas we see it clearly in [S II]