Data for "Updated maps of major oxides and Mg# on the lunar surface"

In this study, we added CE-6 lunar soil sample, and used high spatial resolution SELENE (Kaguya) Multiband Imager (MI) spectra data of 59 m/pixel, combined with a deep learning inversion model based on the one-dimensional convolutional neural network (1D-CNN), to present new maps of the major oxides (Al2O3, CaO, FeO, MgO, TiO2) and Mg # (Mg/ (Mg + Fe) in mole percent) values.