United States
Environmental Protection
Agency
Office of Research and
Development
Washington DC 20460
EPA/600/R-00/098
September 2000
www.epa.gov
oEPA
Projecting Land-Use Change
A Summary of Models for Assessing the
Effects of Community Growth and Change on
Land-Use Patterns
-------
-------
EPA/600/R-00/098
September 2000
www.epa.gov
Projecting Land-Use Change:
A Summary of Models for Assessing the Effects of
Community Growth and Change on Land-Use Patterns
Science Applications International Corporation
11251 Roger Bacon Drive
Reston, VA 20190-5201
Contract #68-07-0011
Project Officer
Susan Schock
U.S. Environmental Protection Agency
Cincinnati, OH 45268
National Exposure Research Laboratory
National Health and Environmental Effects Research Laboratory
National Risk Management Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
-------
Notice
The summaries of land-use change models contained within this report do not reflect the actual
use of these products. Information presented on the function and capabilities of each model
was collected largely through independent research of published materials such as users'
manuals and Internet sites. The model developers have reviewed and verified the information
compiled for their specific products, ensuring the accuracy of the material presented in this
report. Significant efforts were made to explore and incorporate all land-use models currently
available for public use during the project period. However, the population of such models is
continually changing and it is recognized that any compilation will ultimately exclude one or
more relevant models.
Preparation of this document has been funded wholly or in part by the U.S. Environmental
Protection Agency. It has been subjected to the Agency's review process, and has been
approved for publication as NHEERL draft number OD-01-001. Mention of trade names or
commercial products does not constitute endorsement or recommendation for use.
This report should be cited as follows:
U.S. EPA, 2000. Projecting Land-Use Change: A Summary of Models for Assessing the
Effects of Community Growth and Change on Land-Use Patterns. EPA/600/R-00/098. U.S.
Environmental Protection Agency, Office of Research and Development, Cincinnati, OH.
260 pp.
11
-------
Acknowledgments
This report reflects an uncommon degree of collaboration among the three laboratories of the
U.S. Environmental Protection Agency's Office of Research and Development (ORD), as well
as selected program and regional offices across the Agency. Funding was provided by the
Administrator's Office of Congressional and Intergovernmental Relations (OCIR) and ORD's
National Exposure Research Laboratory (NERL), National Health and Environmental Effects
Research Laboratory (NHEERL), and National Risk Management Research Laboratory
(NRMRL). Laura Jackson of NHEERL directed the development of this report, with support
from Jim Kreissl of NRMRL. Many thanks go to Linda Rimer for lead sponsorship through
OCIR's Sustainable Urban Environments Task Force.
Science Applications International Corporation (SAIC), under contract to ORD, provided key
technical support. Led by Gary Gaunt, SAIC's staff included Kellie DuBay, Christine Garrow,
Jana Lynott, Mary O'Kicki, Michael Palace, Daniel Sklarew, and Elizabeth Smeda.
In addition, the following people made essential contributions to this report:
Technical Workgroup
(Laura Jackson, U.S. EPA National Health & Environmental Effects Research Laboratory)
(Jim Kreissl, U.S. EPA National Risk Management Research Laboratory)
Ron Matheny, U.S. EPA National Exposure Research Laboratory
Peter Washburn, U.S. EPA Office of Congressional and Intergovernmental Relations (formerly)
Mark Flory, U.S. EPA Office of Congressional and Intergovernmental Relations
Geoff Anderson, U.S. EPA Office of Policy and Reinvention
John Thomas, U.S. EPA Office of Policy (formerly)
Victor McMahan, U.S. EPA Office of Air
Vicki Sandiford, U.S. EPA Office of Air
Mark Wolcott, U.S. EPA Office of Air
Sumner Crosby, U.S. EPA Region III (formerly)
Tom Brody, U.S. EPA Region V
Steve Goranson, U.S. EPA Region V
Michael Gulp, U.S. Dept. of Transportation, Federal Highway Administration
Tom Gunther, U.S. Dept. of Interior, Office of Water and Science
Joe Tassone, Maryland Department of Planning
Bernie Engel, Purdue University
Dick Klosterman, University of Akron
Bob Johnston, University of California-Davis
Bryan Pijanowski, Michigan State University
Eliot Allen, Criterion Planners/Engineers, Inc.
in
-------
Peer Reviewers
Chris Bird, Alachua County, Florida
Angela Harper, County of Henrico, Virginia
Robert Holm, City of Indianapolis, Indiana
John Schlegel, Clark County, Nevada
Ellen Walkowiak, City of De Moines, Iowa
Morgan Grove, USDA Forest Service
Trade Jackson, U.S. EPA Office of Air
Model Developers
Finally, grateful recognition goes to the model developers for their patience and helpfulness in
providing two reviews and final approval of all model information. The name of each model
developer is featured at the top of the respective model fact sheet in Section Five.
IV
-------
Abstract
Many potential clients for land-use change models, such as city and county planners,
community groups, and environmental agencies, need better information on the features,
strengths, and limitations of various model packages. Because of this growing need, the U.S.
Environmental Protection Agency (EPA) has developed a selective summary of 22 leading land-
use change models currently in use or under development. Partners in scoping this effort
include the U.S. Departments of Transportation and Interior, the academic and consulting
communities, and multiple program and regional offices across EPA.
EPA's Office of Research and Development (ORD) initiated the land-use change models
summary in order to improve its ability to assess and mitigate future risk to ecological systems,
human health, and quality of life. Target user groups for this publication are:
• Community planners, citizens, and decision makers who are seeking tools to analyze
future land-use scenarios;
• EPA program office and regional staff who support communities with planning tools and
information for sustainable development; and
• ORD modelers and research planners who are currently assessing land-use models and
gaps in the state of the science.
-------
VI
-------
Contents
Page
Notice ii
Acknowledgments iii
Abstract v
1.0 Using This Guide 1
1.1 What Does this Guide Cover? 1
1.2 Who Should Use this Guide? 2
1.3 How Should I Use this Guide? 3
2.0 Background 5
2.1 Defining Community 5
2.2 Growing Pains 5
2.3 Asking the Right Questions 6
2.4 Using the Right Tools 8
3.0 Decision-Making Tools for a New Planning Approach 11
3.1 Models: An Overview 11
Land-Use Models 11
Transportation Models 12
Economic Models 12
Environmental Impact Models 12
3.2 Geographic Information Systems (GISs) 13
3.3 Integrated Planning and Decision-Making Systems 14
4.0 Selecting the Best Land-Use Model 15
4.1 Step 1. Understanding the Proposal 16
4.2 Step 2. Asking the Right Questions 16
4.3 Step 3. Identifying Information Needs 19
4.4 Step 4. Assessing Internal Capabilities 20
4.5 Step 5. Choosing the Right Model (Using Selection Criteria) 20
5.0 Summarizing the Land-Use Models 27
5.1 Developing This Summary 27
5.2 Land Use Models: Identified by Key Selection Criteria 31
5.3 Land Use Models: General Fact Sheets 35
vn
-------
Appendix A Land Use Models: Comparative Matrices A-1
Appendix B Land Use Models: Technical Fact Sheets B-1
Appendix C Current Trends in Community Growth and Planning C-1
1.0 Suburbanization: A Snapshot C-1
2.0 Changing Trends C-2
3.0 Initiatives C-3
3.1 State and Local Initiatives C-4
3.2 Regional Initiatives C-5
3.3 National Initiatives C-6
4.0 The Future of Community Planning C-9
Appendix D Key Terms and Definitions D-1
Appendix E References E-1
Vlll
-------
Exhibits
Exhibit 1-1. Descriptive Outline of Guide Sections and Appendices 3
Exhibit 2-1. Questions Facing Growing Communities 7
Exhibit 2-2. Selected Summary Reports on Transportation Models 8
Exhibit 3-1. Layers of a GIS 13
Exhibit 4-1. The Five-Step Process for Selecting a Land-Use Change Model 15
Exhibit 4-2. Opportunities for Employing Land-Use Change Modeling - Example 1 17
Exhibit 4-3. Opportunities for Employing Land-Use Change Modeling - Example 2 18
Exhibit 4-4. Opportunities for Employing Land-Use Change Modeling - Example 3 19
Exhibit 4-5. Example Criteria Rating Table 25
Exhibit 5-1. Land-Use Change Models Included in This Guide 29
Exhibit 5-2. Technical Expertise Needed to Use Model 32
Exhibit 5-3. Purchase Cost 32
Exhibit 5-4. Existence of Model Support 33
Exhibit 5-5. Ease of Transferring to Other Locations 34
Exhibit 5-6. How Many Locations Has the Model Been Applied To? 34
Exhibit A-1. Skills/Technical Expertise Comparative Matrix A-1
Exhibit A-2. Hardware Comparative Matrix A-3
Exhibit A-3. Software Comparative Matrix A-5
Exhibit A-4. Cost Comparative Matrix A-7
Exhibit A-5. Urban Land Use Categories Addressed Comparative Matrix A-9
Exhibit A-6. Non-Urban Land Use Categories Addressed Comparative Matrix A-11
Exhibit A-7. Impacts of Community Decisions on Land-Use Patterns
Comparative Matrix A-13
Exhibit A-8. Impacts of Land-Use Patterns on Community Characteristics
Comparative Matrix A-14
Exhibit A-9. Model Utility and Integration Comparative Matrix A-15
Exhibit A-10. Basic Operational Characteristics Comparative Matrix A-16
Exhibit A-11. Spatial And Temporal Capabilities Comparative Matrix A-20
Exhibit C-1. Characteristics Defining Sprawl C-1
Exhibit C-2. Adaptation of Anderson Land-Use Classification System for County or
Small City C-4
IX
-------
-------
1.0 Using This Guide
1.1 What Does this Guide Cover?
This guide provides a summary of 22 computer modeling tools that can be used to assess the impacts of
community actions and policies on land use, and the reciprocal effect of land-use changes on certain
community characteristics. In essence, it provides the avenue through which one can begin the process
of identifying the best land use modeling tool to help accomplish effective Smart Growth planning.
While a host of traditional planning tools (such as zoning and easements) do play a large role in Smart
Growth planning as management tools, these types of tools are not discussed in this guide. Rather, the
guide focuses solely on one important decision-making tool-land-use change models.
The models summarized in this publication are considered today's leading land-use change models
currently in use or under development. By using this guide, the reader will be able to readily determine
the models' applicability, data and resource requirements, strengths and limitations, and costs. Of note is
the fact that each model summarized in this guide has its own unique attributes and differs in ease of use.
The models range from the simple (i.e., they can be used by a nontechnical user on a desktop computer)
to the complex (i.e., they require consulting expertise, detailed data, and sophisticated computer
technologies). Some consider land use in the context of employment and housing activity, while others
explore the suitability of land for various intensities of use. Some models provide statistical output in the
form of tables, while others provide more visual output, such as maps. Because such a diverse array of
land-use change models exists, it is often difficult to identify the most useful one for a given situation, let
alone easily understand its individual applicability, strengths, and limitations. This guide attempts to
make a potentially difficult task much easier by providing background information on why there is a need
for planning models, a brief overview of the types of planning models, steps on how to choose the right
model for your needs, and the essential general and technical information on 22 leading land-use change
models.
Please note that this guide does not cover all models pertaining to growth-related issues, such as
transportation or environmental quality models. The discipline of modeling transportation/land-use
interactions is well established and several compendia already exist that document and compare leading
models of this type (e.g., Southworth, 1995; Miller et al., 1999; Parsons, Brinckerhoff, Quade, and
Douglas, Inc., 1999). As for environmental quality models, a companion document to this one is
currently under development to characterize models projecting the environmental impacts that may result
from land-use change. Many of the models contained within this summary, however, incorporate
sufficient flexibility to evaluate, with appropriate data, these and other growth-related issues.
-1-
-------
1.2 Who Should Use this Guide?
Anyone who is interested in analyzing land-use changes will find this guide beneficial-from federal
agencies that support state and local sustainable development efforts, to state and local organizations that
develop and implement comprehensive plans and growth-management policies, to students and other
interested citizens investigating land-use change. If you find you are grappling with questions
surrounding physical growth and change in your community, but do not know the best way to find the
answers, this guide is for you. Although many types of individuals and groups can use this guide, EPA's
primary intention is to meet the informational needs of the following audiences:
• Land-Use Planners and Decision Makers. On the local level, land-use planners play a key role
in guiding the process of community growth and development. They help a community achieve
its vision for the future by overseeing the development permitting process, ensuring that local
zoning ordinances are followed, and modifying plans to reflect the changing goals of the
community. They require sound information on how different land-use decisions will impact the
community's quality of life. Land-use planners not only use that information to perform their
jobs, they share it with other members of the community to aid in other decision-making
processes. Understanding which models are available and the type of information they can
provide will help land-use planners and decision-makers select the most appropriate tools to
assist in community land-use decisions.
• Citizens. Defining and fulfilling the vision of a community requires the participation
of community citizens. Land-use change models can assist community members in determining
the vision for their community by providing them with the information necessary to support-or
reject-policies affecting land use. Community members, both individuals and organizations,
who are focused on improving the quality of life for the overall community will benefit from an
increased awareness of these decision-making tools.
• Researchers. Researchers include federal representatives and academic staff who support
communities with new planning tools and information. Members of this group often participate
in the development and assessment of land-use tools. In addition to assessing existing tools,
researchers identify gaps to determine which other types of information or capabilities
communities may need to make sound decisions about growth and development. The summary
of land-use change models contained in this guide will provide researchers with a better
understanding of the tools currently available to communities and the direction their research
should take in the future. Supplementary technical information that may be particularly helpful
to researchers can be found in Appendix B.
-2-
-------
1.3 How Should I Use this Guide?
In addition to providing summary information on 22 land-use change models, this guide provides
background information on the need for these models and the key considerations for selecting the
appropriate model for your community. This guide is divided into five sections, plus appendices.
Because this guide was written for a large and varied audience, not all sections and appendices will be
relevant to all users. The table below provides a descriptive outline of the guide and indicates the target
audience for each section/appendix based on the user's knowledge level of land use policy, planning, and
modeling. If you are relatively new to land use policy and planning and don't know much, if anything,
about modeling, you would fall in the limited knowledge category. Some sections provide background
information that may be useful for you. However, if you have moderate to substantial land use policy
and planning experience and know the technical aspects of modeling, your knowledge level would be
considered extensive. A reader with extensive knowledge may wish to avoid the sections covering
background information but not want to miss the more technical information provided in Appendix B.
Use Exhibit 1-1 to determine how to navigate this guide as appropriate for your needs.
Exhibit 1-1. Descriptive Outline of Guide Sections and Appendices
Knowledge Level
Limited
Extensive
Section/Appendix
Section 1.0: Using This Guide
Section 2.0: Background
This section highlights the questions facing growing communities and the
need for land-use models as a tool for better planning decisions and smarter
growth.
Section 3.0: Decision-Making Tools for a New Planning Approach
This section provides a general overview of models-what they are, how they
work, and what type of information they can provide. It describes the
evolution of land-use change models, discusses how geographic information
system (GIS) technology serves as a key component of this emerging
modeling approach, and briefly describes the potential for integrated planning
and decision-making systems.
Section 4.0: Selecting the Best Land-Use Model
-3-
-------
Knowledge Level
Limited
Extensive
Section/Appendix
Selecting the best model for your community involves having a clear
understanding of the project(s) being studied, the questions needing to be
addressed, and the capabilities the community has to answer these questions.
This section presents an overall process for selecting a land-use change
model. It also presents specific criteria for selecting the right model from a
host of choices.
Section 5.0: A Summary of Current Land Use Models
This section explains how 22 land-use change models are summarized in this
guide and how the summary was developed. It includes a sorting of the
models based on 5 key selection criteria and a general fact sheet for each
model. The fact sheets provide contact information, a brief introduction,
reference information, the data and resources needed to run the model, the
type of information produced by the model, a description of the model's
strengths and weaknesses, how to obtain preview copies of the model, and
case study information.
Appendix A. Comparative Matrices on Land-Use Models
These matrices include, for each model, selected information presented in the
general fact sheets in Section 5.0 and in the technical fact sheets in Appendix
B, as well as some additional information, in a format that allows for quick
comparisons between models.
Appendix B. Technical Fact Sheets on Land-Use Models
These fact sheets complement the general fact sheets in Section 5.0. They
include, for each model, information on the spatial and temporal resolution
and scale of the model, input pre-processing requirements, model
assumptions, setting parameters, comparing scenarios, output post-processing
requirements, and next steps for model development.
Appendix C. Current Trends in Community Growth and Planning
This appendix: 1) describes the traditional approaches used in community
planning that have, in conjunction with other factors, led to growth and
development patterns with detrimental impacts on a community's overall
quality of life, 2) explains the wide array of these impacts, and 3) provides an
overview of the initiatives set in motion nationwide to put community
planning on the path toward "smarter," less detrimental growth and
development.
Appendix D. Key Terms and Definitions
-4-
-------
Knowledge Level
Limited
•
Extensive
Section/Appendix
• Appendix E. References
-5-
-------
2.0 Background
Community citizens, planners, and decision makers often struggle to balance the demands of growth with
the desire to preserve the natural environment, unique community characteristics, and other cherished
quality-of-life attributes. To accomplish this balancing act, it is important to ask the right questions
about the benefits and consequences of growth. It is also necessary to have the right tools to integrate
and evaluate a diversity of information as a means of obtaining reliable answers.
This guide focuses on one type of tool-computer models that help explore the combined effects that
social policies, individual behavior, and other drivers have on land-use change. Land-use change models
are an essential component of a comprehensive approach for communities to project and evaluate the
potential consequences of policy decisions and other actions on land-use patterns in their project areas.
In addition, some of these models explore the reciprocal effects of changing land use on selected
community characteristics such as commute distances, availability of open space, and environmental
conditions. This guide can be used to help identify and distinguish individual characteristics of 22
leading land-use change models for selection of the best model for a particular community. The model
information is presented through a series of general and technical fact sheets, as well as through matrices
comparing key features across each model (see Section 5.3 and Appendices A and B).
2.1 Defining Community
The term community, as used throughout this guide, is EPA's definition, and relates to geographically
based areas of varying sizes:
...a geographic area within which different groups and individuals share common interests
related to their homes and businesses, their personal and professional lives, the surrounding
natural landscape and environment, and the local or regional economy. A community can be
one or more local governments, a neighborhood within a small or large city, a large
metropolitan area, a small or large watershed, an airshed, tribal lands, ecosystems of various
scales, or some other specific geographic area with which people identify.
-EPA, Sustainable Development Challenge Grant, Federal Register, Vol. 64, No. 126, July 1, 1999
Defining community in this way draws attention to the fact that many people's interests are at stake when
discussions involve transforming land or existing structures into other forms.
2.2 Growing Pains
Along with tremendous opportunities for growth come welcomed benefits but also potentially complex
problems. Our thriving economy has paved the way for new and more jobs, burgeoning tax revenues,
-6-
-------
added services, and various amenities, but these assets are often accompanied by rapid low-density
development that can be hard to manage. This type of growth is commonly known as sprawl. Low-
density development that leapfrogs outward from city centers, single-use zoning that separates one type
of land use from another, and a heavy reliance on cars for transportation, are all characteristics of sprawl.
Farmland and natural habitat are being converted to highways, housing developments, and strip malls.
Older neighborhoods are being turned into new forms of multi- or single-family dwellings, often
displacing older residents and leading to a deteriorating sense of community. People and businesses are
moving away from the urban core, which can lead to central city and inner suburb decay and the
accompanying isolation of disadvantaged populations left behind. As sprawl moves across the land,
replacing significant amounts of open space, many communities are finding that they are consuming land
at rates significantly higher than their populations are growing. In some cases, costs incurred by such
growth are exceeding the benefits communities are receiving.
So how do communities accommodate this rapid development? How do we estimate all of its direct and
indirect impacts? Furthermore, how will our decisions affect specific land-use patterns? Our
communities? Land-use change models can assist in evaluating alternative futures and potential
consequences of those alternatives. With these models, the user can begin to understand the complex
array of actions and interactions associated with development by projecting the conversion and loss of
land that occurs as a result of development and community policies.
2.3 Asking the Right Questions
The first step toward turning growing pains into community gains is to understand the complex dynamics
that affect the results of development. A community can achieve growth while preserving valued
characteristics, but stakeholders must first identify, ask, and answer questions that will allow them to
make informed choices. Exhibit 2-1 illustrates a range of potentially helpful questions involving land-use
change and its economic, fiscal, social and environmental impacts. When these questions are reviewed
and a community develops ones of its own, it is likely that many will directly or indirectly involve land
use.
In general, there are two direct ways a project may affect land use: 1) by consuming open space, or 2) by
converting existing land uses to alternative ones. A project also may seed indirect actions that could
affect how land uses change. For example, a new road may need to be built or an existing one widened;
additional housing may be needed for new employees, or other businesses may need to be brought in to
support the new development.
The many factors driving land-use change are diverse and often interrelated. Communities desiring to
grow may offer a range of incentives to attract new development, such as tax breaks and subsidies. A
new road or other transportation alternative may open up a new part of town. Or, utility services may
expand to increase the development potential of a specific area. When factors such as these take place
simultaneously, potentially complex situations develop. For example, a series of tax incentives and other
subsidies may lure a new shopping mall to town. Developers and investors may then decide to build the
-7-
-------
mall on the outskirts of town because property
is cheaper or other attractive incentives are
being offered. The location of businesses
away from the urban core results in a loss in
the central city's tax base often leading to
urban decay.
Furthermore, because the mall is built on the
outskirts, agricultural land or other natural
resources may be consumed. To
accommodate the increased traffic flow and
new traffic patterns, the transportation
infrastructure would need to be enhanced.
These transportation efforts, in turn, could
consume more open space and/or provide
pathways to additional lands that are more
attractive to businesses and residential
developers. As a result of these actions, local
property taxes may need to increase in order
to extend the existing infrastructure to these
new developments. As property taxes
increase, landowners that generate inadequate
revenues (e.g., farmers) may be forced to sell
their properties. On the flip side,
communities may consciously or inadvertently
offer disincentives for development through
road tolls, utility fees, taxes, and planning
restrictions. These disincentives could deter
new businesses or favor one development
over another.
As illustrated above, when one land use
changes, it can trigger a change in another,
creating a ripple effect. As a result, the
community's public infrastructure may
become overburdened. Traffic congestion
may increase. Water, sewer, and other
utilities may need bolstering. Pollution is
likely to increase when open space is
converted to impervious surface or more cars
are added to area roads. Important natural
habitats may be lost to strip malls, commercial
Exhibit 2-1. Questions Facing Growing
Communities
Impacts on Costs of Services and Revenues
• How will this project affect the school system?
• Will other public services change as a result of
this project?
• Will infrastructure costs (such as water, sewers,
and roads) change due to this project?
• Will local tax revenue increase as a result of this
project?
• Will the project "pay for itself?
Impacts on Local and Regional Economies
• Will the project make the community more
competitive in a commercial sense and more
attractive to other business?
• What will be the project's impact on the cost of
housing and property values in the community?
• Will the project have a negative effect on other
communities in the area?
Transportation Impacts
• Will the project increase traffic congestion?
• Will the project provide ample and safe parking
for residents/customers/employees?
• Will people who cannot drive be able to get to
jobs, shopping, and services, such as doctors?
Environmental Impacts
• Will the project affect the amount of available
green space, flood plains, and natural habitat?
• How will the project affect the consumption of
energy and other natural resources?
• What will be the project's impact on water and air
quality and quantity?
Social Impacts
• How will the project change the character of the
community?
• How will the project affect disadvantaged
populations?
• How will important community attributes (e.g.,
historic/cultural sites) be affected by the project?
Long-Term Considerations
• What other land-use changes will be encouraged
by this project?
• How will impacts evolve overtime?
Source: Adapted from Why Smart Growth: A Primer, ICMA and
the Smart Growth Network, 1998
-------
buildings, and housing. Newcomers may demand significant expansions of public services.
Unfortunately, these new needs may end up exhausting more money than the new development can
generate.
2.4 Using the Right Tools
Trying to determine how community growth and change affect land-use patterns can be very difficult.
Fortunately, land-use change models are a tool that can make the process easier. However, caution
should be taken when using model results. Models are estimation techniques-they are not an exact
science and their results should be understood in the context of the qualifications, assumptions, and
limitations of the model. Models rely on data and mathematical equations to simulate the "real world."
They are only as good as the quality of data used and the decision rules and assumptions applied.
Models are tools that can help develop policy when used carefully and in conjunction with other
information.
Exhibit 2-2. Selected Summary Reports on
Transportation Models
Traditionally, computer models used in
community planning have focused on regional
economic trends or transportation-related
impacts of economic growth. Such tried-and-
true models have been around for a long time
(see Exhibit 2-2). But now, with the advent of
improved data collection methods and data
accessibility, more affordable and advanced
computer systems and technology (e.g., the
growing popularity of geographic information
systems), and an increasing emphasis on
sustainable community development (i.e.,
Smart Growth), models that address the land-
use consequences and environmental impacts
of growing communities are becoming more
popular. These more spatially explicit models
seek to understand the dynamics of land-use
change in a community setting. This
innovative approach is a relatively new and
rapidly changing field. At present, there is a
lack of easy-to-use, well-organized information
that can assist communities-the predominant
users of land-use change models-in selecting
the most appropriate tool. Hence, the need for
this guide.
Hunt, John Douglas, David S. Kriger, and Eric J.
Miller. 1998. Current Operational Urban Land-Use
Transport Modeling Frameworks. Submitted for
presentation at the 78th Annual Meeting of the
Transportation Research Board.
Parsons Brinckerhoff Quade & Douglas, Inc. 1999.
Land Use Impacts of Transportation: A
Guidebook. Report 423A. National Cooperative
Highway Research Program. Washington, D.C.,
National Academy Press. Transportation
Research Board.
Simmonds, David. 1995. Available Methods for Land-
Use/Transport Interaction Modeling. A white
paper by David Simmonds Consultancy.
Southworth, Frank. 1995. A Technical Review of
Urban Land Use-Transportation Models as Tools
for Evaluating Vehicle Travel Reduction
Strategies. U.S. Department of Energy, Oak
Ridge National Laboratory: Publication number
ORNL-6881.
Texas Transportation Institute. 1998. Travel Model
Improvement Program, Land Use Compendium.
DOT-T-99-03. U.S. Department of Transportation,
U.S. Environmental Protection Agency.
Texas Transportation Institute. 1995. Travel Model
Improvement Program, Land Use Modeling
Conference Proceedings. Final Report. DOT-T-
96-09. U.S. Department of Transportation, U.S.
Environmental Protection Agency, and U.S.
Department of Energy.
-9-
-------
EPA's Office of Research and Development (ORD) is seeking to fill this gap. ORD spearheaded a
comprehensive research team who examined and compared a variety of leading land-use change models
currently being used or under development that could be employed by states and local governments
and/or by EPA in conjunction with other models. The model inventory and summary information
presented here is the result of this endeavor.
-10-
-------
-11-
-------
3.0 Decision-Making Tools for a
New Planning Approach
Communities across the nation are choosing to complement traditional planning approaches with
analytical decision-making tools to help them plan their sustainable futures. Today, with recent
technological advances, a number of technologically-based tools are available to communities to use
for assessing the impacts of various planning decisions and to help balance the demands of growth,
environmental sustainability, and quality-of-life needs. Technologically-based tools such as models
and geographic information systems (GISs) can provide increased clarity on probable or alternative
outcomes, and thus enable decision-makers to more effectively use traditional planning tools. This
section provides a brief overview of models and GISs, as well as integrated planning and decision-
making systems which are part of the next evolution of modeling capabilities.
3.1 Models: An Overview
Models of many kinds have been used by a host of diverse professionals. Transportation engineers use
models to proj ect the number of commuters that will travel by car versus those who will make their trips
by transit; economists use models to represent the flow of dollars within a regional economy; and
biologists use models to describe the impact water pollutants will have on living organisms. In essence,
a model is a simplified representation of a real-life system. By representing reality with only those
variables that truly affect the behavior of the system, and by clarifying the relationships between those
variables, the assumed "real world" is broken down into a form amenable to analysis (Taha, 1976).
Models can range from simple spreadsheets that provide order-of-magnitude estimates to highly complex
simulations that require the use of a super-computer. Simple models provide rapid estimates with
minimal effort and required data input. Technically-complex models provide the greatest level of
accuracy, but they are usually much more costly in terms of data needs, hardware and software demands,
and required professional expertise.
The four main types of planning models (land-use, transportation, economic, and environmental impact)
are briefly described below. The models summarized in this document are considered land-use models
but many have elements of the other three model types, as well as geographic information system (GIS)
components.
Land-Use Models
Land-use models often incorporate a variety of land use categories as inputs and, thereby, can account
for different subclassifications of urban and nonurban land use such as commercial, industrial, and
agricultural, and even more detailed subclassifications such as density of residential use and type of
commercial/industrial development (see the Anderson land classification system in Appendix C, Exhibit
C-2). Some models offer an environmental approach and project, for example, the impact of
-12-
-------
transportation and land-use choices on air and water quality. Many of the more user-friendly models are
integrated with GISs to become spatially explicit decision-support systems with relational database
technology.
The common denominator of the models summarized in this guide is that each projects changes in land
use. Understanding these models comprehensively, however, requires a basic understanding of the
components of transportation and urban economic models-the foundations of many of the land-use
models described in this report. Two major federal actions in the early 1990s added momentum to the
integration of transportation and economic models and to the development of land-use models. In 1990,
Congress passed the Clean Air Act amendments, which mandated that metropolitan areas look at the
relationships between transportation and air quality. One year later, Congress passed the Intermodal
Surface Transportation Efficiency Act (ISTEA) of 1991, TEA-21's precursor surface transportation law.
ISTEA required transportation planners to consider the likely effect of transportation policy decisions on
land use and development, and the consistency of transportation plans and programs with provisions of
all applicable short- and long-term land-use development plans (Deakin, 1995).
Transportation Models
In the United States, both urban transportation and economic modeling began in earnest in the mid-
1950s. Today, modern transportation models use some variant of the Urban Transportation Planning
System (UTPS) models, which are characterized by a four-step, single-destination, separable-purpose,
and daily trip-based approach. Using the following four steps, transportation models answer questions
about future travel patterns:
• Trip Generation: How many trips will be made?
• Trip Distribution: Where will the trips be?
• Mode Split: Which modes (automobile, transit, cycle, or on foot) will be used?
• Traffic Assignment: What routes will be used and at what time of day will the trips be taken?
(Beimborn et al., 1996)
Economic Models
Urban economic models project employment, population, wage rates, rents, incomes, and prices, among
other variables, for different geographic areas. Until recently, these models tended to be used in for
projecting economic growth rather than estimating the impacts of transportation policies on land use
markets.
Environmental Impact Models
Many different types of models have been developed to assess the impacts that both natural and
anthropogenic changes can have on the environment. These models range from projecting the long-range
transport of pesticides to evaluating the impacts of vehicular emissions on air quality. More recently,
models have been developed to address the effects that human induced land use changes can have on
different aspects of the environment including surface water quality, groundwater recharge and pollution,
habitat fragmentation, wildlife loss, floral and faunal community composition, and impaired ecosystem
function. A great many of these models are watershed models, or models that examine the relationship
-13-
-------
between land uses in the watershed and water quality including nutrient loading, fecal coliform loadings
and urban storm water runoff.
3.2 Geographic Information Systems (GISs)
Accompanying the proliferation of personal computers is the explosion in the acceptance and use of
geographic information systems (GISs). GISs combine land-use mapping capabilities with relational
databases and statistical analysis tools to enable planners to link numerous types of information, rapidly
perform sophisticated analyses, and effectively communicate complex information with maps and
graphical reports. They are powerful tools that could be a component of a model or modeling system.
GISs are used to explain events, visualize trends, project outcomes, and strategize long-term planning
goals. For example, using a GIS, a planner can map population density along a transportation corridor
and evaluate whether there will be sufficient ridership to support public transit. Information on slope,
soil conditions, and distance from a stream can be overlayed to determine land most suitable for
development and land that should be preserved for environmental protection.
GISs help users to study and synthesize the factors contributing to sprawl in a particular community and
identify potential solutions. Data from numerous departments, such as land-use planning, transportation,
and economic development, can be pulled together, trends and interactions can be analyzed and
visualized, and, as a result, better decisions can be made. For example, data such as the primary land
uses of an area can be combined with the existing and planned major roads and highways and the planned
residential and commercial developments to project how much open space will be lost, how much traffic
will increase, and where more development will likely occur. A number of models featured in this guide
are linked to GISs, enabling improved spatial analysis and the ability to map trends and projections.
A GIS is composed of numerous digital
maps that display information on various
characteristics of the region under study. These
maps, also called layers, coverages, or themes,
are layered one on top of the other according to
precise location references common to each
map (see Exhibit 3-1 for a regional-scale
example). These layers, in turn, are linked to
data tables of additional information about a
particular map feature, be it a street, ZIP Code,
or census tract. Information displayed on
different layers can be compared and analyzed
in combination. A user can zoom in on a
particular area and cut the desired location from
all layers in the system. The power of a GIS
lies in this ability to separate information in
layers and combine it with other layers of
information to support decision making. Thus,
Exhibit 3-1. Layers of a GIS
Cities Layer
Rivers Layer
Protected Areas Layer
GIS
Cities - Rivers - Protected Areas
-14-
-------
a GIS is much more than a software application; it is a decision-support process (Foote and Lynch,
1997).
Many modern GIS software packages offer user-friendly, Windows-based environments. Among the
software packages popular with planning professionals are Atlas GIS, Maplnfo, and Environmental
Systems Research Institute, Inc.'s Arc View and Arclnfo GIS. Courses in GIS are widely available at
universities and community colleges across the country.
3.3 Integrated Planning and Decision-Making Systems
Integrated planning and decision-making systems (IPDMSs) promise to be the next state-of-the-art tool
for modeling land-use change interactions. An IPDMS can pull together all of the modeling capabilities
found in various models into one integrated modeling system. An integral part of the power of the
IPDMSs is their employment of "expert systems." Expert systems are decision-support systems that
model human reasoning by codifying the experience and judgment of experts and then programming a set
of procedures that computers can execute when faced with relevant problems (Sumner, 1992). Thus, an
IPDMS could combine travel demand, urban economic, and fiscal and environmental analysis into one
powerful modeling system. Further, such a system could incorporate the mapping capabilities of a GIS
to provide visual depictions of model results. Though an IPDMS that examines land use can not be
identified at this time, it is conceivable that such a system could be developed to support a community's
visioning and goal setting efforts, survey a community on visual preferences, identify and map
community assets, and project future impacts to quality of life based on a range of potential development
scenarios.
