Concept of subhalo mass

Suchira Sarkar
  • 19 Jul '23

Dear Dylan,
Is a subhalo exactly equal to a single galaxy along with its stars, gas and dark matter halo ? In other words, can it contain more than one or part of another galaxy? In some of the SKIRT SDSS images, there seems to be some signature of interaction.

-Suchira

Dylan Nelson
  • 19 Jul '23

During the early phases of a galaxy merger, Subfind will (usually) identify the two galaxies as two separate subhalos.

However, at some point (i.e. the end of the merger / coalescence), these two subhalos must eventually become one. As this begins to occur, I imagine it is possible that one subhalo appears to contain "a part" of another.

Note: some SKIRT images (and images in general) are made using "all FoF halo particles", thus two subhalos of a merging pair would both be visible. However, some images are made using "subhalo particles only", i.e. the image of each galaxy in the pair would appear to (for the most part) be missing the companion. Which makes more sense would depend on your science application.

Suchira Sarkar
  • 1
  • 13 Dec '23

Dear Dylan,
I am trying to get a sample of disk galaxies beyond stellar mass >=10**11 solar mass. I wanted to get a measure of V_rot/stellar velocity dispersion for each galaxy to decide if it is a disk galaxy or not. I realized that stellar velocity dispersion is not readily available among the fields. So, I need to calculate that myself.
For a measure of rotation velocity, can we use SubhaloVmax? It is not clear to me how it is calculated. I guess, it is the maximum of the rotation curve determined by the gravity of stars+gas+DM halo? Or, is it a sum of rotation velocities from all particles?

Dylan Nelson
  • 13 Dec '23

Yes as per the documentation the SubhaloVmax field is "Maximum value of the spherically-averaged rotation curve. All available particle types (e.g. gas, stars, DM, and SMBHs) are included in this calculation."

I would treat this only as a first/rough estimate, likely you want to calculate something in more detail.

If you are e.g. trying to reproduce an observational selection, it may be important to calculate a V_rot in a way more similar to observations (not in 3D, not based on total mass, etc).

  • Page 1 of 1