The Mass Composition of Cosmic Rays above 10 Ev* 1. the Scientific Motivation for Studying the Highest Energy Cosmic Rays

Our knowledge of the mass composition of cosmic rays is deficient at all energies above 10 17 eV. Here, systematic differences between different measurements are discussed and, in particular, it is argued that there is no compelling evidence to support the common assumption that vast majority of the cosmic rays of the highest energies are protons. Our knowledge of the mass needs to be improved if we are to resolve uncertainties about the energy spectrum. Improvement is also needed for proper interpretation of data on the arrival direction distribution of cosmic ray. Since the recognition in 1966, by Greisen and by Zatsepin and Kuzmin, that protons with energies above 4 × 10 19 eV would interact with the cosmic microwave radiation, there has been great interest in measuring the spectrum, arrival direction distribution and mass composition of ul-tra high-energy cosmic rays (UHECR), defined as those cosmic rays having energies above 10 19 eV. Specifically they pointed out that if the sources of the highest energy protons are universally distributed , there should be a steepening of the energy spectrum in the range from 4 to 10×10 19 eV. This feature has become known as the GZK 'cut-off' but the sharpness of the steepening expected depends on unknown factors such as the evolution and production spectrum of the sources. If the UHECR are mainly Fe nuclei then the spectrum is also expected to steepen, but it is harder to predict the character of this feature as the relevant diffuse infrared photon field is poorly known: steepening is expected to set in at higher energy. 2. Importance of mass composition for accurate estimates of the energy spectrum The energy spectrum of cosmic rays has frequently been inferred from observations of signals in arrays of scintillators or water-Cherenkov detectors (see Nagano and Watson [1] for a review). In practice, the energy is derived from a measurement of the detector response at (typically) 600 m from the shower axis using the results of detailed Monte Carlo calculations. For example , in the most recent reports of spectra from the AGASA [2] and Haverah Park groups [3], the QGSJET model of high energy interactions has been adopted. Additionally, in the AGASA work, a study of the impact of different models was made and it was found, for example, using QGSJET98 in the Corsika propagation code, that the energy estimates for protons or iron nuclei differ …