Frontiers in Physics (Oct 2020)

Initial Value Dependence of Urban Population's Growth-Rate Distribution and the Long-Term Growth

  • Atushi Ishikawa,
  • Shouji Fujimoto,
  • Arturo Ramos,
  • Takayuki Mizuno,
  • Takayuki Mizuno,
  • Takayuki Mizuno

DOI
https://doi.org/10.3389/fphy.2020.00302
Journal volume & issue
Vol. 8

Abstract

Read online

This paper uses census municipal population data for the United States, Italy, and Spain to analyze the statistical properties of their 10-year growth (short-term property). As a result, it was confirmed that the smaller the initial urban population is, the greater the probability that the urban population will decrease and that the probability that the urban population will increase does not depend on the initial urban population. We also observed the statistical properties of long-term growth of urban populations in each country over 100 years. Specifically, we identified the following properties by observing the geometric mean of logarithmically equal sized bins of the oldest urban population in the data used in the analysis. (1) The average urban population increases or decreases exponentially with time. (2) The smaller the initial average urban population, the smaller the exponent, which can be negative in Italy and Spain. (3) When the average urban population is large, exponential growth may stop. We showed that these long-term properties are derived from the short-term property by random sampling simulations from real data.

Keywords