Data Specifications#
Available data#
The data hosted here is split into across seven suites, five from Symphony, one from Milky Way-est, and one from EDEN. The five Symphony suites are referred to as SymphonyLMC, SymphonyMilkyWay, SymphonyGroup, SymphonyLCluster, and SymphonyCluster, Milky Way-est is refrred to as MWest, and EDEN is referred to as EDEN_MilkyWay_8K. High resolution version of some SymphonyMilkyWay and EDEN_MilkyWay_8K halos are also available in suites named SymphonyMilkyWayHR and EDEN_MilkyWay_16K, respectively.
Each halo in these suites has several processed datasets describing its subhalos and particles. Some suites have additional, unique data products. More detailed descriptions of each dataset for experts can be found below.
halos- Lightweight dataset giving the evolution of subhalos over time according to the Rockstar and Symgind halo finders. This is the main dataset that most users will be interested in. It is light-weight enough to use on a personal computer.particles- Particles of the halo and all its subhalos, tracked over time.trees- Full Rockstar + consistent-trees merger treees for the entire high-resolution region around the halo.full_snapshots- Full Gadget-2 snapshots for the entire zoom-in simulation. These are only provided at a few snapshots.
No particle-tracking routines have been run on SymphonyCluster. This means that there are no Symfind catalogues for this suite and no particles dataset.
The trees and particles datasets rely on the halos dataset also being present.
Dataset details:
halos: This dataset contains information host halos and subhalos, including the full main-branch evolution of these objects. Disrupted subhalos and “splashback” subhalos are also included. Separate catalogues are provided, constructed with (a) a combination of Rockstar (Behroozi et al. 2013a) and consistent-trees (Behroozi et al. 2013b) and (b) Symfind (Mansfield et al., in prep). Numerous numerical artifacts in the Rockstar catalogs have been fixed. A custom format is used for these halos, as descibed in the Working with Subhalos and the Symlib Documentation pages. This dataset is light-weight enough to analyze on a personal computer, is relatively easy to use, and will be sufficient for the majority of halo-based research tasks.trees: This dataset contains merger trees for all objects in the high-resolution region of the box, not just the subhalos near the central host. These trees were generated with Rockstar (Behroozi et al. 2013a) and consistent-trees (Behroozi et al. 2013b), but have been converted to a more efficient and easy-to-use format and have been heavily annotated with several custom routines, as described in the Working with Merger Trees and the Symlib Documentation pages. This dataset is large enough that it will be difficult to analyze on a personal computer, and it will be unnecessary for most halo-based analysis tasks, even those that require looking at the evolution of objects over time. You should use this dataset only if you want to look at poorly-resolved subhalos, objects far from the host, or completed mergers that occured between future subhalos prior to their infall into the host.particles: This dataset contains all the particles associated with the host and its subhalos. Particles are organized according to which subhalo they orbited prior to becoming a subhalo (see Mansfield et al., 2023), making it easy to follow the evolution of the mass around subhalos over time. This sata is stored in a custom, compressed format and will need to be read with library functions (see Working with Particles and the Symlib Documentation).full_snapshots: This dataset contains full Gadget-2 snapshots for the entire zoom-in simulation at \(a(t)=0.2,\ 0.33,\ 0.5\ 0.67, 1.00\). These snapshots can be read with any software that can read Gadget-2 snapshots. IDs are 32-bit integers, positions and velocities use 32-bit floats. High-reoslution particles are stored at type-1 particles, and lower resolution particles are stored as successively higher particle types.
Accessing Data#
Data is stored as a series of .tar files at the password-protected s3df.slac.stanford.edu/data/kipac/symphony.
If you’re comfortable working with halo data already, you’ll want to see the Quickstart page. Otherwise, check out the Symlib pages to learn how to use the data.