-15-
-------
4.0 Selecting the Best Land-Use Model
There will be times when city or county planners, government officials, and concerned citizens will have
to understand the complex issues surrounding development and growth in their community and make
informed choices about which alternatives to pursue. The scenarios presented in Exhibits 4-2 through
4-4 depict some common situations communities are dealing with today. When faced with such
situations-either as the decision-maker or one who seeks to inform or influence the decision-making
process-it is worth considering the five-step approach to selecting the best land-use model illustrated
in Exhibit 4-1. This approach begins by clearly defining the whole scenario involved in the decision
making (e.g., a proposal for a new development project) and concludes with criteria for model selection.
The appropriate model will vary from scenario to scenario and from community to community. The
model needed today may be completely different than the model needed for tomorrow's project. This
variation between projects requires an understanding of the diversity of models available in order to
determine which is the best model to use in a given situation.
Exhibit 4-1. The Five-Step Process for
Selecting a Land-Use Change Model
Step 1: Understanding the Proposal
Step 2: Asking the Right Questions
Step 3: Identifying Information Needs
Step 4: Assessing Internal Capabilities
Step 5: Choosing the Right Model
-16-
-------
The most critical component of an effective community decision-making modeling tool is its ability to
address specific analytical needs and to model alternative planning approaches. The model's competence
in integrating and analyzing diverse data to quantify the benefits of a particular approach, followed by its
ability to project the outcomes of that approach with the desired level of precision, are invaluable when
trying to build community consensus, understand the costs and benefits of a proposed action, and move
new plans toward implementation.
It is important to note that while current land-use change models serve as powerful tools to assist
communities in making sound planning decisions, many are limited in their capabilities. Few, if any,
attempt to fully model all of the various factors affecting land-development patterns, and none can
account for all of the factors that affect land markets and travel behavior. Thus, no model can perfectly
replicate reality. But, as technology continues to advance and land-use changes continue to be of
concern, more-and better-models will be developed to answer the specific policy questions of citizens,
land-use planners, and decision makers.
4.1 Step 1. Understanding the Proposal
The first step in selecting a model is comprehending the project's proposal. Is the consideration a new
discount chain store that wants to locate on the edge of town? The impacts of expanding infrastructure
to a new area? The consequences of a revised business tax strategy? Or the desire to preserve
environmental resources? When reviewing the proposal, it is important to determine the project's entire
geographic scope. What are the project's land boundaries? The proposal must be stated as succinctly,
objectively, and as accurately as possible in order to begin the decision-making process. In addition, it is
imperative that all involved decision-making parties have a common understanding of the proposal.
4.2 Step 2. Asking the Right Questions
Once the proposal has been clearly defined, the project scope must be broken down into detailed pieces.
At this point, essential questions must be asked in order to have a complete understanding of the full
costs and benefits of the proposal. Consideration of the direct and indirect, synergistic, and cumulative
impacts the project will bring to the community is needed. The scenarios in Exhibits 4-2 though 4-4 list
some of the land-use questions that should be considered, though the examples are far from complete.
Each community's situation is unique, so the questions from project to project will be unique as well.
-17-
-------
Exhibit 4-2. Opportunities for Employing Land-Use Change Modeling - Example 1
RURAL-EDGE Weighs Implications of a Proposed Indoor Shopping Mall
RURAL-EDGE, a community of 25,000 people nestled in the middle of a large county, is comprised
primarily of choice agricultural land. The area is known for its apple industry and livestock
production-most production occurs at farms of 50 to 500 acres throughout the county. RURAL-
EDGE provides the industrial support to package and ship these agricultural products. Although
located less than 120 miles from a major metropolitan area (i.e., an area exceeding 1 million people),
RURAL-EDGE has managed to escape the dramatic population increases of communities closer in.
Until recently, that is. Prior to 1990, RURAL-EDGE was connected to the metropolitan area by a
single two-lane highway; however, in 1990, efforts to widen the existing highway were concluded and
a new highway was constructed. RURAL-EDGE was now much more accessible, turning the rural
area into a prime destination for urbanites who were seeking out rural life, not minding their longer
commutes. Soon RURAL-EDGE began to be recognized as a hub city for the region. Local leaders
began offering a range of incentives and subsidies to attract new business. Out-of-towners, in turn,
began exploring land-investment opportunities. An entity from a neighboring state recently
approached the city and county planners with a proposal to build a large indoor shopping mall at the
interchange of the new roads. Affiliated with the shopping mall would be a telecommuting center for
area workers.
Currently, the city and county planners are under pressure by government leaders to endorse the
development proposal because they believe it will bring needed jobs and revenues to the area.
Several local citizen groups are opposed, citing increased traffic congestion, loss of important historic
and environmental resources, and an overall economic cost for the area. The citizen groups are
threatening to sue if the project moves forward. The city and county planners know they need to
weigh clearly all of the direct and indirect costs and benefits of the proposal, and have the data to do
so. They also realize that they need some sort of tool to integrate and analyze all of the information
so it makes sense. They believe that a land-use change model will provide the kind of integrative
analytical tool that will help them answer some of the following questions:
• How much land will the project consume?
• Will the project convert existing commercial and retail land uses, or will it consume existing
open space?
• Will prime agricultural land or other sensitive areas be lost due to the construction of this
project?
• How many people will the project attract as shoppers and telecommuters? How many jobs
will the project create? Where will these people come from? Will additional land uses need
to be modified to accommodate the flow of traffic and the anticipated number of shoppers,
new residents, and employees?
• What are the potential indirect land-use changes as a result of the project?
Will existing roads need to be widened or new roads built?
Will additional housing need to be constructed?
Will the project encourage other new businesses?
• What other land-use changes will be encouraged as a result of this project?
• Will this project have a negative effect on other parts of the community? For example, will
the project shift activities (e.g., shopping, housing) from one part of the area to another?
• Will the project affect the amount of undisturbed open space and available parks?
-18-
-------
Exhibit 4-3. Opportunities for Employing Land-Use Change Modeling - Example 2
METROBIG Weighs Transportation Alternatives' Effects
on Community Land-Use Pattern
METROBIG is a large U.S. city that has enjoyed substantial prosperity and economic growth due to
a booming high-technology industry. Unemployment is at an all-time low, out migration from the
central city has slowed, and an improved tax base has given local decision makers a bit more
spending flexibility. But with this economic development success has come relentless traffic
congestion. The population has grown 20 percent in the last decade, while the acreage dedicated to
urban uses has grown three times the population growth during the same period of time.
Transportation system capacity has grown a mere 8 percent, frustrating commuters on a daily basis.
Citizens are demanding solutions to gridlock. Some rally behind increased investment in roadway
expansion, while others protest any transportation project that is not commuter rail.
Getting citizens and their elected officials to come to agreement on where to direct limited public
resources in addressing transportation alternatives is no small matter given the fact that
METROBIG's land area traverses two states and a number of local jurisdictions. Planners from a
regional planning organization believe that integrated transportation and land-use modeling may help
the localities reach consensus on regional planning goals and the joint implementation steps that
must be taken for sound land-use planning.
At a recent series of vision-planning workshops, community members identified potential solutions.
Regional planners are searching for the right land-use modeling tools that will help them determine
the best mix of these potential planning options and answer the following questions:
• What impact would increasing the roadway capacity by half on the city's beltway have on
land markets inside and outside the beltway? Would this increased capacity meet capacity
needs 20 years into the future or would the investment merely induce demand and lead to
similar gridlock?
• What impact would a circumferential rail transit around the beltway have on land markets
inside and outside the beltway? Would this investment meet the travel demands?
• How do these two transportation investments compare in terms of overall induced land
consumption and impacts on air and water quality?
• How will the opening of a new commuter rail station in the central city affect land markets?
• If traffic congestion becomes so bad in the outer suburbs that lack sufficient transit service,
will people move back to the central city and inner suburbs? What impact will there be on
these land markets?
• What impact will there be in terms of traffic movement and land markets should congested
streets be replaced by tree-lined boulevards with fewer traffic lights?
• Can the use of public dollars toward central city home-ownership tax credits further reverse
population decline and bring more families and jobs back to the city?
-19-
-------
Exhibit 4-4. Opportunities for Employing Land-Use Change Modeling - Example 3
WISETOWN Plans Future Development
WISETOWN is located in a beautiful mountain valley along the banks of a pristine river. Since its
settlement in 1802, it has been an important tourist destination. In addition to its natural beauty, it is
filled with important historic sites associated with the town's popularity as a mountain spa and ski
area. The town has quite a few historic hotels, churches, spa-related structures, and a railroad depot.
Population pressures from neighboring jurisdictions and increased access provided by several new
roads and a small airport have begun to threaten the special character of WISETOWN. Recognizing
the importance of maintaining its tourist attractions and the qualities that draw visitors, the
WISETOWN mayor recently convened a community-wide visioning meeting. As a result of this three-
day workshop, WISETOWN community members and leaders prepared a strategic plan that outlined
areas they wanted to preserve from development. Participants identified historic sites, cultural
attributes, and special environmental areas.
To implement the plan and to provide improved guidance for future development, community leaders
need to take the results of the planning and visioning exercise and organize them to support future
analyses. These leaders are interested in finding a tool that will help them with this effort. They hear
that there are land-use change models that may help them accomplish their goals and answer the
following questions:
• Where are our precious community resources located?
• Given that we want to preserve wetlands, forested slopes, riparian buffer strips, and flood
plain areas, what land is left for development?
• Of the remaining land, what parcels should be set aside because they contain a
historically-or culturally-important resource?
• Where are our steep slopes, loose soils, and otherwise non-buildable lands located?
• How will a proposed development affect the areas that have been set apart from
development?
• What is the proximity of a proposed project to a protected area?
• Where are the best places to locate future development?
4.3 Step 3. Identifying Information Needs
Once the questions are formulated in step 2, a determination should be made as to what types of data
are needed to answer the questions. If one question is where to locate a proposed project to minimize
impacts on sensitive environmental areas, information is needed on where those areas are. If another
question is the potential consequences of a new mall on surrounding land uses, information may be
needed on trends from similar projects, projections on the number of shoppers and employees,
and estimates of potential daily traffic. At this stage, it may not be possible to identify all of the
informational needs (some answers will come after a decision has been made as to which model to use),
but the process of identifying how well internal capabilities can find answers to the questions should
begin.
-20-
-------
4.4 Step 4. Assessing Internal Capabilities
The next step in selecting a model requires a clear understanding of what internal capabilities can be
accessed to acquire and use the model. This includes assessing the following:
• Financial resources. How much can be afforded?
• Staff resources. What is the extent and talent of staff available to use the tool? Does additional
help or consulting expertise need to be hired?
• Computer resources. Do the proper hardware, software, and computing power resources exist
on-site to run models?
It is important to be realistic in this assessment since a shortage of resources could result in ineffective
installation and maintenance of the modeling tool-essentially rendering it useless.
4.5 Step 5. Choosing the Right Model (Using Selection Criteria)
Once the first four "background" steps have been completed, the final step is assessing and selecting
the best model to meet identified needs. Before choosing, however, each option should be thoroughly
analyzed against selection criteria. Thirteen primary selection criteria are provided and explained below
as guidance. They are listed in an order that follows the likely thought process of a community that is
considering a range of models. This is not an all-inclusive list of selection criteria. Other criteria may be
important to consider based on the particular results of steps 1 through 4. The criteria may be weighted,
based on level of importance, to guide the decision-making process. See Exhibit 4-5, at the end of this
section, for an example approach to weighting the criteria.
• Relevancy. Does the model provide pertinent information that meets the analytical needs of the
community?
For a land-use model to be relevant and of value to a community, it must be able to model and
project outcomes for scenarios that relate to the community and its needs. The first step in
determining the relevancy of a model to community needs is to ask which land-use change will
be evaluated by the study. Keep in mind that some models can evaluate several different types of
land-use changes, while other models are limited to only one or two types.
The next step is to carefully identify the questions or issues that will be addressed by the study.
Careful definition of the questions is essential in determining the boundaries of a study (e.g.,
topical, spatial, temporal) and the general types of information required to run the model. Every
attempt should be made to break the larger questions into smaller, more quantifiable ones.
Supporting documentation on the model, as well as a detailed description of the model's data
output, should provide the necessary information to determine if the model can answer the
-21-
-------
questions. It may be helpful to review the data outputs and capabilities of various land-use
models in order to better identify and clarify the types of questions for potential evaluation.
Resources. Are the model and the computer requirements (hardware, software) and staff
(number of people and their time) needed to support the system within the community's budget
and infrastructure?
The resources required to use a model include cost of the model and associated computer
requirements and the staff time to implement the model. To determine the full cost of a model,
conduct an accurate accounting of the associated costs needed to acquire and maintain the model,
measured in both dollars and time. Consider the purchase price of the model, as well as any
additional hardware and software computer requirements needed to support the model. Also
consider any long-term maintenance costs associated with the use of the model and associated
computer resources.
It is important also to factor in the amount of time and labor needed to run the model. Some
models require full-time attention from dedicated staff and/or consultants, while others provide
more user-friendly software tools that someone with minimal experience can run from a desktop
computer. If the staff involved in the project are not volunteers, then their salaries, or the
appropriate percentage of their salary based on anticipated labor hours needed to perform the
study, should be incorporated into the overall operating costs of the model. In addition to any
staff that may be supervising or supporting the study, expert consultants may be required to run
the model and interpret the results, depending on the complexity of the model chosen. If an
outside consultant is required, these additional consulting fees must also be added to the cost of
the model.
Finally, when considering cost, it is necessary to evaluate whether it would be more cost-efficient
to hire an outside consulting firm to perform the study than to purchase the model. Government
agencies may be able to save money by seeking assistance from another agency with adequate
resources. In general, the more sophisticated a model is, the more expensive it is to obtain, tailor
to local conditions, and operate. In any proposed project, the cost of the models used must be
weighed against the level of precision necessary to meet the project's objectives. If the types of
analyses desired are needed on a regular basis, it may be most worthwhile to purchase the model
and its components and hire skilled operators as permanent in-house technical staff.
Model Support. Do the model developers, or does the model itself, provide sufficient support
needed to understand and implement the model (e.g., model documentation, user discussion
groups, training)?
Computer models, like any other computer hardware or software products, often have varying
levels of support for end-users. Typically, models offer documentation and a user's guide to help
understand how to load and run the model. Other levels of service also may be offered, including
the potential to join users' groups, take workshops or electronic tutorials, view an Internet web
-22-
-------
site for additional information, and contact help lines. A careful assessment of in-house
capabilities is needed to determine which kind of model support would be necessary. Depending
on the outcome of such an assessment, this could be an important criterion for consideration.
Technical Expertise. Does the community have the technical expertise required to use, calibrate,
and interpret the results of the model?
In general, the more sophisticated a model is, the more technical expertise will be required to
operate the model and interpret the results. An expensive, complex, sophisticated model is of no
value to the community if the community lacks the ability to use the model or understand its data
output. Before selecting a model, a community must understand the level of technical expertise
required to maintain and operate the model in order to determine if the model can be maintained
in house or if the services of a consultant will be needed.
Data Requirements. Does the community have, or can they obtain, the data necessary to run the
model?
Many land-use change models are data intensive and/or require a certain scale of data to provide
reliable results. For example, a model may require that land-use data be on a scale that can be
provided only by aerial photography, not satellite imagery. Some models operate best with
locally-based data inputs. Unfortunately, much available data are aggregated to a county,
regional, or larger area. Disaggregation of such data may be impossible and/or severely
compromise data quality. Collection of local data may require a significant resource
commitment. In some instances, the necessary temporal scale of data is not available. It is
important to conduct a realistic assessment of existing data resources (including time period, and
spatial coverage and resolution) and/or a user's ability to collect new data. Always remember
that the selected model will be constrained by the data available.
Accuracy. Are the projections generated by the model reliable to a degree that is useful to the
community?
The term "accuracy" can be interpreted in different ways. In general, it refers to how close the
model comes to reality, and how well the model answers the questions posed. Complex models
usually take into account more variables (i.e., they contain a greater level of detail) and,
therefore, can provide more specific results and can more successfully simulate true conditions
than simplified models that rely on many averages and assumptions. Accuracy also involves the
"goodness-of-fit" of model results when compared against known outcomes of given scenarios.
Some model developers have conducted accuracy analyses by "back-casting" projections through
a recent historical period, and comparing the results with what actually transpired. The more
important that known accuracy is to a study, the more advisable it is to use a model where
goodness of fit has been evaluated.
-23-
-------
Additional factors for consideration include the resolution and the temporal capability of the
model (see below). The more accurate a model is, the more useful the results and potentially the
more defensible if challenged. When a model is intended to provide the basis for key land use
policies and decisions that may greatly impact citizens or businesses, the user should take
accuracy into consideration to evaluate whether the model results could withstand challenges by
affected community members.
Resolution. What amount of land and what level of detail can be modeled in a single scenario?
Resolution refers to the minimum unit of land that the model recognizes in its functions. Some
models can simulate land use down to the parcel level, while others may be limited to larger
areas (e.g., larger than a certain number of acres, full city- or county-level). High resolution
(e.g., square feet) is useful when the study area is small and generalizations or averages would
render differences between land areas within the overall study with less clarity. Low resolution
(e.g., acres) is useful when the study area is large, averages would provide adequate information,
and collection of highly detailed data would create a volume of information so large that it would
impede a thorough analysis.
Temporal Capabilities. Can the model project outcomes for multiple time periods?
When evaluating a model, it is important to determine the level of flexibility a model provides in
temporal resolution and extent. The term temporal capabilities refers to the time periods the
model examines and the length of each of these time periods. For example, a model may project
housing needs for the next 50 years, breaking the results down by 10-year increments. In some
models, these time periods may be fixed. If there is a need to examine trends over different time
periods and at different intervals, this type of model probably is not best.
Versatility. Can the model project outcomes for multiple variables (i.e., land use, transportation,
employment, housing, and environmental)?
The versatility of a model refers to the model's ability to evaluate, integrate, and link multiple
variables such as land use, transportation, employment, and housing. Consider versatility once it
is clear how complex the proposal is that is being evaluated. Generally, the more versatile a
model is, the more complicated it is. As a model becomes more complex, the data requirements
and technical expertise needed to operate the model increase. When selecting a model, it is
necessary to be aware of the types of issues that need to be evaluated and the cost-effectiveness
of investing in a model that can evaluate multiple variables. When looking at the versatility of a
model, it is important to consider two fundamental selection criteria: relevancy and cost.
Linkage Potential. Can the model be linked to other models currently in use by, or of interest to,
the community?
-24-
-------
The linkage potential of a model refers to the ability of a model to join with other tools, including
geographic information systems (GISs), other models, or presentation software. A model with
high linkage potential is desirable, since it allows the user to connect the data outputs to other
software that could help further analyze and/or present the information in a different or more
useful way. To date, no single model exists that can perform all community planning functions;
it is very likely that it will be necessary to link economic, transportation, and land use models
together, then visualize the results by incorporating the output into a GIS.
Public Accessibility. Can the model be run in an interactive public environment and display the
results in a manner that is comprehensible to the general public?
A model is publicly accessible if it can be approached and understood by the general public. If
data output is presented in an easy-to-comprehend manner, such as a graph or bar chart,
the results can reach a wider audience. Using a model in a public forum or meeting to
demonstrate the outcomes of different scenarios can be a powerful way to educate the public and
generate support for a proposed policy or plan.
Transferability. Can the model be applied to locations other than the one(s) for which it was
developed?
A model may have been designed for a particular location, and therefore may require intensive
efforts to adapt it for use in another. Site-specific information that may require modification
includes land use, environmental, and economic policies; land-use categories; available data and
resources; time periods; and spatial extent (e.g., regional, local, neighborhood). The type of
information that can or must be changed will depend on the model, as will the level of effort
necessary to make the changes, such as having to re-calibrate underlying statistical equations,
change input parameters, and modify model assumptions. Such efforts can be costly due to the
time and the technical expertise required for each adaptation. If resources are minimal, it is wise
to select a model that can be easily transferred. For several land use models, the technical fact
sheets in Appendix B provide details on the efforts (reflected in the pre-processing, calibration,
assumptions, and setting parameters information) required for adaptation to a new locale.
Third-Party Use. How extensively has this model been used in "real-world" situations?
Some land-use change models are under development, or have been used primarily in academic
settings, while others have been used more extensively in community settings. A model used in a
community setting, however, does not necessarily mean that the model is better than one with
more limited use; one should carefully consider all the selection criteria to select the best model.
Usually, it is best to select a model with a proven track record, especially if it has been used for
communities having a similar size or similar situations. Also, if a model has been used
extensively, there should be documented case studies about the efficacy of the model and
opportunities to consult with end-users.
-25-
-------
Exhibit 4-5. Example Criteria Rating Table
Criteria
Relevancy
Resources
Model Support
Technical Expertise
Data Requirements
Accuracy
Resolution
Temporal Capabilities
Versatility
Linkage Potential
Public Accessibility
Transferability
Third-Party Use
Total Score
Weight1
(W)
^m
Rating Score: 10 = High Match
(RS) 1 = Low Match
Alternatives
Model #1
RS
•
RSxW
Model #2
RS
•
RSxW
Model #3
RS
•
RSxW
1 Weights may be numeric (e.g., 1=low weight, 5=high weight) or qualitative (e.g., low, medium, high).
A numeric weight enables users to multiply the weight and the rating score to determine a numeric
value for each model. For example, if relevancy was given a weight of "5" and Model 1 scored a 2 for
a rating, Model 2 scored 5 and Model 3 scored 10, their respective numeric totals would be 10, 25, and
50. Clearly, Model 3 is best for relevancy. After all criteria are applied, the total scores for each model
are tallied.
Source: Adapted from Chang and Kelly, 1995.
-26-
-------
-27-
-------
5.0 Summarizing the Land-Use Models
Given the variety and scope of land-use models, finding and sorting through information and identifying
a model can be a daunting task. To make the task more manageable, this guide provides summary
information on 22 leading land-use change models (see Exhibit 5.1). The goal of the summary
information is to enable a sufficiently clear and thorough understanding of the models to allow readers to
make an informed selection of the model that is most appropriate for their particular resources and needs.
The models are summarized in this guide in four ways:
• Key selection criteria (see Section 5.2).
The models are sorted and listed in tables according to a few key selection criteria.
• General fact sheets (see Section 5.3).
The general fact sheets provide a brief but thorough overview of the fundamental features of
each model. See Section 5.3 for more details on the particular features covered.
• Comparative matrices (see Appendix A).
The comparative matrices include for each model selected information from the general and
technical fact sheets, as well as some additional information not found in the fact sheets, in a
format that allows for quick comparisons between the models.
• Technical fact sheets (see Appendix B).
The more technical information for each model is provided in these fact sheets, including the
geographic and temporal scale of the model, model assumptions, parameters, post-processing
requirements, and the next steps for development.
The information contained in the fact sheets and matrices is based on reference materials and model
developer reviews, not on direct use and experience with the models. All information was verified by the
model developers or representatives of the developers.
5.1 Developing This Summary
EPA's first step in developing this inventory of land-use change models was to convene a workgroup
comprising academic modelers and federal staff who serve as community liaisons. This workgroup
framed the guide's content and format according to perceived user needs. EPA then contracted with
Science Applications International Corporation (SAIC) to collect appropriate information and compile
the findings. Using the workgroup's list of 14 model features for characterization, SAIC developed a
standard work sheet to examine identified models. Model features researched by SAIC included the
following:
-28-
-------
• Model name • Strengths
• Contact information for the model developer • Limitations
• Description • Required expertise
• Resource requirements • Geographic and temporal scale
• Theoretical framework • Transferability to multiple locations
• Data inputs • Capability of linking with other models
• Model outputs • Gaps that need to be addressed.
SAIC conducted independent research on these factors using journal publications, Internet web sites,
user's guides, and demonstration models. Researchers used information collected from these sources to
complete a summary worksheet for each model. These worksheets formed the basis for the narrative fact
sheets (both general and technical) and comparative matrices found in this guide.
Upon completion, SAIC provided the worksheets and fact sheets to the model developers, or appropriate
designated colleagues, for verification of technical accuracy and supplemental information. All
information on models presented in this guide was subject to this review process and was approved by
the model developers.
-29-
-------
Exhibit 5-1. Land-Use Change Models Included in This Guide
Model
Developer
Purpose
California Urban
Futures (CUF) Model:
CUF-1
John Landis, Institute of Urban
and Regional Planning, University
of California at Berkeley
Provides a framework for simulating
how growth and development policies
might alter the location, pattern, and
intensity of urban development
California Urban
Futures (CUF) Model:
CUF-2
John Landis, Institute of Urban
and Regional Planning, University
of California at Berkeley
Same as CUF-1 (CUF-2 addressed
some of the theoretical holes of
CUF-1)
California Urban and
Biodiversity Analysis
Model (CURBA)
John Landis, Institute of Urban
and Regional Planning, University
of California at Berkeley
Evaluates the possible effects of
alternative urban growth patterns and
policies on biodiversity and natural
habitat quality
DELTA
(formally DSCMODE)
David Simmonds Consultancy
Projects changes in urban areas,
including the location of households,
population, employment, and the
amount of real estate development
Disaggregated
Residential Allocation
Model of Household
Location and
the Employment
Allocation Model
(DRAM/EMPAL)
S.H. Putman and Associates, Inc.
Projects the interactions and
distribution of employment and
housing in a specified geographic
area
Growth Simulation
Model (GSM)
Maryland Department of Planning,
Baltimore, Maryland.
Contact: Joe Tassone
Projects population growth and new
development effects on land use/land
cover under alternative land
management
INDEX8
Criterion Planners/Engineers, Inc.
Measures the characteristics and
performance of land-use plans and
urban designs with "indicators"
derived from community goals and
policies
IRPUD Model
(formally Dortmund)
Michael Wegener, Institute of
Spatial Planning, University of
Dortmund, Germany
Projects the impacts of long-range
economic and technological change
on housing, transportation, public
policies, land uses, and infrastructure
Land Transformation
Model (LTM)
Dr. Bryan C. Pijanowski,
Michigan State University
Integrates a variety of land use
change driving variables to project
impact on land use on a watershed
level
-30-
-------
Model
Developer
Purpose
Land-Use Change
Analysis System
(LUCAS)
Michael W. Berry, et al.,
Department of Computer
Sciences, University of
Tennessee
Examines the impact of human
activities on land use and the
subsequent impacts on environmental
and natural resource sustainability
Markov Model of
Residential Vacancy
Transfer
Philip Emmi and Lena Magnusson
Explores changes in demand for
various types of residential housing
within a community
MEPLAN
Marcial Echenique & Partners
Limited. Contact: Ian Williams
Helps communities analyze the inter-
related effects of land use and
transportation and is designed to
compare proposed plans/policies
METROSIM
Alex Anas & Associates
Uses an economic approach
forecasting interdependent effects of
transportation and land use systems
and of land use and transport policies
Sub-Area Allocation
Model-Improved
Method (SAM-IM)
(formally LAM)
Planning Technologies, LLC
Creates new land use scenarios that
reflect alternative development
concepts for the future
The SLEUTH Model
(formally Clarke
Cellular Automata)
Keith C. Clarke, Department of
Geography, University of
California at Santa Barbara
Projects urban growth and examines
how new urban areas consume
surrounding land and impact the
natural environment
Smart Growth INDEX®
Criterion Planners/Engineers, Inc.
(with Fehr & Peers Associates,
Inc.)
Evaluates transportation and land-use
alternatives and assesses their impact
on travel demand, land consumption,
housing and employment density, and
pollution emissions
Smart Places
Electric Power Research Institute
(EPRI). Contact: Paul Radcliffe
Assists communities in the simulation
and evaluation of land-use
development and transportation
alternatives using indicators of
environmental performance
TRANUS
Modelistica
Analyzes the effects of land-use and
transportation policies or combinations
of policies on the location of various
activities and the land market
UGrow
Wilson W. Orr, Prescott College
Projects long-term changes to
communities in response to changes
in transportation and fiscal policies
-31-
-------
Model
UPLAN
Developer
Robert Johnston, Dept. of
Environmental Science and
Policy, University of California at
Davis
Purpose
Creates alternative development
patterns in response to changes in
development and fiscal scenarios
UrbanSim
Paul Waddell, Daniel J. Evans
School of Public Affairs, University
of Washington
Explores how the interactions
between land use, transportation, and
public policy shape a community's
development trends and affect the
natural environment
What if?
Dr. Richard E. Klosterman (as
Community Analysis and Planning
Systems, Inc)
Supports comprehensive community
land-use planning in regard to
determining land suitability for
development, projecting future land-
use demand, and providing the
capability to allocate the demand to
the most suitable location
5.2 Land Use Models: Identified by Key Selection Criteria
As discussed in Step 5 ("Choosing the Right Model") in Section 4.5, communities may use a range of
selection criteria to use help them single out the model that will work best for them. This section
highlights five key criteria that every community is likely to consider and identifies how each of the
22 land-use models included in this guide fit the criteria. Please note that the models are listed in
alphabetical order, not ranked, and the sorting is based on the worksheet information as provided by the
model developers.
In addition to the five criteria provided here, the comparative matrices in Appendix A provide several
more categories of information for each model. Readers may use the matrices to sort and compare the
models based on the categories of information that best address their particular needs.
-32-
-------
Exhibit 5-2. Technical Expertise Needed to Use Model1
None
CURBA
Markov
METROSIM
SAM-IM
Smart Places
UGrow
Some2
CUF-1
CUF-2
GSM
MEPLAN
SLEUTH
Smart Growth INDEX®
TRANUS
UPLAN
UrbanSim
What if?
Extensive3
DELTA
DRAM/EMPAL
INDEX®
IRPUD
LTM
LUCAS
1 Once the model has been installed/developed.
2 Requires land use planning experience.
3 Requires land use modeling experience.
Exhibit 5-3. Purchase Cost1
Free
LTM
LUCAS
Markov
SLEUTH
Smart Growth INDEX®
UGrow2
UPLAN
UrbanSim
$1 - 5,000
What if?3
$5,007 -
70,000
TRANUS
$70,000+
DRAM/EMPAL
INDEX®
MEPLAN
METROSIM
SAM-IM
Contact
Developer4
CUF-1
CUF-2
CURBA
DELTA
GSM
IRPUD
Smart Places
1 Does not include operating, maintenance, or training costs.
2 Software is free but model developer must adapt model to fit particular community at a cost ranging
from $30,000 - $200,000.
3 Cost varies depending on whether for academic ($250) or professional ($2,495) single user.
Academic and professional site licenses are available.
4 Models in this category are not available for direct purchase; either they were developed in an
academic context, for a particular situation, or are coupled with consulting services. For example, the
CUF models were developed in an academic context and utilized for specific pilot studies. They may
be available for application to other locales, but it is necessary to contact the model developer to
ascertain how to make that happen. Other models, like DELTA, are part of a consulting service and
unavailable for direct purchase-the affiliated consulting firm must be hired.
-33-
-------
Exhibit 5-4. Existence of Model Support
Model
CUF-1
CUF-2
CURBA
DELTA
DRAM/EM PAL
GSM
INDEX®
IRPUD
LTM
LUCAS
Markov
MEPLAN
METROSIM
SAM-IM
SLEUTH
Smart Growth
INDEX®
Smart Places
TRANUS2
UGrow
UPLAN
UrbanSim
What if?
Written
Documentation1
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
IVefc S/te
/
/
/
/
/
/
/
/
/
/
/
/
/
Training
/
/
/
/
/
/
/
/
/
/
/
1 Refers to materials such as a user's manual, published articles, or fact
sheets.
2Also has extensive e-mail support.
-34-
-------
Exhibit 5-5. Ease of Transferring to Other Locations1
Effortless'
Feasible3
METROSIM
Smart Growth INDEX8
TRANUS
CUF-1
CUF-2
CURBA
DELTA
DRAM/EM PAL
GSM
INDEX®
IRPUD
LTM
LUCAS
Markov
MEPLAN
SAM-IM
SLEUTH
Smart Places
UGrow
UPLAN
UrbanSim
What if?
For a detailed explanation of this sorting criterion, see
section 4.5.
2 Requires no modifications in order to transfer.
3 Requires some modifications in order to transfer.
Exhibit 5-6. How Many Locations Has the Model Been Applied To?
1-5
CUF-1
CUF-2
IRPUD
LTM
LUCAS
SAM-IM
Smart Places1
UPLAN
UrbanSim
What if?
6-10
CURBA
DELTA
METROSIM
UGrow
11-20
INDEX®
Markov
SLEUTH
Smart Growth
INDEX®
20+
DRAM/EMPAL
GSM
MEPLAN
TRANUS
1 At present, Smart Places has been applied only once but more than 35 communities have
license agreements to use it.
-35-
-------
5.3 Land Use Models: General Fact Sheets
The general fact sheets, provided alphabetically in this section, include essential information about each
land-use change model, such as: model developer contact information, an overview of the capabilities of
the model, the resources required to operate the model, land uses addressed by the model, the questions
or issues the model can help assess, data inputs required by the model, outputs generated by the model,
model strengths and limitations, model accessibility, and case study information. It is expected, in most
cases, that the general fact sheets alone will provide enough information for readers to make a sound
decision as to whether particular models may be useful to them.
The general fact sheets provide a starting point in investigations about appropriate models. The more
technical information is included in Appendix B (intended primarily for those readers with modeling
experience). After narrowing the choices, potential users should contact the developers for more
complete and up-to-date information. In many cases, models are continuously being refined and
improved, so it is important to seek the latest information on models of interest.
-36-
-------
California Urban Futures (CUF) Model: CUF-1
MODEL DEVELOPER(S):
MAILING ADDRESS:
CONTA CT IN FORM A TION:
WEB SITE:
DOCUMENTATION:
John Landis, Ted Bradshaw, Ted Egan, Peter Hall, Ayse Pamuk,
David Simpson, Qing Shen, Michael Teitz, and Ming Zhao.
