Assessment of a gravity-based approach to constructing future spatial population scenarios

Abstract

Spatially explicit projections of global population are of growing importance in scenario-based assessment of anthropogenic climate change and climate related vulnerability. However, to date there are very few spatially explicit projections of the future distribution of global population, and methods for producing such projections are in their infancy. One of the most sophisticated methods currently available, developed as part of the Greenhouse Gas Initiative at the IIASA (Grübler et al. in Technol Forecast Soc Change 74(7):980-1029, 2007), uses the gravity-based population potential model to produce scenario dependent projections. Population potential is widely used as a descriptive tool and in models of spatial allocation. While it is attractive through its simplicity and ability to summarize complex patterns, the population potential model suffers from several well documented problems. This paper provides an assessment of population potential as a tool for constructing future spatial population scenarios, particularly within the context of the pitfalls associated with the model. Despite improvements over other existing methods, it was found that the model is capable of producing only a narrow range of spatial outcomes, is subject to border effects, is restrictive in its handling of urban and rural population dynamics, and will in most cases misallocate population loss. The potential-allocation approach to producing large-scale spatial population scenarios is promising, but several methodological modifications are necessary to address the model's shortcomings. © 2014 The Author(s).