Next-generation Computing and Data Technologies to probe the Cosmic Metal content

project code : P2022ZLW4T CUP : C53D2301002 0001

This is the webpage of the project "Next-generation Computing and Data Technologies to probe the Cosmic Metal content" (number P2022ZLW4T), funded by a PRIN 2022 PNRR grant of the Italian Ministry of University and Research - MUR (ERC sector PE9), Italy. The project received a funding of Euro 300,000 from MUR; to be executed over the period from 1 December 2023 to 28 February 2026.

It involves two research units: National Institute of Astrophysics - Astronomical Observatory of Trieste (INAF-OATs) and the University of Milano-Bicocca (UniMiB), with the following team members.

Executive Summary

Our project proposes to employ next-generation high-performance computing and advanced algorithms for data analysis, including machine learning (ML) tools, to investigate the unknown metal distributions around galaxies in the early Universe.

One objective of the project is the development of HPC codes, execution of cosmological simulations in petaflop computing platforms, the generation and analysis of large statistical samples of simulated data, and finally the comparison with astronomical observational data. Cosmological Hydrodynamical Simulation has emerged as a powerful tool over the last twenty years, for calculating the non-linear evolution of matter over the age of the Universe, leading to large-scale-structure formation. Simulations attempt to produce galaxies that are similar to the observed ones, in the way comprehending the underlying physical processes in forming the variety of galaxy populations. We will develop efficient numerical codes and include new gas physics to shed light on cosmic evolution.

The other objective of the project is to develop a new toolkit for fast and reliable analysis of big data from absorption spectroscopy. These tools will rely on advanced ML techniques to process spectral data, extract and classify absorption features. A novel GPU accelerated fitting code will be developed to extract physical measurements from the classified features. With a mock maker code for simulations of general-purpose absorption spectra, we will produce large libraries of artificial spectra, useful for the development and training of the toolkit. In addition, we will build on the existing “Astrocook” package for spectral analysis to implement sophisticated techniques for precision spectroscopy, namely the spectro-perfectionism approach to data treatment and the Bayesian information criterion for accurate modeling of the spectral features. The new code libraries will accelerate progress in our understanding of astrophysical problems and will be ready to be deployed for industry-based applications.

Our simulations and analysis tools will be applied to the investigation of the metal distribution around galaxies in the Circumgalactic and Intergalactic Medium (CGM and IGM) at redshifts z~2-4, or ~10-12 Gyr ago. The IGM is the network of large-scale low-density filaments, warm/hot and ionized, containing most (> 60%) of the baryons in the Universe. At the denser intersections, galaxies and clusters build up from the inflow of gas, and evolve by mergers. Star-formation in galaxies produces metals, and processes like galactic outflows carry these metals out, enriching the CGM/IGM. This enriched gas is observed spectroscopically via absorption line systems produced by the ionized metals. We aim to shed light on the multiple baryonic feedback processes that distribute energy/metals in the CGM/IGM which are still poorly understood and debated, in particular the feedback from early supermassive black holes in high-redshift quasars.