Degree
Doctor of Philosophy (PhD)
Department
Geography
Document Type
Dissertation
Abstract
Urban studies have long focused on modeling population density functions and analyzing the theoretical and behavioral foundations of urban structures. Simultaneously, understanding excess commuting, which highlights inefficiencies in urban transportation, is crucial for urban planning and transportation analysis. This dissertation bridges these perspectives by leveraging big data from individual vehicle trips in Tampa, Florida, to explore the relationship between population and employment distribution patterns and measure excess commuting with greater precision.
By addressing the modifiable areal unit problem (MAUP), this study refines the understanding of population and employment distributions. It aggregates data into uniform areal units and thus effectively examines scale and zonal effects. This approach reveals that the exponential density function best fits the population density pattern across various areal units, including census units and designed uniform units such as squares, triangles, and hexagons. While the zonal effect is not significant across uniform units, the scale effect remains evident across all areal units.
Utilizing a combination of Linear Programming (LP) and Integer Linear Programming (ILP) approaches, this research compares minimum and actual commuting patterns to examine excess commuting. The analysis at both the census tract and block group levels, as well as at the individual trip level, demonstrates that smaller spatial units yield shorter minimum commuting times and distances, and highlights the scale effect of MAUP. At the census tract level, the average minimum commuting time and distance are 5.67 minutes and 3.04 kilometers, respectively while at the block group level, these values are 5.05 minutes and 2.68 kilometers. The ILP analysis at the individual trip level shows a significantly lower minimum commuting in time or distance, namely 3.6 minutes and 2.27 kilometers, respectively.
This dissertation employs detailed big data to offer a comprehensive and accurate assessment of urban population density functions and excess commuting. The findings emphasize the importance of using individual data in urban transportation analysis to mitigate the MAUP, provide refined insights into urban structure and transportation inefficiencies, and ultimately contribute to more effective urban planning and policymaking.
Date
10-14-2024
Recommended Citation
Luo, Cehong, "Modeling Urban Population Density Functions and Excess Commuting: A Big Data Approach to the Modifiable Areal Unit Problem" (2024). LSU Doctoral Dissertations. 6592.
https://repository.lsu.edu/gradschool_dissertations/6592
Committee Chair
Fahui Wang