2020ApJ...899...15M

Resolving the phase structure of neutral hydrogen (H I) is crucial for understanding the life cycle of the interstellar medium (ISM). However, accurate measurements of H I temperature and density are limited by the availability of background continuum sources for measuring H I absorption. Here we test the use of deep learning for extracting H I properties over large areas without optical depth information. We train a 1D convolutional neural network using synthetic observations of 3D numerical simulations of the ISM to predict the fraction (f_CNM) of cold neutral medium (CNM) and the correction to the optically thin H I column density for optical depth ( R_HI) from 21 cm emission alone. We restrict our analysis to high Galactic latitudes (| b| > 30^○), where the complexity of spectral line profiles is minimized. We verify that the network accurately predicts f_CNM and R_HI by comparing the results with direct constraints from 21 cm absorption. By applying the network to the GALFA-H I survey, we generate large-area maps of f_CNM and R_HI. Although the overall contribution to the total H I column of CNM-rich structures is small (∼5%), we find that these structures are ubiquitous. Our results are consistent with the picture that small-scale structures observed in 21 cm emission aligned with the magnetic field are dominated by CNM. Finally, we demonstrate that the observed correlation between H I column density and dust reddening (E(B-V)) declines with increasing R_HI, indicating that future efforts to quantify foreground Galactic E(B-V) using H I, even at high latitudes, should increase fidelity by accounting for H I phase structure.