Determination of cosmic-ray primary mass on an event-by-event basis using radio detection

We present a new methodology to discriminate between light and heavy ultra-high energy cosmic-ray primaries on an event-by-event basis using information from the radio detection of extensive air showers at MHz frequencies. Similarly to other methods to determine primary cosmic ray composition, the one presented here is based on comparisons between detected radio signals and Monte Carlo simulations for multiple primary cosmic ray compositions. Unlike other methods that first reconstruct the depth of maximum shower development $X_{\rm max}$ to relate it to the nature of the primaries, we instead infer the cosmic-ray composition directly. The method is most effective in the case of inclined showers that arrive at large zenith angles with respect to the vertical to the ground, where methods based on the determination of $X_{\rm max}$ lose accuracy. We show that a discrimination efficiency between 65% and 80% can be reached for zenith angles $\theta \gtrsim 60^{\circ}$, even when typical uncertainties in radio detection are taken into account, including shower energy uncertainty. Our methodology could in principle be applied in large and sparse radio arrays, designed with the large radio footprint of inclined showers in mind, to significantly increase the statistics of ultra-high energy cosmic-ray composition studies.