Dynamic population distribution patterns: A comparison of three datasets for Tokyo, Japan

Authors: Toshinori Ariga*, National Institute for Environmental Studies Japan, Shih-Lung Shaw, The University of Tennessee
Topics: Urban and Regional Planning, Asia
Keywords: population distribution, Landscan, mobile phone, census
Session Type: Virtual Paper
Day: 4/10/2021
Start / End Time: 9:35 AM / 10:50 AM
Room: Virtual 9
Presentation File: No File Uploaded


Population distribution is an essential consideration to address various social and environmental issues such as public service, energy consumption, and evacuation. Many studies have relied on population census data, which are based on residential locations, to show population distribution patterns that reflect mainly night-time population distributions. Humans are dynamic in nature and we move around different places to fulfill various needs. Population distributions therefore change with time in a day. One important question is to find out to what extent different datasets could affect a study across the locations of various characteristics and at different times in a day to better reflect the dynamic spatial population distribution patterns. This paper compares population distribution patterns in Tokyo, Japan based on three different datasets, which are (1) population census data that mainly reflects night-time population distribution, (2) Oak Ridge National Laboratory’s Landscan data that reflects 24-hour average population, and (3) bi-hourly population estimates derived from mobile phone data. First, all three datasets were partitioned into 1km x 1km grid cells. Next, we computed the differences in population estimates among the three datasets for each grid cell at different times in a day. We then analyzed the differences in population estimates with respect to various types of points of interests to assess potential impacts of using different population datasets on various types of locations and at different times in a day for Tokyo, Japan.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login