Institute of Urban and Regional Development
University of California at Berkeley
Berkeley, CA 94720-1879
Phone: 510-642-5918
Fax: 510-643-9576
E-mail: jlandis@uclink.berkeley.edu
http://www-dcrp.ced.berkeley.edu
Landis, J. D. 1994. The California Urban Futures Model: A New
Generation of Metropolitan Simulation Models. Environment and
Planning, B: Planning and Design, 21: 399-420.
Landis, J. D. 1995. Imagining Land Use Futures: Applying the
California Futures Model. Journal of the American Planning
Association, 61: 438-457.
OVERVIEW
The California Urban Futures Model is known as the CUF Model or CUF-1 (earlier versions
of the model were known as the Bay Area Simulation System [BASS II]). The purpose of the
CUF-1 model is to provide a framework for simulating how growth and development policies
might alter the location, pattern, and intensity of urban development. The model is designed
to consider growth and development policies at various levels of government (e.g., state
government, local government, and special districts). The model was originally developed
to simulate the impacts of alternative regulatory and investment policy initiatives on urban
development in the Northern California Bay Region. Note: CUF-1 has been superceded by
CUF-2 and CURBA.
The CUF-1 model allows the user to:
• Project population growth at a sub-area level (e.g., a city) and then aggregate projected
growth to larger units (e.g., a county),
• Allocate growth to individual sites based on development profitability,
• Incorporate several variables, including spatial accessibility, to determine the location
and density of new development,
• Assemble, organize, manage, and display data describing land development options
with geographic information systems (GIS),
• Incorporate development policies into the growth forecasting process, and
• Simulate new policy scenarios quickly and display results in easy to understand map
forms with various levels of detail.
-37-
-------
The CUF-1 model uses two primary units of analysis, political jurisdictions (incorporated cities
or counties) and developable land units (i.e., undeveloped or underdeveloped areas that may
be developed or redeveloped [DLUs]). First, the model projects population growth based on
city population growth trends and development potential by DLU. The CUF-1 model then
simulates growth of an area by determining how much new development to allocate to each
DLU per model period based on population growth of each city or county, the profitability
potential of each DLU if developed, and user-specified development regulations and/or
incentives. This is accomplished using four related submodels: the bottom-up population
growth submodel, a spatial database, the spatial allocation submodel, and the annexation-
incorporation submodel.
REQUIRED RESOURCES
Purchase Costs
Not available for "off the shelf" purchase. Contact the model developer.
Equipment Needs
CUF-1 requires a UNIX-based workstation with a UNIX operating system, programming
language compilers, and SPSS statistical analysis software.
Staff Requirements and Expertise
Installation and calibration of the model requires experience in SPSS and Arclnfo as well as
land-use modeling expertise. Use of the model requires land-use expertise and general
computer experience.
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
The land-use categories addressed by CUF-1 are user defined and, therefore, could include
any land-use category as appropriate for the study area. Four major land-use categories were
used for the CUF-1 San Francisco Bay and Sacramento areas simulation: agricultural land
type, general plan use category, current land use, and wetlands.
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
No?
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
•
•
•
No?
-38-
-------
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees, taxes,
and incentives)
Yes?
•
•
No?
•
•
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Environmental Quality
School Quality
Crime
Other Quality-of-Life Conditions
Yes?
•1
•1
•
•2
No?
•
•
•
1 Only with some modifications or additions.
2 For water quality, "yes" with some modifications or additions.
For air quality, "no."
Outputs Provided
Output
Quantitative outputs
Graphical outputs
Format
Acreage tabulations and total
Maps of newly-developed areas
INFORMATION NEEDED TO RUN THE MODEL
This model requires the following user input:
• Multiple Arclnfo coverages or themes profiling existing land use, general plan, and
environmental characteristics, as well as jurisdictional boundaries.
• Jurisdiction-level tabular information profiling historical population growth.
-39-
-------
MODEL STRENGTHS AND LIMITATIONS
Strengths
• Easy to use and visual: The CUF-1 model allows users to prepare and evaluate
alternative policy scenarios quickly (a typical simulation can be completed in a matter of
hours) and in easy to read map form at almost any level of detail.
• Expandable: The CUF-1 model is designed as a modular system of related but
independent submodels that can be updated to include new information and theories.
• Policy approach: The CUF-1 model simulates alternative development futures based on
specific policy changes.
Limitations
{A second generation of the CUF model ("CUF-2") has been developed to address several of
the limitations discovered in the original CUF-1 model.}
• Availability: The CUF-1 model is currently unavailable as a product that can be
purchased "off the shelf."
• Singular land use: The CUF-1 model is limited to residential development and does not
include methods for projecting and/or allocating future industrial, commercial, and public
activities. Therefore, sites that are the most profitable to develop are reserved for
residential development (unless explicitly prohibited).
• Lack of "infill" development and urban redevelopment: The CUF-1 model assumes that
almost all population growth will occur at the urban edge.
• Growth allocation primarily depends calculated on development profitability: The CUF-1
model may be insensitive to other factors that impact growth patterns and locations
(e.g., impacts of new infrastructure investments).
• Lack of integration of historical experiences: The CUF-1 model's rules for allocating
future development were not calibrated against historical experience.
LEARNING MORE
Additional References
None.
Availability of Preview Copies of the Model
Potential users must contact the model developer.
Case Studies
Landis, J. D. 1995. "Imagining Land Use Futures: Applying the California Futures Model."
Journal of the American Planning Association, 61: 438-457.
Application Sites
Northern California; Solano and Sonoma counties
-40-
-------
California Urban Futures (CUF) Model
Second Generation: CUF-2
MODEL DEVELOPER(S): John Landis
MAILING ADDRESS:
CONTA CT INFORM A TION:
WEB SITE:
DOCUMENTATION:
Institute of Urban and Regional Development
University of California at Berkeley
Berkeley, CA 94720-1870
Phone: 510-642-5918
Fax: 510-643-9576
E-mail: jlandis@uclink.berkeley.edu
http://www-dcrp.ced.berkeley.edu
Landis, J. D. and M. Zhang. 1998. "The Second Generation of
the California Urban Futures Model: Part I: Model Logic and
Theory." Environment and Planning, B: Planning and Design,
25: 657-666.
Landis J. D., M. Zhang, and M. Zook. 1998. CUFII: The
Second Generation of the California Urban Futures Model. UC
Transportation Center, University of California, Berkeley, CA.
OVERVIEW
The purpose of the California Urban Futures Model Second Generation (CUF-2) model, like the
CUF-1 model, is to provide a framework for simulating how growth and development policies
might alter the location, pattern, and intensity of urban development. (See the evaluation of
the CUF-1 model for a more detailed description of the model's intended use.) The second-
generation was developed to address some of the theoretical holes of the first model.
The CUF-2 model performs many of the functions as the CUF-1 model (see the evaluation of
the CUF-1 model). Several changes were made to the first generation, however. The following
provides a brief description of each of the four main components of the CUF-2 model:
• The activity projection component uses a series of econometric models to project future
population, households, and employment by jurisdiction at 10-year intervals. Although
the future population and households are projected as they are in the CUF-1 model, the
employment projection is a new component of CUF-2.
• The GIS based spatial database generates and updates the location and attributes of
each developable land unit (DLU) and allows the user to visually display the spatial
pattern of growth. In CUF-2, DLUs are one-hectare grid-cells, not (as in CUF-1)
irregularly-shaped polygons.
-41-
-------
• The land use change submodel is calibrated against historical urban land use changes
Independent variables include: local population and employment growth; proximity to
regional job centers; site slope; whether the site is within or beyond city boundaries or
spheres of influence; the uses of surrounding sites; the availability of vacant land; site
proximity to freeway interchanges and transit stations; and site proximity to major
commercial, industrial, and public land uses.
• The model allows for spatial bidding for sites between four types of new development
land uses and three types of redevelopment and use change submodel is calibrated
against historical urban land use changes.
REQUIRED RESOURCES
Purchase Costs
Not available for "off the shelf" purchase.
Equipment Needs
CUF-2 requires a 300 MHz or higher PC or Sun Spare computer with 32 MB of RAM, 2 GB of
hard drive space, color monitor, Soloris and Windows 95 operating systems, SAS statistical
analysis software, and ArcView or Arclnfo.
Staff Requirements and Expertise
Installation and calibration of the model requires experience in SAS and Arclnfo as well as land-
use modeling expertise. Use of the model requires land-use expertise and general computer
experience.
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
The land-use categories addressed by CUF-2 are user defined and, therefore, could include
any land-use category as appropriate for the study area. Seven major land-use categories
were used for the CUF-2 San Francisco Bay Area simulation: undeveloped, single-family
residential, multi-family residential, commercial, industrial, transportation, and public.
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
•
•
•
•
No?
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
•
•
•
No?
-42-
-------
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees, taxes,
and incentives)
Yes?
•
•
•
•1
No?
1 Only for developer impact fees (with modifications), and
municipal sewer and water fees.
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Environmental Quality
School Quality
Crime
Other Quality-of-Life Conditions
Yes?
•1
•
•
•2
No?
•
•
•
1 Only if linked to a travel demand model.
2 For water quality, "yes" potentially or if amended by the user.
For air quality, "no."
Outputs Provided
Output
Quantitative outputs
Graphical outputs
Format
New development an redevelopment
acreage total by land use type
Maps of existing and projected
development (by land use type)
INFORMATION NEEDED TO RUN THE MODEL
This model requires the following user inputs:
Hectare-scale digital maps of urban land uses at two or more points in time.
Multiple Arclnfo coverages or themes profiling existing land uses, general plan, and
environmental characteristics, as well as jurisdictional boundaries.
• Jurisdiction-level tabular information profiling historical population growth.
-43-
-------
MODEL STRENGTHS AND LIMITATIONS
Strengths
• Easy to use and visual: The CUF-2 model allows users to prepare and evaluate
alternative policy scenarios quickly (a typical simulation can be completed in a matter
of hours) and in easy to read map form at almost any level of spatial detail.
• Expandable: The CUF-2 model is designed as a modular system of related but
independent submodels that can be updated to include new information and theories.
• Policy approach: The CUF-2 model simulates alternative development futures based on
specific policy changes.
• Calibrated to past local experience.
Limitations
• Availability: The CUF-2 model is currently unavailable as a product that can be
purchased "off the shelf."
• Data intensive: The CUF-2 model requires much more data than the original CUF-1
model.
• Model calibration requires detailed knowledge of statistics. Results can be spatially
auto-correlated.
LEARNING MORE
Additional References
Landis, J. D. 1995. Imagining Land Use Futures: Applying the California Futures Model.
Journal of the American Planning Association, 61: 438-457.
Landis, J. D. 1994. The California Urban Futures Model: A New Generation of Metropolitan
Simulation Models. Environment and Planning, B: Planning and Design, 21: 399-420.
Availability of Preview Copies of the Model
Potential users must contact the model developer.
Case Studies
For more information contact the model developer.
Application Sites
San Francisco, CA Bay Region.
-44-
-------
California Urban and Biodiversity
Analysis Model (CURBA)
MODEL DEVELOPER(S): John Landis, Michael Reilly, Pablo Monzon, and Chris Cogan.
MAILING ADDRESS: Institute of Urban and Regional Development
University of California at Berkeley
Berkeley, CA 94720-1870
CONTACT INFORMATION: Phone: 510-642-5918
Fax: 510-643-9576
E-mail: jlandis@uclink.berkeley.edu
WEB SITE: http://www-dcrp.ced.berkeley.edu
DOCUMENTATION: Development and Pilot Application of the California Urban and
Biodiversity Analysis (CURBA) Model. University of California at
Berkeley. September 1998.
OVERVIEW
The CURBA model was developed as a tool to help urban planners to evaluate the possible
effects of alternative urban growth patterns and policies on biodiversity and natural habitat
quality. CURBA can help direct urban growth while promoting environmental and ecological
quality.
The CURBA model consists of two major components, an Urban Growth Model and a Policy
Simulation and Evaluation Model. The Urban Growth Model assists the user in calibrating
equations that describe past urbanization patterns and applying the equations to project future
development patterns. The Policy Simulation and Evaluation Model projects how alternative
development policies will affect future urbanization patterns and the associated impacts on
habitat integrity. For example, CURBA can help users investigate the effects of urban growth
on vegetation land cover by type, habitat for various species (e.g., different mammals, reptiles,
and birds), changes in the level of fragmentation, etc. The CURBA model is used in
conjunction with ArcView and various Avenue scripts.
REQUIRED RESOURCES
Purchase Costs
Contact the model developer.
Equipment Needs
CURBA requires a 300 MHz or higher PC with 32 MB of RAM, 300 MB of hard drive space,
color monitor, Windows operating system, SAS or SPSS statistical analysis software, and the
ERSI ArcView Geographic Information System.
-45-
-------
Staff Requirements and Expertise
Installation and calibration of the model requires experience in ArcView and land-use modeling
expertise. Use of the model requires no land-use expertise, but does require familiarity with
ArcView, SAS, or SPSS.
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
The pilot simulation used the following categories: urban (does not differentiate between
commercial, industrial, and other urban land use types), vegetation types (from GAP datasets),
agricultural lands, and other landscape and infrastructure data. These categories are from the
California Farmland Mapping and Monitoring Project (FMMP).
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
No?
•
•
•
•
•
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
•
•
•
No?
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees, taxes,
and incentives)
Yes?
•1
•1
•1
No?
•
1 With calibration.
-46-
-------
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Bio-diversity and Habitat Quality
Environmental Quality
School Quality
Crime
Other Quality-of-Life Conditions
Yes?
•
•
•
No?
•
•
•
•
•
Note: CURBA can evaluate potential impacts of land use
changes on habitat types and species that are reliant on these
habitats.
Outputs Provided
Output
Evaluation results predicting the impacts of projected urban
growth on various habitat types
ESRI "shape files" displayed as bit-mapped images and
graphical displays of analysis results
Format
Maps and tabular
summaries
Polygon and grid files
INFORMATION NEEDED TO RUN THE MODEL
Inputs will depend on the goals and objectives of the user. Typically, the more detailed inputs
the user can provide, the more in-depth analyses can be performed. The following user input
was used for the pilot simulations:
• Land use types from the California Farmland Mapping and Monitoring Project (FMMP)
database (ArcView coverages). [Note: Certain data may require reformatting to
accommodate the GIS.]
• First and second generation GAP Analysis data, including ecoregions, vegetation
classification zones, and vertebrate species habitat.
• Slope and elevation data from the USGS Digital Elevation Model.
• Locations of roads, hydrographic line features, and major water bodies (Census Bureau
TIGER files).
• Jurisdictional boundaries (Census Bureau TIGER files).
-47-
-------
• Wetlands and vernal pools from the National Wetlands Inventory. Various
socioeconomic data (e.g., population and employment levels). FEMA Q3 Floodzones.
MODEL STRENGTHS AND LIMITATIONS
Strengths
• Easy to access and use: Once calibrated, CURBA is run entirely in ArcView and can be
used on a desktop computer. Base data are entirely in the public domain. It is also fast
and flexible.
• Reveals trends and patterns: CURBA allows stakeholders to better understand the
drivers of recent urbanization trends and patterns.
• Projective: CURBA allows users to project future urban growth patterns, the sensitivity
of urban growth to alternative regulatory and environmental policies, and the effects of
projected growth on habitat integrity and quality.
Limitations
• Future projected based on past: CURBA's results are reliant on how well the Urban
Growth Model equations explain historical growth patterns and the extent that the
factors that drove these patterns direct future development.
• All urban growth is equal: CURBA treats all urban developments the same and does not
allow for the possibility of redevelopment. Also, CURBA assumes that all forms and
patterns of urban growth represent the same level of habitat decline.
LEARNING MORE
Additional References
Not provided.
Availability of Preview Copies of the Model
Potential users must contact the developer.
Case Studies
Three environmental conservation scenarios for nine California counties (see Development and
Pilot Application of the California Urban and Biodiversity Analysis (CURBA) Model. University
of California at Berkeley. September, 1998.
Application Sites
Nine California counties:
El Dorado • Placer • Santa Cruz
Monterey • Sacramento • Sonoma
Nevada • San Joquin • Stanislaus
-48-
-------
DELTA
MODEL DEVELOPER(S): David Simmonds Consultancy
MAILING ADDRESS: 10 Jesus Lane
Cambridge CBS 8BA
England
CONTACT INFORMATION: Phone: +44(0)1223316098
Fax: +44(0)1223313893
E-mail: dsc@davidsimmonds.com
WEB SITE: http://www.davidsimmonds.com/land-use.html
DOCUMENTATION: Available from the model developer.
OVERVIEW
The DELTA model projects changes in urban areas, including changes in the location of
households, population, employment and the amount of real estate development. Typically
DELTA is set up to interact with a transport model. With a transport model, DELTA projects
changes in land use that affect the demand for transportation and the impact of changes
in accessibility on a variety of factors, including the location of different activities (e.g.,
households, employment) and the value of buildings. An optional regional level can be added
within DELTA to model the regional economy and migration between urban areas.
REQUIRED RESOURCES
Purchase Costs
DELTA software currently is not available for purchase off-the-shelf as a self-contained
software package. It is included as a component of overall consulting services provided by the
model developer. The price of overall consulting services is set on a project-by-project basis
depending on several factors, including the needs of the client, the size of the study area, and
the amount and condition of the data available for incorporation into the system.
Equipment Needs
DELTA requires an IBM (or compatible) Pentium 200 MHz computer, MS DOS run either from
DOS mode or under Windows 95/98, and DBOS memory manager (distributed with DELTA
model). The model developer also recommends the use of spreadsheets, databases, and
geographic information system (e.g., Maplnfo) to prepare the data needed to run the model
and to analyze the data obtained from the model. A color monitor is recommended.
Staff Requirements and Expertise
The consulting services of the model developer are required to use DELTA.
-49-
-------
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
•
•
•
•
No?
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
No?
•
•
•
•
•
•
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees, taxes,
and incentives)
Yes?
•
•
•
•
No?
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Environmental Quality
School Quality
Crime
Other Quality-of-Life Conditions
Yes?
•1
•1
•
No?
•
•
•
•
1 When DELTA is integrated with transport and environmental
models.
-50-
-------
Outputs Provided
The outputs provided by DELTA are the same type of data used to run the model (i.e.,
forecasts of numbers of households by type and location, jobs by sector and location,
floorspace by category and location, etc.), although the data are updated as a result of running
the model. DELTA is capable of producing output files that can be passed through Windows-
based tools. The outputs should be loaded into a spreadsheet, database, and/or mapping
software for analysis.
INFORMATION NEEDED TO RUN THE MODEL
DELTA was designed to be flexible to meet the needs of the user; therefore, it does not have
rigid input requirements. The inputs required to operate the model are user defined, allowing
the user to alter the inputs for each run of the model. DELTA is both an urban and a regional
model. The DELTA urban model contains six sub-models that address the development
process, demographic change (e.g., household formation) and economic growth, location and
relocation of households and jobs in the property market, car-ownership choices, changes in
employment status (working/non-working) and commuting patterns, and changes in the quality
of residential areas. The DELTA regional model contains an additional three models for
migration between different urban areas, the location of investment/disinvestment, and the
pattern of production and trade.
The inputs to the DELTA urban and/or regional models include fixed-format ASCII files
containing information on the location of households and jobs, car ownership levels, and
floorspace supply and rent for a base year and pre-base years. In addition, the user should
provide variables that define the economic and demographic scenarios to be modeled and
coefficients to describe the behavior of households, businesses, developers, etc. (e.g.,
sensitivity of developers' location decisions to expected profitability, sensitivity of households'
location decisions to rent, the quality of residential areas, etc.).
MODEL STRENGTHS AND LIMITATIONS
Strengths
• The DELTA model is unique in its ability to forecast changes over a series of short
periods. This allows the user to prepare and evaluate future planning and development.
• The DELTA model allows the user to incorporate information to generate specific
conditions into the model.
• The DELTA model provides an integrated software package that can be used as a stand
alone package or could be set up to interact with a wide range of transport models.
Limitations
• The DELTA model currently is unavailable as an off-the-shelf product. Licensing for this
model is on a project specific basis, and include the services of the model developer
-51-
-------
LEARNING MORE
Additional References
Simmonds, D.C. "The design of the DELTA land-use modeling package." Environment and
Planning, B: Planning and Design. Volume 26, pages 665-684. March 1999.
Simmonds, D.C. and Still, Ben. "DELTA/START: Adding Land Use Analysis to Integrated
Transport Models." Paper presented to World Conference on Transport Research,
Antwerp, 1998. "Selected proceedings of the 8th world conference on transport
procedures." Volume 4. Pergamon, an imprint of Elsevier Science. 1999.
Availability of Preview Copies of the Model
Not available.
Case Studies
Case study information is available from the model developer. Brief overviews of where the
model developer created both urban and regional DELTA applications can be found on the
developer's web site.
Application Sites
Greater Manchester and the Trans-Pennine Corridor, England
Edinburgh, Scotland
• Yorkshire, England (in development)
Sardina, Italy (in development)
Uruguay (in development)
-52-
-------
DRAM/EMPAL
(Other names: DRAM/EMPAL was part oflTLUP (Integrated Transportation and Land Use
Package); now it is part of METROPILUS.)
MODEL DEVELOPER(S): S.H. Putman and Associates, Inc.
MAILING ADDRESS: 128 Deakyneville Road
Townsend, DE 19734-9751
CONTACT INFORMATION: Phone: (302)659-3263
Fax: (302) 659-3264
E-mail: putman@pobox.upenn.edu
WEB SITE: http://dolphin.upenn.edu/~yongmin/intro.html
DOCUMENTATION: A user's manual is provided as part of the package.
OVERVIEW
DRAM/EMPAL projects the interactions and distribution of employment and housing in a
specified geographic area. DRAM/EMPAL combines two spatial interaction models: the
Disaggregated Residential Allocation Model (DRAM) and the Employment Allocation Model
(EMPAL) to quantify the interactions between the metropolitan patterns of employment and
population location and the networks of transportation facilities that connect them.
DRAM/EMPAL provides a tool that relates future estimates of the location and type of
employment in an area to their prior distributions, regional growth or decline, and the region's
transportation system.
DRAM/EMPAL formed the two major components of an integrated set of computer models
known as the Integrated Transportation and Land Use Package (ITLUP). Output from
DRAM/EMPAL (i.e..employment and household location and land use, trips generated for
home-to-work, home-to-shop, and work-to-shop) were used with the third component of ITLUP
to perform standard travel demand modeling (including submodels to estimate trip distribution,
modal choice, and traffic assignment). DRAM/EMPAL currently is the most widely used
employment, population, and land use forecasting application; it has been used internationally
in more than 4 dozen metropolitan areas.
DRAM/EMPAL has been incorporated into a new system called METROPILUS, which combines
employment and residence location and land consumption into a single, comprehensive
package operating within an ArcView GIS environment.
REQUIRED RESOURCES
Purchase Costs
$30,000 to $60,000. Purchase cost includes consultant services, maintenance, and training
costs.
-53-
-------
Equipment Needs
DRAM/EM PAL requires a MS Windows 95/98/NT operating system, a Pentium PC, color
monitor, and printer. Spreadsheet, database, statistical, and ArcView geographic information
system software are necessary for data management.
Staff Requirements and Expertise
Requires about one senior modeler with junior support.
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
DRAM/EM PAL models the interrelationships among transportation, location, and land use in
metropolitan areas. Land-use categories are further broken down into use by households by
income group, streets, highways, employment type, and vacant or unusable land. Vacant lands
are user defined and may contain various combinations of agricultural, and restricted land use
types such as parklands, etc.
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
•
•
•
No?
•
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
•
•
•
No?
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees,
taxes, and incentives)
Yes?
•
•
•
•
No?
Note: Any of the above questions may be answered "yes"
when DRAM/EMPAL is linked to the right model, such as
LANCON or METROPILUS. Without linking, most cannot be
answered as such.
-54-
-------
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal
Conditions
Availability of Open Space
Environmental Quality
School Quality
Crime
Other Quality-of-Life Conditions
Yes?
•
•
•
•
No?
•
•
•
Note: Any of the above categories with a "yes" may be
addressed when DRAM/EMPAL is linked to the right model,
such as LANCON or METROPILUS. Without linking, most
cannot be answered as "yes."
Outputs Provided
Output
EMPAL employment projections in each zone by
economic sector
DRAM projects the number of households in each zone
by income level or any other user-defined level
Submodel LANCON land consumption projections in
each zone
METROPILUS operates within GIS
Format
Standard formats such as Excel
or dbf, and ArcView maps
Standard formats such as Excel
or dbf, and ArcView maps
Standard formats such as Excel
or dbf, and ArcView maps
ArcView maps
INFORMATION NEEDED TO RUN THE MODEL
This model requires the following user input:
• Regional level for EMPAL: Target year values of total employment by economic sector.
• Regional level for DRAM: Total population, total person trips by purpose, percent
unemployment by sector, employees per household by household type, matrix of
households by income per employee by sector, jobs per employee, and net regional rate
of employee commuting.
• Analysis zone level for EMPAL (base year): Household by type, employment by sector,
total land area, land area occupied by basic and commercial employment, zone-to-zone
travel times and/or costs.
-55-
-------
• Analysis zone level for DRAM (base year): Households by type, total population, total
employed residents, group quarters population, land area use, land area occupied by
basic and commercial employment, employment by sector, zone-to-zone travel times
and/or costs.
Note: Data for DRAM/EMPAL are required both for the region, or overall modeling domain, and
for each analysis zone. Input/output can be a number of different formats, including Excel or
dbf.
MODEL STRENGTHS AND LIMITATIONS
Strengths
• Has been (and continues to be) used by numerous metropolitan areas and is a robust
model.
• Has ability to introduce constraints or other influences, particularly to account for local
knowledge.
• Input requirements use generally available data sources.
• Calibration of model is relatively easy.
Limitations
• DRAM/EMPAL modeling process is statistical, or aggregate theory based, rather than
disaggregate or micro theory based; a reduced form of logit is used for location.
• Little or no scope to introduce planning policies other than land zoning, except by
specific constraints or attractiveness functions.
• The absence of any mechanism for simulating the land market cleaning process
underlying multi-year infrastructure change.
• The impact of zoning policies cannot be well represented in DRAM/EMPAL. Monetary
and non-monetary incentives to guide land-use development cannot be represented in
DRAM/EMPAL.
• Limited number of independent variables used to make projections may lead to
underestimates of the full impact of some infrastructure improvements.
• The spatial resolution of the zones in DRAM/EMPAL is limited by a number of factors,
the principal factor being availability of data.
• Sensitivity analyses are not possible.
-56-
-------
• In order to achieve relative ease of use the model focuses on aggregate choice behavior
rather than on individual choice behavior.
• The model requires training and experience to run correctly and efficiently. It is not an
off-the-shelf product. It requires initial consultant involvement.
LEARNING MORE
Additional References
Putman, S.H. 1975. Urban land use transportation models: a state-of-the-art summary.
Transportation Research 9:187-202.
Putman, S.H. 1983. Integrated Urban Models. Pion Limited, London. England.
Putman, S.H. 1984. Dynamic properties of static-recursive model systems of transportation
and location. Environment and Planning 16A: 1503-1519.
Putman, S.H. 1986. Future directions for urban systems models: some pointers from empirical
investigations. Hutchinson, B. and Batty, M. (Eds.) Urban Systems Modeling. 91-108.
Elsevier North-Holland, Amsterdam.
Putman, S.H. 1991. Integrated Urban Models 2. Pion Limited, London, England.
Rosenbaum, A.S. and Koenig, B.E. 1997. Evaluation of Modeling Tools for Assessing Land
Use Policies and Strategies. Prepared for USEPA (EPA420-R-97-007).
Texas Transportation Institute. 1998. Land Use Compendium. Prepared for USEPA and
USDOT.
Availability of Preview Copies of the Model
Not available.
Case Studies
• Northeast Illinois Planning Commission (Chicago MPO)
Max Dieber
Phone: (312) 454-0400
• Bart Lewis (model tester)
Chief, Socioeconomic Analysis Division
Atlanta Regional Commission
Phone: (404) 364-2540
-57-
-------
• The following Web site contains summaries of telephone interviews with MPOs who use
DRAM/EMPAL:
www.bts.gov/other/tmip/papers/landuse/compendium/dvrpc_toc.htm
Application Sites
• Southern California • San Diego, CA
• Atlanta Region • Orange County, CA
Boston, MA • Kansas City
Northeast Illinois • Orlando/Kissimmee, FL
• North Central Texas • Phoenix, AZ
Houston-Galveston, TX area • Portland-Vancouver, OR
Sacramento, CA • Colorado Springs, CO
• Seattle, WA • San Antonio, TX
-58-
-------
Growth Simulation Model (GSM)
MODEL DEVELOPER(S): Joe Tassone
MAILING ADDRESS: Maryland Department of Planning
301 W. Preston Street, Room 1101
Baltimore, MD 21201-2385
CONTACT INFORMATION: Phone: (410)767-4500
Fax:
E-mail: JTassone@mdp. state, md. us
WEB SITE: http://www.mdp.state.md.us
DOCUMENTATION: Developing Growth Management Options for Maryland's
Tributary Strategies. Managing Maryland's Growth, Growth and
Watershed Planning Series. Draft, March 1997.
OVERVIEW
The GSM was developed by the Maryland Office of Planning beginning in 1992 to project
population growth and new development effects on land use/land cover nutrient pollution loads,
and small streams under alternative land management strategies To develop these estimates,
the GSM uses population, household, and employment projections to estimate demand for
residential and commercial development. Demand is then distributed to developable land,
based on capacity under existing or alternative zoning, development regulations, and resource
conservation mechanisms; and on information about development patterns and trends. Land
use change to accommodate projected growth is then estimated as a function of management
tools.
REQUIRED RESOURCES
Purchase Costs
The GSM is public domain, but has not yet been adapted as an application that can be
distributed to other users. Contact model developer for details.
Equipment Needs
The GSM is currently set up to run on a UNIX-based workstation or Windows NT. This could be
modified to meet the needs of the user. Computer size and speed are a function of the
database scale of resolution and the geographic extent of the study area. As it is currently
being used, GSM requires a 500 MHz PC, 128 MB RAM, and 133 MHz bus speed. Arclnfo and
a relational database (i.e., Paradox) software are also necessary to run GSM.
Staff Requirements and Expertise
It is necessary that staff running GSM are comfortable with GIS and relational database
software. GSM has not been packaged for less skilled users. The target users of the model
are land use planners and managers, and others interested in land and water resource
-59-
-------
conservation. Calibration and use of model requires experience in land use management and
modeling and programming in relational database applications (e.g., Paradox, Oracle).
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
The GSM can address many different user-defined land-use types. Under current
implementation, the Maryland Office of Planning is using three categories of residential density,
four types of non-residential developed land, four types of natural vegetation cover, and four
categories of agricultural land. These land uses can be modified by the user. The following
basic categories can be accommodated; additional information is provided in Appendix B.
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
•
•
•
•
No?
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
•
•
•
No?
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees,
taxes, and incentives)
Subdivision Regulations
Environmental Regulations
Yes?
•1
•
•
•2
•
•
No?
1 Currently under development.
2 No fiscal policies are pre-set in the model. However, if the
user can provide specifications on the impact of the revenue
source, then the policy can be incorporated.
-60-
-------
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal
Conditions
Availability of Open Space
Environmental Quality
School Quality
Crime
Other Quality-of-Life Conditions
Land use
Stream buffers
Nutrient Pollution Loads
Yes?
•1
•
•
•
•
•
No?
•2
•
•
•
1 Currently under development.
2 Was under development, but has been put on hold due to
lack of resources.
Outputs Provided
GSM generates land use and land use change information in a dataset that can then be tied to
the original land use GIS dataset. Using the information in the GIS form, a varied set of
statistics and graphics can then be generated by the user.
Output
Land use projections
Data Summaries
Format
Arclnfo, Maps
Relational Database files (e.g.,
Spreadsheets
Reports
Graphs
Paradox)
-61-
-------
INFORMATION NEEDED TO RUN THE MODEL
There is a great deal of flexibility regarding the data needed to run GSM. At a minimum, land
use/land cover data and geo-referenced management areas are necessary to operate the
model. An example of some of the data that have been incorporated into model applications
are:
Input
Land use
Soils
Watershed and county boundaries
Streams
Buffer zones
Environmentally sensitive areas
Zoning, growth, and land preservation boundaries
Sewer service boundaries
Population, household, and employment demographics
Preserved land
Format
Arclnfo
Arclnfo
Arclnfo
Arclnfo
Arclnfo
Arclnfo
Arclnfo
Arclnfo
Database
Arclnfo
Additional information is also required about the effects of management alternatives. For
example, typical lot yields in an R-1 zoning district are 3 D.U. / acre, and there are not forest
conservation requirements for subdivisions. This information can be derived from empirical
data or values can be assumed based on potential ranges.
MODEL STRENGTHS AND LIMITATIONS
Strengths
Simulates land use change as a function of population, employment, land use
management techniques, and market preferences.
Can be customized to work a various levels of scale and detail.
Can be used to extrapolate land use change for a much larger geographic area than
that to which it was applied.
Limitations
GSM has to be customized for each application by a skilled programmer, depending on
the scale, resolution, and data used to represent generalized needs.
-62-
-------
Cannot currently map land use change in vector format as output. The user must
currently use geo-referenced spheres or buffer polygons, statistics, and graphs to look
at the change.
LEARNING MORE
Additional References
Chesapeake Bay Program. 1998. Integrating Build-Out Analysis and Water Quality Modeling
to Predict the Environmental Impacts of Alternative Development Scenarios. CBP/TRS
178-97.
Tassone, J. 1998. Smart Growth Options for Maryland's Tributary Strategies. Maryland Office
of Planning. Baltimore, MD.
Maryland office of Planning. 2000. Methods used to Estimate 1997-2020 Land Use Change.
Baltimore, MD.
Population, Socioeconomic, and Land Use Task Force of the Scientific and Technical Advisory
Committee, Chesapeake Bay Program. 1999. Population Growth Land Use Change and
Impacts to Land and Water Resources.
Availability of Preview Copies of the Model
No preview copy is available. Contact model developer for additional information.
