Simulating the spatial distribution of population and emissions to 2100

Abstract

Urbanization and economic development have important implications for many environmental processes including global climate change. Although there is evidence that urbanization depends endogenously on economic variables, long-term forecasts of the spatial distribution of population are often made exogenously and independent of economic conditions. It is common for research concerning long-run projections of global environmental change to use population density as the primary means to spatially distribute emissions projections. However, researchers typically utilize year 1990 cross-sectional population data to distribute their emissions projections for both the short- and long-term, without projecting any changes in population density. Thus, a beta distribution for individual countries/regions is estimated to describe the geographical distribution of population using a one-degree-by-one-degree latitude–longitude global population data set. Cross-sectional country/regional data are then used to estimate an empirical relationship between parameters of the beta distribution and macroeconomic variables as they vary among countries/regions. This conditional beta distribution allows the simulation of a changing distribution of population, including the growth of urban areas, driven by economic forecasts until the year 2100.