Dissertation/Thesis Abstract

Modeling land use change using an eigenvector spatial filtering specification and sprawl measures: A case study of Collin County, Texas
by Sinha, Parmanand, Ph.D., The University of Texas at Dallas, 2015, 121; 3706684
Abstract (Summary)

To date, sprawl measures have been neglected in land use forecasting models. This dissertation focuses on incorporating sprawl measures based on smart growth principles together with adjacency of nearby land uses. The measures have been applied with an allocation process using an Integer Programming (IP) model that maximizes net suitability of land uses for Collin County, Texas. The suitability of land cells from vacant to five mutually exclusive land use categories: (1) Single Family (SF); (2) Multi-Family (MF); (3) Commercial (C); (4) Industrial (I); (5) Open Space (OS) is calculated by using spatial discrete choice analysis employing built environment, socioeconomic and demographic characteristics for 70,129 vacant cells (150m-by-150m). An eigenvector spatial filtering (ESF) specification has been used to describe land uses, which is a relatively new technique that accounts for the presence of spatial autocorrelation. To date eigenvector spatial filtering (ESF) has been applied to a binary dataset and requires shorter and more straightforward computation compared to other estimation techniques but in the case of a multinomial dataset, the computation time significantly increases. To handle large multinomial dataset, the ESF specification resolution used in this study is made coarser than the response variables. This reduces the computation time enormously while giving good goodness of fit. Sprawl measures have been applied comprehensively based on mixed use factors and spatial dependences arising from the proximity between land uses. This dissertation contribution is innovative because spatial proximity between land uses, which has been ignored to date, can be used to control sprawl, resulting in a better mixing of different land uses based on the constraints imposed in a spatial optimization problem.

Indexing (document details)
Advisor: Griffith, Daniel A.
Commitee: Chun, Yongwan, Dean, Denis J., Murdoch, James C.
School: The University of Texas at Dallas
Department: Geospatial Information Sciences
School Location: United States -- Texas
Source: DAI-B 76/10(E), Dissertation Abstracts International
Source Type: DISSERTATION
Subjects: Geographic information science, Transportation planning, Urban planning
Keywords: Eigenvector spatial filtering, Land use change, Multinomial logit, Spatial discrete choice model, Spatial logit, Spatial optimization
Publication Number: 3706684
ISBN: 9781321807363
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy
ProQuest