Case Studies
None
Application Sites
GSM has been utilized to estimate growth patterns across the state of Maryland in over 350
watersheds at a variety of spatial scales ranging from 3rd order streams, to major tributaries of
the Chesapeake Bay, to the entire Chesapeake Bay Watershed (see Population Growth, Land
Use Change, and Impacts to Land and Water Resources).
-63-
-------
INDEX®
MODEL DEVELOPER(S): Criterion Planners/Engineers, Inc.
MAILING ADDRESS: 725 NW Flanders Street, Suite 303
Portland, OR 97209-3539
CONTACT INFORMATION: Phone: 503-224-8606
Fax: 503-224-8702
E-mail: eliot@crit.com
WEB SITE: www.crit.com
DOCUMENTATION: Available on the web site. Also, custom user guides are prepared
for each community-specific version.
OVERVIEW
INDEX® was developed in 1994 to measure the characteristics and performance of land-use
plans and urban designs with "indicators" derived from community goals and policies (e.g.,
measures the degree of transit orientation in a proposed residential subdivision).
REQUIRED RESOURCES
Purchase Costs
Negotiated fee dependant upon scope of community customization; fees range from $15,000 to
$75,000.
Equipment Needs
INDEX® requires a 200 MHz or higher PC with 64 MB of RAM, 150 MB of free hard drive space,
Microsoft Windows 95 or NT operating system, GIS software, color monitor with a minimum
resolution of 800 x 600, and a color printer.
Staff Requirements and Expertise
Installation, calibration, and use of the model requires experience in ArcView GIS as well as
land-use modeling expertise.
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
The number of land-use categories addressed by INDEX® is determined by the number of land-
use categories in each community's unique land-use planning system. Therefore, the actual
land-use categories are defined by the community and can be as detailed or general as
needed. Typically, as few as 6 and as many as 30 categories are used.
-64-
-------
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
•
•
•
•
No?
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
•
•
•
No?
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees,
taxes, and incentives)
Yes?
•
•
No?
•
•
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Environmental Quality
School Quality
Crime
Other Quality of Life Conditions
Yes?
•
•
No?
•
•
•
•
•
Outputs Provided
Outputs are determined by community customization. Selected results are mapped in ArcView
shapefiles and some are stored in an Access database.
-65-
-------
Example Outputs
jobs/housing ratio
residential density
employment density
land-use mix
proximity to community amenities
residential water consumption
criteria air pollutant emissions
street connectivity
transit orientation
parking supply
imperviousness
pedestrian route directness
park and open space availability
greenhouse gas emissions
INFORMATION NEEDED TO RUN THE MODEL
The number and type of inputs required are determined by each community during
customization that includes topical scoping. Example inputs include:
Input
Parcels
Street centerlines
Land-uses
Dwelling units by type
Employment by type
Transit routes and stops
Sidewalks
Bicycle routes
Off-street parking areas
Building footprints
Significant environmental features
Format
ArcView shapefile
ArcView shapefile
ArcView shapefile
ArcView shapefile
ArcView shapefile
ArcView shapefile
ArcView shapefile
ArcView shapefile
ArcView shapefile
ArcView shapefile
ArcView shapefile
MODEL STRENGTHS AND LIMITATIONS
Strengths
• Each copy is customized for a community's unique set of conditions and priorities.
• Integrates the explanatory power of GIS mapping with a comprehensive set of urban
impact measurements.
-66-
-------
• Provides communities with a consistent, efficient tool for evaluating incremental
development proposals and monitoring the implementation of long-range land-use
plans.
Limitations
• Requires detailed GIS data and user expertise.
• Must be in tandem with local four-step travel demand models in order to provide
comprehensive land-use/transportation impact estimates.
• The land-use plan or urban design being evaluated must be created exogenously.
LEARNING MORE
Additional References
Refer to www.crit.com.
Availability of Preview Copies of the Model
Contact the model developer.
Case Studies
Local government operational use of the model does not typically produce narrative case
studies, but instead generates indicator scores for a given plan or design. Federal use of the
model has been documented in the following U.S. EPA Urban and Economic Development
Division reports: Transportation and Environmental Effects of Infill Versus Greenfield
Development, 1998; and Transportation and Environmental Analysis of the Atlantic Steel
Development Proposal, Atlanta, GA, April 1999.
Application Sites
• Atlanta, GA (multiple locations)
Beaverton, OR
Coquitlam, BC
• Ft. Lewis, WA
Kamloops, BC
Montgomery County, MD (multiple locations)
• Orlando, FL
Queens, NY
San Diego, CA (multiple locations)
Sacramento, CA
• West Palm Beach, FL
-67-
-------
IRPUD
MODEL DEVELOPER(S): Michael Wegener
MAILING ADDRESS: Institute of Spatial Planning
University of Dortmund
D-44221 Dortmund
Germany
CONTACT INFORMATION: Phone: +492317552291/2401
Fax: +49 231 755 4788
E-mail: mw@irpud.rp.uni-dortmund.de
WEB SITE: http://irpud.raumplanung.uni-dortmund.de/irpud/index_e.htm
DOCUMENTATION: See the web site listed above.
OVERVIEW
The IRPUD model projects the location decisions of industry, residential developers and
households, the travel patterns that result from location decisions, construction activity and
land-use development, and the impacts of public policies in the fields of industrial development,
housing, public facilities, and transportation within an urban area over a specified amount of
time.
The IRPUD model consists of six integrated submodels that address the following factors:
transportation; changes to population, employment, residential buildings and non-residential
buildings due to biological, technological or long-term socioeconomic trends; public programs;
private construction; regional labor market; and regional housing market. Together, the six
submodels form one comprehensive stand-alone model system.
REQUIRED RESOURCES
Purchase Costs
The IRPUD model is a research model developed through academic research projects funded
by the German National Research Council. It is not available for purchase off-the-shelf as a
self-contained software package. It is presently being used in a research project funded by the
European Commission. In future projects, the IRPUD model will be included as a component of
overall consulting services by the model developer, with the price of the overall consulting
services being set on a project-by-project basis depending on several factors, including the
needs of the client, the size of the study area, and the amount and condition of the data
available for incorporation into the system.
Equipment Needs
The IRPUD model requires a 300 MHz or higher PC with 128 MB or more RAM, 4 GB or more
hard drive space, color monitor (1024x768), color printer, Windows NT operating system,
-68-
-------
programming language compilers (e.g., FORTRAN, C, C++), and statistical analysis software
(e.g., SAS, SPSS). In addition, a geographical information system (e.g., Arclnfo) is used to
prepare the data needed to run the model. The model has its own result representation and
analysis tools.
Staff Requirements and Expertise
The consulting services of the model developer are required to use the IRPUD model. The use
and calibration of the model by the user requires land use modeling expertise.
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
•
•
•
•
No?
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
No?
•
•
•
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees,
taxes, and incentives)
Yes?
•
•
•
•
No?
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Yes?
•
•
No?
•
-69-
-------
Community Characteristic
Environmental Quality
School Quality
Crime
Other Quality-of-Life Conditions
Yes?
•1
No?
•
•
•
1The IRPUD can forecast CO2 emissions as a function of
forecasting transportation-related indicators. Environmental
submodels that calculate traffic noise and air pollution indicators
are under development.
Outputs Provided
The IRPUD model generates graphical and tabular outputs. Graphical outputs are in the form
of trajectories-curves representing the development of a particular model variable or output
indicator over time-or maps. The table below lists examples of indicators that a user can select
for output. Tabular output is in the form of ASCII printer output files. Graphical output is either
on-screen or in WordPerfect WPG format for later post-processing and printing. In addition,
custom-written programs are used to extract model results from the model data base for
mapping.
Example Indicator Outputs Generated by the IRPUD Model:
Percent foreign population
• Trips by trip purpose (work, shopping, education, other)
• Percent population 0-5, 6-14, 15-29, 30-59, 60+ years
• Trips by mode
Households
Mean travel time
• Total employment
Mean travel cost
Non-service, service, retail employment
Car-km per capita per day
Unemployment rate
CO2 emissions by car per capita per day
• Job-labor ratio
CO2 emissions by transport per capita per day
• Total dwellings
• Transport expenses per household per month
Percent single-family dwellings
Public transport expenses per household per month
Housing floor space per capita
Car ownership (cars per 1,000 population)
Mean housing rent per square mile
-70-
-------
INFORMATION NEEDED TO RUN THE MODEL
The user of the IRPUD model must provide four groups of data as inputs: model parameters,
regional data, zonal data and network data. The table below presents the types of inputs that
fall into each of these groups. All user input is in the form of ASCII files.
Model Parameters
These inputs influence the model equations
contained in the IRPUD model.
Demographic parameters
Household parameters
Housing parameters
Technical parameters
Monetary parameters
Preference parameters
Regional Data
The IRPUD model requires information on
the total urban region to project intraregional
changes.
• Employment
• Immigration
• Outmigration
Zonal Data
This information describes activities in the
urban region undergoing analysis during the
base year simulated by the IRPUD model.
Each zone in the urban region must have the
following information:
• Population
• Labor force/unemployment
• Households
• Dwellings
• Households/housing
• Employment/workplaces
• Public facilities
• Land use
• Rents/prices
Network Data
The IRPUD model considers transport
networks with modes car, public
transportation and walk/bicycle. The model
codes the following network information link
by link using a multimodal coding scheme:
• Link type
• From-node
• To-node
• Link length
• Link travel time (public transportation)
• Base speed (road)
For each public transportation line, the
IRPUD model codes the following
information:
• List of nodes
• Peak-hour headway
-71-
-------
MODEL STRENGTHS AND LIMITATIONS
Strengths
• The IRPUD model studies the impacts of policies from the fields of industrial develop-
ment, housing, public facilities, and transportation. The model addresses global policies
(i.e., those that affect urban development in the whole region) and local policies (i.e.,
regulatory or direct zone-specific investment projects).
• The IRPUD model differs from other operational urban models due to its high temporal
resolution and its full integration of land-use transport interaction in each simulation
period. This makes it a truly dynamic model (compared with other approaches, such as
cross-sectional equilibrium approaches).
• The IRPUD model introduces assumptions about human spatial behavior drawn from
time-space geography based on time and cost budgets and satisfying behavior into
urban modeling. This makes the model uniquely suitable to model elastic trip generation
behavior (responsible for much of the growth in mobility in metropolitan regions).
Limitations
• Limitations of the present version of the IRPUD model include its coarse spatial
resolution and its lack of a submodel of urban freight transportation. The model
developer is addressing both limitations in the ongoing revision of the model through the
PROPOLIS project.
LEARNING MORE
Additional References
Wegener, M. 1998. The IRPUD Model: Overview. Dortmund: Institute of Spatial Planning.
http://irpud.raumplanung.uni-dortmund.de/irpud/pro/mod/mod_e.htm.
Wegener, M. 1983. Description of the Dortmund Region Model. Working Paper 8. Dortmund:
Institute of Spatial Planning.
Availability of Preview Copies of the Model
Contact the model developer.
Case Studies
Wegener, M. 1996. Reduction of CO2 emissions of transport by reorganisation of urban
activities. In: Hayashi, Y., Roy, J. (eds.): Land Use. Transport and the Environment.
Dordrecht: Kluwer Academic Publishers, 103-124.
Wegener, M., Mackett, R.L., Simmonds, D.C. 1991. One city, three models: comparison of
land-use/transport simulation models for Dortmund. Transport Reviews 11, 107-29.
-72-
-------
Wegener, M. 1998. Sustainable Urban Spatial Structures. Do We Need to Rebuild Our Cities?
Dortmund: Institute of Spatial Planning, http://irpud.raumplanung.uni-
dortmund.de/irpud/pro/co2/co2_e.htm.
Application Sites
The metropolitan region of Dortmund, Germany
-73-
-------
Land Transformation Model (LTM)
(Other names: LTM-ANN (Artificial Neural Network); LTM-MCE
(Multi-Criteria Evaluation; and LTM-LR (Logistic Regression).)
MODEL DEVELOPER(S): Dr. Bryan C. Pijanowski
MAILING ADDRESS: Michigan State University
231 Natural Science Building
East Lansing, Ml 48824-1115
CONTACT INFORMATION: Phone: (517)432-0039
Fax: (517)432-1054
E-mail: pijanows@msu.edu
WEB SITE: http://www.ltm.msu.edu
DOCUMENT A TION: http://www. Itm. msu.edu/document
http://www.ncgia.ucsb.edu/conf/landuse97/papers/pijanowski_bry
an/paper, html
OVERVIEW
Development of the Land Transformation Model (LTM) began in 1995 and is ongoing. The
model uses landscape ecology principles, patterns of interactions to simulate land use change
process, to forecast land use change. Though the model can be used in any definable region,
precedence is given to watersheds as the spatial extent in LTM applications. Conceptually,
the LTM contains six interacting modules: 1) policy framework; 2) driving variables; 3) land
transformation; 4) intensity of use; 5) processes and distributions; and 6) assessment
endpoints. The pilot model was developed for Michigan's Saginaw Bay Watershed and
contains two of the six LTM modules; driving variables and land transformation. The pilot model
integrates a variety of land use change driving variables, such as population growth, agricultural
sustainability, transportation, and farmland preservation policies for the watershed.
REQUIRED RESOURCES
Purchase Costs
Contact the model developer. It is likely that there will be no associated costs to obtain the
model and its associated routines.
Equipment Needs
LTM requires a 300 MHz or higher PC or Sun Spare with a minimum of 256 MB RAM, Windows
NT or Sun Solaris operating system, a color monitor with a minimum resolution of 1024 x 768
and a color printer. Spreadsheet (Excel), database (MS Access), statistical (S-Plus and SAS),
programming language compiler ('C'), GIS (ArcView or Arclnfo), and Stuttgart Neural Network
Simulator (SNNS) software are necessary for data management.
-74-
-------
Staff Requirements and Expertise
Calibration and use of the model requires expertise in land-use modeling and "C" language
programming as well as SNNS neural network batch files.
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
The LTM can address up to eight different land-use types.
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
No?
•
•
•
•
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
•
•
No?
•
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees, taxes,
and incentives)
Yes?
•
•
No?
•
•
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Environmental Quality
School Quality
Yes?
•
No?
•
•
•
•
-75-
-------
Community Characteristic
Crime
Other Quality-of-Life Conditions
Yes?
No?
•
•
Outputs Provided
Output of the LTM includes a time series of projected land uses in the watershed at user
specified time steps.
Output
Land use projection maps
Data Summaries
Format
GIS(ArclnfoGRID)
Excel
INFORMATION NEEDED TO RUN THE MODEL
To operate the model, a community must have a GIS data base that contains basic land use
information. At a minimum, the following input data are needed:
Input
Previous land use
Roads, highways, streets
Surface water (rivers, lakes, etc.)
Elevation
Public lands
Population
Per capita use requirements
Format
ArclnfoGRID
Arclnfo Lines
Arclnfo lines or polygons
ArclnfoGRID
ArclnfoGRID
ArclnfoGRID
ArclnfoGRID
MODEL STRENGTHS AND LIMITATIONS
Strengths
• GIS outputs provide stakeholders and resource managers with easy to understand
results.
• Allows users to explore various types of inputs that are parameterized using a GIS.
• Coupled to a neural network software package that learns how historical changes in use
are driven by various social, political and environmental factors.
-76-
-------
Limitations
• The "drivers" are not dynamic; projective ability is around 35% for 100m x 100m cell
size.
• Takes several large C programs to couple the GIS and neural network simulation
software.
• Environmental process models that are being used require large amounts of memory
(around 2 GB of RAM for a 5- to 7-county area.
• The model requires training and experience to run. It is not a commercial off-the-shelf
product. It was developed to be used by a researcher working with resource managers.
LEARNING MORE
Additional References
Several papers were in press at the time of publication of this report. Refer to web site for
additional information.
Availability of Preview Copies of the Model
Contact the model developer.
Case Studies
Several projects utilizing the LTM are underway, refer to web site for more information.
Application Sites
Saginaw Bay, Michigan watershed
-77-
-------
Land-Use Change Analysis System (LUCAS)
MODEL DEVELOPER(S): Michael W. Berry, Richard O. Flamm, Brett C. Hazen, Rhonda M.
Maclntyre and Karen S. Minser.
MAILING ADDRESS: Department of Computer Sciences
114AyresHall
University of Tennessee
Knoxville, TN 37996-1301
CONTACT INFORMATION: Phone: (865)974-3838
Fax: (865) 974-4404
E-mail: berry@cs.utk.edu
WEB SITE: http://www.cs.utk.edu/~lucas
DOCUMENTATION: http://www.cs.utk.edu/~lucas/publications/pblications.html
OVERVIEW
LUCAS was developed in 1994 to examine the impact of human activities on land use and
the subsequent impacts on environmental and natural resource sustainability. LUCAS
stores, displays and analyzes map layers derived from remotely-sensed images, census
and ownership maps, topographical maps, and outputs from econometric models using the
Geographic Resources Analysis Support System (GRASS), a public-domain GIS. Simulations
using LUCAS generate new maps of land cover representing the amount of land-cover change.
Issues such as biodiversity conservation, conservation goals, long-term landscape integrity,
changes in real estate values, species abundance, and land-ownership characteristics can be
addressed by LUCAS.
REQUIRED RESOURCES
Purchase Costs
The model is public domain and is distributed upon request.
Equipment Needs
LUCAS requires a UNIX-based workstation (e.g., Sun SPARC station), Microsoft Windows with
the OSF/Motif library toolkit (version1.21), a color monitor and a color printer. GIS (GRASS) and
spreadsheet software is necessary to analyze the results.
Staff Requirements and Expertise
Calibration and use of the model requires expertise in land-use modeling and "C++" language
programming.
-78-
-------
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
Land-use is modeled as a multi-variate function of land cover change. This is a spatially-explicit
modeling approach. LUCAS can address many different land-use types.
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
•
No?
•
•
•
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
•
•
•
No?
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees, taxes,
and incentives)
Yes?
•
•
•
No?
•
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Environmental Quality
School Quality
Crime
Other Quality-of-Life Conditions
Yes?
•
•
No?
•
•
•
•
•
-79-
-------
Outputs Provided
Output of the LUCAS includes a time series of projected land uses in the watershed at user
specified time steps. For each land cover type, the following output information is provided.
Output
Area
Amount of edge
Edge/area ratio
Mean patch size
Number of patches
Cumulative frequency distribution of patches by size
Proportion of land cover
Amount of total edge
Standard deviation of patch size
Size of largest patch
Mean patch shapes
Format
Statistical (SAS) and graphical
Statistical (SAS) and graphical
Statistical (SAS) and graphical
Statistical (SAS) and graphical
Statistical (SAS) and graphical
Statistical (SAS) and graphical
Statistical (SAS) and graphical
Statistical (SAS) and graphical
Statistical (SAS) and graphical
Statistical (SAS) and graphical
Statistical (SAS) and graphical
INFORMATION NEEDED TO RUN THE MODEL
To operate the model, a community must have the following GIS information. The GIS used by
LUCAS is GRASS, but most commercial GIS software can readily convert their files to the
GRASS format.
Input
Transportation networks (access and transportation costs)
Slope and elevation (indicators of land-use potential)
Ownership (land holder characteristics)
Land cover (vegetation)
Population density
Format
GRASS grid
GRASS grid
GRASS grid
GRASS grid
GRASS grid
MODEL STRENGTHS AND LIMITATIONS
Strengths
• LUCAS provides a graphical user interface that is intuitive and easily understood by
users with a wide range of technical abilities and experience.
• The system provides a flexible and interactive computing environment for landscape
management studies.
-80-
-------
Limitations
• As a non-commercial GIS package, many bugs still exist in the GRASS software. Also,
some of the features of GRASS are not well-documented.
• The model requires training and experience to calibrate. It is not a commercial off-the-
shelf product and was developed to be used by a researcher working with resource
managers.
LEARNING MORE
Additional References
M.W. Berry, B.C. Hazen, R.L. Maclntyre, and R.O. Flamm. 1996. Lucas: A System for
Modeling Land-Use Change. IEEE Computational Science & Engineering, Vol. 3, No. 1,
1996, pp. 24-35.
M. W. Berry, R.O. Flamm, B.C. Hazen, and R.L. Maclntyre. 1994. The Land-Use Change
Analysis System (LUCAS) for Evaluating Landscape Management Decisions. Technical
Report CS-94-238, University of Tennessee, Knoxville, TN, December 1994.
R.L. Maclntyre, B.C. Hazen, and M.W. Berry. 1994. The Design of the Land-Use Change
Analysis System (LUCAS): Part I - Graphical User Interface. Technical Report CS-94-266,
University of Tennessee, Knoxville, TN.
Also, please refer to web site.
Availability of Preview Copies of the Model
Refer to web site and contact the model developer for additional information as needed.
Case Studies
LUCAS was developed and implemented for the Little Tennessee River Basin and for the Hoh
and Dungeness watersheds in the Olympic Peninsula.
Application Sites
Little Tennessee River Basin, Tennessee
Olympic Peninsula, Washington
-81-
-------
Markov Model of Residential Vacancy Transfer
MODEL DEVELOPER(S): Philip Emmi and Lena Magnusson
MAILING ADDRESS: Professor Philip Emmi
Geography Department
University of Utah
260 S. Central Campus Drive, Rm 270
Salt Lake City, UT 84112-9155
CONTACT INFORMATION: Phone: (801)581-5562
Fax: (801)581-8219
E-mail: pcemmi@geog.utah.edu
WEB SITE: www.geog.utah.edu/faculty/emmi.html
DOCUMENTATION: Philip C. Emmi and Lena Magnusson. 1995. Opportunity and
mobility in urban housing markets. Progress in Planning, 43(1):
^^^^^^^^^^^^^^8^^^^^^^^^^^^^^^^^^^^^^^^
OVERVIEW
The Markov Model of Residential Vacancy Transfer explores changes in demand for various
types of residential housing (e.g., high-density apartments or condos, single-family detached
dwellings, etc.) within a community as various subpopulations, such as single adults, young
families, and empty nesters transition through the community from one housing environment
to another. The model could be applied to help plan new residential zoning and development
based on existing demographics and population pressures, or to identify where certain
residential sectors or areas might decline without coordinated efforts to accommodate
demographic changes. The model would be particularly useful for small towns and cities on the
metropolitan fringe seeking to establish, and redeveloping inner cities trying to re-establish, a
sustainable housing landscape throughout existing and future residents' lifetimes.
REQUIRED RESOURCES
The model is a mathematical formula based on linear algebra and ordinary differential
equations-it is not a discrete software package. Pre-processing of inputs and developing the
formula can be implemented with a variety of statistical analysis software (e.g., SAS, SPLUS,
SPSS).
Purchase Costs
None.
Equipment Needs
Any type of computer can be used; there are no specific hardware or software requirements.
-82-
-------
Staff Requirements and Expertise
Extensive.
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
No?
•
•
•
•
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
No?
•
•
•
•
•
•
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees,
taxes, and incentives)
Yes?
•
•
No?
•
•
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Environmental Quality
School Quality
Crime
Other Quality of Life Conditions
Yes?
•
No?
•
•
•
•
•
•
-83-
-------
Outputs Provided
Output
A simulation of intra-urban household moves between
residential sectors across a census or projection period
Probabilities of residential mobility by household type
as a function of the sectoral distribution of vacancy
initiations, the pattern of housing sector interaction and
the sectoral distribution of households
A measure of housing sector interaction in terms of the
probability of a vacancy introduced into sector "1" being
associated with a residential move in sector "J"
Format
Explicit numerical output
Numerical matrix
Format type not available
INFORMATION NEEDED TO RUN THE MODEL
Input
Format
Residential addresses, both past and present,
classified into "internally homogeneous" housing
sectors (e.g., single-family residences, retiree
apartment complexes, single-parent public housing,
etc.) within a specified area
Any electronic format that can be
read by the user's statistical
analysis software
Households and their past and present addresses, the
community, or area, as derived either from sequential
census records or a survey of recent household
creations, conclusions, and moves
Any electronic format that can be
read by the user's statistical
analysis software
Note: Emmi (1994) relaxes the requirement for "internally homogeneous" sectors to ease
exposition and facilitate inter-urban comparisons.
MODEL STRENGTHS AND LIMITATIONS
Strengths
• The model simulates impacts of new vacancies on urban residential relocations and the
accommodation of new entrants (immigrants, newly formed households) into the
housing market.
• The model also simulates impacts of newly created vacancies on the residential mobility
for various urban sub-groups (e.g., single professionals, young families, wealthy empty
n esters).
• The model achieves a high level of projective accuracy.
-84-
-------
Limitations
• The model depends on a stable, semi-closed system of residential moves between
census years.
• The model does not explicitly simulate land-use changes.
• The model examines only discrete sectors of the residential housing market.
• By focusing on vacancy transfers, the model speaks only indirectly to behaviorally-
based adjustments in housing accommodations.
LEARNING MORE
Additional References
Emmi, P.C. 1990. A model of monopolistic competition among sectors of a metropolitan
housing market. Netherland Journal of Housing and Environmental Research 5: 87-103.
Emmi, P.C. and Magnusson, L. 1988. Residential vacancy chain model of an urban: Exercises
in impact and needs assessment. Scandinavian Housing and Planning Research 5: 129-
145.
Emmi, P.C. and Magnusson, L. 1993. Intrasectoral homogeneity and the accuracy of
multisectoral models. Ann. Reg. Sci. 27:343-363.
Emmi, P.C. and Magnusson, L. 1994. The projective accuracy of residential chain vacancy
chain models. Urban Studies 31 (7): 1117-1131.
Emmi, P.C. and Magnusson, L. 1995(b). Further evidence on the accuracy of residential
vacancy chain models. Urban Studies 32(8): 1361-1367.
Availability of Preview Copies of the Model
Not applicable because the model has not been implemented as its own software. Authors'
papers, however, provide examples and procedures for implementing this model.
Case Studies
Case-study information is summarized in several academic papers listed above.
Application Sites
Application sites are identified in the academic papers listed above. The sites include:
3 Swedish cities
• 42 U.S. metropolitan areas
-85-
-------
MEPLAN
MODEL DEVELOPER(S): Developed by Marcial Echenique/lan Williams
MAILING ADDRESS: Marcial Echenique and Partners (ME&P)
49-51 High Street
Trumpington, Cambridge CB2 2HZ
England
CONTACT INFORMATION: Phone: +44(0)1223840704
Fax: +44(0)1223840384
E-mail: admin@meap.co.uk
WEB SITE: www. meap.co.uk
DOCUMENTATION: User, technical, and programming manuals are available from
the developer.
OVERVIEW
MEPLAN was developed to help communities analyze the inter-related effects of land use and
transportation policies. Specifically, MEPLAN can: 1) determine the effects of transport on the
choices of location by residents, employers, developers, and others; 2) determine how land use
and economic activity create the demand for transport; and 3) project and evaluate the many
impacts that planning decisions will have on land use and transport. MEPLAN is an integrated
software package of modules that enables users to look at supply and demand in both land and
transport, and compare the effects of one policy with those of another policy. These
comparisons may include physical changes, economic benefits, and social and environmental
indicators. MEPLAN also incorporates a forecast capability to assess: 1) what is likely to
happen through time given certain assumptions about economic growth, 2) how the most likely
'reference' scenario would be altered as a result of specific policy measures in a future year, 3)
how prices will be affected (e.g., house prices, cost of living, production costs, etc.)
Three main modules are provided in the model: 1) Land use/economic module (LUS), 2)
Transport module (TAS), and 3) Economic evaluation module (EVAL). LUS combines a spatial
model with a variable relationship input-output model to project where factors will locate and
what the pattern of trades will be between zones. TAS examines modal split, route assignment,
and capacity restraint. An interface program (FRED) between TAS and LUS deals with the two-
way interactions between these two modalities. FRED enables MEPLAN to estimate the
number and distribution of trips or flows directly from the results of the land-use model. FRED
also is able to calculate the reverse interaction-how transport changes affect the pattern of land
uses. EVAL combines the results of LUS, TAS, and FRED and compares them to alternative
plans or to a base-case scenario.
-86-
-------
REQUIRED RESOURCES
Purchase Costs
MEPLAN costs about $25,000 to purchase and about $640 per day to train someone to operate
and use the model. Annual estimated maintenance costs are approximately 10% of the
purchase price.
Equipment Needs
MEPLAN requires a 200 MHz or higher PC with 64 MB of RAM, Microsoft Windows NT
operating system, and an associated graphics system such as MEPLUS/Maplnfo.
Staff Requirements and Expertise
Installation, calibration and use of the model requires training in MEPLAN as well as experience
in land-use and transportation modeling. A typical small team to operate MEPLAN will consist of
a planner, a transport engineer, and an economist. Specialists in computing should not be
needed. Training in the use of MEPLAN is provided as a consulting service by the developer.
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
The land-use categories addressed by MEPLAN are user defined and therefore could include
any land-use category.
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
•
•
•
•
No?
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
•
•
•
No?
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees, taxes,
and incentives)
Yes?
•
•
•
•
No?
-87-
-------
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Environmental Quality
School Quality
Crime
Other Quality-of-Life Conditions
Yes?
•
•
•
•
No?
•
•
•
Outputs Provided
Specific outputs are dependent on how the user has set up the model within the analytical
framework it provides and what the model is being used to test. MEPLAN could potentially
generate the following outputs:
Employment by sector
Population by income group
Households by car ownership group
Land area by activity
Floor space by activity
Price by floorspace/land type
INFORMATION NEEDED TO RUN THE MODEL
MEPLAN was designed to be flexible to meet the needs of the user; therefore, it does not have
rigid input requirements. Because MEPLAN provides a framework that can be adapted to suit a
variety of user needs, the inputs required to operate the model are user defined and can be
altered for each run of the model. The MEPLAN framework incorporates a land/use economic
model and a transport model, so information (inputs) on land use, floor space, supply and
demand for land and buildings, prices for space, pattern of prices, availability of public
transport, ownership of cars, road and rail infrastructure, trip types, and related information
would be useful.
-88-
-------
MODEL STRENGTHS AND LIMITATIONS
Strengths
• MEPLAN comes close to modeling interrelated variables in cities.
• The model allows analysis of different kinds of policies.
• The highly synthetic nature of the model allows most of the description of even the base
situation to be estimated within the model, reducing the reliance on observed data.
• It is possible to implement the MEPLAN with very little data other than that for the base
year. MEPLAN includes floor space zoning restrictions in the spatial choice formulation
as well as development costs, which allows for the impact of zoning policies to be
represented.
• In MEPLAN, the development of floor space is projected, in part, on the basis of
development costs. Therefore, policies that offer monetary incentives to developers to
build in targeted zones or at specified densities could be represented in terms of
decreased development costs.
Limitations
• MEPLAN is quite data-intensive. Calibration process can be difficult and time consuming
if base year observed data is lacking or inconsistent.
• Validation of base year results may be problematic if suitable observed data is lacking.
• There is no way to integrate data for years subsequent to the base year. (The model
could be recalibrated to a later year if suitable data are available, or a partial update
could be done with a subset of more up-to-date data).
LEARNING MORE
Additional References
Research into Practice: the Work of the Martin Centre in Urban & Regional Modelling. Special
issue of Environment & Planning B: Planning & Design, Vol. 21(5), pages 513-650, 1994.
Availability of Preview Copies of the Model
Not available.
-89-
-------
Case Studies
A wide range of case study project summaries are available from the model developer.
Application Sites
London, England • Naples, Italy
South East of England • Bolzano, Italy
Cambridge, UK • Madrid, Spain
Santiago, Chile • San Sebastian, Spain
Sao Paulo, Brazil • Basque Region, Spain
Bilbao, Spain • Colombia (national model)
• Tokyo, Japan • Chile (national model)
Helsinki, Finland • Sweden (national model)
Caracas, Venezuela • Sao Paulo state, Brazil
Sacramento, USA • Central Region of Chile
-90-
-------
METROSIM
MODEL DEVELOPER(S): Alex Anas & Associates
MAILING ADDRESS: 151 Rollingwood Street
Williamsville, NY 14221-1855
CONTACT INFORMATION: Phone: 716-688-5816
Fax: 716-688-5816
E-mail: aanas@adelphia.net
URL: http://www.acsu.buffalo.edu/~alexanas
WEB SITE: http://www.bts.gov/tmip/papers/landuse/compendiurn/dvrpc_ch1.
htm#1.4.3
http://www.nsf.gov/cgi-bin/showaward?award=9816816
DOCUMENTATION: Please e-mail Alex Anas to receive an electronic copy of
documentation
OVERVIEW
METROSIM is an operational large scale computer simulation model that uses an economic
approach to forecast the interdependent effects of transportation and land use systems and of
land use and transport policies at the metropolitan level. METROSIM is used to evaluate
transportation projects and travel related changes, land use controls, employment growth
scenarios, income growth and other polices or forecast changes.
METROSIM can be used to obtain quantitative forecasts of travel flows, employment changes,
congestion levels, new construction of residential and commercial buildings, land use changes,
etc. The user can specify land use constraints and zoning regulations in the model. The user
can also obtain benefit-cost ratios for projects or policy interventions simulated by METROSIM.
METROSIM can produce a one-shot long run equilibrium forecast for transportation and land
use in a metropolitan area, or METROSIM can operate in annual increments and produce
yearly changes to transportation and land use from the existing situation until convergence to a
steady state is achieved.
REQUIRED RESOURCES
Purchase Costs
To purchase: $20,000-$30,000
To run the software: $2,500 for three initial runs or negotiable. Full reports are included.
To maintain the model software: User support for $5,000-$10,000/year
Train to use the software: $10,000 one time
-91-
-------
Equipment Needs
METROSIM requires a 300 MHz or higher computer (PC, Macintosh, or Sun Spare) with 128
MB of RAM, color monitor, FORTRAN or C programming language compilers, Excel or Access
for data management, SAS or SPSS statistical analysis software, and Arclnfo or Maplnfo. It
can be adapted to any operating system, but UNIX is preferred.
Staff Requirements and Expertise
No technical land-use experience is required-only general computer experience is necessary.
Participation of the model developer is necessary for its use.
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
The categories and number of land uses addressed by the model are user defined with no limit.
It all depends on the level of available data, but at least two residential, two non-residential, and
three vacant uses are recommended.
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
•
•
•
•
No?
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
•
•
•
No?
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees, taxes,
and incentives)
Yes?
•
•
•
•
No?
-92-
-------
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic)
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Environmental Quality
School Quality
Crime
Other Quality-of-Life Conditions
Yes?
