Project: Simulating the Euclid Universe

Prace Call: 17th
ID: 2018184475, Leader: Romain Teyssier
Affiliation: Universitat Zurich, CH
Research Field: Universe Sciences
Collaborators: Doug Potter Universitat Zurich CH , Joachim Stadel Universitat Zurich CH
Resource Awarded: 136 Mil. core hours on Piz Daint


The Euclid satellite will be launched by the European Space Agency shortly after 2020 to map the entire Universe with more than a billion galaxies. This will provide an unprecedented measurement of the galaxy and mass distribution with sub-percent accuracy and a unique opportunity to discover new physics. Such an ambitious measurement requires a precise theoretical prediction of the observables. The challenge resides mostly at small scales, where Euclid will provide invaluable new information, complementary to other cosmology probes, but where the theory requires a full nonlinear treatment of collisionless fluid dynamics. Only N-body simulations with very large volumes and particle numbers can provide both large scale statistics and small scale accuracy. These simulations are used for two main purposes: 1- generate mock galaxy catalogues that we can compare directly to observations and assess the quality of the entire data processing pipeline, 2- provide emulators of the observables as a function of a multi-dimensional cosmological parameter space, in order to perform the likelihood analysis of the data. This proposal aims at: 1- producing a mock galaxy catalogue for the Euclid collaboration, including massive neutrinos and 2- producing a matter power spectrum emulator on an extended parameter space that include the massive neutrinos mass. Objective 1 will be met by simulating our fiducial model in a 4 Gpc periodic box with 4.4T particles, and by post-processing the simulation data on the light cone to obtain the final galaxy catalogue. Objective 2 will be met by performing 100 simulations of moderate size on a predefined set of parameters, and by using Uncertainty Quantification techniques to provide an accurate interpolator across parameter space. The products of this project will be made available to the entire Euclid community. For the first time, we will be able to assess the importance of nonlinear scale in determining the most likely cosmological parameters, compared to other cosmological probes.