A global test of factorization for nucleon-nucleon, gamma p and gamma gamma scattering

The purpose of this note is to show that the cross section factorization relation $\sigma_{nn}(s)/\sigma_{\gamma p}(s) = \sigma_{\gamma p}(s)/ \sigma_{\gamma\gamma}(s)$ is satisfied experimentally in the energy domain $8\le\sqrt s\le 2000$ GeV, where the $\sigma$'s are total cross sections and $nn$ denotes the even portion of the $pp$ and $\pbar p$ total cross section. A convenient phenomenological paramaterization for a global simultaneous fit to the $pp$, $\pbar p$, $\gamma p$ and $\gamma\gamma$ total cross section data together with the $\rho$-value data for $pp$ and $\pbar p$ is provided by using real analytic amplitudes. Within experimental errors, we show that factorization is satisfied when we unfold the published $\gamma\gamma$ data which had averaged the cross sections obtained by using the two different PHOJET and PYTHIA Monte Carlo results. Our analysis clearly favors the PHOJET results and suggests that the additive quark model, together with vector meson dominance, allows one to compute $\sigma_{\gamma p}(s)$ and $\sigma_{\gamma\gamma}(s)$ from $\sigma_{nn}(s)$ with essentially no free parameters. The universal $\rho$-value predicted by our fit, {\em i.e.,} $\rho_{nn} = \rho_{\gamma p} = \rho_{\gamma\gamma}$, is compared to the $\rho$-value obtained by a QCD-inspired analysis of $\pbar p$ and $pp$ data, including the p-air cross sections from cosmic rays. The $\rho$-values obtained from the two techniques are essentially indistinguishable in the energy region $8\le\sqrt s\le 2000$ GeV, giving us increased confidence in our parameterization of the cross sections needed for the factorization relation.