On improving analytical models of cosmic reionization for matching numerical simulation

TL;DR

This paper develops methods to enhance analytical models of cosmic reionization by training them on numerical simulations, enabling fast generation of large mock catalogs that match simulation statistics for observational studies.

Contribution

It introduces a new code that trains analytical models on numerical simulations to produce large, accurate mock catalogs efficiently.

Findings

The analytical model can be trained to match numerical simulation statistics.

Mock catalogs generated are large and computationally inexpensive.

The method improves the utility of analytical models for observational predictions.

Abstract

The methods for studying the epoch of cosmic reionization vary from full radiative transfer simulations to purely analytical models. While numerical approaches are computationally expensive and are not suitable for generating many mock catalogs, analytical methods are based on assumptions and approximations. We explore the interconnection between both methods. First, we ask how the analytical framework of excursion set formalism can be used for statistical analysis of numerical simulations and visual representation of the morphology of ionization fronts. Second, we explore the methods of training the analytical model on a given numerical simulation. We present a new code which emerged from this study. Its main application is to match the analytical model with a numerical simulation. Then, it allows one to generate mock reionization catalogs with volumes exceeding the original simulation quickly and computationally inexpensively, meanwhile reproducing large scale statistical properties. These mock catalogs are particularly useful for CMB polarization and 21cm experiments, where large volumes are required to simulate the observed signal.