Empirical models for the energy distribution of the sun, as seen after atmospheric scattering, show a strong correlation on an annual or month-by-month basis to observed data. When applied to cases where the requirement is for a real-time solar energy distribution, such as in the optimisation of the flux distributions in imaging concentrators, these models prove insufficient. In this paper we present results illustrating trends in observed solar profiles that are invariant to changes in location that lay the framework to a definitive solar model. We show how using this information, a more complete understanding of the effect of a change in the spatial energy distributions of the sun can have on the size of the spatial energy distribution in the absorber plane of a (linear Fresnel) concentrating collector. (C) 2003 Elsevier Ltd. All rights reserved.