MUSE cubes of 30 Dor (Tarantula) in the LMC

Collaboration of Will Henney with Mabel Valerdi
Initial plan [2021-06-16 Wed]
- Mabel has reported difficulty with the first stages of data reduction:
  - She tried to follow similar steps to those in NGC 346 notebook but it was taking unreasonable amount of time (hours?)
  - This may be due to low RAM on her machine (8GB versus 64GB on mine)
- So Will is going to carry out the initial steps and produce smaller files that Mabel’s machine will be able to manage:
  1. [X] Download the data cubes from ESO archive for the 4 fields of 30Dor (A to D)
    - Note that the data cubes will not be stored in this repo (they are too big!). But we should store the download shell scripts to make this step reproducible
  2. [X] Fit and remove continuum pixel-by-pixel
  3. [X] Divide the wavelength range into chunks of about 800 Å and write out smaller cubes
    - For instance, the range 6100 → 6900 is more than enough for analyzing the Raman wings of Hα
    - The full range is 4700 Å and the data cubes are about 3GB, so our smaller files should be about 6 times smaller, or 0.5GB

What to put in a paper [2022-04-21 Thu]

Case study of neutral gas associated with 3 molecular clouds in the center of Tarantula Nebula
Massive YSOs could feature in it
- Except it turns out that the known YSO is not the source of the [Fe III] jet :(
- Our main YSO with Raman emission is associated with MC 10 molecular cloud
  - Source 7B from van Gelder 2020
  - Except that the Raman emission does not peak at this source either!
  - Looks like Raman emission is just form the molecular cloud, not from any particular YSO
- The jet source is just S of there in a region without much molecular gas
Mystery molecular lines
- 9114 and 8152 angstrom
- both have similar spatial distribution, which seems deeper than [C I] 8727 in some respects, but weirdly different
  - see the rgb map in the notebook

Original files

I downloaded what looked to be the combined cubes for each of the 4 positions (with overview images)

A
ADP.2016-07-14T14:17:17.826.fits

B
ADP.2016-07-19T06:16:41.316.fits

C
ADP.2016-07-19T11:27:30.987.fits

D
ADP.2016-07-15T13:22:09.432.fits
Layout of the fields in RA, Dec is
```
D C
B A 
    
```
Will’s local copy of the files is at
- /Users/will/Work/Muse-Hii-Data/30Dor/2021-06/
Overview images and log files are copied to the data folder of this repo
- ../data/lmc-30dor/
- The PNG files are white light images of each field, together with

Cuts through interesting regions

The 30 Dor-10 GMC in the N
- This has ver small offset between ionized and neutral species
The big clump to the SW
- This has a nice wall that can be used for a cut through the perpendicular ionization front, akin to in the Orion Bar
The triangle cloud
- Strong, limb-brightened C I 8727 emission
The close-in extinction cloud
- Bright in the mystery 8152 line

[Fe III] lines and a possible jet

Some notes in this Craft doc
- Open locally in Craft
The jet seems to be coming from the same source as the YSO that has the peak Raman lines

The Hβ Raman wing

This is seen best on the blue side
Here are some example profiles
Note the absorption notch in the profile at observed wavelength of 4845 +/- 0.5
- This is quite possibly another UV absorption line
- [ ] We can try and work out what the UV rest wavelength must be

Scattering and extinction

DIBs

We should measure the 5781 EW maps, and correlate it with the reddening
There is also the DIB in the red Raman wing

Scattered light

We have the broad 4686 line emission (WR blue bump)
- Quote from Lopez-Sanchez:2010y
```
The blend of the broad He Il 14686, CITI/CIV 14650 and NII 14640 emission lines constitutes the blue WR bump; it mainly originates in WN stars with a minor contribution of WC stars.
        
```
- This is seen all over the field - probably various WN stars and maybe contribution from hot O stars too
  - There seems to be scattering associated with the ionized gas and the neutral gas too
  - Very little evidence of limb brightening
We have the 5411 absorption line (albeit potentially contaminated by [Fe III] 5412 emission). Also, it is emission in some cases

We have the WR red bump 5800 broad emission (broad C IV)

Quote from Lopez-Sanchez:2010y

The blend of the C III \lambda 5698 and CIV \lambda 5808 broad emission lines constitutes the red WR bump. CIV \lambda 5808 is the strongest emission line in WC stars, but it is barely seen in WN stars.

