Solving large scale structure in ten easy steps with COLA

By S. Tassev, M. Zaldarriaga, Daniel J. Eisenstein. 2013

281 citations recorded by the source corpus.

We present the COmoving Lagrangian Acceleration (COLA) method: an N-body method for solving for Large Scale Structure (LSS) in a frame that is comoving with observers following trajectories calculated in Lagrangian Perturbation Theory (LPT). Unlike standard N-body methods, the COLA method can straightforwardly trade accuracy at small-scales in order to gain computational speed without sacrificing accuracy at large scales. This is especially useful for cheaply generating large ensembles of accurate mock halo catalogs required to study galaxy clustering and weak lensing, as those catalogs are essential for performing detailed error analysis for ongoing and future surveys of LSS. As an illustration, we ran a COLA-based N-body code on a box of size 100 Mpc/h with particles of mass ≈ 5 × 109M⊙/h. Running the code with only 10 timesteps was sufficient to obtain an accurate description of halo statistics down to halo masses of at least 1011M⊙/h. This is only at a modest speed penalty when compared to mocks obtained with LPT. A standard detailed N-body run is orders of magnitude slower than our COLA-based code. The speed-up we obtain with COLA is due to the fact that we calculate the large-scale dynamics exactly using LPT, while letting the N-body code solve for the small scales, without requiring it to capture exactly the internal dynamics of halos. Achieving a similar level of accuracy in halo statistics without the COLA method requires at least 3 times more timesteps than when COLA is employed.

Subfields: Artificial Intelligence, Astronomy and Astrophysics, Nuclear and High Energy Physics

View Paper: https://www.semanticscholar.org/paper/92a585d637355a381c5d919802eb44c870f9413c

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Astronomy and Astrophysics

In Nuclear and High Energy Physics

Citations in the corpus

nbodykit: An Open-source, Massively Parallel Toolkit for Large-scale Structure (2017, methodology): “The rise of particle mesh simulation methods (Merz et al. 2005; Tassev et al. 2013; White et al. 2014; Feng et al. 2016) has offered a computationally cheaper alternative to running full N -body simulations.”
Improved WiggleZ Dark Energy Survey Distance Measurements to z = 1 with Reconstruction of the Baryonic Acoustic Feature (2014, unknown): “We use 3 × 1296 grids per dimension to calculate the gravitational force with enough spatial resolution (Tassev et al. 2013).”
Modelling baryon acoustic oscillations with perturbation theory and stochastic halo biasing (2013, methodology): “…-body simulations to account for a change in the cosmological parameters (Angulo & White 2010), compute covariance matrices from a set of small-volume simulations (Schneider et al. 2011), orincluding 2LPT within the Vlasov equations solver to speed upN -body codes (COLA, Tassev et al. 2013).”
Ingredients for 21 cm Intensity Mapping (2018, methodology): “On the other hand, HI HOD, i.e. painting HI on top of dark matter halos from either N-body or fast numerical simulations like COLA (Tassev et al. 2013), can produce highly accurate results.”
A volumetric deep Convolutional Neural Network for simulation of dark matter halo catalogues (2018, background): “…of large scale structure which attempt to create simulations of a satisfactory accuracy at minimal computational cost (Bond & Myers 1996a; Monaco et al. 2013; Tassev et al. 2013; Izard et al. 2016; Feng et al. 2016; Avila et al. 2015; Kitaura et al. 2014; Chuang et al. 2015; White et al. 2014).”
The clustering of the SDSS main galaxy sample – II. Mock galaxy catalogues and a measurement of the growth of structure from redshift space distortions at z = 0.15 (2014, methodology): “For this, we have created the code PICOLA, a highly developed, planar–parallel implementation of the COLA method of Tassev et al. (2013); this implementation is described in Howlett et. al.”
The clustering of the SDSS DR7 Main Galaxy Sample I: a 4 per cent distance measure at z=0.15 (2014, methodology): “These were created using a newly-developed code PICOLA, a highly-developed, planar-parallelised implementation of the COLA method of Tassev et al. (2013).”
CosmoFlow: Using Deep Learning to Learn the Universe at Scale (2018, methodology): “A slightly modified version of pycola9 [36], [37] is used to create the dark matter simulations.”