An improved algorithm for reducing reflectometry data involving divergent beams or non-flat samples

Reflectometry is a powerful technique for determining many physical quantities of stratified media, including length scales, densities and magnetism. However, experimentally neutron reflectometry in particular suffers from the relatively feeble brilliance of the sources compared with those of X-rays, for example. In this paper, a simple modification of existing data-reduction methods is demonstrated, allowing quantitative improvements in the quality of the data. Using the same algorithm, reflections from non-flat surfaces can be treated, leading to a full recovery of the resolution. The method involves re-binning of the data in the linear coordinates of the raw data, which leads to substantial gains in statistical quality, equivalent to a significant flux increase, and also improved resolution.