Graduate Thesis Or Dissertation
 

Urban development in coastal Oregon : discrete-choice estimation with spatial autocorrelation

Public Deposited

Downloadable Content

Download PDF
https://ir.library.oregonstate.edu/concern/graduate_thesis_or_dissertations/r781wk20p

Descriptions

Attribute NameValues
Creator
Abstract
  • Development along the Oregon coast is continuing in areas that are known to be hazardous, in spite of strict zoning and other laws. The coast commonly sees erosion that can wash away cliffs and undermine structures as well as accretion that can bury houses in sand. However the much more acute risk is the potential for a tsunami triggered by an earthquake in the ocean. This paper examines the factors influencing development along the Oregon coast using satellite imagery and other spatial data. An econometric model for land use change at the pixel level is developed. The scale at which the data (satellite imagery) are collected is different from the scale at which development occurs, leading to spatial correlation among pixels that are geographically close to each other. Estimating a standard probit model in this case leads to parameter estimates that are inconsistent. Incorporating a spatial lag of the dependent variable will account for the spatial autocorrelation but makes maximum likelihood estimation nearly impossible so a Bayesian approach is used instead. A Gibb's sampling algorithm is implemented to estimate the conditional distribution of each parameter in the model, from which parameter estimates can be derived. The Bayesian spatial probit estimation is very computationally intensive however the results indicate that the spatial lag is a crucial part of the model. This approach also allows projections of the spatial pattern of future development, unlike estimation methods that sample the data to remove autocorrelation among observations. The spatial lag model is then used to project future development patterns in several regions along the Oregon coast and explore potential applications of this projection model. In particular, predicted development patterns in the tsunami hazard zone are examined in the Waldport area, on the central coast of Oregon.
License
Resource Type
Date Available
Date Issued
Degree Level
Degree Name
Degree Field
Degree Grantor
Commencement Year
Advisor
Committee Member
Academic Affiliation
Non-Academic Affiliation
Subject
Rights Statement
Publisher
Peer Reviewed
Language
Replaces

Relationships

Parents:

This work has no parents.

In Collection:

Items

Thumbnail Title Date Uploaded Visibility Actions
file details Todd_thesis.pdf 2017-08-19 Public Download