•
•
•
•1
•
•
•
No?
1 For water quality only, not for air quality.
Outputs Provided
Output
Basic industry distribution by zone and by type of basic industry.
Non-basic industry distribution by zone and by type of non-basic
industry.
Residential real estate distribution by type and zone.
Non-residential real estate distribution by type and by zone.
Vacant land distribution by type and by zone.
Households
Travel (commuting and non-work)
Traffic assignment on the network
Rents and market prices for each type of real estate by zone
Vacancy rates for each type of real estate
Graphical outputs available-this is a special feature which can be
provided upon request at a modest fee.
Format
As desired.
As desired.
As desired.
As desired.
As desired.
As desired.
As desired.
As desired.
As desired.
As desired.
As desired.
INFORMATION NEEDED TO RUN THE MODEL
This model requires the following user input:
• CTPP elements 1, 2, and 3, and Bureau of the Census STF1A and STF3A files.
• Transportation network's link-node description in Maplnfo, Arclnfo, or other format.
• Land use by type of land use desired in the model.
-93-
-------
• If available, regional input/output model
• If available, land and property values by zone and land use type. Please note that file
formats are not important because files can be easily converted from one format to
another.
MODEL STRENGTHS AND LIMITATIONS
Strengths
• It is a model firmly rooted in economics and recognizes how market forces operate in
shaping and changing land use
• It is very rapid on the computer and does not rely on approximate solutions.
• It deals explicitly with land use policy and land use change
Limitations
• Currently does not have a GIS interface but that can be easily developed at a small fee.
LEARNING MORE
Additional References
See http://www.acsu.buffalo.edu/~alexanas.
Availability of Preview Copies of the Model
Potential users must contact the developer.
Case Studies
• Alex Anas (model tester)
151 Rollingwood Street
Williamsville, NY 14221-1115
Phone: 716-688-5816
Fax: 716-688-5816
E-mail: aanas@adelphia.net
URL: http://www.acsu.buffalo.edu/~alexanas
-94-
-------
• Kazem Oryani (used model in operational context)
URS Consultants, Inc.
1 Penn Plaza, Suite 610
New York, NY 10119-0698
Phone: 212-736-4444
E-mail: Kazem_Oryani@urscorp.com
• Contact Alex Anas for information and reports on the following specific applications:
- New York Region to test effects of region-wide transport scenarios
- Harlem Line Corridor
- Staten Island MIS alternatives
Application Sites
Chicago, IL
Houston, TX
Harlem Line Corridor, New York City, NY
• New York City, NY Region
• Pittsburgh, PA
• Staten Island, NY
• San Diego, CA
-95-
-------
SAM-IM
(Sub-Area Allocation Model-Improved Method)
MODEL DEVELOPER(S): Planning Technologies, LLC
MAILING ADDRESS: 1816 Lomas Boulevard, NW, Suite A
Albuquerque, NM 87104-1206
CONTACT INFORMATION: Phone: (505)243-8088
Fax: (505) 243-5655
E-mail: plantech@all4gis.com
WEB SITE: http://www.all4gis.com
DOCUMENTATION: Planning Technologies, LLC. July 1999. SAM-IM User's Guide
(developed for exclusive use by the Maricopa Association of
Governments).
OVERVIEW
The Sub-Area Allocation Model-Improved Method (SAM-IM) is a land use allocation and
forecasting model developed for the Maricopa (Arizona) Association of Governments to support
official adopted forecasts of land use in the region for use by the transportation and air quality
programs in Maricopa County. Another version of this model includes the Land Use Analysis
Model (LAM) that was also developed by Planning Technologies, LLC for the Middle Rio
Grande Council of Governments (Albuquerque, New Mexico). The Albuquerque application
was used to analyze the implications associated with various approaches to regional growth
control. Planning Technologies, LLC develops and implements each version of this model
based on the needs and unique features of the users, usually planners and other decision-
makers affiliated with governmental organizations such as associations/councils of
governments.
SAM-IM uses geographic information systems (GIS) to access data, including acreage,
population, and employment, to investigate existing land use relationships and construct future
land use scenarios. The basic concept behind SAM-IM was inspired by earlier work conducted
at the University of California-Berkeley by Dr. John Landis. SAM-IM is essentially a modeling
platform for generating forecasts and allocations of land use according to methods similar to
those which have appeared in the literature under Dr. Landis' name.
SAM-IM projects future land use patterns based on a simulation of growth and development in
a region. Urbanization is simulated by evaluating lands available to absorb growth based on
various site suitability characteristics, such as potential net profit as measured by various
surrogates such as highway access, proximity to infrastructure, consistency with zoning and
general plans, local development policies, etc. Land found by the model to be most likely to be
developed, based on its native site suitability characteristics, is "built."
-96-
-------
The user defines the specific allocation methodology in SAM-IM, so the application suite can
support a variety of specific theories of urban growth that fall into this general approach
framework. SAM-IM offers considerable adaptability and flexibility: it can be altered to fit a
locality's own definitions of land uses and can be programmed to reflect whatever principles or
factors are thought (by the user) to most influence growth and development. Virtually any
factor that can be represented geographically (as an ArcView shape file) can be taken into
consideration.
The land use (and other) layers in SAM-IM are all represented by ArcView shape files. The
allocation mechanism itself, however, treats geographies as high-resolution cellular (raster)
features, using ArcView's Spatial Analyst extension. SAM-IM provides automated routines to
convert between vector (polygon) and raster (also called grid, pixels or cells) representations of
land use and other themes (e.g., flood plains, redevelopment districts, retirement communities,
etc).
SAM-IM provides planners with a number of capabilities, including:
• Analyzing Land Use Plans: Analysis of land use plans, whether it be a database of
existing land use, or a proposed plan, with respect to measures of sustainability. SAM-
IM provides a number of measures of relationships between different types of land use,
such as measures of job-housing balance, for either a region or for individual
communities in a region.
• Creating and Editing Land Use Plans: The application is equipped with editors to
support the development of new urban forms that can reflect different philosophies
about growth and development. The new urban forms may promote infill development
in the established urban area; they may promote corridor development around
transportation facilities; or they may reflect current development trends. Whatever the
urban form concept, SAM-IM provides the planner with a toolbox of editing and drawing
tools to create the database to reflect the growth management strategy.
• Creating Site Evaluations: The application is equipped with a toolbox that make the
creation of site evaluation themes (i.e., themes that describe the characteristics of land
and its suitability to be developed) easy. Special tools in the toolbox let planners create
themes that, for example, convert conventional shape file themes (e.g., flood plains) to
grid, measure feature proximity (e.g., to highways or infrastructure), provide
neighborhood measures (e.g., market within 3 miles of a site), etc.
• Projecting Future Land Use Patterns: Forecasting of growth, or more accurately, to
disaggregate or allocate a spatial growth forecast for a region. SAM-IM does so by
giving the planner a way to represent the value of land and the probability that it will be
developed. SAM-IM simulates the development of a region in a way that is consistent
with overall regional projections, consistent with existing land use, and consistent with
the land use plan.
• Supporting Other Urban Model Systems: The SAM-IM platform includes a "geographic
calculator" that lets users easily create and format data sets used by other modeling
systems to complete an urban growth analysis, including transportation models such as
those built for EMME/2.
-97-
-------
The result of a SAM-IM simulation is a land use map which depicts land use throughout the
region for some future year.
REQUIRED RESOURCES
Purchase Costs
SAM-IM software currently is not available for purchase off-the-shelf as a self-contained
software package. It is included as a component of overall consulting services provided by the
model developer. The price of overall consulting services is set on a project-by-project basis
depending on several factors, including the needs of the client, the size of the study area, and
the amount and condition of the data available for incorporation into the system. Typically, the
total cost ranges from $30,000-100,000.
Equipment Needs
SAM-IM requires a 400 MHz or higher IBM (or compatible) computer with 128 MB of RAM and
2 GB free hard drive space, an MS Windows NT or MS Windows 95 operating system, color
monitor and printer, and ArcView 3.2 (with the Special Analyst Extention). Visual BASIC
compiler not required but is helpful. Statistical software necessary for calibration only.
Staff Requirements and Expertise
The consulting services of the model developer and in-house staff that understand ArcView are
required to use SAM-IM, and similar models created by the model developer.
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
SAM-IM is capable of including any kind of land use data. The data included in the model
depends on what land uses the user would like to include, and what data is available.
Therefore, the user defines the actual land use categories (up tp 40), which can be as detailed
or as general as needed.
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other (e.g., business
park, institutional)
Yes?
•
•
•
•
•
No?
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
•
•
•
No?
-98-
-------
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees, taxes,
and incentives)
Yes?
•
•
•
No?
•
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Environmental Quality
School Quality
Crime
Other Quality-of-Life Conditions
Yes?
•
•
No?
•
•
•
•
•
Outputs Provided
Output
Distribution of future land uses
Format
GIS maps
INFORMATION NEEDED TO RUN THE MODEL
Before SAM-IM can perform an allocation, the user must provide certain inputs into the model.
There are also optional inputs that the user can incorporate into the model, but are not
necessary for the model to run.
-99-
-------
Input
Dwelling unit density for each land use
type (to generate a regional forecast)
Employment density for each land use
type (to generate a regional forecast)
Existing land use layer
Proposed land use layer
Proposed project information
Other areas of concern (e.g.,
protected habitat, stream buffers)
Format
Text file
One-digit Standard
Industrial Classification
(SIC) Codes
GIS coverage
GIS coverage
GIS coverage
GIS coverage
Required/Optional
Required
Required
Required
Required
Optional
Optional
MODEL STRENGTHS AND LIMITATIONS
Strengths
• SAM-IM contains an editor feature that allows users to edit land use files while
maintaining planar polygon topology that ArcView shape files conventionally do not
support. Planners can add, delete, copy, paste, and split land use polygons and can
recede them with new uses and densities with the help of on-screen forms. Planners
can also create an on-screen encyclopedia of known project types and densities with
which to draw on when coding land use themes.
• SAM-IM contains a toolbox of functions to make site evaluation themes that control the
forecasting process easy to construct. The functions significantly extend ArcView's own
capabilities for cellular representation of geography (known as "grid") offered in the
Spatial Analyst. They let users build cellular land use grids, grids from any other
conventional polygon theme on any attribute, compute new grids, compute new grids
through look-up tables, measure proximities, feature buffers, and neighborhood sums
etc. All of these features are of particular interest when creating a consolidated set of
factors that influence development probability and potential.
• SAM-IM lets users create and forecast new land use "scenarios." Scenarios typically
represent alternative land use plans, growth policies, and target years. So time-series
of long-range forecasts over extended periods of time, by five year intervals, can be
supported.
• The allocation process in SAM-IM will not violate land use policies for a region, for
example the projection of future land uses will not violate the uses and densities
adopted as part of zoning or general plan maps. This feature was especially highly
regarded in Maricopa County, whereby previous models generated forecasts that were
in conflict with adopted planning policy.
• SAM-IM contains a "geographic calculator" feature by which planners can generate new
datasets used by other urban models, such as transportation models. The calculator
summarizes land use according to any zone-based geography of interest, such as traffic
-100-
-------
analysis zones, municipal boundaries, school district boundaries, etc. In addition, it can
compute new socioeconomic variables from equations (e.g., population) from dwelling
unit descriptions and assumptions about dwelling unit vacancy rates and persons-per-
households all expressed geographically), automatically recognizing and manipulating
the geographic unit for which the terms of the equation are expressed in. Virtually any
output file format can be created, including dBASE files, delimited ASCII files, and fixed
format ASCII files for use in other programs.
SAM-IM provides numerous measures of relationships between different types of land
use, such as measures of job-housing balance, for either a region or for individual
communities in a region.
SAM-IM can also operate on a "microscopic level." It can provide a user with a forecast
for an area smaller than an acre.
Communities can use SAM-IM to evaluate alternative land use scenarios against local
performance indicators developed by the community.
Limitations
• SAM-IM is targeted for major agencies responsible for forecasting land use, with mature
GIS support capabilities, systems, and databases and a significant degree of GIS
expertise on-staff.
• SAM-IM can be too complicated for a community to use without assistance from the
model developer. It has a steep initial learning curve, which can affect the technical
expertise and resources needed to use this model.
• Although SAM-IM addresses a wide variety of land uses, its treatment of mixed uses
and redevelopment is presently cumbersome. These areas have been targeted for
improvement in the future.
LEARNING MORE
Additional References
Presentation at the ESRI International Users Conference, June, 2000; San Diego, CA.
Availability of Preview Copies of the Model
Contact the developer to arrange for an on-site presentation and demonstration.
Case Studies
Not specified.
Application Sites
Maricopa County, AZ
• Albuquerque, NM metropolitan area
-101-
-------
SLEUTH
MODEL DEVELOPER(S) Keith C. Clarke
MAILING ADDRESS: Geography Department
University of California, Santa Barbara
Ellison Hall 3620
Santa Barbara, CA 93106-4060
CONTACT INFORMATION: Phone: (850)893-7961
Fax: (850) 893-3146
E-mail: kclarke@geog.ucsb.edu
WEB SITE: http://www.geog.ucsb.edu/projects/gig
DOCUMENTATION: http://www.ncgia.ucsb.edu/projects/gig/model_document.htm
All available documentation on SLEUTH can be found on the
above web site.
OVERVIEW
The SLEUTH (Slope, Land use, Exclusion, Urban, Transportation, Hillshading) model,
commonly known as the Clarke Cellular Automata Urban Growth Model or as the Clarke Urban
Growth Model, is intended to simulate urban growth in order to aid in understanding how
expanding urban areas consume their surrounding land, and the environmental impact this has
on the local environment. SLEUTH derives its name from the six types of data inputs: slope,
land use, urban, exclusion, transportation, and hillshading. This model simulates the transition
from non-urban to urban land-use using a grid of cells (cellular automaton) each of whose land-
use state is dependent upon local factors (e.g., roads, existing urban areas, topography),
temporal factors, and random factors.
REQUIRED RESOURCES
Purchase Costs
None - may be downloaded for free at http://www.ncgia.ucsb.edu/projects/gig/download.htm
Equipment Needs
The SLEUTH model requires a PC, workstation, or mainframe with a UNIX operating system
and gnu C compiler (gcc). X-Windows is required for graphical version.
Staff Requirements and Expertise
Installation and calibration of the model requires experience in UNIX operating system, text
editor, and gnu C compiler (gcc), as well as land-use modeling expertise. Use of the model
requires land-use expertise and familiarity with UNIX operating system, text editor, and gnu C
compiler (gcc). Familiarity with X-Windows and X-libraries also important for graphical
versions.
-102-
-------
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
The SLEUTH model assumes two land use maps and a set of predefined land use categories
with names assigned by the user (e.g., a numeric value, such as 6, in the land cover file to
represent forest nonurban land uses). The model handles any combination of user-defined
land-use categories, including those represented in the tables below.
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
•
•
•
•
No?
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
•
•
•
No?
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees, taxes,
and incentives)
Yes?
•
•
•
•
No?
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Environmental Quality
School Quality
Crime
Other Quality-of-Life Conditions
Yes?
•1
•
•1
No?
•
•
•
•
1 Any fiscal or environmental impact which can be estimated as
a function of urbanized area could be developed for the output
of this model, but the model does not do so directly.
-103-
-------
Outputs Provided
The model provides outputs as a set of GIF image files that can be merged into an animation or
brought into a GIS as data layers. Resolution of output images depends on the resolution of
the input data.
Output
Snapshot of a particular year
Cumulative image that results from multiple runs and show a
of urbanization for a given year (i.e., Monte Carlo image that
Monte Carlo probability runs)
probability
results from
A set of best fit metric between modeled and real data for calibrating the
model
Actual values of model output for control years averaged over the number
of model simulations
The standard deviations of the average actual values
Ending coefficient values
The start and stop times for an entire model execution
Format
GIF image file
GIF image file
—
—
-
-
-
INFORMATION NEEDED TO RUN THE MODEL
Input
Slope: Derived from average from a DEM (Digital Elevation Model)
Excluded areas: Water bodies and other land where urbanization cannot
occur.
Roads/transportation network
Seed
Background
ASCII "schedule" files control when images are read. These include:
• Urban.dat. • Landuse.dat (deltatron)
• Roads.dat • Landuse. classes (deltatron)
Format
GIF
GIF
GIF
GIF
GIF
MODEL STRENGTHS AND LIMITATIONS
Strengths
• Concurrently simulates four types of growth (spontaneous, diffusive, organic, and road-
influenced)
• Provides both graphical and statistical outputs.
• Incorporates momentum of booms and busts using threshold multiplier with subsequent
temporal decay.
• Allows for relatively simple alternative scenario projection.
-104-
-------
Limitations
• The model does not explicitly deal with population, policies and economic impacts on
land use change, except in terms of growth around roads.
LEARNING MORE
Additional References
Clarke, K.C., Hoppen, S. and Gaydos, L. 1997. A self-modifying cellular automaton model of
historical urbanization in the San Francisco Bay Area. Environment and Planning B:
Planning and Design, vol. 24, pp. 247-261.
Clarke, K. C. and Gaydos, L. 1997. Long Term Urban Growth Projection Using A Cellular
Automaton Model and GIS: Applications in San Francisco and Washington/Baltimore.
International Journal of GIS, Special Issue of Population Modeling and Development,
(Under Review).
Clarke, K.C., Hoppen, S., Gaydos, L. 1996. Methods and Techniques for Rigorous Calibration
of a Cellular Automaton Model of Urban Growth. Third International Conference/Workshop
on Integrating GIS and Environmental Modeling, Santa Fe, New Mexico, January 21-25,
1996. Santa Barbara: National Center for Geographic Information and Analysis.
Kirtland D., DeCola L, Gaydos L, Acevedo W., Clarke K., Bell C. 1994. An analysis of human-
induced land transformations in the San Francisco Bay/Sacramento area. World Resource
Review, vol. 6(2); pp 206-217. (URL: http://edcdgs9.cr.usgs.gov/urban/pubs.html)
Kramer, J. 1996. Integration of A GIS with a Local Scale Self-Modifying Urban Growth Model in
Southeastern Orange County, New York. M.A. Thesis. Hunter College-CUNY.
Availability of Preview Copies of the Model
The SLEUTH Model can be downloaded from
http://www.ncgia.ucsb.edu/projects/gig/download.htm
Case Studies
• Leonard Gaydos, Chief (tested model)
Research, Technology and Applications
USGS Western Mapping Center
345 Middlefield Road, MS 531
Menlo Park, CA 94025-3591
Phone: 650-329-4330
Fax: 650-329-4710
Email: lgaydos@usgs.gov
-105-
-------
Application Sites
• Baltimore-Washington, DC • Chicago-Milwaukee
• Chester County, PA • Detroit, Ml
Orange County, CA • Greater New York Area
Santa Barbara, CA • Mid-Atlantic Interstate Area
San Francisco, CA • Middle Rio Grande Basin, NM
Sterling Forest, NY • Philadelphia-Wilmington, PA
• Utah Front Range, UT
-106-
-------
Smart Growth INDEX1
MODEL DEVELOPER(S): Criterion Planners/Engineers, Inc.
(with Fehr & Peers Associates, Inc.)
MAILING ADDRESS: 725 NW Flanders Street, Suite 303
Portland, OR 97209-3539
CONTACT INFORMATION: Phone: 503-224-8606
Fax: 503-224-8702
E-mail: eliot@crit.com
WEB SITE: www.crit.com
DOCUMENTATION: Available on the web site.
OVERVIEW
Smart Growth INDEX® was developed in 1998 to help communities evaluate alternative land-
use and transportation scenarios, including regional growth management plans, land-use and
transportation plans, and land development proposals. It is a customizable, GIS-based tool that
evaluates the different scenarios by scoring their outcomes using a set of environmental
indicators. Smart Growth INDEX® can operate in two basic modes: it can provide forecasts
over time and parcel-based "snapshots" at a single point in time. An underlying assumption of
the model is that population and employment growth are directly related to a locale's
accessibility to transportation and infrastructure services. It contains an internal travel demand
submodel that allows the user to estimate transportation outcomes from land-use changes
without the use of a traditional four-step transportation model. Smart Growth INDEX® is a
sketch-level planning tool intended to provide a preliminary analysis of alternative growth
scenarios, land-use plans, and urban designs to aid a community in defining alternatives that
warrant a closer look.
REQUIRED RESOURCES
Purchase Costs
Initial versions of Smart Growth INDEX® became available in mid-1999 at no cost through the
U.S. EPA Urban and Economic Development Division. Enhanced versions currently are under
development. For final pricing, contact developer.
Equipment Needs
Smart Growth INDEX® requires a 300 MHz or higher PC with 64 MB of RAM, 500 MB of free
hard drive space, Microsoft Windows 95 or NT operating system, GIS software, color monitor
with a minimum resolution of 1024 x 768, and a color printer.
-107-
-------
Staff Requirements and Expertise
Installation, calibration, and use of the model requires experience in GIS as well as land-use
and transportation-based modeling.
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
The number of land-use categories addressed by Smart Growth INDEX® is determined by the
number of land-use categories in each community's unique land-use planning system.
Therefore, the actual land-use categories are defined by the community and can be as detailed
or general as needed. Typically, 6-30 categories are used.
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
•
•
•
•
No?
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
•
•
•
No?
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees,
taxes, and incentives)
Yes?
•
•
•
No?
•
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Environmental Quality
Yes?
•
•
•
No?
•
-108-
-------
Community Characteristic
School Quality
Crime
Other Quality of Life Conditions
Yes?
•
No?
•
•
Outputs Provided
The output files are dependent upon the type of analysis being performed (i.e., forecast or
snapshot). For a forecast analysis, the performance indicators (output files) are scored for
interval and the designated horizon years. For a snapshot analysis, a similar set of indicators
are scored for a single point in time. All tabular results are provided in dbf file format. Selected
results are mapped in ESRI shapefiles and some are stored in an Access database.
Forecast Analysis Outputs
growth compactness • jobs/housed workers balance
residential population • vehicle miles and hours traveled
employment density • vehicle hours of delay
land-use mix • mode split
incentive area • auto travel cost
housing density • air pollution
housing transit proximity • climate change
employment transit proximity
residential energy and water
consumed
Snapshot Analysis
population density •
land-use mix •
residential density •
diversity of housing type •
housing proximity to transit •
jobs/housed workers balance •
employment density •
employment proximity to transit •
street connectivity •
energy consumption •
criteria air pollutant emissions •
Outputs
park space availability
housing proximity to recreation
open space
pedestrian orientation
pedestrian route directness
vehicle miles traveled
vehicle trips
street network density
auto travel costs
residential water consumption
greenhouse gas emissions
-109-
-------
INFORMATION NEEDED TO RUN THE MODEL
To operate the model, a community must have a GIS database containing basic land use,
transportation, housing, and employment information. At a minimum, the following inputs are
required:
Input
Existing housing by type
Existing employment by job count and location
Current and/or proposed land-use plan designations by class
Street centerlines by functional class
Transit lines by type
Format
ESRI shapefile
ESRI shapefile
ESRI shapefile
ESRI shapefile
ESRI shapefile
If the model is operated in tandem with TRANSCUD or MINUTP, two common travel demand
models, then a traffic analysis zone (ESRI shapefile) also is required.
Optional inputs include the following:
Input
Features that constrain urbanization; e.g., agricultural soils, steep slopes
Incentives areas for urbanization; e.g., transit corridors, brownfields
Existing and planned infrastructure service areas and facilities
Existing traffic counts by street segments
Local jurisdiction boundaries
Other subarea boundaries, such as traffic analysis zones or census
tracts
Format
ESRI shapefile
ESRI shapefile
ESRI shapefile
ESRI shapefile
ESRI shapefile
ESRI shapefile
If the model is operated in tandem with TRANSCUD or MINUTP, then cross-classification
matrix of persons per household and auto ownership by housing type can also serve as input.
MODEL STRENGTHS AND LIMITATIONS
Strengths
• Smart Growth INDEX® provides communities with a consistent, efficient tool for
evaluating alternative growth scenarios, land-use plans, and urban designs.
• The model integrates the explanatory power of GIS mapping with a comprehensive set
of urban impact measurements.
-110-
-------
The model can either be operated with an internal abbreviated four-step travel demand
sub-model or be linked to the TRANSCAD or MINUTP travel demand models.
Limitations
• The model requires detailed GIS data and user expertise and land-use and
transportation modeling expertise.
• The parameters and assumptions that control the spatial allocation of the growth
forecast do not include the effects of land prices.
• The method used for spatially allocating the growth forecast is "hard coded" and cannot
be readily modified by users.
LEARNING MORE
Additional References
Refer to www.crit.com.
Availability of Preview Copies of the Model
Contact the model developer.
Case Studies
Not available.
Application Sites
Sacramento, CA • st Paul, MN
• Newark, DE . Wildwood, MO
• Gainesville, FL • Wilmington, NC
Indianapolis, IN . Las Vegas, NV
• Covington, LA . Charleston, SC
• Boston, MA . Houston, TX
• Baltimore, MD . San Antonio, TX
' Carol County, MD . Burlington, VT
• Westminster, MD . Milwaukee, Wl
-111-
-------
Smart Places
MODEL DEVELOPER(S): Electric Power Research Institute (EPRI)
Contact: Paul Radcliffe
MAILING ADDRESS: 3412 Hillview Avenue
Palo Alto, CA 94304-1344
CONTACT INFORMATION: Phone: (650)855-2720
Fax: (650) 855-2619
E-mail: pradclif@epri.com
WEB SITE: http://www.epri.com orwww.smartplaces.com
DOCUMENTATION: The user's manual and the developer's manual can be obtained
from Paul Radcliffe (EPRI).
OVERVIEW
Smart Places, Version 4.0, is a geographic decision support system created to assist
communities in the design and evaluation of land-use development alternatives with user-
selected criteria. Smart Places provides an interactive computer tool that allow communities to
explore and design alternative development plans, and evaluate environmental and economic
impacts of a proposed design. The model allows users to define the parameters to be used for
evaluation. Smart Places is intended to be customized for specific locations and design
objectives and can be applied to small rural towns as well as large urban areas.
Smart Places is an extension of Environmental Systems Research Institute's (ESRI) ArcView
geographic information systems software, and includes the following tools:
• Tools to explore data and related documentation allowing users to review data layers
and retrieve documentation while constructing land-use design scenarios. Users can
browse and select design backdrops such as land-use designations, economic
characteristics, transportation networks, or aerial photographs.
• Tools for scenario design allowing users to construct geographic designs displaying
familiar landmarks, aerial photographs, and proposed features. Users can assign land
use categories with a table of attributes (e.g., total area of building, number of
employees, and estimated energy consumption) for each category.
• Tools for scenario evaluation allowing users to define the scope of the scenario to be
considered for evaluation. For example, users can evaluate the entire design or
interactively select land-use categories, regions, or other boundaries within the design.
Smart Places then uses development features (e.g., energy or water) with evaluation
models to conduct a cause and effect analysis to determine the viability and impact of a
design alternative.
-112-
-------
• Tools for production of design scenarios and evaluation results allow the user to plot
data, export graphics, and store data layers for scenario design and analysis reports.
Results can be printed using distinctive graphical layouts including data layers, images,
symbols, and explanatory text.
REQUIRED RESOURCES
Purchase Costs
Contact Paul Radcliffe (EPRI) for cost information.
Equipment Needs
Smart Places requires a 120 MHz or faster PC with a Pentium chip, Windows 95/98 or higher
operating system, a minimum of 32 MB of RAM, 1 GB free hard disk space, a CD-ROM drive,
and ERSI ArcView Geographic Information System software. A color monitor is recommended
but not required.
Staff Requirements and Expertise
Use of the Smart Places model requires no land use experience and only general computer
experience. Making significant changes in customizing the model, however, may prove difficult
for an inexperienced user.
Smart Places has scaled menus that accommodate four expertise levels:
(1) New user - people using the program for the first time, or with minimal experience with
the program.
(2) Experienced user- people with considerable experience in using the program.
(3) ArcView user - "experienced users" that also have considerable experience using
ArcView.
(4) Manager - developers or other technical people with the capability to make changes in
the program code.
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
Because Smart Places is customized to accommodate the user's own data, the number and
type of land-use categories will vary from application to application. Smart Places users define
their own land use categories as appropriate for their study area and there is no limit the
number of land use categories. Smart Places can accommodate one or more land use
categories (e.g., residential, commercial, etc.) as themes associated with a particular project.
-113-
-------
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
•
•
•
•
No?
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
•
•
•
No?
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees, taxes,
and incentives)
Yes?
•1
•1
•1
•1
No?
1 Smart Places can be customized to evaluate the impact of
changes in land use patterns based on user-specified criteria.
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Environmental Quality
School Quality
Crime
Other Quality-of-Life Conditions
Yes?
•1
•1
•1
•1
•1
•1
•1
No?
1 Smart Places can be customized to evaluate these impacts
based on user-specified criteria.
-114-
-------
Outputs Provided
Output
Evaluation results comparing scenario results with
assigned goals and boundaries for each evaluation
parameter
ESRI "shape files"
Format
Report
Displayed as bit-mapped images
and graphical displays of analysis
results
INFORMATION NEEDED TO RUN THE MODEL
Inputs will depend on the goals and objectives of the user. Typically, the more detailed inputs
the user can provide, the more in-depth analyses can be performed. Depending on the user's
objectives, likely data inputs include:
• Standard ArcView coverages with information on: 1) natural features; 2) infrastructure
plans; 3) existing land use patterns; and 4) approved comprehensive plans or zoning
ordinances (ArcView files and other data files). Data may include attributes related to
features stored in a variety of file formats.
• User generated analysis models (models can be written in several programming
languages, including Avenue, C++, and Visual Basic) used to evaluate alternative
designs.
MODEL STRENGTHS AND LIMITATIONS
Strengths
• Easy to use: The Smart Places graphical user interface supports four distinct user
levels, from new users to programmer/developer. Users can prepare and evaluate
alternative land use scenarios using simple, narrative, pull-down menus.
• Customizable: Smart Places is intended to be customized by the user, allowing the
system to model specific locations and design objectives.
• System integration: Smart Places provides the user the ability to link to models written in
Avenue or other programming languages, including C++ and Visual Basic.
Limitations
• Not a self-contained system: Smart Places is an extension of ESRI's ArcView GIS.
Users must have ArcView to use Smart Places.
-115-
-------
• Lack of sophisticated modeling: Smart Places provides an open system that allows
users to evaluate the impacts of alternative development scenarios. The users must
provide data inputs and evaluation models.
• Expertise: Some background in GIS is desirable for all user levels beyond "new users."
LEARNING MORE
Additional References
DOE Internet site for Center of Excellence for Sustainable Development.
http://www.sustainable.doe.gov/
Electric Power Research Institute, enPhysic Denver Inc. November 1998. Smart Places E
Series User Manual, Version E4E.
Public Technology Inc.. September 1997. Denver Smart Places Sustainable Development
Project (Case study report)
Smart Places Strategic Decision Support Software. A paper presented at the EPA Brownfields
Conference, 1998.
Availability of Preview Copies of the Model
Potential users must contact Paul Radcliffe, EPRI. A partnership is being formed with Public
Technology, Inc. to bring the model to cities in the United States.
Case Studies
• Preston White (tested model)
University of Virginia
Charlottesville, VA 22906
Phone: (804) 982-4561
Fax: (804) 982-2678
E-mail: kpw8h@virginia.edu
• Matthew Foster (used model in operational context)
Pueblo of Sandia
P.O. Box 6008
Bernalillo, NM 87004
Phone: (505) 867-3317
Application Sites
Denver, CO
38 other sites have license agreements to use
-116-
-------
TRANUS
MODEL DEVELOPER(S):
MAILING ADDRESS:
CONTA CT INFORM A TION:
WEB SITE:
DOCUMENTATION:
Modelistica
P.O. Box47709
Caracas 1041A
Venezuela
Phone: +58(2)761-5432
Fax: +58(2)761-7354
E-mail: info@modelistica.com
http://www.modelistica.com/tranus/
The most up to date documentation may be obtained by
contacting the developer by e-mail at info@modelistica.com.
The web page contains a large amount of material that
complements the basic documentation.
OVERVIEW
The Tranus Integrated Land Use and Transport Planning System (TRANUS) model is intended
to assist transportation and land use planners in simulating and evaluating transportation,
economics and/or environmental policies at an urban, regional or national scale. The integrated
land use-transport modeling makes possible an assessment of the implications of transport
policies on the location and interaction of activities, and its effects on the land market.
Compared to transport-only models, it provides projections to the future, based on the growth
and location of activities instead of increasing trip matrixes by applying growth factors not
directly related to changes in accessibility or functionality of the study area.
TRANUS reproduces the behavior of the different agents in the urban or regional area; the way
they relate and interact, their consumption patterns, and the consequent use of the supply of
land, floorspace, transport services and transport infrastructure. The model also allows for the
definition of any number of 'scenarios' with corresponding policies and projects to simulate. A
base case scenario is used to compare results and obtain the probable effects of applying
particular policies and projects. The model calculates many indicators to evaluate these
effects, from social, economic, to financial and energy points of view.
-117-
-------
REQUIRED RESOURCES
Purchase Costs
Purchase the model software: $7,500
Run the model software: Included
Maintain the model software: One-year guarantee included; one-year extensions may be
acquired at 50% of cost which means the that user group
receives a new versions with a software guarantee and
technical support.
Train to use the model A typical 2-week, full-time course is in the order of $8,000 plus
software: expenses
A common scheme that has been adopted in the past is that the agency interested in applying
the model provides the data and Modelistica performs the calibration of the model and even
projections into the future based on alternative scenarios. If training is added, the agency can
continue with the work on its own. This kind of turn-key operations may be arranged at very
competitive rates.
Equipment Needs
TRANUS requires a PC with enough RAM to run a Windows 95/98/NT operating system (64
MB or higher advisable), and free hard drive space of 10 MB for programs plus 20 MB or more,
depending on the size and complexity of the model implementation. A color monitor is required
and a color printer is useful, as is a digitizer but only if a network has to be digitized. In terms of
software, Windows-based spreadsheets word processors and presentation programs; statistical
analysis software (e.g., SAS, SPSS); a geographic information system (e.g., ArcView, Arclnfo).;
and a logit calibration program (e.g., Alogit or Hielow) are useful but not essential.
