(Authors: Juliana Kwan, Katrin Heitmann, Salman Habib, Nikhil Padmanabhan, Adrian Pope, Nicholas Frontiere, Hal Finkel)

LSST will measure the galaxy power spectrum to an accuracy of a few percent. The broadband shape of the galaxy power spectrum alone will provide strong constraints on cosmological parameters such as n_s, Om_m, Om_b. Furthermore, the galaxy power spectrum is an important ingredient in many other areas of LSST science highlights, including Baryon Acoustic Oscillation measurements and galaxy-galaxy weak lensing. Achieving this level of precision will require new tools for data analysis that are informed from fully nonlinear models of structure formation. Prescriptions based on the halo model will not be adequate enough for LSST data, as shown in our paper, the error between the halo model approach and the N-body simulation data can be as much as 20% on scales k > 0.1 1/Mpc. Moreover, the bias between the clustering of dark matter and galaxies is not known a priori and may be scale dependent and non-trivial to model. To facilitate the analysis of future survey data at a sufficiently high precision, we have made available a fast prediction scheme, known as an emulator, for the fully nonlinear galaxy power spectrum as measured from an N-body simulation.

Our emulator makes a prediction for the nonlinear galaxy power spectrum based on the halo occupation model. We use halo catalogs from a LCDM N-body simulation and populate these with galaxies using a halo occupation distribution (HOD) model (Z. Zheng, I. Zehavi, D. Eisenstein, D. Weinberg, and Y. P. Jing, ApJ, 2009, 707, 554). Although this particular form of the HOD is more specific to a particular galaxy type ('SDSS LRG'), other HOD models can easily be implemented, as required. Our current emulator gives an estimate of the galaxy power spectrum between 0.01 <= k <= 1 1/Mpc from 0 <= z <= 1 to within 3%. It runs on a single processor in a few seconds and only requires the GSL library to compile and run.

See arXiv:1311.6444 for details. The emulator can be downloaded from: www.hep.anl.gov/cosmology/CosmicEmu/emu.html