Public Data Access Overview / Discussion Forum

For example, SubfindID=1275053 at Snap=4. It is a high-redshift galaxy that has no descendent. How could I know its Subfind ID at later time, e.g. higher snapshot or low redshift. If it doesn't have descendent, I can't use merger tree to find its SubfindID at later time. And because it doesn't have descendent, it cannot be merged into other galaxy, which means that it exists at later time. How could I find it?

Hi Jinning,

Which simulation do you refer to, for this ID/snapshot combination?

Hi Dylan,

Sorry I forgot to mention it. It is TNG-100

You're right that SubLink wasn't able to find a progenitor for this halo (it is very small). One alternative is to do a manual search based on dark matter IDs. It has 19 DM particles with IDs:

[1415388371, 1415351971, 1415335471, 1415335671, 1415334771,
       1415332771, 1415334871, 1415335371, 1415388271, 1415388671,
       1415336471, 1415336871, 1415334971, 1415388171, 1415388071,
       1415334571, 1415390271, 1415387771, 1415336671]

if you load all the DM IDs at redshift zero, then you can find which halos these belong to. The corresponding snapshot indices for each particle above are:

[3944470590, 1583264666, 1583608028, 1583497655, 1583608061,
       3944446520, 3944447455, 1583226856, 1583506148, 3144831334,
       1583608042, 1583607864, 1583607897, 1583525575, 1583607946,
       1583244894, 3944471445, 3944470340, 1583608017]

and if you map these back into halo IDs, they are:

[     -1,     541,     541,     541,     541,      -1,      -1,
           541,     541, 1020414,     541,     541,     541,     541,
           541,     541,      -1,      -1,     541]

so if you wanted a best guess for the z=0 descendant, I would say halo 541. (You can then try to follow this halo backwards in time, if you want the descendant at another redshift). (You could use the same technique for any small/disappearing halo).

It seems like using Particles ID is reliable method. That helps me a lot! Thank you!

@Dylan Nelson said:
You're right that SubLink wasn't able to find a progenitor for this halo (it is very small). One alternative is to do a manual search based on dark matter IDs. It has 19 DM particles with IDs:

[1415388371, 1415351971, 1415335471, 1415335671, 1415334771,
       1415332771, 1415334871, 1415335371, 1415388271, 1415388671,
       1415336471, 1415336871, 1415334971, 1415388171, 1415388071,
       1415334571, 1415390271, 1415387771, 1415336671]

if you load all the DM IDs at redshift zero, then you can find which halos these belong to. The corresponding snapshot indices for each particle above are:

[3944470590, 1583264666, 1583608028, 1583497655, 1583608061,
       3944446520, 3944447455, 1583226856, 1583506148, 3144831334,
       1583608042, 1583607864, 1583607897, 1583525575, 1583607946,
       1583244894, 3944471445, 3944470340, 1583608017]

and if you map these back into halo IDs, they are:

[     -1,     541,     541,     541,     541,      -1,      -1,
           541,     541, 1020414,     541,     541,     541,     541,
           541,     541,      -1,      -1,     541]

so if you wanted a best guess for the z=0 descendant, I would say halo 541. (You can then try to follow this halo backwards in time, if you want the descendant at another redshift). (You could use the same technique for any small/disappearing halo).

Hi Dylan! Can I ask how did you map these particleIDs into halo/subhaloIDs ?

You can use the offsets. In particular, the Group/SnapByType and Subhalo/SnapByType fields.

The biggest offset, which is smaller than your particle index, gives you the subhalo/halo it belongs to, so long as the difference between that offset and the particle index is within the length of the subhalo/halo.