Staff Requirements and Expertise
Installation and calibration of the model requires experience with Office suites and GIS, as well
as land-use modeling expertise. Use of the model requires land-use expertise and familiarity
with Office suites and GIS.
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
There are no fixed categories to TRANUS. Any number of land use categories may be defined,
limited only by available information and the objectives of the project. The model is flexible to
be set with the categories normally used in the planning offices or master plans of the study
area. They are introduced into TRANUS with their familiar names and measured in any
convenient units. The modeler also defines the functional relationships between the categories.
Consequently, the potential user does not have to look anywhere for the definitions and
descriptions, since he or she would have defined them. What the model does with the
categories, relations and data is fully described in the documentation.
TRANUS is an Activity Location, Land-use, and Transport model. Not only land-use categories
need to be defined, but also activities like population (by income groups or household size)
employment (by type), education (employment and students) and any other relevant activity in
the study area. Relationships like which activities consume which types of land may be
specified. Restrictions on land consumption or on activities may be specified to reflect land use
-118-
-------
policies. Attracting variables may be added to represent environmental quality, historical values,
social aspects, crime rates, or any other.
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
•
•
•
•
No?
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
•
•
•
No?
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees, taxes,
and incentives)
Yes?
•
•
•
•
No?
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Environmental Quality
School Quality
Crime
Other Quality-of-Life Conditions
Yes?
•
•
•
•1
•
•
•
No?
1 Addresses changes in criteria pollutants, changes in
greenhouse gases, energy consumption, and emissions from
buildings. In the case of environmental indicators, TRANUS
provides the inputs required for other models, such as CUFM,
Burden, etc.
-119-
-------
Outputs Provided
TRANUS produces many more outputs than the sample included below. Many matrices are
provided: trips by category, mode, or operator. Matrices of costs, distances, time, and
disutilities. Matrices of trips that use selected links, selected assignment results by cordon, link
type, V/C relation and so on.
All transport outputs are graphically represented in vector format by the graphic user interface
of TRANUS. Any Windows program can process the graphic outputs of TRANUS. Graphics
may be imported to word processing or presentation programs (Word, Excel, PowerPoint, etc.)
through cut-and-paste.
Output
Format
All paths between each O-D pair for each
transport mode and combination of modes
through transfers.
Vector images produced by the Graphic
User Interface, which can be customized
in color, scale, labels and many other
features and pasted to any Windows
application: word processors,
spreadsheets, and other graphic
programs for further customizing and
printing. Optional ASCII file produced
with the sequence of modes and links in
each path.
General assignment results for each link: total
volume in demand units (passengers or Tons),
vehicles, equivalent vehicles, V/C ratio and level
of service.
Numeric and graphic outputs available
with the graphic TRANUS User Shell
system in the windows environment, and
tab delimited files generated for
spreadsheets and database programs.
Detailed assignment results for each link:
volume by each mode and route, speed and
waiting times in the link under congestion
conditions, demand/capacity ratio for transit
vehicles.
Numeric and graphic outputs available
with the graphic TRANUS User Shell
system in the windows environment, and
tab delimited files generated for
spreadsheets and database programs.
Many indicators about the performance of the
transport system to feed evaluation processes,
organized by the three agents: users, operators
and administrators. For users, TRANUS reports
global demand by mode and transport category,
total and average travels time, distance, cost
and disutility by transport category. For
operators, TRANUS reports the number of
passengers (or Tons) passenger-Km (or Ton-
Km) their income (from tariffs), their operating
costs and revenue in the simulation period. For
administrators of the transport infrastructure,
TRANUS reports maintaining costs and incomes
for tolls, road pricing or any other charges.
Tab delimited files generated for
spreadsheets and database programs.
-120-
-------
Output
A database file, with a database format,
containing all results of the transport model. The
planner can process the file at will with any
database program to extract specific information
that is not standard output of IRAN US.
Routes profile, with demand-supply information
for each route in each link: number of
passengers boarding the route in the link,
waiting time for boarding passengers, demand/
capacity of transit vehicles in each link, available
seats and so on.
Activity location and land use consumption
outputs.
Format
Tab delimited files for spreadsheets and
database programs.
Tab delimited files for spreadsheets and
database programs.
Tab delimited files for spreadsheets and
database programs.
INFORMATION NEEDED TO RUN THE MODEL
This model requires the following user input:
• Network nodes, links and routes (comma delimited text files generated by any GIS
program can be imported to TRANUS). Alternatively, small networks can be
interactively created in the graphic window of TRANUS with the mouse.
• Transport variables (introduced in the database with the Windows-like menus and
commands provided by the TRANUS User Shell).
• Activity location and land use data by zone (copied from any worksheet or GIS and
pasted to the TRANUS database).
• Activity location and land use variables (introduced in the database with the Windows-
like menus and commands provided by the TRANUS User Shell or copy-and-paste from
spreadsheet).
MODEL STRENGTHS AND LIMITATIONS
Strengths
• TRANUS is one of the very few integrated land use and transport models commercially
available, backed by a sound history of practical applications in many countries.
• TRANUS is extremely user-friendly, with a powerful graphical Windows-based interface.
The interface is supported by a dynamic object-oriented database, with GIS interface
capabilities. From the beginning, the model was developed for use on personal
computers, and it pioneered the area.
-121-
-------
TRANUS' 'site-license' allows installing the software in any computer of the institution,
which makes it easy for work groups and to use the model for teaching.
TRANUS has extensive email support that is provided with the license.
The model can be applied to a large variety of case studies, ranging from very simple
urban or regional models to highly sophisticated national or regional input-output
models.
TRANUS is backed by a continuous research and development process. Every year a
new main version of TRANUS is released, with many upgrades to the main version
produced monthly, based on the experience of applications and suggestions or requests
of users, including Modelistica. An ambitious 'Millennium Edition' is soon to be released.
Limitations
• TRANUS is not a traffic model yet. The model was designed for planning, for strategic
decisions about main policies: railroads, metro systems, new highways, master plans,
input-output relations and so on. After many years of development and applications,
Modelistica expanded the model to cope with detailed urban applications, including a
sophisticated representation of transit systems (routes, frequencies, transfers, and
queuing theory to calculate waiting times) with capacity restriction and dynamic
assignment. However TRANUS does not cover all aspects of traffic models, such as
calculating signal times or intersection movements.
• The graphic interface of TRANUS provides full graphic information of transport data and
simulation results. However, for the graphical display of activities and land use data and
results, a GIS software such as ArcView, TransCAD or Maplnfo to map the results of
the model is required.
• TRANUS includes a sophisticated model of the supply-side of the estate market
including redevelopment. However, this new feature requires further testing. An
agreement between Modelistica and the University of California is underway t o advance
this area.
LEARNING MORE
Additional References
De la Barra. 1989. Integrated Land Use and Transport Modeling. Cambridge University Press.
(Available in most bookstores and Internet facilities such as www.amazon.com, Barnes &
Noble, Heffers, etc.)
Modelistica. Mathematical and Algorithmic Structure of TRANUS (see
www.modelistica.com/tranus/tranus2.htm)
-122-
-------
Modelistica. 1993. Multidimensional Path Search and Assignment. Paper presented at the 21st
PTRC Summer Annual Conference, Manchester, England.
Modelistica. 1994. Improved Logit Formulations for Integrated Land Use, Transport, and
Environmental Models. Paper presented at the Royal Institute of Technology, Stockholm,
Sweden.
Modelistica. 1996. Dual Graph Representation of Transport Networks. Transportation
Research, Vol. 30 No. 3, pp. 209-216, Exeter, England.
Availability of Preview Copies of the Model
Potential users must contact the model developer.
Case Studies
• Professor Robert Johnston (tested model)
University of California Davis
1 Shields Avenue
Davis, CA 95616-5200
Phone: (916) 752-6580
Fax: (916) 752-3350
Email: rajohnston@ucdavis.edu
• Patrick Costinett (used model in an operational context)
Parsons Brinckerhoff
8011 NE 121st Street
Kirkland, WA 98034-5833
Phone: (206) 557-8588
Fax: (425) 820-9358
Email: pcostinett@worldnet.att.net
Application Sites (this is only a sample)
• Year: 1999
Project: Detailed Transport Demand Study for the Bogota Metro Systems,
Columbia
Consultant: Cal & Mayor Consultants, Bogota, Columbia
Client: Urban development Institute of Santa Fe, Bogota, Columbia
• Year: 1998-99
Project: Land Use and Transportation Model for the city of Valencia, Spain.
Consultant: TEMA Consultants, Spain
Client: Local government of Valencia, Spain (Ayuntamiento)
-123-
-------
Year: 1998-99
Project: Land Use and Transportation Model for the Baltimore Metropolitan Areas,
USA
Consultant: Modelistica
Client: Baltimore Metropolitan Council, USA
Year: 1997-98
Project: An Input-Output and Transport Model for the State of Oregon
Consultant: Parsons Brinckerhoff, USA
Client: Oregon Department of Transportation
-124-
-------
UGrow
MODEL DEVELOPER(S): Wilson W. Orr
MAILING ADDRESS: Prescott College
220 Grove Avenue
Prescott, AZ 86301-2990
CONTACT INFORMATION: Phone: (520)717-6070
Fax: (520) 717-6073
E-mail: worr@prescott.edu
WEB SITE: http://www.prescott.edu
DOCUMENTATION: UGrow an Urban Growth Model, Prescott College, Sustainability
and Global Change Program, Prescott, AZ 8630-2990,
(520) 717-6070. The model is documented within the model
diagrams and equations itself.
OVERVIEW
UGrow is a system dynamics suite of models for urban policy design and testing. Numeric
(system dynamics), spatial (GIS - maps) and 3-Dimensional (fly through visualization) are tools
which are integrated to serve a community's needs. UGrow is part of an overall process of
working with community leaders to identify drivers of change in the region, adapting the core
UGrow model to address those drivers, and then testing a variety of future scenarios based on
changes in local development policy, input conditions, or external variables.
UGrow is PC-based running over 300 equations, which define the basic interdisciplinary
relationships among the economic, social, and environmental sectors of a community. The
model runs from 1950 to 2100 with pauses at years 1990 and 2030 for policy adjustments. It is
designed to test proposed policies and can be stopped at any year to produce the community
status as a scenario responding to the proposed policy(s). There are presently ten policy option
categories which encourage/discourage efficiencies in, for instance: housing density, energy
consumption, transportation, land use/land cover, and business activity. Each of these may be
adjusted for "intensity," representing the strength with which the policy is implemented. From
the inputs and various policy options, the model produces a variety of future scenarios and
projects variables groups into sectors such as: Quality of Life, Economics and Business,
Housing, Population, Land Use, Transportation, Climate Change Impacts, and Energy. The
numerical output is then used to generate GIS-format maps of the "future communities."
REQUIRED RESOURCES
Purchase Costs
Model software is free. However, for the model(s) to be useful to a community, the developer
must adapt the model(s) to the community's particular data. This costs anywhere from $30,000
-125-
-------
for a quick visual (spacial model) to $200,000 to include a more sophisticated numeric model
and multiple facilitated public meetings.
Equipment Needs
UGrow requires an IBM PC with Windows 95, Arclnfo, ArcView, and Powersim modeling
software. Minimal speed, RAM, and hard drive space is necessary.
Staff Requirements and Expertise
The installation and calibration of the UGrow model requires land use modeling expertise, as
well as system dynamics, GIS, and facilitation experience. The model is to be used by trained
system dynamics and GIS modelers. At the end of the process, a "flight simulator" can be
created that would require no modeling experience and only general computer experience.
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
The UGrow model, at present, addresses only four land-use categories; however, the model is
quite adaptable based on the community's modeling needs.
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
•
•
•
•
No?
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
•
•
•
No?
Questions Answered
(1) The model is not capable of addressing the effects on land-use patterns from changes in
any of the following community actions. The general core of the model focuses on land use
changes.
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees, taxes,
and incentives)
Yes?
No?
•
•
•
•
-126-
-------
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Environmental Quality
School Quality
Crime
Other Quality-of-Life Conditions
Yes?
•
•
•1
No?
•
•
•
•
1 Only includes effects of changes on greenhouse gases.
Outputs Provided
Output
Data tables within the modeling software
Graphs within the modeling software
Format
Powersim
Powersim
INFORMATION NEEDED TO RUN THE MODEL
The UGrow model requires the following user input:
• Spatial
• Social
• Housing
• Economic
• Vehicle Use
• Environment
• Public
MODEL STRENGTHS AND LIMITATIONS
Strengths
• A whole system approach
• Highly adaptable to specific concerns
• Model emphasizes building understanding over generating answers
-127-
-------
Limitations
• Requires year-long adaptation with consulting team
• Extensive model advisory committee input needed
• Not a projection tool
LEARNING MORE
Additional References
Not provided.
Availability of Preview Copies of the Model
Potential users must contact model developer.
Case Studies
• Wilson W. Orr (tested model and used model in operational context)
Prescott College
220 Grove Avenue
Prescott, AZ 86301-2290
Phone: (520)717-6070
Fax: (520)717-6073
E-mail: worr@prescott.edu
Application Sites
Eastern Pima County, AZ (Tucson area)
Gallatin County, MT
Santa Barbara, CA
• Yavapai County, AZ
Oahu and Maui, HI
• Green Bay, Wl
Cayuga County, NY (on the waiting list)
-128-
-------
UPLAN
MODEL DEVELOPER(S):
MAILING ADDRESS:
CONTA CT IN FORM A TION:
WEB SITE:
DOCUMENTATION:
Developed by Robert Johnston at University of California, Davis;
built by David Shabazian
UPLAN; Robert Johnston
Dept. of Environmental Science and Policy
University of California Davis
1 Shields Avenue
Davis, CA 95616-5200
Phone: 530-582-0700
Fax: 530-582-0707
E-mail: rajohnston@ucdavis.edu
Will be on the Information Center for the Environment's web site
at http:// ice.ucdavis.edu in the near future.
Contact the model developer.
OVERVIEW
The UPLAN Urban Growth Model ("UPLAN") provides a land use evaluation and change
analysis based on general land-use plans, population and employment projections,
characteristics of housing, and other user-defined conditions. It is an integrated package
of user-specified attractions that enable users to: 1) conduct a land suitability analysis, and
2) project future land use demand. UPLAN helps communities create alternative visions for
their area's future by mapping alternative development patterns determined by local land
development policies. Some of the policies and decisions UPLAN addresses include
establishing various criteria to "weight" the suitability of different locations for a particular land
use, incorporating various land use planning and zoning considerations and other allocation
scenarios, and defining various growth scenarios. The model can also be used to determine
various environmental and social constraints to growth by modifying the criteria and the
associated weights.
The UPLAN model allows the user to develop specific parameters in the form of grids in which
to model future land uses. The model allows the user to generate attraction grids, exclusion
grids, general plan grids, and existing urban grids. Attractive grids are locations for future
development (i.e., near to freeway ramps); exclusion grids, list areas where development
should not occur (i.e. parks, waterways etc.); general plan grid is a composite grid of the
general plan land use maps from the users region; and existing urban grid provides the current
land use conditions. Each grid applies user-defined decision criteria (e.g., identifying and
weighting grid factors), to derive study-area conditions. These decision criteria are applied to
land use information stored in geographic information system (GIS) data files to create maps
and reports showing where future development may occur.
-129-
-------
REQUIRED RESOURCES
Purchase Costs
The UPLAN software is available for free from the model developer.
Equipment Needs
UPLAN requires a 300 MHz or higher IBM PC (or other computer type that can run Windows
95/98 or NT 4.0), 32 MB of RAM, several hundred MBs (a few GBs even better) of hard drive
space, and a color monitor (recommend at least 21" monitor). Color printer and plotter
recommended, but not required. Software requirements include PC ArcView (need PC Arclnfo
if wish to prepare data layers for local applications) and Excel or SAS for data exchange with
other models.
Staff Requirements and Expertise
Use of the model requires land-use expertise and basic knowledge of ArcView. Also, users will
need to be able to program in Avenue to adapt the model to their data sets and policy needs. A
three-county application in New Mexico by inexperienced users required about 3 person months
of ArcView user, plus about 2 weeks of consulting by an Avenue programmer.
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
UPLAN can accommodate up to six different types of land uses, as is, (industry; high- and low-
density commercial; and high-, medium-, and low-density residential) and can be modified
to accommodate any number of categories. Because UPLAN can be customized to
accommodate the user's own data, the number and type of land-use subcategories will vary
from application to application.
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
•
•
•
•
No?
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
•
•
•
No?
-130-
-------
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action)
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees, taxes,
and incentives)
Yes?
•
•
•
•1
No?
1 Can be done only for developer impact fees, property taxes,
municipal sewer and water fees, subsidies, and parking fees.
UPLAN is policy-oriented, scenario-based land use change projective tool. It shows the
likely impacts of different user-supplied assumptions concerning land suitability, land use
demands. Market mechanisms, such as subsidies and taxes can be coded in as subareas
with positive or negative values.
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Environmental Quality
School Quality
Crime
Other Quality-of-Life Conditions
Yes?
•1
•
•
•2
No?
•
•
•
1 Not directly, but the land uses can be taken into the zone files
of any travel model.
2 In terms of air quality, changes in criteria pollutants and
greenhouse gases can not be estimated directly but can be
done by feeding the land uses into any travel model. In terms
of water quality, a model is under development for nutrient
loading or sedimentation in surface waters and other nonpoint
source water pollution.
-131-
-------
Outputs Provided
Output
Grid Maps
Analysis Report
Assumptions Report
Image Files
Format
ArcView layouts (GIS)
Report
Report
ArcView layouts (GIS)
INFORMATION NEEDED TO RUN THE MODEL
The inputs listed below are desired, but not required for UPLAN since the model can be run
with default values. The more detailed specific inputs the user provides, the more accurate the
analysis. The desired inputs include:
• Demographic and land use factors, population projections, persons per households,
assumed housing densities per land use, average parcel size for each density class,
employment by type, assumed employment density. (Hand entered by system user).
• Regional General Plans, and incorporated city areas provided by users Local Planning
Organization. (Hand entered by system user).
• All roadway and intersection data obtained from U.S. Geological Survey (USGS) digital
line graphs. Data must be entered for each development scenario.
• Major waterways, lakes, and rivers data obtained from U.S. Geological Survey
hydrography digital line graphs. (Hand entered by system user).
• Major infrastructure locations, i.e., airports. (Hand entered by system user digitized
data)
• Existing urban lands for the base year of 1990, general land use plans, light rail station
locations, and slope data obtained from Digital Elevation Model USGS.
MODEL STRENGTHS AND LIMITATIONS
Strengths
• Easy to use: UPLAN allows users to prepare and evaluate alternative suitability, growth,
and allocation scenarios by specific prompts generated by the program.
• Customizable: UPLAN incorporates information provided by the users and applies its
decision-tools to currently available GIS and non-GIS data, allowing the system to be
customized to many different geographic areas and conditions.
-132-
-------
Integrated system: UPLAN provides an integrated software package that incorporates
user-provided GIS and other data as a foundation and applies various
evaluation/decision-tools (e.g., land use projection) to the underlying data. UPLAN uses
currently available GIS data to prepare maps and reports showing the outcomes of
alternative development scenarios on future land use patterns.
The default six land use types (industrial; commercial hi-density and low density; and
three residential densities) permit the evaluation of the impacts of the future land use
pattern on runoff, water pollution, habitats, and costs from flooding and wildfires. Data
grids can be a small as the data permit, generally 25 m grid cells.
Limitations
• Lack of sophisticated modeling: UPLAN provides a way for end users to visualize (as
maps) the impacts of alternative development scenarios on future land use patterns.
The users must provide existing urban land-use and digital land-use plans to the UPLAN
system as inputs, as well as other normally available data layers. UPLAN does not
provide the sophisticated modeling capability and/or theoretical basis to examine the
interrelated factors of fiscal policies, and other planning decisions on the amount and
type of future development and land use change that will occur. The attractiveness
criteria are pseudo-economic, in that they represent land value and accessibility.
LEARNING MORE
Additional References
Not provided.
Availability of Preview Copies of the Model
Available for free from the model developer.
Case Studies
A paper on the Sacramento, CA application is available from the model developer.
Application Sites
Sacramento, CA region
Espanola region of New Mexico
-133-
-------
UrbanSim
MODEL DEVELOPER(S): Paul Waddell: model design and project lead
Michael Noth, Alan Borning: software architecture
MAILING ADDRESS: Daniel J. Evans School of Public Affairs
University of Washington
P.O. Box 353055
Seattle, WA 98195-3055
CONTACT INFORMATION: Phone: 206-221-4161
Fax: 206-685-9044
E-mail: pwaddell@u.washington.edu
WEB SITE: http://www.urbansim.org/
DOCUMENTATION: Available via the web site.
OVERVIEW
UrbanSim is a software-based system designed to be used for integrated planning and analysis
of urban development, incorporating the interactions between land use, transportation, and
public policy. It is designed to interface to existing travel modeling procedures, including both
current four-step as well as newer activity-based travel models. It is currently being extended to
address environmental impacts of development by simulating land cover, water demand and
nutrient emissions. Specifically, the model:
• Simulates the key decision makers and choices impacting urban development; in
particular, the mobility and location choices of households and businesses, and the
development choices of developers;
• Explicitly accounts for land, structures (houses and commercial buildings), and
occupants (households and businesses);
• Simulates urban development as a dynamic process over time and space, as opposed
to a cross-sectional or equilibrium approach;
• Simulates the land market as the interaction of demand (locational preferences of
businesses and households) and supply (existing vacant space, new construction, and
redevelopment), with prices adjusting in response to short-term imbalances between
supply and demand (vacancy rates);
• Incorporates governmental policy assumptions explicitly, and evaluates policy impacts
by modeling market response;
-134-
-------
• Is designed for high levels of spatial and activity disaggregation, currently using a 150
meter grid; and
• Addresses both greenfield development and redevelopment or intensification.
REQUIRED RESOURCES
Purchase Costs
None - may be downloaded for free at http://www.urbansim.org/.
Equipment Needs
UrbanSim requires a 333 MHz or higher computer with 128 MB of RAM, 2+ GBs of free hard
drive space (can be less depending on study area grid resolution). It runs on Windows 95/98,
Windows NT 4.0/2000, Linux or UNIX, using Java JKD 1.3.
Staff Requirements and Expertise
Calibration of the model requires knowledge of statistical software to perform multiple
regression and logit model estimation using external econometric software such as Alogit or
Limdep. Further work on calibration tools may make the use of external software unnecessary
in the future. Use of the model requires land-use and transportation planning expertise and
general computer experience. The user interface for the model is intended for relatively non-
technical users.
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
The land uses are user defined with typically 10 or more urban categories, but there is no
internal limit on the number of urban or non-urban categories.
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
•
•
•
•
No?
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
•
•
•
No?
-135-
-------
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees, taxes,
and incentives)
Yes?
•
•
•
,,1,2
No?
1 Road tolls, parking fees, fuel and sales taxes, vehicle miles
traveled, registration fees included through interaction with
travel models.
2 Includes fiscal policies such as Urban Growth Boundaries and
regulations on development of environmentally sensitive lands.
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Environmental Quality
School Quality
Crime
Other Quality-of-Life Conditions
Yes?
•1
•2
•3
•4
No?
•
•
1 Included through interaction with travel models.
2 The model provides the necessary outputs to estimate these
impacts, but does not currently estimate them.
3 Model components are currently being implemented to
simulate land cover change, nutrient loading and water
demand.
4 The model can represent density, accessibility,
socioeconomic mix, and land use mix. Any variable that can be
measured and simulated for future years can be incorporated
into the bid functions for housing and location choice. (Crime
and school quality have not yet been included due to lack of
model specifications to predict these values into the future, but
could in principle be added based on relationships to other
endogenous variables.)
-136-
-------
Outputs Provided
All outputs are currently by zone, but can be made available at parcel or grid cell level.
Output
Households by type (income, size, age of head, children, workers)
Businesses and employment by type (sector)
Acres by land use (real estate development type)
Housing units and building square footage by type
Prices of land, housing and commercial space by type
Development projects simulated; new and redevelopment; conversion
of land by type
Format
ASCII
ASCII
ASCII
ASCII
ASCII
ASCII
INFORMATION NEEDED TO RUN THE MODEL
Input
Parcels
Business establishments
Household data from Census (STF3A and PUMS)
Environmentally sensitive layers; e.g., wetlands, floodplains, high
slopes, fault zones; and Urban Growth Boundaries or other policy
boundaries
Zones used in travel modeling
Travel impedance from travel models (peak times and logsums)
Format
ArcView shape file
and attribute table
ASCII
ASCII
ArcView
ArcView
ASCII
MODEL STRENGTHS AND LIMITATIONS
Strengths
• Dynamic behavioral foundation is used that makes the model more transparent and
explainable to users and decision-makers; reflects real-world processes that make the
model easier to evolve and to interface to other process models such as environmental
models.
• High degree of spatial resolution: Currently uses spatial grid of 150 meters for interface
with environmental models such as land cover.
• Model and source code are entirely open source: They are freely available for use and
modification, and can be downloaded from the web site. This is intended to facilitate
collaborative use and further development.
-137-
-------
A visualization component has been designed into the model architecture, and is now
operational. This provides integrated 2 and 3-dimensional mapping, in addition to charts
and graphs for interpreting and comparing model results, and for diagnosis during
model development and testing.
Limitations
• The model currently has high data requirements; data mining and synthetic data
cleaning tools are currently being designed to facilitate working with messy data.
• The model has been recently developed, so experience is limited to current applications
in Hawaii, Oregon, Utah and Washington.
• The model is being rapidly evolved, with the first major release based on a complete
redesign of the software architecture in the second quarter, 2000.
LEARNING MORE
Additional References
Alberti, M. and P. Waddell. (forthcoming). An Integrated Urban Development and Ecological
Simulation Model. Integrated Assessments.
L. Denise Pinnel, Matthew Dockrey, A.J. Bernheim Brush, and Alan Borning. (Forthcoming).
Design of Visualizations for Urban Modeling. Proceedings of VISSYM '00 - Joint
Eurographics - IEEE TCVG Symposium on Visualization, Amsterdam, May 2000.
Waddell, P. 2000. A behavioral simulation model for metropolitan policy analysis and planning:
residential location and housing market components of UrbanSim. Environment and
Planning B: Planning and Design 2000, volume 27(2): 247-263.
Waddell, P. (forthcoming). Between Politics and Planning: UrbanSim as a Decision Support
System for Metropolitan Planning. Building Urban Planning Support Systems: Combining
Information, Models, Visualization, Klosterman, R. and R. Brail eds. Center for Urban
Policy Research.
Waddell, P. (forthcoming). Monitoring and Simulating Land Capacity at the Parcel Level.
Monitoring Land Supply with Geographic Information Systems: Theory, Practice and
Parcel-Based Approaches, Vernez-Moudon, A. and M. Hubner, eds., John Wiley & Sons,
Inc.: New York.
Wegener, M., P. Waddell and I. Salomon, (forthcoming). Sustainable Lifestyles?
Microsimulation of Household Formation, Housing Choice and Travel Behavior.
Proceedings of the National Science Foundation-European Science Foundation Conference
on Social Change and Sustainable Transportation, Berkeley, California.
-138-
-------
Availability of Preview Copies of the Model
The UrbanSim model and user's manual are available at http://www.urbansim.org/.
Case Studies
• Bud Reiff
Lane Council of Governments
99 E. Broadway, Suite 400
Eugene, OR 97401-3111
Phone: 541-682-4283
Peter Donner
Governor's Office
State of Utah
116 State Capitol
Salt Lake City, UT 84114
Phone: 801-538-1027
Application Sites
Honolulu, HI
Eugene-Springfield, OR
Greater Wasatch Front area (Salt Lake City), UT
• Puget Sound (Seattle, WA)
-139-
-------
What if?
MODEL DEVELOPER(S): Developed by Dr. Richard E. Klosterman (as Community
Analysis and Planning Systems, Inc.)
MAILING ADDRESS: Community Analysis and Planning Systems, Inc. (CAPS, Inc.)
78 Hickory Lane
Hudson, OH 44236-2707
CONTACT INFORMATION: Phone: 330-650-9087
Fax: 330-650-9087
E-mail: lnfo@What-if.com
WEB SITE: www.What-if.com
DOCUMENTATION: Available in hard copy from model developer.
OVERVIEW
What if? was developed in 1997 to support communities in many aspects of the land-use
planning process. What if? provides an integrated package of modules that enable users to:
1) conduct a land suitability analysis (Suitability Module), 2) project future land-use demand
(Growth Module), and 3) allocate projected demand to the most suitable location (Allocation
Module). What if? helps communities create alternative visions for their area's future by
mapping alternative development patterns determined by local land development policies.
Some of the policies and decisions What if? address include establishing various criteria to
weigh the suitability of different locations for a particular land use, incorporating various land-
use planning and zoning considerations and other allocation scenarios, and defining various
growth scenarios.
REQUIRED RESOURCES
Purchase Costs
What if? is offered in two different contexts: 1) to provide support to communities, and 2) to
provide a teaching tool in academic settings. The professional price is $2,495 for a single user;
site licenses are available. The academic price of What if? is $250.00 for a single user; site
licenses are available. A free demonstration CD for a real community can be obtained from
CAPS, Inc..
Equipment Needs
What if? requires a 300 MHz Processor Intel Pentium II or above, 64 MB of RAM, 1 GB of free
hard-disk space, a CD-ROM drive, a monitor that is SVGA compatible or better, and an MS
Windows 95, 98, or NT 4.0 operating system, but no additional software. What if? is a fully
self-contained software package.
-140-
-------
Staff Requirements and Expertise
Use of What if? requires at the minimum an ability to work with ArcView and similar packages
and a familiarity with local and use planning principles and procedures.
INFORMATION PROVIDED BY THE MODEL
Land Uses Addressed
What if? is capable of including any kinds of land-use data available from the community.
Therefore, the actual land-use categories are defined by the community and can be as detailed
or general as needed.
Urban Land-Use
Categories
Residential
Commercial
Mixed- Use
Industrial
Other
Yes?
•
•
•
•
•
No?
Nonurban Land-Use
Categories
Agricultural
Forest
Wetlands
Water
Preservation
Park Land
Yes?
•
•
•
•
•
•
No?
Questions Answered
(1) The model is capable of addressing the effects on land-use patterns from changes in the
following community actions:
Community Action
Transportation Infrastructure
Local Zoning
City/County Master Plans
Local Fiscal Policies (e.g., fees,
taxes, and incentives)
Yes?
•
•
•
No?
•
(2) The model is capable of addressing the effects of changing land-use patterns on the
following community characteristics:
Community Characteristic
Travel Demand
Local Government Fiscal Conditions
Availability of Open Space
Environmental Quality
Yes?
•
No?
•
•
•
-141-
-------
Community Characteristic
School Quality
Crime
Other Quality-of-Life Conditions
Yes?
No?
•
•
•
Outputs Provided
Output
Suitability Analysis Maps
Suitability Analysis Report
Suitability Analysis Assumptions Report
Growth Analysis Results Report
Growth Analysis Assumptions Report
Allocation Map
Allocation Analysis Results Report
Allocation Analysis Assumptions Report
Format
GIS maps
Report
Report
Report
Report
GIS map
Report
Report
INFORMATION NEEDED TO RUN THE MODEL
The following inputs are desired, but not required to use, for What if?. The more detailed inputs
the user can provide, the more robust analyses can be performed.
Input
Standard GIS coverages including information on
1) natural features, 2) infrastructure plans, 3) existing
land-use patterns, and 4) approved comprehensive
plans or zoning ordinances. These layers are
combined to create homogeneous land units (i.e.,
uniform analysis zones or UAZs).
Growth projections for number of households, assumed
vacancy and loss rates, assumed housing densities per
land use, employment by type, assumed employment
density
Alternative development scenarios that are pre-defined
by the community using What if?
Land-use classifications that are pre-defined by the
community using What if?
Infrastructure plans that are pre-defined by the
community using What if?
Format
ArcView, Arclnfo, and any other
program that can generate files in
ESRI shapefile format
Hand entered by system user
Hand entered by system user
Hand entered by system user
Hand entered by system user
-142-
-------
MODEL STRENGTHS AND LIMITATIONS
Strengths
• Easy to use: What if? allows users to prepare and evaluate suitability, growth, and
allocation scenarios by using only Windows standard buttons, check boxes, and text
boxes.
• Customizable: What if? incorporates information provided by the users and applies its
decision tools to currently available GIS and non-GIS data, allowing the system to be
customized to many different geographic areas and conditions.
• Integrated system: What if? provides an integrated software package that incorporates
user-provided GIS and other data as a foundation and applies various
evaluation/decision tools (e.g., suitability analysis and land-use projection and allocation)
to the underlying data. What if? uses currently available GIS data to prepare maps and
reports showing the outcomes of alternative development scenarios on future land-use
patterns.
• Self-contained system: What if? is self-contained and requires no additional GIS or
non-GIS software, although the user must be able to incorporate GIS layers (e.g., ESRI
coverages or shapefiles) as input to the system. If desired, What if? inputs and outputs
can be used with ArcView and any other package that works with ESRI shapefiles.
• Flexible data requirements: What if? is fairly easy to use with minimum data
requirements. Users need only to provide inputs for existing land-use data to provide
the basics for running What if? scenarios, although the system is able to accommodate
more detailed information and data layers the study area may have available.
Limitations
• Lack of sophisticated modeling: What if? provides a way for end users to visualize (as
maps) the impacts of alternative development scenarios on future land-use patterns.
The users must provide the scenarios to the What if? system as inputs. What is? does
not provide the sophisticated modeling capability and/or theoretical basis to examine the
interrelated factors of transportation infrastructure, fiscal policies, and other planning
decisions on the amount and type of future development and land-use changes that
occur.
• What if? does not include measures of spatial interaction.
• Does not employ random utility or discrete choice theory to explain and project the
behavior of various urban actors. Does not represent the interlinked markets for land,
housing, nonresidential uses, labor and infrastructure, or provide any procedures for
"market clearing" and price adjustment in the face of changes in demand and/or supply.
-143-
-------
• Does not explicitly model the behavior of actors such as households, businesses, and
developers.
LEARNING MORE
Additional References
Klosterman, Richard E. 1999. The What if? Collaborative Planning Support System.
