Mapping Circumgalactic Medium Observations to Theory Using Machine Learning

TL;DR

This paper develops a machine learning framework using random forests to predict the physical conditions of the circumgalactic medium from quasar absorption lines, trained on synthetic data from cosmological simulations, improving understanding of CGM properties.

Contribution

It introduces a novel machine learning approach that maps CGM observables to physical conditions without simplifying assumptions, validated on simulated data.

Findings

Random forest models accurately predict overdensity, temperature, and metallicity.

Feature importance analysis reveals key observables influencing predictions.

Normalising flow enhances the model's ability to capture scatter in physical conditions.

Abstract

We present a random forest framework for predicting circumgalactic medium (CGM) physical conditions from quasar absorption line observables, trained on a sample of Voigt profile-fit synthetic absorbers from the Simba cosmological simulation. Traditionally, extracting physical conditions from CGM absorber observations involves simplifying assumptions such as uniform single-phase clouds, but by using a cosmological simulation we bypass such assumptions to better capture the complex relationship between CGM observables and underlying gas conditions. We train random forest models on synthetic spectra for HI and selected metal lines around galaxies across a range of star formation rates, stellar masses, and impact parameters, to predict absorber overdensities, temperatures, and metallicities. The models reproduce the true values from Simba well, with normalised transverse standard deviations of $0.50-0.54$ dex in overdensity, $0.32-0.54$ dex in temperature, and $0.49-0.53$ dex in metallicity predicted from metal lines (not HI), across all ions. Examining the feature importance, the random forest indicates that the overdensity is most informed by the absorber column density, the temperature is driven by the line width, and the metallicity is most sensitive to the specific star formation rate. Alternatively examining feature importance by removing one observable at a time, the overdensity and metallicity appear to be more driven by the impact parameter. We introduce a normalising flow approach in order to ensure the scatter in the true physical conditions is accurately spanned by the network. The trained models are available online.