Deep Learning Prediction of Quasars Broad Ly$\alpha$ Emission Line

TL;DR

This paper presents a deep learning model trained on SDSS quasar spectra to accurately predict the Ly$ ext{α}$ emission line's flux and shape, aiding studies of DLA absorbers and the intergalactic medium.

Contribution

The study introduces a neural network approach for predicting quasar Ly$ ext{α}$ emission lines using multiple broad emission lines as input, achieving high precision and low bias.

Findings

Predicts Ly$ ext{α}$ flux with 6-12% precision

Can estimate HI column density in DLA systems

Useful for studying the epoch of reionization

Abstract

We have employed deep neural network, or deep learning to predict the flux and the shape of the broad Ly$\alpha$ emission lines in the spectra of quasars. We use 17870 high signal-to-noise ratio (SNR > 15) quasar spectra from the Sloan Digital Sky Survey (SDSS) Data Release 14 (DR14) to train the model and evaluate its performance. The SiIV, CIV, and CIII] broad emission lines are used as the input to the neural network, and the model returns the predicted Ly$\alpha$ emission line as the output. We found that our neural network model predicts quasars continua around the Ly$\alpha$ spectral region with $\sim$6 - 12% precision and $\lesssim$1% bias. Our model can be used to estimate the HI column density of eclipsing and ghostly damped Ly$\alpha$ (DLA) absorbers as the presence of the DLA absorption in these systems strongly contaminates the flux and the shape of the quasar continuum around Ly$\alpha$ spectral region. The model could also be used to study the state of the intergalactic medium during the epoch of reionization.