Environment and Planning, B: Planning and Design. 26: 393-408.
What if? Evaluation Copy of the software.
Availability of Preview Copies of the Model
May be obtained from the model developer.
Case Studies
• Professor Richard K. Brail
Department of Urban Planning and Policy Development
Rutgers University
P.O. Box 5078
New Brunswick, NJ 08903-5078
Phone: (732) 932-2591, x731
E-mail: rbrail@rci.rutgers.edu
Ms. Jane Dembner, AICP, Principal
LDR International, Inc.
Quarry Park Place
9175 Guilford Road, Suite 100
Columbia, MD 21046-1861
Phone: (410)792-4360
E-mail: dembner@ldr-int.com
Application Sites
Hamilton County, OH
Medina County, OH
Summit County, OH
-144-
-------
Appendix A
Land Use Models: Comparative Matrices
-------
-------
Exhibit A-1. Skills/Technical Expertise Comparative Matrix
Model Name
CUF-1
CUF-2
CURBA
DELTA
DRAM/EMPAL
GSM
INDEX
IRPUD
LTM
LUCAS
Markov
MEPLAN
METROSIM
SAM-IM
SLEUTH
Smart Growth
INDEX
Smart Places
TRANUS
Target User Group
Nontechnical community planning
participants
Nontechnical community planning
participants
Land use planners, policy makers,
and environmentalists
Politicians, policy makers, planners
Regional transportation and land-
use planners
Land Resource Managers
Community planning participants
Regional transportation and land-
use planners, researchers
Watershed stakeholders (resource
managers, landowners, planners)
Land resource managers
Demographers, residential
planners, developers, policy
makers
Planners, transportation engineers,
economists
Planners, transportation engineers,
economists
Land-use planners and forecasters
Academic and government
researchers, planners
Community planning participants
Planners (land use, transportation,
environmental), community groups
Transportation and land use
planners and academics
Technical Expertise
for Usage
(1 [none] - 3
[extensive]) '
2
2
1
3
3
2
3
3
3
3
1
2
1
1 (but there is a
learning curve/training
required)
2
2
1
2
Consultant
Expertise
Required?
No
No
No
Yes
Yes, cost to
purchase
includes
consultant
services
No
Yes
Yes
Yes
Yes
No
No
Yes
Yes
No
No
No
No
Computer Skills
for Usage
(1 [general] - 3
[extensive]) 2
3
3
2
1
2
2
2
1
3
3
2
1
1
2
2
2
1
2
A-1
-------
Model Name
UGrow
UPLAN
UrbanSim
What if?
Target User Group
Academic and government
researchers, planners, policy
makers
Nontechnical community planning
participants
Planners (land use, transportation,
environmental), community groups
Nontechnical community planning
participants
Technical Expertise
for Usage
(1 [none] - 3
[extensive]) '
1
2
2
2
Consultant
Expertise
Required?
No
No
No
No
Computer Skills
for Usage
(1 [general] - 3
[extensive]) 2
1
2
1
1
1 (1) No experience required; (2) land use experience; (3) land use modeling experience
2 (1) General computer experience; (2) familiarity with specific software applications; (3) programming skills
A-2
-------
Exhibit A-2. Hardware Comparative Matrix
Model Name
CUF-1
CUF-2
CURBA
DELTA
DRAM/EMPAL
GSM
INDEX
IRPUD
LTM
LUCAS
Markov
MEPLAN
METROSIM
SAM-IM
SLEUTH
Smart Growth
INDEX
Smart Places
TRANUS
Type Computer
Required
workstation
Sun Spare or PC
PC
PC
PC
PC
PC
IBM Pentium III PC
Sun Spare or PC
UNIX-based
workstation(e.g., Sun
Spare station 10)
Any
PC
Any
PC
PC, workstation, or
mainframe
PC
Pentium PC
PC
CPU Required
(MHz)
Not specified
300
300+
Pentium 200
Pentium
500
200
300+
300
Not specified
Not specified
200+
300+
400
Not applicable
300
120
Not specified -- the
faster the better
Minimum Disk Space
Required/RAM
(MB)
Not specified
2 GB/32
1GB/32
Depends on model
dimensions
Not specified
Not specified/1 28+
150/64
4+GB/128+
Not specified/256
Not specified
Not specified
500/64
Not specified/1 28
2GB/128+
Not applicable
500/64
1 GB/32
30/64
Peripherals Needed?
Not specified
Color monitor
Color monitor
Color monitor
recommended
Color monitor and
color printer
Color plotter
Color monitor and
color printer
Color monitor with
minimum resolution of
1024x768 and
color printer
Color monitor with
minimum resolution of
1024x768 and
color printer
Color monitor and
color printer
Not specified
Color monitor
Color monitor
Color monitor and
color printer
None
Color monitor (1 024 x
768 min. resolution) and
color printer
CD-ROM drive required;
color monitor
recommended
Color monitor required;
colored printer and a
digitizer are useful
A-3
-------
Model Name
UGrow
UPLAN
UrbanSim
What if?
Type Computer
Required
PC
PC
Any
PC
CPU Required
(MHz)
Not specified - a
minimal amount is
necessary
300
333
300
Minimum Disk Space
Required/RAM
(MB)
Not specified - a minimal
amount is necessary
Several hundred/32
2GB/128
1 GB/64
Peripherals Needed?
None
Color monitor required;
21 " monitor, color
printer, and plotter are
recommended.
None
None
A-4
-------
Exhibit A-3. Software Comparative Matrix
Model Name
CUF-1
CUF-2
CURBA
DELTA
DRAM/EMPAL
GSM
INDEX
IRPUD
LTM
LUCAS
Markov
MEPLAN
METROSIM
Operating
System
UNIX
MS Windows
95, Sun Soloris
MS Windows
MS DOS
(either under
DOS mode or
Windows 95/
98)
MS Windows
95/98 or NT
MS Windows
NT or UNIX
MS Windows
95 or NT
MS Windows
NT
MS Windows
NT or Sun
Soloris
MS Windows
with OSF/Motif
toolkit
Any
MS Windows
NT
Any; UNIX
preferred but
not required
Program
Compiler
Needed?
(Y/N)
N
N
N
N
N
N
N
Y; FORTRAN,
C, C++
Y
Y
Not specified
N
Y; FORTRAN
orC
Data
Management
Tools
Not specified
Not specified
Not specified
Spreadsheets
and data bases
highly
recommended
Spreadsheets
and data bases
Paradox or
Oracle
None
None
Spreadsheets
and data bases
Spreadsheets
Recommended
for users
Any for data
preparation;
Maplnfo 4.5 and
ACCESS 95 for
MEPLUS
Excel or Access
Statistical
Software
Needed?
(Y/N)
Y; SPSS
Y;SAS
Y; SAS or
SPSS
Not specified
N
N
N
Y
Y; S-Plus and
SAS
Y;SAS
Recommended
for developers
Y: any will do
Y; SAS or
SPSS
CIS Software
Needed?
(Y/N)
Not specified
Y; Arclnfo or
Arc View
Y; ArcView
Highly
recommended
Y; ArcView
Y; Arclnfo or
other CIS
Y; ArcView
Y; Arclnfo
Y; Arclnfo or
ArcView
Y; GRASS
Not specified
Y
Y; Arclnfo or
Maplnfo
Other
Not specified
Not specified
None
DBOS memory
manager
(distributed
with DELTA
model)
Developer
participation
None
None
None
Stuttgart
Neural Networi
Simulator
Not specified
Not specified
Any word
processor;
Maplnfo is
needed if
MEPLUS is
being used to
process
MEPLAN
results
None
A-5
-------
Model Name
SAM-IM
SLEUTH
Smart Growth
INDEX
Smart Places
TRANUS
UGrow
UPLAN
UrbanSim
What if?
Operating
System
MS Windows
NT or 95
UNIX
MS Windows
95 or NT
MS Windows
95/98 or
higher
MS Windows
95, 98 or NT;
or Mac with
Windows
emulation
MS Windows
95
Windows 95/98
or NT
MS Windows
95/98,
Windows NT
4.0/2000,
Linux, or UNIX
MS Windows
95/98 or NT
4.0
Program
Compiler
Needed?
(Y/N)
N; although
Visual BASIC
is helpful
Y; gnu C
compiler (gcc)
N
N
N
N
N
Y; JAVA JKD
1.3
N
Data
Management
Tools
None
Not specified
None
None
Windows-based
spreadsheets,
word processors,
and presentation
programs very
useful, but not
essential
None
Excel for data
exchange with
other models.
Not specified
None
Statistical
Software
Needed?
(Y/N)
Y;for
calibration only,
if want to do it
Not specified
N
N
SAS or SPSS
can be useful,
but not
essential
N
Y; SAS for data
exchange with
other models.
Not specified
N
CIS Software
Needed?
(Y/N)
Y; ArcView 3.2
with the Spatial
Analyst
Extention
Not specified
Y; any local
system
Y; ArcView
ArcView or
Arclnfo can be
very useful, but
not essential
Y; Arclnfo,
ArcView
Y; ArcView
(need Arclnfo
to prepare data
layers for local
application)
Not specified
N
Other
None
X-Windows
required for
graphical
version
None
None
Logit
calibration
program can
be useful (e.g.,
Alogit, Hielow)
Powersim
modeling
software
None
Not specified
None
A-6
-------
Exhibit A-4. Cost Comparative Matrix
Model Name
CUF-1
CUF-2
CURBA
DELTA
DRAM/EMPAL
GSM
INDEX
IRPUD
LTM
LUCAS
Markov
MEPLAN
METROSIM
SAM-IM
SLEUTH
Smart Growth
INDEX
Smart Places
Purchase Cost
Not available for "off-
the-shelf" purchase.
Contact developer.
Not available for "off-
the-shelf" purchase.
Contact developer.
Not specified
Contact developer
$30,000-$60,000
which includes training
and consulting
services
Not applicable - not
yet adapted as an
application for
distribution
$15,000- $75,000
Contact developer
Contact developer
(likely no cost)
No cost
No cost
$25,000
$20,000-$30,000
Contact developer
(average $30,000 -
$100,000 total cost)
No cost
Initial version available
at no cost through
U.S. EPA Urban and
Economic
Development Division;
enhanced versions
currently under
development
Contact developer
Operating Cost
Not specified
Not specified
Not specified
Contact developer
Not specified, but
requires about 1
senior modeler with
junior support
Not applicable - not
yet adapted as an
application for
distribution
Typically 1 -8 person
hours
Contact developer
Contact developer
Not specified
No cost
Not available
$2,500 for three initial
runs (negotiable). Full
reports are included.
Contact developer
Not specified
Typically 1-8
person-hours; cost
dependent upon
salary rate for staff or
consultant labor.
Contact developer
Maintenance Cost
Not specified
Not specified
Not specified
Contact developer
Not specified
Not applicable - not
yet adapted as an
application for
distribution
Typically 4-6 person
weeks
Contact developer
Contact developer
Not specified
No cost
10% of purchase
price annually
$5,000 -$10,000/yr
Contact developer
Not specified
Typically 4-6
person-weeks/yr.;
cost dependent upon
salary rate for staff or
consultant labor.
Contact developer
Training Costs
Not specified
Not specified
Not specified
Contact developer
Included with
purchase cost
Not applicable - not
yet adapted as an
application for
distribution
Typically 2-3 person
days
Contact developer
Contact developer
Not specified
No cost
About $640 per day
$10,000 one-time
Contact developer
Not specified
Typically 2-3
person-days; cost
dependent upon
salary rate for staff or
consultant labor.
Contact developer
A-7
-------
Model Name
TRANUS
UGrow
UPLAN
UrbanSim
What if?
Purchase Cost
$7,500
Software is free.
However, to be useful,
the developer must
adapt the model which
can cost $30,000 -
$200,000.
No cost
No cost
For a single user, the
professional price is
$2,495 and the
academic price is
$250. Professional
and academic site
licenses are available.
Operating Cost
Included in purchase
cost
Not specified
Not specified
Not specified
No cost
Maintenance Cost
1 -year guarantee
included in purchase
cost
Not specified
Not specified
Not specified
$1,000/year
Training Costs
$8,000 plus expenses
for 2-week, full-time
course
Not specified
Not specified
Not specified
Not available
A-8
-------
Exhibit A-5. Urban Land Use Categories Addressed Comparative Matrix
Model Name
CUF-1
CUF-2
CURBA
DELTA
DRAM/
EMPAL
GSM
INDEX
IRPUD
LTM
LUCAS
Markov
MEPLAN
METROSIM
User
Defined?
Limits?
No; limited to
residential
No; four
"new" land
uses and three
redevelopment
land uses
Urban Land-Use Categories
Residential
Yes
Single-family;
multi-family
Commercial
No
Yes
Mixed Use
No
Not
considered
separately
from
residential or
commercial
land uses
Industrial
No
Yes
Other
Open space
Residential,
commercial,
and industrial
redevelopment
No; all urban development considered together
Yes; no
limitations
No
Yes; no
limitations
Yes; typically
6-30
categories
No
Yes; can
accommodate
up to 8 land
uses
Yes
Yes;
residential
sector only
Yes; no
limitations
Yes; no
limitations
User defined,
so all potential
urban
categories
By household
income
User defined,
so all potential
categories
User defined,
so all potential
categories
Yes
By density
By density
Owner/renter,
Single/
multi-family,
Size of home
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
urban
categories
By
employment
type
User defined,
so all potential
categories
User defined,
so all potential
categories
Yes
No
Yes
No
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
urban
categories
No
User defined,
so all potential
categories
User defined,
so all potential
categories
Yes
No
No
No
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
urban
categories
By
employment
type
User defined,
so all potential
categories
User defined,
so all potential
categories
Yes
No
No
No
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
urban
categories
Vacant
developable,
vacant
undevelopable
User defined,
so all potential
categories
User defined,
so all potential
categories
Yes
No
No
No
User defined,
so all potential
categories
User defined,
so all potential
categories
A-9
-------
Model Name
SAM-IM
SLEUTH
Smart
Growth
INDEX
Smart
Places
TRANUS
UGrow
UPLAN
UrbanSim
What if?
User
Defined?
Limits?
Yes; limit of 40
categories
Yes; no
limitations
Yes; typically
6-30
categories
Yes; no
limitations
Yes; no
limitations
Yes; no
limitations
Yes; no
limitations
Yes; no
limitations
Yes; can
accommodate
up to 15
different land
uses
Urban Land-Use Categories
Residential
User defined,
so all potential
categories
User defined,
so all potential
categories
By density
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
Commercial
User defined,
so all potential
categories
User defined,
so all potential
categories
Office, retail,
service
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
Mixed Use
User defined,
so all potential
categories
User defined,
so all potential
categories
Can be
customized to
accommodate
user's data
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
Industrial
User defined,
so all potential
categories
User defined,
so all potential
categories
Light/heavy,
Brownfields,
Enterprise
zones
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
Other
User defined,
so all potential
categories
User defined,
so all potential
categories
Can be
customized to
accommodate
user's data
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
User defined,
so all potential
categories
A-10
-------
Exhibit A-6. Non-Urban Land Use Categories Addressed Comparative Matrix
Model Name
CUF-1
CUF-2
CURBA
DELTA
DRAM/EMPAL
GSM
INDEX
IRPUD
LTM
LUCAS
Markov
MEPLAN
User
Defined?
Limits?
Yes;
determined by
availability of
input map
layers
Yes;
determined by
availability of
input map
layers
Yes;
determined by
availability of
input map
layers
No
Yes
Yes; no
limitations
Yes; typically
6-30
categories
No
Yes; can
accommodate
up to 8 land
uses
Yes
No
Yes; no
limitations
Nonurban Land-Use Categories
Agriculture
Yes
Yes
Yes
No
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
Yes
Yes
Yes
No
User defined,
so all
potential
categories
Forest
Yes
Yes
Yes
No
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
Yes
Yes
Yes
No
User defined,
so all
potential
categories
Wetlands
Yes
Yes
Yes
No
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
No
Yes
Yes
No
User defined,
so all
potential
categories
Water
Yes
Yes
Yes
No
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
Yes
Yes
Yes
No
User defined,
so all
potential
categories
Preservation
As identified
by user
As identified
by user
As identified
by user
No
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
No
No
Yes
No
User defined,
so all
potential
categories
Parkland
Yes
Yes
Yes
No
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
No
Yes
Yes
No
User defined,
so all
potential
categories
A-ll
-------
Model Name
METROSIM
SAM-IM
SLEUTH
Smart Growth
INDEX
Smart Places
TRANUS
UGrow
UPLAN
UrbanSim
What if?
User
Defined?
Limits?
Yes; no
limitations
Yes; limit of
40 categories
Yes; no
limitations
Yes; typically
6-30
categories
Yes; no
limitations
Yes; no
limitations
Yes; no
limitations
Yes; no
limitations
Yes; no
limitations
Yes; can
accommodate
up to 15 land
use types
Nonurban Land-Use Categories
Agriculture
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
Forest
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
Wetlands
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
Water
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
Preservation
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
Parkland
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
User defined,
so all
potential
categories
A-12
-------
Exhibit A-7. Impacts of Community Decisions on Land-Use Patterns
Comparative Matrix
Model Name
CUF-1
CUF-2
CURBA
DELTA
DRAM/
EMPAL2
GSM
INDEX
IRPUD
LTM
LUCAS
Markov
MEPLAN
METROSIM
SAM-IM
SLEUTH
Smart
Growth
INDEX
Smart
Places5
TRANUS
UGrow
UPLAN
UrbanSim
What if?
Transportation
infrastructure
/
/
/
/
/3
/
/
/
/
/
/
/
/
/
/
/
/
Local
zoning
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
City&
county
master
plans
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
Other
local
fiscal
policies
/
/4
/
/
/
/
/
/
Developer
impact
fees
/
/
/
/
/
/
/
/
Property
taxes
/
/
/
/
/
/
Municipal
sewer and
water fees
/
/
/
/
/
/
/
/
Subsidies
/
/
/
/
/
/
Road
tolls
/'
/
/
/
/
/
/6
Parking
fees
/'
/
/
/
/
/
/
/
/6
Fuel
and
sales
taxes
/'
/
/
/
/
/
/
/6
VMT
/'
/
/
/
/
/
/
/6
Registration
fees
/'
/
/
/
/
/
/
/6
1 Yes, but only if these can be modeled in an associated transport model integrated with DELTA.
2 Any may be addressed by DRAM/EMPAL when linked to the right model. Without linking, most can not be.
3 Under development
4 No fiscal policies are pre-set in the model. However, if the user can provide specifications on the impact of the revenue source, then the policy can
be incorporated.
5 Smart Places can be customized to evaluate the impact of changes in land-use patterns based on user-supplied criteria.
6 Included only through interaction with travel models.
A-13
-------
Exhibit A-8. Impacts of Land-Use Patterns on Community Characteristics Comparative Matrix
Model Name
CUF-1
CUF-2
CURBA
DELTA
DRAM/
EMPAL
GSM
INDEX
IRPUD
LTM
LUCAS
Markov
MEPLAN
METROSIM6
SAM-IM
SLEUTH
Smart Growth
INDEX
Smart Places8
TRANUS
UGrow
UPLAN
UrbanSim
What if?
Travel
demand
/2
/
/3
/
/
/
/
/
/
/
/
/9
/
Changes in
infrastructure
costs
/
/
/
/
/
/
/
/
/
Changes
in local
tax
revenue
/
/
/
/
/7
/
/
/
/
Other
fiscal
impacts
/
/
/
/
/7
/
/
/
/
Open
Space
/
/
/
/
/
/
/5
/
/
/7
/
/
/
/
/
/
/
Nutrient
loading
/'
/'
/'
/
/
/
/7
/
/
/10
/
/1
Increases
in storm
water
runoff
/'
/'
/'
/
/
/
/7
/
/
/1
Other
nonpoint
source
water
pollution
/'
/'
/'
/7
/
/10
/1
Other water
quality
impacts
/'
/'
/'
/
/
/7
/
/
/1
Changes in
criteria
pollutants
/'
/'
/'
/2
/
/
/
/
/7
/
/
/
/
/1
Changes in
greenhouse
gasses
/'
/'
/'
/2
/
/
/
/
/7
/
/
/
/
/1
Other air
quality
impacts
/J
/J
/J
/2
/4
/7
/
/
/1
1 Model does not directly address these issues. However, model results may be applicable as inputs into appropriate impact models to determine effects of
urbanization and land use change on other systems.
2 When DELTA is integrated with transport and environmental models.
3 Currently under development.
4 The IRPUD can forecast C02 emissions as a function of forecasting transportation-related indicators. Environmental submodels that calculate traffic noise and air
pollution indicators are under development.
5 Open space is addressed in MEPLAN only when linked to the right model.
6 Effects on air and water pollution can be treated if METROSIM is interfaced with any add-on environmental package.
7 Any fiscal or environmental impact which can be estimated as a function of urbanized area could be developed for the output of SLEUTH, but the model does not do
so directly.
8 Smart Places can be customized to evaluate these impacts based on user-specified criteria.
9 UPLAN does not address travel demand directly but can when linked to any travel model.
10 UPLAN is under development to address nutrient loading or sedimentation in surface waters and other nonpoint pollution.
A-14
-------
Exhibit A-9. Model Utility and Integration Comparative Matrix
Model Name
CUF-1
CUF-2
CURBA
DELTA
DRAM/EMPAL
GSM
INDEX
IRPUD
LTM
LUCAS
Markov
MEPLAN
METROSIM
SAM-IM
SLEUTH
Smart Growth INDEX
Smart Places
TRANUS
UGrow
UPLAN
UrbanSim
What if?
Relative Ease of Linking to
Other Models
(1 [easy] - 3 [hard]) '
2
2
2
2
2
2
2
2
1 (environmental process models)
3
2
2
1-2 (depends on package linked to)
2
2
2
2
2
3
2
2
2
Relative Ease of Transferring to
Other Locations
(1 [easy] - 3 [hard]) '
2
2
2
2
2
2
2
2
2
2
2
2
1
2
2
1
2
1
2
2
2
2
Number of Locations to Which
Model Has Been Applied2
1
1
>10
6
40+
350
>10
1
1-5
1-5
>10
25+
6 (includes earlier versions)
2
13
18
1 (35+ other sites have license
agreements to use)
35+
6
2
4
3
1 (1) Effortless; (2) feasible with a manageable amount of modifications required; (3) impossible or impractical, would require a great deal of effort.
2 The spacial scales of the locations vary and include regions/watersheds, large and small cities/ towns, and neighborhoods.
A-15
-------
Exhibit A-10. Basic Operational Characteristics Comparative Matrix
Model
Name
CUF-1
CUF-2
CURBA
DELTA
DRAM/
EMPAL
GSM
Model Type
• Urban growth
• Land use
change
• Urban growth
• Urban
economic/
land use
market
• Urban
statistical
• Spatial
interaction
• Aggregate
logit
•CIS
Thematic Scope
• Urban
development
evaluation and
simulation
• Urban
development
evaluation and
simulation
• Urban growth
• Environmental
and ecological
quality
• Urban and
regional
economics
• Housing
• Employment
• Development
• Resource Land
• Conservation
• Watershed
Management
Underlying
Math Structure
• Deterministic
• Deterministic
• Stochastic
• Stochastic
• Deterministic
• Stochastic
Not specified
Operational
Methods
• Regression
• Multinominal
logit
• Regression
• Binomial logit
• Regression
• Markov chains
• Multinominal
logit methods
• Cobb-Douglas
utility functions
• Elasticity-based
responses
• Matrix
adjustment
methods
• Multinominal
logit
• Regression
Not specified
Technical
Expertise for
Calibration
(1 [general] -
3 [extensive]) '
2
3
2
3
3
3
Relative Ease
of Calibration
(1 [easy] -3
[hard])1
3
3
2
3
1
2
Measure of
Confidence or
Goodness of
Fit? (Y/N )
N
Y
Y
N
Y
Not applicable
A-16
-------
Model
Name
INDEX
IRPUD
LTM
LUCAS
Markov
MEPLAN
Model Type
•CIS
• Urban Impact
• Travel demand
model
• Urban
economic/land
use market
models
•CIS
• Urban Impact
• Neural
network
•CIS
• Markov chain
• Travel demand
• Urban
economic/land
use market
• Hedonic
Thematic Scope
• Land use
• Transportation
• Housing
• Employment
• Natural
Environment
• Transportation
• Economics
• Technological
impacts
• Land use
• Ecology
integrity
• Economic
sustainability
• Land use
• Environmental
impacts
• Socioeconomic
• Residential
housing
• Mobility
• Spatial
economic-
based
input/output
Underlying
Math Structure
• Deterministic
• Probabilistic
• Stochastic
• Empirical
• Stochastic
• Stochastic
• Stochastic
Operational
Methods
• Causal
inference
• Correlation
• Linear
programming
• Network
analysis
• Time-series
• Markov chains
• Multinominal
logit methods
• Microsimulation
• Variant of
inclusive value
method
• Markov chains
• Regression
• Artificial neural
networks
• Time series
• Linear
programming
• Markov chains/
transition
matrices
• Multinominal
logit
• Regression
• Multinominal
logit
• Network
analysis
Technical
Expertise for
Calibration
(1 [general] -
3 [extensive]) '
2
3
3
3
3
3
Relative Ease
of Calibration
(1 [easy] -3
[hard])1
2
3
3
3
2
2
Measure of
Confidence or
Goodness of
Fit? (Y/N )
Not applicable
Not specified
Y
Y
Y
Y
A-17
-------
Model
Name
METROSIM
SAM-IM
SLEUTH
Smart
Growth
INDEX
Smart
Places
Model Type
• Travel demand
• Markov chain
• Urban
economic/land
use market
• Hedonic
• Discrete
choice method
•CIS
• Cellular
automata
•CIS
• Urban impact
• Travel demand
•CIS
Thematic Scope
• Land use
• Metropolitan
economy
• Urban growth
• Transportation
• Economics
• Environmental
impacts
• Urban growth
• Environmental
impacts
• Land use
• Transportation
• Housing
• Employment
• Infrastructure
• Environment
• Land-use
• Economics
• Environmental
impacts
Underlying
Math Structure
• Deterministic
• Stochastic
• Empirical/sem
i-Empirical
• Deterministic
• Stochastic
• Empirical/sem
i-Empirical
• Stochastic
• Deterministic
• Deterministic
Operational
Methods
• Markov chains
• Multinominal
logit methods
• Network
analysis
• Regression
• Time-series
• Dynamic
economic
general
equilibrium
analysis
• Cellular
automata
• Multinominal
logit methods
• Regression
• Cellular
automata
• Time-series
• Monte Carlo
imaging
• Causal
inference
• Correlation
• Linear
programming
• Multinominal
logit
• Network
analysis
• Time series
• Causal
inference
Technical
Expertise for
Calibration
(1 [general] -
3 [extensive]) '
1
1 (but there is a
learning
curve/training
required)
3
3
2
Relative Ease
of Calibration
(1 [easy] -3
[hard])1
1-2
3
2
1
2
(calibration is
not required)
Measure of
Confidence or
Goodness of
Fit? (Y/N )
Y; if desired
N; model
doesn't provide
but statistical
packages used
in the calibratior
do
Y
N
N
A-18
-------
Model
Name
TRANUS
UGrow
UPLAN
UrbanSim
What if?
Model Type
•CIS
• Urban impact
• Travel demand
• Urban
economic/land
use market
• Hedonic
• Systems
dynamics
•CIS
• Urban impact
• Random utility
logit
• Urban
economic/
land use
market
•CIS
• Hedonic
•CIS
Thematic Scope
• Transportation
• Economics
• Environmental
impacts
• Private and
public
infrastructure
• Land use
• Transportation
• Land-use
evaluation and
change analysis
• Land-use
• Transportation
• Economics
• Environmental
impacts
• Land-use
evaluation and
change analysis
Underlying
Math Structure
• Stochastic
• Deterministic
• Deterministic
• Empirical/sem
i-Empirical
• Deterministic
Operational
Methods
• Causal
inference
• Multinominal
logit
• Network
analysis
• Time-series
• Discrete choice
analysis
• Decision theory
• Random utility
theory
• Input-output
analysis
• Algorithms
Casual inference
System dynamics
Not specified
• Expert systems
• Multinominal
logit
• Regression
• Monte Carlo
simulation
• Mapping (CIS)
Technical
Expertise for
Calibration
(1 [general] -
3 [extensive]) '
3
3
Not applicable
3
Not applicable
Relative Ease
of Calibration
(1 [easy] -3
[hard])1
2
3
Calibration not
required
2
Calibration not
required
Measure of
Confidence or
Goodness of
Fit? (Y/N )
Y
N
N
Y
N
1 (1) Parameters can be recalibrated using options embedded in the software for the model.; (2) Parameters can be recalculated using
methods/instruction cited in model documentation or by altering input files; (3) Parameters can only be recalibrated using original programming with
no guidance from the model developers (e.g., no documentation), or parameters are hardwired and cannot be recalibrated.
A-19
-------
Exhibit A-11. Spatial And Temporal Capabilities Comparative Matrix
Model Name
CUF-1
CUF-2
CURBA
DELTA
DRAM/EMPAL
GSM
INDEX
IRPUD
LTM
LUCAS
Markov
MEPLAN
METROSIM
SAM-IM
SLEUTH
Smart Growth
INDEX
Spatial Resolution
User defined, but
generally 1 acre or larger
One-hectare(1 00x1 00m)
grid cells
One-hectare(1 00x1 00m)
grid cells
User defined, but intended
to work with strategic
rather than very detailed
zones
Census tracts for some
data; regional level for
economic data
User defined
User defined
Revised version of model
will allow about 300 zones.
Parcel (30m x 30m), plat
(1 00m x 100m), block
(300m x 300m), and local
(1 km x 1 km)
User defined; a single grid
cell or pixel may be
defined to 90m x 90m.
One or more households
User defined; can vary
from a few hundred
meters to whole countries,
depending on study
User defined
User defined
User defined
User defined between
5-1 00 acres
Spatial Extent
Customized for user needs
Customized for user needs
Scalable and can be
customized for user needs
Customized for user needs,
typically applicable to cities with
populations of 250,000+
Customized for user needs, but
preferably metropolitan areas
with a population of at least
200,000
User defined
User defined (depends on the
extent of local CIS)
Local or regional level
User defined (precedence
given to watersheds)
User defined
Not applicable
User defined; has been used to
represent cities in regional
context to entire countries
User defined
User defined
User defined
Community or region, depends
on the extent of local CIS
Temporal
Resolution
5 year
5 year
User defined
1 year increments
recommended, but
can be longer
5 year
User defined
Yearly
User defined
5 or 10 year
5 year
Usually 3-5 years
User defined, but
five years is
common
Yearly or some
aggregation of
years such as 2
years, 5 years or
decades
User defined
Yearly
Yearly
Temporal Extent
(Future and Past)
5+ years into the future
5+ years into the future
User defined into the
future
User defined into the
future
40 years into the future
User defined
User defined (depends
on available data)
User defined into the
future
20-50 years into the
future; can hindcast
into the past
100 years into the future
Limited by Census data
User defined
Any number of time
periods can be
accommodated, but
more than 30 not
recommended
User defined into the
future
As far into the past or
future as available
data will allow
20 years into the future
A-20
-------
Model Name
Smart Places
TRANUS
UGrow
UPLAN
UrbanSim
What if?
Spatial Resolution
User defined
User defined, but too
many zones can become
a nuisance
Depends on available
CIS data
Low density residential
represented in 10 acre
parcel size (200 m
cells),while all other land
uses are represented by
Vz acre parcel size
(50 m cells)
User defined, current
application have used
1 50 m resolution
User defined, but best
suited for sizes larger than
single parcel
Spatial Extent
User defined
User defined
User defined
User defined
User defined
User defined
Temporal
Resolution
Not applicable
User defined, but 5
years is common
Yearly
User defined
Yearly, but option
for arbitrary time
intervals.
User defined
Temporal Extent
(Future and Past)
Not applicable
User defined
1,950-2,100
User defined
User defined into the
future
Up to 4 projection
periods
A-21
-------
A-22
-------
Appendix B
Land Use Models: Technical Fact Sheets
-------
-------
California Urban Futures (CUF) Model: CUF-1
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
CUF-1 may be customized to meet study area needs of the end user. The typical county in the
San Francisco Bay area simulation had 30,000 to 50,000 DLUs. DLUs range in size from one
acre to several hundred acres. The spacial scale is based on the resolution of the data
provided by the end user, but it is generally one acre or larger. The flexibility of the CUF model
allows it to be used at a regional or county level. Because the CUF model allocates growth on
a site-by-site basis, it is well suited to examining environmental impacts associated with
development alternatives.
Temporal Resolution and Extent
The CUF-1 model generates 5-year population growth forecasts and maps.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
Requires population trend data to calibrate growth models.
Requires development cost data to estimate potential profitability of site development.
Input Pre-Processing Requirements
Input data may possibly require pre-processing, but pre-processing requirements not specified.
Model Assumptions
The CUF model enables users to build their own set of assumptions about land use suitability;
projected growth demands (future population); and regulatory and investment policies.
Setting Parameters
Several modules of CUF require the end user to set various input parameters to evaluate
alternative scenarios. The required parameters are listed as follows:
• The Bottom-up Population Growth Submodel
- Dependent variable: Population levels by city or county five years ago
- Independent variables: City size class (i.e., very small, small, medium, medium
large, and large); whether the city or county has a population, housing, or
development cap; whether the city is prevented from expanding because it is land-
locked (by neighboring communities) or water-locked; gross population density by
city for the previous five-year period; a variable of current statewide population
weighted by the land area of each city in the previous five-year period; the numerical
B-l
-------
change in "basic" employment in the county during the previous five years; and,
each county's share of region-wide population growth (lagged five years) and
weighted by the state population in the current year.
• The Spatial Database
- A series of user-defined map layers that describe attributes such as the
environmental, land-use, zoning, current density, and accessibility characteristics of
the sites in the study region
• The Spatial Allocation Submodel
- Several parameters related to the total profit a home builder would expect to realize
on the construction of as many new homes as could be accommodated on a
particular DLU, including new home sale price; raw land price; hard construction
costs; site improvement costs; service extension costs; development, impact,
service hookup, and planning fees; delay and holding costs; and extraordinary
infrastructure capacity costs, exactions, and impact mitigation costs
Comparing Scenarios
The CUF model addresses multiple scenarios, allowing the user to determine the impacts of
alternative growth policies. For example, three regional growth policy scenarios for the
Northern California Bay Region were evaluated using the model: a business as usual scenario,
a maximum environmental protection scenario, and a compact cities scenario. The scenarios
are user defined and used by the CUF model to determine future land use patterns.
