Multivariate Predictors of Lyman Continuum Escape. I. A Survival Analysis of the Low-redshift Lyman Continuum Survey*
Anne E. Jaskot, Anneliese C. Silveyra, Anna Plantinga, Sophia R. Flury, Matthew Hayes, John Chisholm, Timothy M. Heckman, L. Pentericci, D. Schaerer, Maxime Trebitsch, Anne Verhamme, Cody Carr, Henry C. Ferguson, Zhiyuan Ji, Mauro Giavalisco, Alaina Henry, R. Marques-Chaves, Göran Östlin, Alberto Saldana-Lopez, Claudia Scarlata, G. Worseck, Xinfeng Xu
Abstract
Abstract To understand how galaxies reionized the Universe, we must determine how the escape fraction of Lyman continuum (LyC) photons ( f esc ) depends on galaxy properties. Using the z ∼ 0.3 Low-redshift Lyman Continuum Survey (LzLCS), we develop and analyze new multivariate predictors of f esc . These predictions use the Cox proportional hazards model, a survival analysis technique that incorporates both detections and upper limits. Our best model predicts the LzLCS f esc detections with an rms scatter of 0.31 dex, better than single-variable correlations. According to ranking techniques, the most important predictors of f esc are the equivalent width (EW) of Lyman-series absorption lines and the UV dust attenuation, which track line-of-sight absorption due to H i and dust. The H i absorption EW is uniquely crucial for predicting f esc for the strongest LyC emitters, which show properties similar to weaker LyC emitters and whose high f esc may therefore result from favorable orientation. In the absence of H i information, star formation rate surface density (Σ SFR ) and [O iii ]/[O ii ] ratio are the most predictive variables and highlight the connection between feedback and f esc . We generate a model suitable for z > 6, which uses only the UV slope, Σ SFR , and [O iii ]/[O ii ]. We find that Σ SFR is more important in predicting f esc at higher stellar masses, whereas [O iii ]/[O ii ] plays a greater role at lower masses. We also analyze predictions for other parameters, such as the ionizing-to-nonionizing flux ratio and Ly α escape fraction. These multivariate models represent a promising tool for predicting f esc at high redshift.