Galaxies drastically blowing up between snapshots

DErrick Carr
  • 28 Sep

I am currently working on a project to investigate the evolution of compact star-forming galaxies. I created a sample and noticed something that I am trying to understand. For more than just a sporadic amount of galaxies, I will visualize a range of snapshots which precede and flank the snapshot I select the galaxy in. I can then see that as you increase in snapshot number, the galaxy gas will become slowly smaller and smaller and then within one snapshot will gain a substantial amount of stellar mass. In this situation, is the galaxy a satellite and slowly being absorbed by the central, at which the SubfindID for that galaxy now becomes the central? An example of a galaxy that does this for the visualization tool:

(Snapshot number, ID)
79, 442874
80, 445136

Thanks

Dylan Nelson
  • 29 Sep

Do you mean it grows smaller/larger in terms of a value in the group catalog, or in terms of an actual image?

I'd suggest to make a series of images (i..e of stellar mass), and this may help understand if it is a satellite galaxy merging, as you say.

If you are following the merger tree forwards in time, then this behavior makes sense for a merger.

DErrick Carr
  • 1
  • 29 Sep

The stellar mass/gas of the subhalo grows, sometimes by a 10 or even more. The group catalog mass stays roughly the same. I do have the series of images (albiet for gas and not stellar mass) and it's clear that the galaxy becomes something else.

With that said, I think I want to be able to select galaxies that do NOT merge into other galaxies (or better said, I only want galaxies that do not merge into other galaxy). Is that possible to do using the sublink.loadTree function?

Dylan Nelson
  • 29 Sep

Typically someone would look at a low redshift galaxy (i.e. at the last snapshot) and apply a criterion to the number of mergers above a given mass ratio, in the past. This lets you select galaxies which had "quiet assembly histories" i.e. few or no important mergers.

You can do the same thing if you're selecting at high redshift, and thinking about the future. One idea would be: follow the object to z=0 (using the "onlyMDB" option in loadTree, or the "Root Descendant" information in the tree itself). Then, you can load the main progenitor branch backwards in time. If your original subhalo is not on the MPB, that means it is on a smaller branch, i.e. was a subdominant merger into the final system, and so you would exclude it.

  • Page 1 of 1