A testrun on merger trees part II

—    —

  previous

next  

Summary:TBA

This is the continuation of the post from 2021-01-07. The following questions are still open will be discussed in this post.

1. Verify the approach of connection halos by the provided haloid and descID by comparing directly with the numbers of particle assigned to the halos between snapshots.
2. How does ROCKSTAR assign the descID when running?
3. Calculate halo mass function and accretion rates.

Task1: Merger tree verification

The example of merger tree treeID13

In this analysis we scanned through the entire box of Cholla with side-lenght 50\(h^{-1}\)Mpc and a resolution of \(256^3\). We approximate the halo mass of Cholla buy gathering all dark matter particles within a certain radius (e.g. \(5\times\) the virial radius \(R_{vir}\)). Then we multiplied the number of particel \(n_{particle}\) by the mass of one dark matter particle in the simulation being \(m_{DM}=5.407\times 10^8\) \([h^{-1}M_{\odot}]\). We compare this result with the halo mass the halo finder ROCKSTAR provided. In Figure 1 we show the offset between the two halo masses on the x-axis for ROCKSTAR and on the y-axis for Cholla. The differences are up to \(\pm50\%\) in masses. Figure 2 gives a detail list of the data plotted in Figure 1.

The figure visualizes the offset between the halos from merger tree with ID13 provided by ROCKSTAR (x-axis) in comparision what we can approximate when summing up dark matter particles within $5\times$ the virial radius $R_{vir}$ (y-axis). The number of particles are given by the color bar

In this table we list all main progenitor halos for the merger tree treeID13. The halo mass provided by ROCKSTAR is shown in column mhalo and the approximated in column m_particles. In column n_particles we shown the number of dark matter particles we found to determine the halo mass. In column mhalo-m_particle(%) we show the difference between the halo masses from ROCKSTAR and Cholla in \%. The halo masses are given in $[h^{-1}M_{\odot}$ and the positions columns X, Y, and Z in $[h^{-1}comvkpc]$.

Although the difference in halo masses are significant high, we can still confirm that ROCKSTAR sets the descIDs correctly. In Figure 3 we illustrated that by tracking the position of the main progenitor halos of treeID13 between \(2.19<z<10.33\).

Positions of the main progenitor halos for the merger tree treeID13 througout redshift.

Positions of all main progenitors at z=4.15

Positions of the main progenitor halos at z=4.15. The black dots show halos which have less then 10% difference in the halo masses for ROCKSTAR and Cholla.

Halo mass function (HMF) of all main progenitors at z=4.15

Halo mass function of the main progenitor halos at z=4.15 for ROCKSTAR (red line), Cholla (blue line), and halos with less than 10% difference in halo mass (black dashed line).

Position of all main progenitors between 2<z<11

Positions of the main progenitor halos for the merger tree treeID13 througout redshift.

Halo mass function (HMF) of all main progenitors between 2<z<11

Halo mass function of the main progenitor halos bewteen 2<z<11 for ROCKSTAR (red line), Cholla (blue line), and halos with less than 10% difference in halo mass (black dashed line).

  A testrun on merger trees

A testrun on merger trees part III  

—    —