A Model for Producing Spatially Explicit Future Population Scenarios

Graduate Thesis Or Dissertation

Citations

Citeable URL: https://scholar.colorado.edu/concern/graduate_thesis_or_dissertations/vm40xr673

Abstract

Spatially explicit projections of global population are of growing importance in the large-scale scenario-based assessment of, for example, the spatial distribution of land use, demand for food and water, energy usage, and emissions. In this research we present a method for producing large-scale, plausible spatial population scenarios. We begin with the assessment of one of the most sophisticated existing methodologies developed at IIASA and oriented around the gravity-based population potential model. We then propose and test significant modifications to the procedure, and apply the modified approach in a test against historical United States data. We conclude by producing a small set of 100-year spatial scenarios for the continental United Stated based on the SRES A2 storyline. We find that the IIASA methodology is subject to border effects, can produce limited patterns of spatial change, and is limited in its capacity to classify and appropriately model urban and rural patterns of change. Through the course of our work we propose modifications to address these and other problems. The unique contributions of this research include a method for calibrating the model to historical data to capture and replicate observed patterns of spatial population change and a method for removing boundary effects from a gravity-based spatial-allocation model with moving windows.

Creator