SWUTC Research Project Description
The Transportation-Related Causes and Consequences of Land Use Change
University: University of Texas at Austin
Principal Investigator:
Chandra Bhat
Department of Civil and Environmental Engineering
(512) 471-4535
Project Monitor:
Johanna Zmud
Senior Transportation Policy Researcher
RAND Corporation
200 South Hayes Street
Arlington, VA 22202-5050
Funding Source: USDOT and State of Texas General Revenue Funds
Total Project Cost: $79,285
Project Number: 600451-00063
Date Started: 5/1/12
Estimated Completion Date: 6/30/13
Project Summary
Project Abstract:
This study focuses on developing models for predicting shares of land-use by extending the geographic scope of individual landowner level analysis. By recognizing that the shares of land-use at the grid level are a result of aggregation of decisions of individual land owners, the current research will develop an approach to estimate the parameters of the underlying process model at the individual level. Furthermore, the proposed model will explicitly consider spatial dynamics caused by interdependence among individual landowners of proximally located properties. The resulting model can be embedded within an agent based microsimulation system to study the impact of localized as well as large-scale transportation and land-use policies.
Project Objectives:
It is clear that land-use patterns in any given region significantly impact the travel patterns of people residing in or traveling to that region. Thus, it is important to model future land-use patterns as a function of potential land-use and transportation policies to promote sustainable travel patterns. In this proposed project, the objective is to estimate a grid-level land-use model that explicitly accounts for the underlying individual landowner level behavioral processes using the City of Austin land use database for multiple years (1995, 2000, 2003, and 2006).
The resulting process model can be embedded within an agent-based microsimulation system to study the effects of localized as well as large-scale transportation and land-use policies. The proposed formulation will explicitly consider spatial dynamics caused by interdependence among individual landowners that lead to the land-use decision of one landowner affecting those of the landowners of proximally located properties.
Task Descriptions:
TASK 1: Assemble land-use and transportation data bases
In this task, we will assemble land-use and transportation data bases from Austin obtained using a combination of aerial photography and real-estate appraisal information. It is very likely that we will have to append data from several secondary data sources to ensure that we include all the exogenous factors that might impact land use decisions. Given that data availability has been one of the main constraints for undertaking a similar study in the past, this task will take up a significant amount of time.
TASK 2: Preparation of data and sample for analysis
In this task, we will process the land-use and transportation data bases from Austin, and undertake exploratory analysis of the data using simple statistical methods. We will descriptively examine and structure the data in a form suitable for estimation. This will be a critical first step in our process of developing the land-use model. In particular, the descriptive statistics will help inform the specification of the dependent variables as well as the independent variables. Also, the task will involve computing several additional measures such as distance to nearest highway, travel time to the city centre, and accessibility by transit and non-motorized modes.
TASK 3: Develop the model formulation
The problem with the existing land-use models is that while aggregate pattern-based models do not have any behavioral connection to how landowners make decisions, the process-based individual models have limited geographic scope. In this study, we will formulate an approach to extend the geographic scope of process-based models by using grids as the unit of analysis. Within each grid, there can be multiple types of land-uses, since a grid is an aggregation of parcels. But by knowing the shares of the land-use in each grid, and expressly recognizing these shares as the result of the aggregation of decisions of individual landowners, we will develop an approach to estimate the parameters of the underlying process model at the individual level.
TASK 4: Model estimation and specification analysis
Estimate a land-use model that explicitly considers spatial dynamics caused by interdependence among individual landowners. This will include estimation code development and testing a variety of different spatial proximity specifications and model specifications. The final model specification will be based on a systematic process of eliminating variables found to be statistically insignificant, intuitive considerations, parsimony in specification, and results from earlier studies. Several distance based proximity specifications, different variable specifications, and functional forms of variables as well as interaction variables will be examined. At the time of estimation, a clear structure for embedding the model within an agent-based microsimulation setting will also be developed.
TASK 5: Compute elasticities for policy analysis
Compute elasticities to evaluate the impacts of exogenous transportation, socioeconomic, and policy variables. This task will help quantify the impact of different policies on future land-use patterns and facilitate policy design.
TASK 6: Create land use maps for future years
Create land use maps for several future years based on the model predictions, and evaluate the potential effects of alternative land-use and transportation policies. As part of this effort, promising land-use planning strategies to promote sustainable travel patterns will be identified and documented.
TASK 7: Prepare and submit final report
This task will prepare and deliver a final research report documenting the research performed, including the literature synthesis, methodology, data analysis, results, and recommendations.
Implementation of Research Outcomes:
It is important to model future land-use patterns as a function of potential land-use and transportation policies to promote sustainable travel patterns. This research effort estimated a grid-level land-use model that explicitly accounts for the underlying individual landowner level behavioral processes using the City of Austin land use database for multiple years.
Products developed by this research include:
Presentation: A New Spatial Multiple Discrete-Continuous Modeling Approach to Land Use Change Analysis, Subodh K. Dubey and Chandra R. Bhat presented to an Invited Seminar, Indian Institute of Technology, Kanpur, India, June 2013.
Presentation: A New Spatial Multiple Discrete-Continuous Modeling Approach to Land Use Change Analysis, Subodh K. Dubey and Chandra R. Bhat presented to the International Choice Modelling Conference, Sydney, Australia, July 2013.
Presentation: A New Spatial Multiple Discrete-Continuous Modeling Approach to Land Use Change Analysis, Subodh K. Dubey and Chandra R. Bhat. To be presented at the 60th Annual North American Meetings of the Regional Science Association International, Atlanta, GA, November 2013.
Presentation: On Incorporating Spatial Dependence in a Multiple Discrete-Continuous Choice Model: Formulation and Estimation Approach, Chandra R. Bhat and Subadh K. Dubey, to be presented at the 93rd Annual meeting of the Transportation Research Board, Washington, DC, January 2014.
Journal Article in Review: A New Spatial Multiple Discrete-Continuous Modeling Approach to Land Use Change Analysis, Subodh K. Dubey and Chandra R. Bhat submitted to Journal of Regional Science.
Impacts/Benefits of Implementation:|
The results from this research will allow analysts to more accurately examine the effects of land use and zoning policies to avoid (or slow down) low density developments. They will also aid in identifying land-use and transportation policies that can curb greenhouse gas emissions.
Web Links: