The High Resolution Fly's Eye (HiRes) experiment has observed the Greisen-Zatsepin-Kuzmin suppression (called the GZK cutoff) with a statistical significance of five standard deviations. HiRes' measurement of the flux of ultrahigh energy (UHE) cosmic rays shows a sharp suppression at an energy of $6 \times 10^{19}$ eV, consistent with the expected cutoff energy. We observe the ``ankle'' of the cosmic-ray energy spectrum as well, at an energy of $4 \times 10^{18}$ eV. We describe the experiment, data collection, analysis, and estimate the systematic uncertainties. The results are presented and the calculation of the statistical significance of our observation is described.

We have studied several sources of systematic uncertainty in calculating the aperture of the High Resolution Fly's Eye experiment (HiRes) in monocular mode, primarily as they affect the HiRes-II site. The energy dependent aperture is determined with detailed Monte Carlo simulations of the air showers and the detector response. We have studied the effects of changes to the input energy spectrum and composition used in the simulation. A realistic shape of the input spectrum is used in our analysis in order to avoid biases in the aperture estimate due to the limited detector resolution. We have examined the effect of exchanging our input spectrum with a simple E^{-3} power law in the "ankle" region. Uncertainties in the input composition are shown to be significant for energies below about 10^{18} eV for data from the HiRes-II detector. Another source of uncertainties is the choice of the hadronic interaction model in the air shower generator. We compare the aperture estimate for two different models: QGSJet01 and SIBYLL 2.1. We also describe the implications of employing an atmospheric database with hourly measurements of the aerosol component, instead of using an average as has been used in our previously published measurements of the monocular spectra.

Air fluorescence detectors traditionally determine the dominant chemical composit ion of the ultrahigh energy cosmic ray flux by comparing the averaged slant depth of the shower maximum, $X_{max}$, as a function of energy to the slant depths expect ed for various hypothesized primaries. In this paper, we present a method to make a direct measurement of the expected mean number of protons and iron by comparing the shap es of the expected $X_{max}$ distributions to the distribution for data. The advantages of this method includes the use of information of the full distribution and its ability to calculate a flux for various cosmic ray compositi ons. The same method can be expanded to marginalize uncertainties due to choice of spectra, hadronic models and atmospheric parameters. We demonstrate the technique with independent simulated data samples from a parent sample of protons and iron. We accurately predict the number of protons and iron in the parent sample and show that the uncertainties are meaningful.