My colleague Annabel and I recently gave a talk as part of the Engineering Mathematics Seminar Series at the Department of Engineering Mathematics, University of Bristol. It was called Ambient populations: Developing robust estimates and presented our work on estimating the ambient population for uses in crime analysis as well as our ideas for new methods to integrate various datasets to create a more accurate measure of the ambient population. The slides are available here and the abstract is:

There is typically an abundance of information about where people live (the ‘residential’ population), but much less information regarding the non-residential, or ‘ambient’ population. The ability to produce estimates of the ambient population is integral to the management and planning of urban areas and is a precursor to the development of insights into socioeconomic and environmental issues that impact cities. In this presentation we provide examples that illustrate why a robust quantification of the ambient population is essential to properly understand many social phenomena, drawing particularly on the analysis of crime patterns. We also review a number of new data sources that have emerged in recent years that might help to shed light on the ambient population and propose a means of incorporating disparate data to estimate the ambient population using geographically-weighted regression.