This is much more restricted to a nebula around the WR sat the top. Presumably this is a WC star

We have the WR 6680 broad emission Not sure what makes this: He I maybe?

Reddening

We have multiple reddening indicators:
- H I 6563/4861 gives red–blue
- H I 9229/6563 gives IR-red
- [Ar III] 7751/7136 gives local red slope
  - [ ] Note that this ratio could be improved by doing a better sky subtraction
    - Or at least making it consistent between the 4 different sub-fields
    - This would help eliminate a jump in some of the diagnostic plots
There is a clump at the bottom that shows anomalously low IR-red reddening

Summary of results from line extraction and mosaicking

[2021-08-03 Tue] Added section on O++ lines
[2022-04-22 Fri] Added section on neutral lines

O++ ORLs vs CELs

We find that the ORL/CEL ratio increases towards the center of cluster and also towards the WR star
[X] Should look at correlation with [Ar IV] / [O III] and with [O III] / [O II], since they should be much better measures of excitation
- Yes this hows a good correlation, especially with [Ar IV] / [O III] - see notebook

Mystery neutral lines

Nominal wavelengths are 8152 and 9114
Potential line IDs [2022-04-22 Fri]
- Did an ADS search on 9114 and got hit on Jorge García thesis
- This has the following IDs
  - 8150.57 Si I
  - 9094.83 C I
  - 9111.81 C I
  - 9113.70 Cl II[]
  - 9123.60 [C1 II]
- So the Si I line looks very likely for 8152
  - I had actually been thinking of Si I earlier
- The other lines need more thinking about for 9114
  - On the face of it Cl II seems very unlikely
  - C I looks more promising, but we have to see if we see the other component at 9095, which in principle should be stronger
  - Looks like I had identified a Ca I 9095 line

Sky emission and absorption spectrum

Absorption lines and DIBs

The DIBs are tracing the foreground absorption of something, presumably small molecules
- They can be correlated with the reddening for instance
- There are 3 possible DIBs that I have spotted so far
  - The classic 5781 DIB, which gets to an absorption depth of about 0.1
    - This seems to be slightly correlated with reddening
      - At least I can see the pattern of the stapler cloud in panel
  - There is a weak DIB at 6612 in the Raman wing, close to the absorption feature 6634
  - There is a strange feature at 5900 that seems to be DIB-like
    - We have to be careful with a couple of weak Na I sky lines at 5890 and 5896, but they do not get in the way much
    - Actually it turns out that the absorption line is the Na I line
      - These are the Na I D lines
    - There are N III photospheric lines at 5896.1, 5901.2, 5918.5
      - although that is a very miscellaneous collection of lines
      - If we take the same multiplet then we get 5896.1 5913.5 5954.4
Previous work on this
- van-Loon:2013z studies both DIBs and Na I (and Ca II) in 30 Doradus
  - They have better spectral resolution than we do
  - But they have terrible spatial resolution since they are only looking at stars
  - They separate out the Galactic and LMC contributions to the DIBs
    - We can see this also for DIB 5781, despite our poor spectral resolution
  - For Na I, they find multiple kinematic components

Full spectrum of selected regions

[2022-06-23 Thu] I am losing track of all the lines that we want to study in this project
So I am going to try and make a plot of the full spectrum for selected regions of the nebula
This is similar to what I did previously for the Orion Nebula
- ../../OrionMuse/full-sky-spectrum.pdf
- Which has proven to be very useful in the years since
Steps to carry out:
1. [ ] Define some regions in ds9
2. [ ] Extract the 1D spectra for each region and each wavelength section
3. [ ] Plot the spectra all on one figure
4. [ ] Compare with the Orion spectra

Choose regions for spectrum extraction

I am working with the original panels, starting with panel A, which is SE (lower right)

Location Within Wider Region

This is from the MCELS images

See ~/Work/MCELS/

HST images

These come from http://www.robgendlerastropics.com

Comparison with molecular gas

Alma observations

[2021-11-14 Sun] There are two sets of data that have good resolution (1 arcsec or better) that I have found so far
1. The molecular cloud 30 Dor-10 at the far N edge of our MUSE field
  - covered in 12CO and 13CO among other lines
  - described in Indebetouw:2020x
  - these are the highest resolution observations but they cover only a small area that overlaps with MUSE
  - The integrated and peak maps are in ../big-data/30-Dor-Radio/
    - Alma-2013.1.00346.S-30_doradus_13CO21-peak.fits
    - Alma-2013.1.00346.S-30_doradus_13CO21-sum.fits
2. Three fields that cover the entire area except for the SW corner
  - The data is enormous - too big to put on Dropbox
  - Originals stored locally in /Users/will/Work/Alma-Data/LMC-30-Dor/2019.1.00843.S/
  - No paper published so far
  - The best line I have found so far is in spectral window 25 (spw25)
    - 12CO v=0 2-1
    - Extract from the FITS header:
```
CTYPE3  = 'FREQ    '                                                            
CRVAL3  =   2.302808527260E+11                                                  
CDELT3  =   6.103865753174E+04                                                  
CRPIX3  =   1.000000000000E+00                                                  
CUNIT3  = 'Hz      '
...
RESTFRQ =   2.305380000000E+11 /Rest Frequency (Hz)                             
SPECSYS = 'LSRK    '           /Spectral reference frame                        
ALTRVAL =   3.343952551877E+05 /Alternate frequency reference value             
ALTRPIX =   1.000000000000E+00 /Alternate frequency reference pixel             
VELREF  =                  257 /1 LSR, 2 HEL, 3 OBS, +256 Radio       
                
```
      - Apparently, ALTRVAL is the velocity of pixel 1:
        2.99792458e8 (2.305380000000E+11 - 2.302808527260E+11) / 2.305380000000E+11 = 334395.255188 m
        
        so that checks out. Units are m/s
      - so velocity pixels are 2.99792458e8 6.103865753174E+04 / 2.305380000000E+11 = 79.3748933989 m/s
      - So in principal, we could change the FITS header to have
        CTYPE3 = 'VRAD' CRVAL3 = 3.343952551877E+05 CDELT3 = -79.3748933989 CRPIX3 = 1.000000000000E+00 CUNIT3 = 'm/s'
        
        This says that we have “Radio Velocity”, which is defined in Greisen:2006a as
  - Weaker lines:
    - spw29: 2.203986840000E+11 Hz 13CO v=0 2-1
    - spw31: 2.182221920000E+11 Hz H2CO Formaldehyde - very weak
      - Other H2CO bands are spw33 and spw35 but they are even weaker
    - spw37: 2.195603580000E+11 Hz C180 v=0 2-1 - very weak
    - So for some fields I am only downloading spw25 and spw29 to save time

Making moment maps and channel maps of CO

This is done in the notebook ../notebooks/05-10-LMC-30dor-molecular-maps.ipynb

Combining ALMA maps with lower-resolution observations

This would be good to fill in the noisiest regions of the ALMA map, where there is not much signal
To do this properly is hard
- See Kong:2018a
But we will do it a simpler way:
- Take the low-resolution observations I_A (synthesized beam width W_A)
- And the high-resolution observations I_B (synthesized beam width W_B < W_A)
- Combine as I_A - (I_B * K) + I_B
  - Where * is convolution
  - And K is smoothing kernel of width W = (W_A^2 - W_B^2)^1/2
One possible problem is that the sensitivity of the two instruments could be different
- In which case we would need to apply a scaling factor to one or the other
We have the option of doing it channel-by-channel or of just the integrated map

