In this paper we demonstrate how the spatial distribution of optical emission rates within an auroral arc may be recovered from rocket photometer measurements made in a tomographic spin scan mode. We describe the tomographic inversion procedures required to recover this information and the implementation of two inversion algorithms that are particularly well suited for dealing with the problem of noise in the observational data. These algorithms are based upon the algebraic reconstruction technique and utilize "least-squares" and "maximum-probability" iterative relaxation methods. The performance of the inversion algorithms and the limitations of the rocket tomography technique are assessed using various sets of simulated rocket measurements that were generated from "known" auroral emission-rate distributions. The simulations are used to investigate how the quality of the tomographic recovery may be influenced by various factors such as noise in the data, rocket penetration of the auroral form, background sources of emission, smearing due to the photometer field of view, and temporal variations in the auroral form.