Using convolutional neural networks to predict galaxy metallicity from three-colour images. Issue 4 (1st February 2019)

• Record Type: Journal Article
• Authors: Wu, John F; Boada, Steven
• Is Part Of: Monthly notices of the Royal Astronomical Society. Volume 484:Issue 4(2019)
• Page Start: 4683
• Page End: 4694
• Publication Date: 2019-02-01
• Subjects: methods: data analysis -- surveys -- galaxies: evolution -- galaxies: general; Astronomy -- Periodicals; Periodicals; 520.5
• DOI: 10.1093/mnras/stz333
• Languages: English
• ISSNs: 0035-8711

Abstract

Abstract: We train a deep residual convolutional neural network (CNN) to predict the gas-phase metallicity ( Z ) of galaxies derived from spectroscopic information ($Z \equiv 12 + \log (\rm O/H)$ ) using only three-band gri images from the Sloan Digital Sky Survey. When trained and tested on 128 × 128-pixel images, the root mean squared error (RMSE) of Z pred − Z true is only 0.085 dex, vastly outperforming a trained random forest algorithm on the same data set (RMSE = 0.130 dex). The amount of scatter in Z pred − Z true decreases with increasing image resolution in an intuitive manner. We are able to use CNN-predicted Z pred and independently measured stellar masses to recover a mass–metallicity relation with 0.10 dex scatter. Because our predicted MZR shows no more scatter than the empirical MZR, the difference between Z pred and Z true cannot be due to purely random error. This suggests that the CNN has learned a representation of the gas-phase metallicity, from the optical imaging, beyond what is accessible with oxygen spectral lines.