Hi Dylan,
I am wondering is there a way to access the mass loss rate due to stellar evolution between every two snapshot for one subhalo. I implement the following method but I don't know if there are any data field in IllustrisTNG which records the mass loss.
I count the IDs of the disappeared stellar particles between two successive snapshots. I assume all the masses from the disappeared stellar particle IDs are totally destroyed.
Then, I get the intersection of IDs from two snapshots. I calculate the corresponding mass reduction from particles IDs in the intersection.
Thank you!
Dylan Nelson
15 Aug '19
Hi David,
You're right, this isn't recorded any way at the group catalog level, so you'll have to derive it from the star particles themselves.
What I would probably do is look at the subhalo at snapshot N and load all Masses of all stars. Comparison to GFM_InitialMass will give the integrated mass loss since birth. You could compute the average using the ages, or else for the instantaneous, you should load also the IDs of those stars, then the IDs of all stars in snapshot N-1, find the same set of stars (all must exist at the previous snapshot, since star particles do not disappear). The difference of their Masses sum would be the mass loss between these two times.
David Wang
15 Aug '19
Thanks, Dylan!
One more question: as you said, since the stellar ID never disappeared in the simulation. If some IDs is missing in the catalog of a subhalo, it will mean that these IDs probably are stripped away. Is my assumption correct here? I am curious about if a stellar particle is totally destroyed by supernova explosion, is IllustrisTNG always keep that the mass of that ID 0 and not remove it from the simulation?
Dylan Nelson
16 Aug '19
Hi David,
You're right that a star with a given ID might not be in the same subhalo (i.e. its progenitor or descendant) at a different snapshot, since it could have been stripped for instance. But it will be somewhere in the simulation volume, thus why you should in general search over all stars (not such a problem, as there aren't too many stars, as compared to e.g. gas or DM).
The physical model for SN explosions in Illustris[TNG] does not ever completely destroy star particles, because star particles represent entire populations, not just single stars.
Gail Zasowski
7 Jul
Hi! I hope it's ok to reopen this question with a related follow-up. I'm curious whether i) the "mass loss" is restricted to SNe & winds and the remaining stellar mass includes stellar remnants (WDs, BHs, etc), or ii) stars evolved past their lifetimes are considered completely removed. I understand that the Portinari+1998 tables were used for some yields and stellar evolution properties, and those do include ejecta/remnant masses, but I haven't been able to find any explicit statement on how much of that was propagated into the Illustris/TNG framework. Thanks in advance!
Dylan Nelson
7 Jul
The most precise statements will be the paragraphs of relevant from the Pillepich+18 TNG methods paper. It should say "exactly" where and how the Portinari inputs are used. Then the answer(s) to your question are really: whatever assumptions/choices were made in Portinari.
Gail Zasowski
7 Jul
Thanks, Dylan. The relevant paragraphs from Pillepich+18 only discuss the inputs in the context of chemical yields, not whether the ejecta mass or the total mass was removed when the star particle. (I did ask Pillepich about this before posting here and she said she wasn't sure.) Portinari provides numerous tables with sufficient information to track remnant masses, but it's not clear which of them were adopted. I'll keep looking through some of the older Illustris papers for clarification.
Hi Dylan,
I am wondering is there a way to access the mass loss rate due to stellar evolution between every two snapshot for one subhalo. I implement the following method but I don't know if there are any data field in IllustrisTNG which records the mass loss.
Thank you!
Hi David,
You're right, this isn't recorded any way at the group catalog level, so you'll have to derive it from the star particles themselves.
What I would probably do is look at the subhalo at snapshot
N
and load allMasses
of all stars. Comparison toGFM_InitialMass
will give the integrated mass loss since birth. You could compute the average using the ages, or else for the instantaneous, you should load also the IDs of those stars, then the IDs of all stars in snapshotN-1
, find the same set of stars (all must exist at the previous snapshot, since star particles do not disappear). The difference of theirMasses
sum would be the mass loss between these two times.Thanks, Dylan!
One more question: as you said, since the stellar ID never disappeared in the simulation. If some IDs is missing in the catalog of a subhalo, it will mean that these IDs probably are stripped away. Is my assumption correct here? I am curious about if a stellar particle is totally destroyed by supernova explosion, is IllustrisTNG always keep that the mass of that ID 0 and not remove it from the simulation?
Hi David,
You're right that a star with a given ID might not be in the same subhalo (i.e. its progenitor or descendant) at a different snapshot, since it could have been stripped for instance. But it will be somewhere in the simulation volume, thus why you should in general search over all stars (not such a problem, as there aren't too many stars, as compared to e.g. gas or DM).
The physical model for SN explosions in Illustris[TNG] does not ever completely destroy star particles, because star particles represent entire populations, not just single stars.
Hi! I hope it's ok to reopen this question with a related follow-up. I'm curious whether i) the "mass loss" is restricted to SNe & winds and the remaining stellar mass includes stellar remnants (WDs, BHs, etc), or ii) stars evolved past their lifetimes are considered completely removed. I understand that the Portinari+1998 tables were used for some yields and stellar evolution properties, and those do include ejecta/remnant masses, but I haven't been able to find any explicit statement on how much of that was propagated into the Illustris/TNG framework. Thanks in advance!
The most precise statements will be the paragraphs of relevant from the Pillepich+18 TNG methods paper. It should say "exactly" where and how the Portinari inputs are used. Then the answer(s) to your question are really: whatever assumptions/choices were made in Portinari.
Thanks, Dylan. The relevant paragraphs from Pillepich+18 only discuss the inputs in the context of chemical yields, not whether the ejecta mass or the total mass was removed when the star particle. (I did ask Pillepich about this before posting here and she said she wasn't sure.) Portinari provides numerous tables with sufficient information to track remnant masses, but it's not clear which of them were adopted. I'll keep looking through some of the older Illustris papers for clarification.