Quantum fluctuations in the early Universe laid the foundation for the cosmic large-scale structure we observe today. The eventual state of these quantum fluctuations, frozen in at inflation, defines the Universe's absolute and immutable "initial conditions" (ICs). In traditional cosmological simulations, the spatial arrangement of phases in the ICs are random, with focus instead placed on the statistical properties of the initial field, often assuming it to be Gaussian. Prominent simulations such as EAGLE, ILLUSTRIS, and MAGNETICUM operate under these assumptions, generating statistically representative random patches of the Universe. However, such simulations are inherently limited when compared to the real Universe, as they can only provide statistical summaries and average behaviours, rather than directly replicating the specific structures we observe. In contrast, accurately determining the true ICs of our Universe—those that have the correct spatial arrangement of phases—breaks these limitations. When these ICs are evolved through numerical simulators, they digitally reconstruct the observed cosmic structure, precisely replicating the positions of galaxy clusters, filaments and voids. In fact, such reconstructions go beyond just matching the observed structure spatially; they also recover the velocities, masses, and internal properties of these structures, offering insights into their formation histories and the environments in which they developed.
This project, MANTICORE, seeks to infer the ICs of our Universe from observational data, with the ambitious goal of reconstructing the largest and most accurate digital realisation of the Universe to date. This endeavour will provide an unparalleled tool for exploration and discovery in our understanding of cosmic structures.
As part of our ongoing NAISS Large Compute (NAISS 2023/3-20) and NASISS Medium Storage (NAISS 2024/6-148) calls, MANTICORE has running on the Dardel machine. We wish to continue and complete the MANTICORE run using time from the upcoming NAISS Large Compute and LUMI Sweden calls.