the indra suite of simulations and origami evolution of ... · the indra suite of simulations and...

The Indra Suite of Simulations and

ORIGAMI Evolution of Structures

Bridget Falck (ICG Portsmouth)

with Tamás Budavári (JHU), Shaun Cole (Durham), Daniel Crankshaw (JHU), László Dobos (Eötvös), Adrian

Jenkins (Durham), Gerard Lemson (MPA), Nuala McCullagh (JHU), Mark Neyrinck (JHU), Alex Szalay

(JHU), Jie Wang (NAOC)

and Kazuya Koyama (ICG), Baojiu Li (Durham), Lucas Lombriser (ICG), and Gong-bo Zhao (ICG)

Current N-body Simulations

September 3, 2013 Bridget Falck 2

The Indra Simulations

Suite of dark matter N-body simulations

512 different random instances, WMAP7 cosmology

each 1 Gpc/h-sided box

10243 particles per simulation

About 1 PB of data!

Dynamically loaded into a SQL database

Available to the public

Particle data:

All particle positions and velocities for all 64 snapshots of each simulation run

Halo catalogs:

Standard Friends-Of-Friends (and others), linked to particles

Fourier modes:

Density grid for 500 time steps of each run


Indra-enabled Science

Covariance of the matter power spectrum

Baryon Acoustic Oscillations in real and redshift space

Cluster and void statistics

Large-scale structure morphology/topology

Galaxy formation and evolution with halo catalogs/merger trees

Mock galaxy catalogs, and calibration of faster methods

Logarithmic/Gaussianized density variable correlations

Detailed studies of the velocity fields (vorticity, multi-streams)

…

… and anything to do with computing ensemble averages, estimating uncertainties over large scales…


Mock Catalogs with Indra

Mock galaxy surveys required to test and calibrate observations

Indra’s large volume and 512 simulations will provide many mocks, and 64 snapshots (with particles) allow merger trees

Spatial indexing in the database

enables very fast on-the-fly light cones

Standard FOF catalog (like Millennium DB), but new halo catalogs can be created from particle positions and velocities

Halo catalogs linked to particle tables


ORIGAMI: Finding folds in phase-space


The ORIGAMI method

• ORIGAMI finds shell-crossing by looking for particles out of order with respect to their original configuration

• Halo particles have undergone shell-crossing along 3 orthogonal axes, filaments along 2, walls 1, and voids 0

In this 1D example,

particles flow toward an

initial overdensity at the

origin, eventually creating

a fold in phase-space

September 3, 2013 7 Bridget Falck

Particle Density by ORIGAMI Morphology


void

filament

wall

halo

Spherical Overdensity halos

ORIGAMI morphology


Comparison to SO

General Relativity

Chameleon f(R) Gravity


Halo Environment

General Relativity

Chameleon f(R) Gravity


Void Regions

ORIGAMI Voids


• Connect only low-

density cores first to

avoid percolation

• Void boundaries are

local density peaks, but

not well defined by

shell-crossing

• Voids occupy ~80% of

the volume of the

universe but only

~20% of the mass

ORIGAMI Current/Future Directions

• ORIGAMI Voids: – Investigate the percolation of single-stream regions

– Include density-based criteria for void catalog

• Application to Modified Gravity simulations – Morphological dependence of screening mechanism

• Both chameleon (f(R)) and Vainshtein (DGP)

– Predictions for future surveys?

• ORIGAMI morphology can easily be added to Indra


Current Status of Indra

Initial Conditions generated and first 100 runs completed – Will be able to re-simulate a few boxes with higher resolution

(Jenkins, 1306.5968; Jenkins & Booth, 1306.5771)

Database design being optimized – Efficient storage with SQL Arrays, and fast searches with

spatial indexing and partitioning the data on the disk

Developing common queries, analysis functions, on-the-fly visualization and loading, and data-handling tools (1PB is a lot!)


Summary

The Indra ensemble of Gpc/h simulations will produce ~ 1 PB of N-body data for precision cosmology

Prototype for future data-intensive simulations/suites

ORIGAMI identifies structures by detecting folds in phase-space (Falck, Neyrinck, & Szalay 2012; 1201.2353)

Applications to modified gravity simulations, halo environments, voids, …

Indra will eventually be made available to the public – stay tuned!


the indra suite of simulations and origami evolution of ... · the indra suite of simulations and...

Documents