Impact Methodologies

Land Development Patterns

Forecasting Methods

Method 1. Proximity-Based Forecasting

This approach forecasts land development based on proximity to highway or transit facilities. It is based on regression or other statistical analysis of historical development data that relates land development to proximity to major transportation facilities such as arterials, interstate highways, or rail transit stations.

• Sanchez (1999) uses remote sensing and statistical models to relate highways to land development in Oregon. A growth trend analysis is performed using a GIS to overlay the extent of urban development for cities over time (derived from aerial photography). Regression analysis is used to relate development to proximity to highway projects. In conjunction with case studies, the combined results are being used to prepare a methodology that the Oregon DOT can apply during the highway impact assessment process when considering potential land use changes.

• The Land Transformation Model (Pijanowski) uses a probability-based approach to predict urban land conversion in Michigan. The primary focus of the project is on coupling land use change and hydrogeologic and geochemical processes. Land conversion is forecast in part based on proximity to urban infrastructure.

Method 2. Delphi/Expert Panel

A group of "experts" including local officials, developers, academics, etc., is gathered to predict the land development impacts of a project.

• Applications of this approach include U.S. 301 in Maryland (Porter et al., 1997); Santa Clara County, CA (Cavalli-Sforza and Ortolano, 1984); and guidance provided by the Wisconsin DOT (WisDOT, 1993).

Method 3. Accessibility-Based Forecasting

Quantitative accessibility measures, derived from travel demand models or other sources, are used to estimate changes in development.

• Hirschman and Henderson (1990) use a gravity-based accessibility measure to forecast land use impacts of highway improvements. A gravity model is developed to predict residential location decisions based on employment accessibility. Employment accessibility for residential zones is calculated based on highway network travel time data from the travel demand model.

Method 4. Simple Land Use Models

Simple land use models consist of sets of equations to forecast land development by zone, using a limited amount of data for model calibration and inputs. While accessibility is a primary driver of development in these models, other factors may also be included to the extent that data are available.

• HLFM II+ is a relatively simple forecasting model that relies on spatial interaction (accessibility). It is designed for smaller MPOs with small budgets and staff for land use modeling. An example of its application to Vancouver, Canada is described in Rice (1998).

• The Capital District Transportation Commission developed a land use model for Albany, NY (see Albany case study). The model consists of a set of three equations to predict incremental residential and employment development. Development is forecast based on accessibility, historical development trends, home prices, and property tax rates.

Method 5. Complex Land Use Models

More complex land use models have been developed to model a larger range of factors and relationships that affect land development. These models may contain both a transportation and a land use modeling component, or they may consist of a land use model that interfaces with an existing regional travel demand model. They typically cover an entire metropolitan region and consist of a zonal structure, similar to travel demand models. Consistent with regionwide forecasts of population and employment, they allocate development to each zone based on transportation accessibility, land prices, available land by development type, and/or other parameters. The models are typically calibrated using historical data on land development, prices, transportation accessibility, and other factors. DRAM/EMPAL is the most widely applied model in the U.S. Other examples of recent U.S. applications include:

• UrbanSim (Waddell, 1998) is an integrated transportation-land use model that is being applied in Portland, Oregon; Salt Lake City, Utah; and Honolulu, Hawaii to test alternative regional land development scenarios and transportation policies. The model implements a perspective on urban development that represents a dynamic process resulting from the interaction of many actors making decisions within the urban markets for land, housing, non-residential space and transportation. By treating urban development as the interaction between market behavior and governmental actions UrbanSim is designed to maximize reality, thereby increasing its utility for assessing the impacts of alternative governmental plans and policies related to land use and transportation.

• The Baltimore Metropolitan Council (Liu, 1998) is modeling the impacts of Smart Growth policies using TRANUS, an integrated transportation/land use model. TRANUS has also been applied by the Oregon Department of Transportation for a statewide modeling project (cf., Donnelly and Upton 1998, or Oregon Department of Transportation Modeling Program.)

• MEPLAN has been applied in Sacramento, CA to test the impacts of a variety of regional transportation and land use scenarios (see Sacramento case study). MEPLAN also has been widely applied in Europe to model metropolitan land use patterns (see the SPARTACUS case study).

• Other models and applications are discussed in Parsons Brinckerhoff Quade and Douglas (1999).

[TOP]