Maps from Okada:2019r

The mm maps are from APEX
- 12CO 2-1 has HPBW of 27.1 arcsec
- Sampled to 10 arcsec pixels
https://vizier.cds.unistra.fr/viz-bin/VizieR?-source=J/A%2BA/621/A62
- That is where you can download all the maps from

Kinematics of the CO emission

In many areas, there seems to be a pronounced gradient with radius: bluer at small radii, redder at large radii
The velocity range of the CO emission seems similar to that of the optical lines
Point by point comparison between CO and [S II] suggests that the velocities are at least partially correlated.
- But I need to look at this more systematically

Conversion from LSRK to Barycentric

Radio observations tend to use Local Standard of Rest frame LSRK

I have a long discussion of this in ~/Dropbox/KeckProplyd/keck-raman.org

I traced it back to Gordon:1976a - here is a quote:

The conventional reference frame used for galactic studies is essentially
that of standard solar motion. The convention of Local Standard of Rest	    
(LSR) assumes the sun to move at the rounded velocity of 20.0 km/sec toward  
18h RA and 30° DEC (1900.0).

Optical observations tend to use heliocentric or barycentric (almost the same)

Here I will calculate the conversion between the two of them for 30 Dor

from pathlib import Path
from astropy.wcs import WCS
import astropy.coordinates as coord
from astropy.time import Time
import astropy.units as u
from astropy.io import fits
import numpy as np

# Choose any of the MUSE images to get the coordinates from
fitspath = Path.cwd().parent / "data" / "lmc-30dor-ABCD-oiii-4959-bin01-sum.fits"
hdu =  fits.open(fitspath)["DATA"]

# Get coordinates from header and convert to Galactic
w = WCS(hdu.header)
ra, dec = w.wcs.crval
c0 = coord.SkyCoord(ra, dec, unit=u.deg).galactic

# Definition of LSRK velocity in Cartesian Galactic frame (km/s)
U, V, W = 10.27, 15.32, 7.74

# Galactic longitude and latitude in radians
lll, bbb = c0.l.radian, c0.b.radian

# Dot product with unit vector to get the V(LSR) - V(HEL)
vlsr = U*np.cos(lll)*np.cos(bbb) + V*np.sin(lll)*np.cos(bbb) + W*np.sin(bbb)

print(f"V(LSR) - V(HEL) = {vlsr:.2f} km/s")

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

lmc-30dor.org

lmc-30dor.org

MUSE cubes of 30 Dor (Tarantula) in the LMC

What to put in a paper [2022-04-21 Thu]

Original files

Cuts through interesting regions

[Fe III] lines and a possible jet

The Hβ Raman wing

Scattering and extinction

DIBs

Scattered light

Reddening

Summary of results from line extraction and mosaicking

O++ ORLs vs CELs

Mystery neutral lines

Sky emission and absorption spectrum

Absorption lines and DIBs

Full spectrum of selected regions

Choose regions for spectrum extraction

Location Within Wider Region

HST images

Comparison with molecular gas

Alma observations

Making moment maps and channel maps of CO

Combining ALMA maps with lower-resolution observations

Maps from Okada:2019r

Kinematics of the CO emission

Conversion from LSRK to Barycentric

Files

lmc-30dor.org

Latest commit

History

lmc-30dor.org

File metadata and controls

MUSE cubes of 30 Dor (Tarantula) in the LMC

What to put in a paper [2022-04-21 Thu]

Original files

Cuts through interesting regions

[Fe III] lines and a possible jet

The Hβ Raman wing

Scattering and extinction

DIBs

Scattered light

Reddening

Summary of results from line extraction and mosaicking

O++ ORLs vs CELs

Mystery neutral lines

Sky emission and absorption spectrum

Absorption lines and DIBs

Full spectrum of selected regions

Choose regions for spectrum extraction

Location Within Wider Region

HST images

Comparison with molecular gas

Alma observations

Making moment maps and channel maps of CO

Combining ALMA maps with lower-resolution observations

Maps from Okada:2019r

Kinematics of the CO emission

Conversion from LSRK to Barycentric