SubhaloParent and SubhaloFlag Questions

Eddie Aljamal
  • 1
  • 11 Aug

I have a few questions about the hierarchical structure in subhalos given in the SubhaloParent tag. When you extract the bound subhalo particles (stars specifically using il.snapshots.loadSubhalo) of the parent halo, do these include the particles of all daughter subhalos of this subhalo?

The other questions regard a proper analysis of subhalos:
1) When calculating the number of satellite galaxies for a host halo, does a proper analysis include only subhalos with SubhaloParent = 0 and SubhaloFlag != 0?
2) When calculating the total stellar mass of all satellite halos, does a proper analysis include only subhalos with SubhaloParent = 0 and SubhaloFlag != 0? I calculate the total stellar mass in satellites, by going to all non-central subhalos hosted by a halo and extracting all stellar particles in that halo (using il.snapshots.loadSubhalo) and summing these. I ask this because I am worried that including subhalos with SubhaloParent != 0 in this calculation, double counts the stellar mass in subhalos with SubhaloParent != 0 (the question above), and because subhalos with SubhaloFlag == 0 do not represent satellite galaxies.

Thank you so much.
Eddie

Dylan Nelson
  • 18 Aug

Yes, e.g. loading a halo loads all the particles/cells of all its subhalos. Likewise for hierarchical subhalos. Note that this occurrence is very rare at the resolutions of IllustrisTNG. Most subhalos do not have child subhalos.

(1) It depends what the purpose of the calculation is. For a general theoretical answer, I would exclude SubhaloFlag == 0 subhalos, as these do not correspond to cosmologically infallen substructure. If you were trying to compare to data, e.g. counts of bright blobs, you could include either or both. I would just check if there is any difference between these, I expect it might be minimal.

(2) I think the differences should be very small, since (as mentioned above), hierarchical subhalos are rare. You can verify what star particles are included in e.g. SubhaloMassType[4] by looking at a single case and comparing the values to values you obtain by summing up various subsets of particle masses.

Eddie Aljamal
  • 18 Aug

These are edge cases for the analysis, and they do make only a very small difference halo-to-halo and a negligible difference to the scaling relations. Thank you so much for the clear answers Dylan, this was very useful.

  • Page 1 of 1