Statistical interpretation of the correlations between forward and backward hadrons at collider energies

Abstract

Given a multiplicity distribution belonging to the class of probability distributions which are superpositions of Poisson distributions whose two components are independently (binomially) distributed, we derive joint and conditional probabilities for the two components. Specializing to the negative binomial case, we can explain the linearity and magnitude of slope and intercept of the forward-backward correlation in a way compatible with the KNO plot for the multiplicity data provided that the final particles are produced in clusters. Generalization to allow for coherent emission allows one to put limits on the amount of coherence, a result not known from high precision fits to multiplicity.