The results on ultrahigh-energy cosmic rays (UHECR) mass composition obtained with the Telescope Array surface detector are presented. The analysis employs the Boosted Decision tree (BDT) multivariate analysis built upon 14 observables related to both the properties of the shower front and the lateral distribution function. The multivariate classifier is trained with Monte-Carlo sets of events induced by the primary protons and iron. An average atomic mass of UHECR is presented for energies $10^{18.0}–10^{20.0}\mathrm{eV}$. The average atomic mass of primary particles shows no significant energy dependence and corresponds to $⟨\ln A⟩=2.0\pm 0.1(\mathrm{stat}.)\pm 0.44(\mathrm{syst}.)$. The result is compared to the mass composition obtained by the Telescope Array with ${\mathrm{X}}_{\max }$ technique along with the results of other experiments. Possible systematic errors of the method are discussed.

• Received 14 August 2018
• Revised 26 December 2018

DOI:https://doi.org/10.1103/PhysRevD.99.022002