meeting 2024 12 16 n4874 - JacobPilawa/TriaxSchwarzschild_wiki_5 GitHub Wiki

Context

  • I've gone back to the beginnings of the N4874 work and have continued on. We were still quite early on in the processing of the data, but here are a few reminders/key details:

  • General Information:

    • D = 99.1 +/- 5.8 Mpc (Jensen+21)
    • MASSIVE XVIII:
      • Slow rotator
      • R_e = 33.59" (16.14 kpc), eps = 0.07, K = 8.46 --> M_k = -26.52
      • sigma_e = 258 km/s, sigma_c = 251 km/s
    • MASSIVE XIV:
      • lambda_e = 0.07,
      • PA_{Mitchell} = 335 +/- 6
      • Note that GMOS Kin PA Is not listed and thus no misalignment angle is reported. The paper says that the algorithm used for kin PA determination was inconclusive on 2/20 of the galaxies, and NGC 4874 is one of those
      • PA_{Phot} = 40.6 +/- 5.0. This PA is "taken from NSA"
      • Note that MASSIVE X does claim a large misalignment between phot PA and Mitchell of ~66 deg misaligned

  • GMOS Specific:

    • PSF FWHM = 0.58"
    • Two sets of data:
      • Symmetric w/ 112 bins
      • Asymmetric w/ 110 bins
    • Currently using spectral mask from BOX, but it does look like this could be improved a little bit
    • Information on Angles:
      • In orbit modeling input files: +X = 40 deg. E of N., +Y = 130 deg. E of N.
      • In "paperversion" figures of kinematic moments: +X = -140 deg. E of N., +Y = -50 deg. E of N.
      • For all other files: +X = 40 deg. E of N., +Y = 130 deg. E of N.

  • Mitchell Specific:

    • Data PA = 145 deg E of N
    • 82 bins in the unfolded data which are already processed to nGh = 6. I don't have to touch this data at all.

  • MGE Details:

    • Probably the biggest open question currently.
    • MASSIVE XIV States: "NGC 4874 were not targeted by our program because they have archival HST observations."
      • Looks like this program is this one with PropID = 11711 (Blakeslee).
      • There's quite a bit of archival imaging for NGC4874, with a pretty detailed color analysis in a few different filters. There's two different sets of F110W data which is what I've been using for my other galaxies up to this point.
    • Does this plan sound fine?
      • There are two different images (from two different orbits) of NGC4874 in the HLA. It looks like these data products should be sky-subtracted (check this out for more information on the data products) and ready for MGE fitting. I can try to fit MGEs to the two images and see how the resulting fits compare? Alternatively I could try to combine the two images and fit the resulting combined image. I might try both since I doubt this will change anything.
      • I will likely have to create my own mask for the data. I've never done this before so could use some guidance!

  • Other Measurements:

Plots and Progress

Files on Box

GMOS
  • First, here are some plots from BOX showing (what I think are) the final sets of kinematics showing the Gh6Mh6 kinematics from Irina.
1D 2D Asymmetric Binning 2D Symmetric Binning Replotting the Actual Data
  • There's a note on BOX that I want to paste here concerning the chip gap:
    • Chip gap rescaling was improved on June 1, 2016.
    • Kinematic fits to older spectra (e.g., 20160202) are unaffected, as the chip gaps fall outside the fitting region for pPXF.
    • Why rescale?
    • The science data are staggered between two wavelength settings (separated by 50 Angstroms), such that in the regions near the chip gaps for each setting, only ~50% of the science frames are represented. Since total brightness can vary from exposure to exposure (e.g., clouds, seeing), these regions are often offset in normalization from the rest of the coadded spectrum. The IDL routine GMOSCOADD (McConnell) attempts to identify the chip gap regions and rescale them to match the adjacent pieces of the spectrum.
Mitchell
Mitchell Bins Mitchell V Profile My Version

Spectral Fitting Diagnostics

First, here's the comparison of the fits for different spectra plotted in differnet colors, with the residuals. It looks like I might need to slightly adjust the sky lines/masks. Note that this is a *very large plot* and takes a minute to load.
Plot
  • I've also broken out the kinematics from the fits above and plotted them here, both in terms of central values with errors and the actual values themselves. For a first pass at these data, the results are looking really robust:
Gh6 vs. Gh8 (moments) Gh6 vs. Gh10 (moments) Gh8 vs. Gh10 (moments) Gh6 vs. Gh8 (values) Gh6 vs. Gh10 (values) Gh8 vs. Gh10 (values)

Comparing to Existing Files

  • There are some old versions of the kinematics on BOX, so here are a few plots showing my results from above compared to the existing kinematic measurements.
Gh6 vs. Gh6 Gh6 vs. Gh8 Gh6 vs. Gh10

Radial Profiles

  • Note that I currently have the dummy moments set to 0 with error = 1.0, so they look a bit funky in those plots.
Gh6 Gh8 Gh10
  • Here's some plots directly from BOX showing the folded/unfolded Mitchell kinematics:
Col1 Col2 Col3
  • And the equivalent plot (only for the symmetric binning) for GMOS + Mitchell:
Plot

Combined Map Orbit Inputs

  • Lastly, here are some examples of the 2d maps for each moment for the fits above:
    • Note that these have not been point-symmetrized and correspond to the Gh6 set of fits from above.
moment Zoomed Slightly Less Zoom Less Zoom No Zoom
V
sigma
h3
h4
h5
h6

A quick symmetric binning scheme test

  • Haven't had the chance to fully test this yet but I wanted to quickly attempt fitting the symmetric binning scheme (which ended up being published). Here's a really quick comparison:
Plot
And here's the spectra, fits, and residuals for this fit
Plot
images/241112/spec_fits_sym-1.png

MGE Details

  • I've tried to do some preliminary MGE fitting but there are definitely some choices that need to be made, and the fits are currently a bit messy so I didn't get as far as I had hoped.

  • I think the major questions are:

    • Which filter do we want to use? --> if F110W, there are two sets of observations; I could fit them separately I suppose or perhaps combine them?
    • I don't think there is currently a mask that exists for this galaxy. What's the best approach to masking the data?
    • The galaxy is sitting right at the edge of the HST images. Will this impact anything?

  • You can see the different images here with PropID = 11711 (Blakeslee).

  • Anyways, here are some diagnostics from some preliminary MGE fitting attempts. Note that I'm currently doing a very simple mask of roughly a square region centered on the galaxy.

    • Note that the image I am currently fitting is the 03 visit of NGC4874 in the F110W filter.
    • I think that the centering of the galaxy is currently off by ~1 pixel in both x and y which I need to look into a bit.
    • Also, I think that the current lack of mask at the boundary is causing there to be some significant distortion in the outer parts of the galaxy (judging by the very very red or very very blue portions of the residuals).
Pixels Restricted Wide Resid. Zoom Resid. Wide Contours Zoomed Contours Major/Minor SB Deprojected Lenc
0.0
0.5
1.0
1.5
2.0
5.0
⚠️ **GitHub.com Fallback** ⚠️