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-Processing Requirements
Not specified.
NEXT STEPS FOR MODEL DEVELOPMENT
A second generation of the CUF model ("CUF-2") has been developed to address several of
the limitations discovered in the original CUF-1 model. See the separate evaluation of the
CUF-2 model in this series for more information.
B-2
-------
California Urban Futures (CUF) Model
Second Generation: CUF-2
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
The CUF-2 model can be customized to meet study area needs of the end user. The basic
units of analysis (DLUs) consist of one-hectare (100 m x 100 m) grid cells. The flexibility of
the CUF-2 model allows it to be used at a regional or county level. Because the CUF model
allocates growth on a site-by-site basis, it is well suited to examining environmental impacts
associated with development alternatives.
Temporal Resolution and Extent
The CUF-2 model generates 5-year population growth forecasts.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
CUF-2 incorporates a statistical model. User must initially calibrate a series of multi-nomial
logit parameters which they can then re-use or modify. Testing of alternative infrastructure
investments may require additional digitizing or GIS preparation.
Input Pre-Processing Requirements
Input data may possibly require pre-processing, but pre-processing requirements not specified.
Model Assumptions
The CUF-2 model enables users to build their own set of assumptions about land use
suitability; projected growth demands (future population); and regulatory and investment
policies.
Setting Parameters
CUF-2 initially requires the user to statistically calibrate the parameters of a non-ordinal, multi-
nomial logit model of urban land use change. Once calibrated, parameters may be re-used and
modified. Policy simulations are tested using a "check list" of constraints and policies.
B-3
-------
Comparing Scenarios
The CUF model addresses multiple scenarios, allowing the user to determine the impacts of
alternative growth policies. The scenarios are user defined and used by the model to determine
future land use patterns.
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-Processing Requirements
Not specified.
NEXT STEPS FOR MODEL DEVELOPMENT
Not specified.
B-4
-------
California Urban and Biodiversity
Analysis Model (CURBA)
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
The CURBA model is scalable and can be customized to meet study area needs of the end
user. CURBA's basic unit of analysis and minimum mapping unit is one-hectare (100 m by
100 m). The typical study area is one or more counties.
Temporal Resolution and Extent
CURBA's pilot simulations were run for 15 years (1995-2010). The pilot simulations did not
have intermediate time steps. Instead, the model showed changes from the initial time period
(1995) and the end point of the simulation (2010). Simulations may be run with different end
points, however.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
Requires external population or household growth projections reported at the county or
municipal level.
Input Pre-Processing Requirements
Input data may possibly require pre-processing. Certain data may require reformatting to
accommodate the GIS.
Model Assumptions
CURBA assumes that factors and forces that determined past urban growth patterns and
trends will also determine future patterns and trends. All urban land uses are assumed to be
homogeneous.
Setting Parameters
CURBA initially requires the user to statistically calibrate the parameters of a binomial logit
model of urban growth. Once calibrated, parameters may be re-used and modified. Policy
simulations are tested using a "pick list" of constraints and policies.
Comparing Scenarios
CURBA allows the user to address multiple scenarios by changing various assumptions and
saving the results of alternative scenarios for comparison.
B-5
-------
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-Processing Requirements
None.
NEXT STEPS FOR MODEL DEVELOPMENT
Model is currently being refined to reduce the effects of spatial auto-correlation.
B-6
-------
DELTA
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
The DELTA model contains an urban level and a regional level. The urban DELTA model
typically applies to cities with a population of 250,000 or greater. The regional DELTA model
applies to areas containing a number of distinct sub-regional economies (e.g., labor market
areas).
The model is intended to work with strategic areas, versus very detailed areas. The spatial
resolution of the model is based on the resolution of the data provided by the user.
Temporal Resolution and Extent
DELTA can be run for any historical period for which data are available. It can be run as far
into the future as the user requires, though accuracy clearly diminishes with longer time
horizons.
The model developer recommends that DELTA'S scenarios run in a yearly increment. The
transport model integrated with DELTA can be run with longer increments.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
The DELTA package is intended to assist in the testing and evaluation of a wide range of
transport and land use policies. DELTA itself is a land-use modeling system; it is designed to
interface with any appropriate transport model to create a full land-use/transport interaction
model. Existing transport models may be used with the DELTA model.
The full DELTA package consists of the following:
• An urban level, containing sub-models for:
- Urban development processes;
- Changes in urban area quality;
- Demographic (household) transitions and economic growth;
- Car ownership;
- Location/real estate market (e.g., location of households and jobs, real estate
values, and occupation/vacancy); and
- Labor market and commuting.
B-7
-------
• A regional level, containing sub-models for:
- Migration between urban areas;
- Investment levels/locations; and
- Production and trade.
Full DELTA applications contain both urban and regional levels. Partial applications can be
built at either the urban or regional level. The system offers a high level of flexibility for more or
less detailed model designs (e.g., few or many categories of households and of employment).
As with all land use models, much of the work associated with using the DELTA system is
related to collecting local data and testing the "realism" of the model under local conditions.
The DELTA package is designed to operate in short (e.g., 1- to 2-year) time increments. Land
uses respond to transport change, modeled in the chosen transport model, as well as to land
use policies that are modeled within DELTA. Outputs from DELTA - particularly numbers of
residents and numbers of jobs - provide inputs to the transport model, which can be run at the
end of each DELTA time increment or at less frequent intervals.
Input Pre-Processing Requirements
DELTA may require input pre-processing, depending on the condition and format of the user's
data. All inputs must be in DELTA-specific, fixed formats (e.g., fixed-format ASCII files).
Model Assumptions
The model assumes that development is new construction. Redevelopment and reuse/change
of use are not easily captured by the existing DELTA system. With the exception of
development, DELTA enables users to build their own set of assumptions about land use
parameters.
Setting Parameters
Behavioral coefficients for each sub-model and variables to determine demographic and
economic scenario(s) must be set in order for DELTA to operate.
Comparing Scenarios
The user may define scenarios by specifying different inputs (e.g., population or economic
growth rates). DELTA will project the local effect(s) resulting from the chosen inputs for a
particular scenario.
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-Processing Requirements
The model developer encourages users to load outputs from DELTA into their preferred
spreadsheets and/or mapping software for analysis.
B-8
-------
NEXT STEPS FOR MODEL DEVELOPMENT
The model developer plans to improve the interactions between the urban and regional levels of
the model. In addition, the developer would like to facilitate the interfacing of the DELTA
system with other models, including different transport models and models for specific
processes.
B-9
-------
DRAM/EMPAL
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
DRAM/EMPAL may be customized to meet user needs but should cover areas with a
population of at least 200,000 persons. The smallest metropolitan area modeled with DRAM/
EMPAL was Colorado Springs (pop. 312,000), though during the models' development the area
of Hazelton, PA (pop 90,000) was modeled, while the largest was Los Angeles (pop.
6,132,000). The spatial scale is based on census tracks.
Temporal Resolution and Extent
DRAM/EMPAL's temporal extent has been extended 40 years into the future. Forecasts are
done in 5-year increments with output from one forecast year becoming the input for the next.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
The employment allocation sub-model, EMPAL, forecasts the location of future employment by
economic sector to spatially contiguous zones overlaying the metropolitan area. EMPAL does
this by taking into account some of the following variables: zone-specific employment levels
(total and by economic sector) for a specified time; number of households (population) in each
zone, by income level, for a specified time; regional level of target year employment (growth
trends); travel time between zones, or other zone-specific measures of accessibility to the work
force; and total land area of each zone. EMPAL uses all of this information to estimate the
likelihood of a site for future employment primarily based on how often it was selected in the
past given the distribution of households and ease of getting to the zone from other zones.
DRAM forecasts the future location of households given this distribution of employment and the
attractiveness (including accessibility) of the zones. To do this, DRAM considers the following
variables: employment, by type, in each zone [from EMPAL]; impedance (travel time and cost)
between zones; percent of households, by type, per zone; and various land uses (vacant
developable land [acres], developed land [%], residential acres [%]). DRAM also contains trip
distribution models and can project home-to-work, home-to-shop, and work-to-shop trips. An
additional submodel within DRAM, called LANCON, calculates land consumption in the forecast
period using multiple regression to combine base year data with a forecast.
Calibration of these models is performed with a submodel, CALIB, and consists of estimating
the historical relationships between employment in each zone and the above-mentioned
variables.
B-10
-------
Input Pre-Processing Requirements
Data provided by the agency must be converted into standard formats that can be used by the
models.
Model Assumptions
The central assumption is that changes in transportation facilities that result in significant
changes in relative travel times will, over time, have corresponding impact on the future
distribution of employment and residential locations. Another assumption of the model is that
activities (employment and households) are complex nonlinear functions of accessibility to other
activities.
Estimating Parameters
All of the equation coefficients must be calibrated in order for the model to operate. These
parameters can be re-calibrated using options embedded in software for the model.
Recalibration takes just a few minutes per activity type by computer time. Significant time, on
the order of months to 2 years depending on the size of the network, may be required if
researchers are involved in defining the variables and obtaining the data. CALIB is the
calibration program that is used to estimate the equation coefficients in both DRAM and
EMPAL. In addition to maximum likelihood estimates of the equation coefficients, CALIB
provides goodness-of-fit statistics, asymptotic t-tests of the statistical significance of the
coefficients, and point elasticities for sensitivity analysis.
The procedure used for estimation of the parameters is a gradient search procedure. This
automatic calibration program is an innovative feature that makes the modeling system unique
among its rivals and is one of the prominent features of the model system. Many land-use
modeling efforts with other models could not be applied because the model system could not be
calibrated properly. CALIB produces estimates of parameters in a systematic way, making it
possible to compare values with those of similar regions as an additional degree of comfort for
modeling and policy analysis. Furthermore, DRAM/EMPAL provides a quantitative measure of
the goodness of fit.
Comparing Scenarios
DRAM/EMPAL addresses multiple scenarios, such as the application of a wide range of
transportation demand management (TDM) strategies to different forms of polynucleated urban
development. Additionally, land-use controls and scenarios on growth rates and military base
closures have been developed.
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-processing Requirements
No output post-processing requirements were identified.
B-ll
-------
NEXT STEPS FOR MODEL DEVELOPMENT
The DRAM/EMPAL models have recently been extended into a new system called
METROPILUS. This is an evolution of the DRAM/EMPAL package. According to the
developer, it combines employment and residence location and land consumption in a single
comprehensive package embedded in a GIS environment.
The model uses a location surplus notion to arrive at the DRAM formulation. METROPILUS
increases its reliability through the addition of a lagged variable of households in DRAM. "Land
value" in the attractiveness measure of DRAM has also been added. This proposed "land
value" is relative house prices in the form of a multi-variate house index giving consideration to
single- and multi-family structures.
Implementation is available in phases. First, a data platform is selected to facilitate model
component relationships and access to a common database. ArcView is the GIS-based data
structure that current DRAM/EMPAL-CALIB uses to interact between model system
components, as well as to access mapping and statistical routines. This ArcView-based
DRAM/EMPAL package will be an intermediate product. METROPILUS uses this data
structure and will eventually contain a reformulation of DRAM and EMPAL with the location
surplus notion as mentioned above. This will allow enhancement to component models without
sacrificing unaffected routines and submodels.
B-12
-------
Growth Simulation Model (GSM)
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
Because of GSM's flexibility, no land is too small or too large to be analyzed using GSM. The
dominant constraints are data availability and the financial and staff resources of the entity
conducting the analysis. For smaller areas, a greater level of spatial resolution and landuse
change algorithms can be incorporated, depending on data availability. For larger areas, a
coarser analysis based on limited data is preferable, due to likely constraints on computer
resources and the type of data available.
Temporal Resolution and Extent
GSM has no limits on temporal resolution and extent. The dominant constraint is data
availability. GSM is currently run based on the target year for which population and
employment figures are available, typically 10 to 30 years.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
The GSM inventories the supply of develop-able land by measuring the capacity of each area to
accommodate new development activity. This is accomplished using geographic information
system (GIS) data on landscape features, linked to a variety of geo-referenced data on local
land use plans and management programs.
In summary, undeveloped land and land developed below capacity represent the "supply" of
developable land in each watershed. Zoning, subdivision, and other regulations and programs
affecting development and / or preservation of land are used to estimate the capacities of
different types of developable land for new development (e.g., the number of new households
that could be accommodated on each type of land under existing or hypothetical programs). A
variety of land attributes associated with each develop-able parcel (e.g., distance from or
proximity to interstate highways, schools, retail services, and undeveloped land) are used to
rate the probability of conversion.
Demand for new development within each small area is distributed by the model to types of
developable land based on probability of conversion, capacity, and/or county-specific
information on recent development patterns and trends which reflect the market for land in the
area. Land use / land cover associated with hypothetically developed parcels is then:
• Converted to a specific type of development to accommodate projected growth;
• Maintained in or converted to forested or other vegetated cover to meet requirements of
forest conservation programs, stream buffer protection ordinances, or open space
preservation programs; or
B-13
-------
• Allowed to remain in its existing condition. This occurs when there is not enough
demand for new development to utilize existing capacity.
The model produces an inventory of future land use resulting from the changes.
Input Pre-Processing Requirements
GSM currently receives input as Arclnfo maps, associated Arclnfo attribute data, and relational
databases. GSM could be modified to accept data in any relational database that could be tied
to Arclnfo. Data inputs must be georeferenced and associated with geographic areas or
relational database linkages.
Model Assumptions
Not specified.
Setting Parameters
The amount of effort required to set parameters in GSM is dependent on how the model is
being used. GSM currently requires staff with experience in Arclnfo and relational databases to
set parameters.
Comparing Scenarios
Multiple scenarios can be compared by running multiple simulations, varying the input data and
parameters. The statistical, graphical, and mapped output can then be compared.
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-processing Requirements
Output results are currently converted to statistics, graphs, and maps. This requires staff
knowledgeable in Arclnfo, relational databases, and graphics packages. GSM is flexible, and
can be modified to meet the output needs of the user.
NEXT STEPS FOR MODEL DEVELOPMENT
The model was developed to examine and address land use and growth patterns for the state
of Maryland. The model is currently being adapted to run in conjunction with transportation
models.
B-14
-------
INDEX®
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
The spatial resolution and extent of INDEX® is user defined; it can be a neighborhood,
community, or region depending on the extent of the local GIS. The model is suited to evaluate
small or large areas - the smallest scale is usually parcels or building footprints.
Temporal Resolution and Extent
The temporal extent of INDEX® is user defined depending on availability of exogenous historic
and/or projected data. The model runs scenarios on yearly time intervals.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
INDEX® uses an exogenous population forecast and user-selected policy constraints and
incentives to determine where growth should occur in the community. The results of this
analysis are scored using a set of environmental performance indicators.
Model Assumptions
None.
Input Pre-Processing Requirements
None.
Setting Parameters
Determined at the time of each community customization. Typical examples include rates of
water consumption and solid waste generation by dwelling type; and air pollutant and
greenhouse gas emission coefficients according to local generation resource mix and building
type. As a sketch-level planning tool, the model is not intended to be precisely calibrated prior
to running. Its customization usually includes the ability to set and vary a few basic local
parameters, e.g., persons per household by dwelling type.
Comparing Scenarios
INDEX® addresses multiple scenarios such as user-selected alternative land-use plans and
urban designs created exogenously.
B-15
-------
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-Processing Requirements
None.
NEXT STEPS FOR MODEL DEVELOPMENT
Each new community version usually incorporates enhanced functionalities.
B-16
-------
IRPUD
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
The IRPUD model can examine location and mobility decisions in a metropolitan area on a local
level and on a regional level. It subdivides the study area into zones that are connected with
each other by transportation networks (e.g., public transportation lines and roads). The present
implementation of the model works with only 30 zones. However, the ongoing revision of the
model will have some 300 zones.
Temporal Resolution and Extent
A simulation period for the IRPUD model is one or more years. The IRPUD model can be run
for any historical period for which data are available. The model can be run as far into the
future as the user requires. The accuracy of large-range projections depends on the reliability
of the assumptions by the user about socio-economic conditions and transport infrastructure
scenarios.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
The IRPUD models four major groups of stock variables: population, employment, residential
buildings (housing) and non-residential buildings (industrial and commercial workplaces and
public facilities). The actors representing these stocks are individuals or households, workers,
housing investors and firms. These actors interact on five submarkets of urban development.
The five submarkets treated in the model and the market transactions occurring on them are
provided below:
• Labor market: new jobs and redundancies;
• Non-residential building market: new firms and firm relocations;
• Housing market: immigration, out-migration, new households and moves;
• Land and construction market: changes of land use through new construction,
modernization or demolition; and
• Transport market: trips.
On each submarket, supply and demand interact and result in market transactions. Choice in
the submarkets is constrained by supply (e.g., jobs, vacant housing, vacant land, vacant
B-17
-------
industrial or commercial floor space) and guided by attractiveness, which in general terms is an
actor-specific aggregate of neighborhood quality, accessibility and price.
The user must obtain the following data for use as inputs to the model: forecasts of regional
employment and population subject to long-term economic and demographic trends; or policies
in the fields of industrial development, housing, public facilities and transport.
To implement the IRPUD model, the user must follow four basic steps:
• Data acquisition from various sources;
• Model calibration;
• Model validation; and
• Policy testing.
The user can spatially and graphically analyze the output of the model both during the
simulation and offline using custom-written presentation software.
Input Pre-Processing Requirements
All inputs must be in model-specific, fixed format ASCII files. However, utility programs exist to
convert input data from other software formats, such as spreadsheet (e.g., Excel) and GIS
(e.g., Arclnfo) to model input formats.
Model Assumptions
The IRPUD model makes several assumptions within each of the six submodels: transport,
ageing, public programs, private construction, labor market, and housing market.
Transport
Several assumptions go into the equations used in this submodel, including the following
assumptions about car availability:
• Car ownership is a function of household travel budgets and other transport
expenditures;
• All cars owned by a household are available for work trips;
• Cars not used for work trips are available for shopping and service/social trips; and
• Cars not used for other trips are available for school trips by students with a driving
licence.
Ageing
This submodel of the IRPUD model simulates changes of zonal stock variables (i.e., population,
employment, residential buildings, and non-residential buildings) which are assumed to result
from biological, technological or long-term socio-economic trends or originating outside of the
B-18
-------
model (i.e., which are not treated as decision-based). Assumptions used in this submodel
include:
• Development of birth and death rates, household formation rates, etc.;
• Development of total migration into and out of the study region;
• Development of technological parameters, such as housing technology and costs, car
technology (e.g., miles per gallon); and
• Development of incomes, household housing and travel budgets, consumer price
indices, etc.
Public Programs
This submodel is used to enter user-specified policies in the fields of land use, housing, non-
residential buildings, public facilities and transport into the model. The user is responsible for
the consistency and plausibility of the policies, however, the model observes consistent rules for
the implementation of the policy (e.g., adds parking facilities for residential and non-residential
buildings or public facilities).
Private Construction
This submodel models the investment and locational behavior of private developers. For each
type of building, or submarket, the model first determines the overall demand for new
construction or rehabilitation based on the vacancy rate in that submarket left over from the
previous simulation period. Then the model allocates that total demand to suitable vacant land
(or to existing building stock in the case of rehabilitation) based on multiattribute measures of
attractiveness indicating the profitability of investing at that location. Land prices and prices/
rents for buildings are adjusted in each period in response to vacancy rates in the submarket.
Labor Market
New hirings, redundancies and changes of jobs are modeled as a function of growth and
decline of industries in local submarkets (determined partly by overall socio-economic
assumptions and partly by location and relocation behavior of firms in the previous simulation
period). This submodel assumes that workers looking for jobs consider their residential location
when choosing a job.
Housing Market
The Housing Market Submodel simulates intraregional migration decisions of households as
search processes in the regional housing market. In this submodel, households are assumed
to adapt their aspiration levels for housing to the supply conditions on the market. Transactions
in the housing market consist of starter households (i.e., newly formed households),
immigrations or out-migration, and moves. Households consider a move if their housing
satisfaction (a multiattribute utility comprising their satisfaction with the dwelling, satisfaction
with the neighborhood and satisfaction with the price or rent) is significantly lower than available
alternatives in the housing market. The housing market is the first submodel implemented as a
stochastic Monte-Carlo microsimulation - it is intended to convert other submodels to
microsimulation in the future.
B-19
-------
Setting Parameters
There are several ways to determine the values of model parameters. Socio-economic and
technical parameters are projected based on other studies, available statistics and expert
judgment. Behavioral parameters are, as far as possible, calibrated using available behavioral
data (e.g., from travel surveys and housing market statistics). Where no sufficiently
disaggregate data are available, behavioral parameters are set by stated-preference techniques
and expert judgment and validated by comparing the model results with available aggregate
information.
Comparing Scenarios
The user may define scenarios by specifying different inputs (e.g., economic growth rates by
sector, immigration and out-migration rates, policies in the fields of land use, housing and public
facilities, and transport). The IRPUD model will project the local and/or regional effect(s)
resulting from the chosen inputs for a particular scenario.
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-Processing Requirements
Graphical output is either on-screen or in WordPerfect WPG format for later post-processing
and printing. In addition, custom-written programs are used to extract model results from the
model data base for mapping.
NEXT STEPS FOR MODEL DEVELOPMENT
Future development of the IRPUD model will follow three directions:
• The model is used for a case study of the PROPOLIS project of the European
Commission, in which a number of urban land-use transport models are used to conduct
comparable scenario runs for a number of European cities (see
http://www.ltcon.fi/propolis). For this study, the spatial resolution of the IRPUD model
will be extended to about 300 zones.
• Also in the context of the PROPOLIS project, the model developer is adding
environmental submodels to the IRPUD model that calculate traffic noise and air
pollution indicators.
• In a separate strand of development, the IRPUD model is being gradually converted
from its present aggregate form into a fully disaggregate form based on Monte-Carlo
microsimulation, following the example of the existing Housing Market Submodel. In the
course of this conversion also the existing transport model will be converted to activity-
based microsimulation similar to, but much less complex than the TRANSIMS traffic
microsimulation developed by the Los Alamos National Laboratory for the U.S.
Department of Transportation.
B-20
-------
Land Transformation Model (LTM)
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
Spatial extent of the LTM can be any definable region; however, because future developments
will focus on coupling land use change and hydrogeologic and geochemical processes,
precedence is given to watersheds as the spatial extent in LTM applications. Four different
resolution classes are used in LTM: parcel (30 m x 30 m); plat (100 m x 100 m); block (300 m x
300 m); and local (1 km x 1 km).
Temporal Resolution and Extent
LTM can provide forecasts for 20 to 50 years into the future, generally at 5- or 10-year
intervals. The model can also be used to hindcast into the past in order to couple to
environmental impact models.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
The LTM utilizes a set of spacial interaction rules, which are organized into an object class
hierarchy. The model is coded within a GIS with graphical user interfaces that allow users to
change model parameters. Each module of LTM applies user-defined parameters (e.g.,
identifying and weighting land use features) to base data to derive study-area conditions. The
model uses GIS to make spatial calculations between drivers of land use change and cells
being considered for land transition. The values resulting from these calculations are converted
to relative land transition probabilities. Though each module of LTM performs a specific
function, all modules and submodules are not recognized to be mutually exclusive.
• The Policy Framework module organizes the goals for the watershed's stakeholders
(e.g., resource managers, and private and corporate landowners). Stakeholder goals
may include: control of pollutant inputs, ecological restoration, or habitat preservation.
• The Driving Variables module contains three general categories: Management Authority;
Socioeconomic; and Environmental. Management Authority is an important component
since federally and state-owned lands need to be excluded from development.
Socioeconomic driving variables include population change, economics of land
ownership, transportation, agricultural economics and locations of employment.
Environmental driving variables of land transformation are abiotic and biotic.
• The Land Transformation module is characterized by change in land use and land
cover. Land uses considered at the most general level are: urban, agricultural/pasture,
forest, wetlands, open water, barren and non-forested vegetation. Land cover types that
B-21
-------
are considered include: types of agriculture (row crops versus non-row crops),
deciduous and coniferous forests, and non-forested vegetation.
• The Intensity of Use module considers land management practices, resource use and
human activities. Intensity of use can be measured as chemical inputs to the land to
increase its productivity, chemical inputs as it results from human activities (e.g., salting
of roads) and natural resource use.
• The Processes and Distribution module characterizes groundwater and surface water
flows, chemical and sediment transport across land and through rivers and streams,
geochemical interactions and fluxes.
• The Assessment Endpoints module provides indicators of ecological integrity and
economic sustainability and are used to quantify the nature of changes in landscapes.
Input Pre-Processing Requirements
Land-use and features in the watershed are characterized as a grid of cells made using
ArcView and the LTM software. To understand the grids, SNNS needs to have a pattern file.
To generate a patter file that could be understood by the neural network, a "C" program was
written to convert the files.
Model Assumptions
The LTM uses landscape ecology principles to forecast land use change and to describe the
influence of land use change on ecosystem integrity and economic sustainability of large
regions
Setting Parameters
Calibration of the model requires expertise in land-use modeling and "C" language
programming as well as SNNS neural network batch files. Minimally, either per capita use
requirements or the number of cells that will transition during each time step need to be set in
order for the model to operate. Recalibration of the model can take several days or weeks
depending upon the level of precision required.
Comparing Scenarios
Multiple scenarios can be developed and compared. Detailed information regarding the
development of multiple scenarios was unavailable at the time of document publications. One
possibility is recalibration of the model; these protocols have not been developed yet.
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-processing Requirements
The expertise of an experienced modeler is needed to process the outputs since considerable
amounts of variable sensitivity analysis are required to understand the spatial scale of
interactions.
B-22
-------
NEXT STEPS FOR MODEL DEVELOPMENT
The pilot LTM was developed for the Saginaw Bay Watershed and incorporated two of the six
conceptual modules. Future versions of the model will, eventually, incorporate all six modules.
Currently, the LTM is being used to provide a preliminary assessment of the impact of 12 years
of development and land-use change in the Grand Traverse Bay, Ml watershed. Additionally, a
model based on the LTM is being developed to examine the changes in land cover patterns as
a function of socioeconomic changes, dispersed development, and subsequent changes in the
spatial patterns of land ownership in the forested regions of Michigan, Minnesota, and
Wisconsin.
B-23
-------
Land-Use Change Analysis System (LUCAS)
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
The spatial resolution and extent analyzed by LUCAS is defined by the scale of maps used for
input. A single grid cell, or pixel, may be defined to a resolution of 90m x 90m.
Temporal Resolution and Extent
LUCAS can provide forecasts for up to 100 years into the future at 5-year intervals.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
LUCAS is a spatially explicit modular system consisting of three modules linked by a common
database of driving variables. The first module contains the socioeconomic models that are
used to derive the transition probabilities associated with changes in land cover. These
probabilities are computed as a function of socioeconomic driving variables including:
transportation networks; slope and elevation; ownership; land cover; and population density.
The second module contains the landscape-change model and receives as its input the
transition matrix produced in module 1 (the socioeconomic models) and accesses the same
spatial database of driving variables. A single iteration of the landscape-change model
produces a map of land cover that reflects socioeconomic motivations behind human land-use
decision making. The third module of LUCAS contains the impact models. These models
utilize the land-cover maps produced by the landscape-change module to estimate the impacts
to selected environmental and resource-supply variables. Environmental variables include the
amount and spatial arrangement of habitat for selected species and changes in water quality
caused by human use. LUCAS was developed with the potential for additional ecological
impact or socioeconomic modules to be added at a later date.
Input Pre-Processing Requirements
All of the inputs maps must be raster files: discrete grids, or matrices, of numeric values,
corresponding to a square parcel of land called a grid cell (pixel). Maps can be stored in an
uncompressed (32-bit) format for each pixel or in the run-length encoding format. Each pixel in
each map is assigned to one of the data categories for each land use type. For example,
categories for vegetation data type could include: forest; unvegetated; or grassy/bushy. Maps
for a specific geographic region are compiled into a mapset. The resolution, or pixel size, for
each map within a mapset must be the same.
Model Assumptions
LUCAS assumes that landscape properties such as fragmentation, connectivity, spatial
dynamics, and the degree of dominance of habitat types, are influenced by market processes,
human institutions, landowner knowledge, and ecological processes. The concept of transition,
B-24
-------
or change, usually in land cover, from a given state to a new state, is central to the model.
Transition probabilities are derived empirically through a time series analysis of changes in land
cover, while considering road networks, population density, and physical attributes of the land
landscape.
Setting Parameters
Calibration of the model requires expertise in land-use modeling and "C++" language
programming.
Comparing Scenarios
Multiple scenarios can be compared by running multiple simulations varying the land-cover
dependent variable. The graphical and statistical outputs generated from each simulation can
then be compared.
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-processing Requirements
There are two types of outputs generated by LUCAS: statistical and graphical. The GRASS
map layers produced can be stored and re-analyzed in an iterative fashion. To use and
understand the statistical outputs, experience in land-use modeling and spreadsheet analysis is
necessary.
NEXT STEPS FOR MODEL DEVELOPMENT
LUCAS has been developed and implemented in the Little Tennessee River Basin and the
Olympic Peninsula. For additional information regarding further development of LUCAS, refer
to www.cs.utk.edu/~lucas.
B-25
-------
Markov Model of Residential Vacancy Transfer
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
Spatial resolution: 1 (or more) household(s).
Spatial extent: This model defines change in terms of movement between residential land-use
sectors that need not have a specific spatial extent.
Temporal Resolution and Extent
Usually 3 to 5 years; limited by Census data (U.S. Bureau of the Census).
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
The Markov chain model measures residential land-use changes in terms of the emergence or
disappearance of vacancies within various housing sectors (e.g., high-density rental
apartments, detached single-family dwellings, etc.). To do so, users must use linear algebra or
data analysis software of their choice to develop a matrix that characterizes the rate at which
vacancies by households in one sector (e.g., empty nesters moving out of their single family
home) result in vacancies in another sector (e.g., young family moving out of their apartment
and into the empty nesters' former abode). As residences are added or removed from the
housing market, one vacancy emerges when another is filled. The model projects how the
resulting vacancies cascade through the residential housing market.
Users who are familiar with operations research or Markov chain modeling should be able to
use the matrix to examine consequences of housing sector changes on vacancies potentially
far removed from the actual site of initial change. Others may need to have a computer
scientist with quantitative modeling experience to design a basic user interface or run a
numerical simulation to utilize the model.
Input Pre-Processing Requirements
A residential vacancy change matrix must be derived from the number of households changing
housing sectors overtime. This highly mathematical procedure is described in detail in Emmi
and Magnusson (1994), as well as by other authors. The procedure could be automated by a
programmer who is experienced in linear algebra, differential equations, and related stochastic
methods.
Model Assumptions
The model assumes that the urban housing market can be delineated from its "rural hinterland."
Internally homogenous housing sectors also are required, although the developers have
B-26
-------
relaxed this assumption. The model also assumes that the number of vacancies created and
transferred into a sector are equivalent to the number transferred from and absorbed out of that
sector, and that the transfer probabilities are constant (stationary). The model is also
dependent on the assumptions of all Markov models:
• Markovicity: The current state of the system depends on the immediately previous
state, and none earlier.
• Stationarity: transition probabilities don't change over time.
• Homogeneity: all state changes within a given sector are subject to a statistically
identical set of transition probabilities.
Setting Parameters
The following parameters must be set or calibrated in order for the model to operate:
• Sectoral categories
• Vacancy transition probabilities from one sector to the next
• Net housing created in each sector over each time step
Comparing Scenarios
The model addresses multiple scenarios as follows: Direct and indirect effects of different
distributions of vacancy initiations (i.e., new construction) on households' mobility and housing
demands across sectors.
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-Processing Requirements
None.
NEXT STEPS FOR MODEL DEVELOPMENT
Efforts are being pursued to integrate this model into the residential mobility component of an
integrated land use and transportation planning model. The dual solution to a linear
programming interpretation of the Markov model is being used to explore the (shadow) housing
price implications of pursuing each of several alternative urban growth scenarios.
B-27
-------
MEPLAN
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
MEPLAN may be customized to meet user needs and has been used to represent cities, as well
as complete countries. Depending on the scale and purpose of the study, the model zone can
vary from a few hundred metres in diameter to an entire country.
Temporal Resolution and Extent
Since MEPLAN is highly customizable to meet user needs, the temporal resolution is user-
specified. The output data can be modeled for any year in the past or present as well as run for
any time interval; however, 5-year increments are commonly used.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
MEPLAN is a stochastic model that is used to project and evaluate the many impacts that
planning decisions will have on land use and transport. To do this, it considers supply and
demand in both land use and transport and assumes that land use and transport affect each
other at all levels. For example, land-use activities, such as industrial development, retailing,
and residential expansion, create demands for industrial land, retail floorspace, and housing. In
turn, basic supply and demand theory influences prices for space in each land-use category. It
is the pattern of prices that influences where people live and work, which affects transportation
needs.
MEPLAN has been implemented as an integrated software package with multiple, functionally
distinct models and submodels. All of the input parameters and land-use categories are user
defined; MEPLAN uses a framework approach within which the user can set up categories and
the characteristics of the relationships between categories to represent whatever is required.
This maximizes the generality of the model.
The following basic steps generally are needed to implement MEPLAN:
• Data acquisition from various sources
• Model calibration
• Model validation
• Policy testing.
The output of the model can be spatially and graphically analyzed through MAPINFO network
and zone plots.
B-28
-------
Input Pre-Processing Requirements
The input pre-processing requirements vary and differ for each run of the model.
Model Assumptions
The developers of MEPLAN did not want to build in any pre-conceived ideas of what the user
wants to model and, therefore, did not specify any model assumptions. Working within a broad
framework, users may define their own assumptions as they implement the MEPLAN
framework.
Setting Parameters
The model parameters can be recalibrated by altering the input files for the model. A significant
amount of computing is necessary during the setting up and calibration of MEPLAN.
Comparing Scenarios
MEPLAN can address multiple scenarios, but the scenarios have to be coded separately as
individual scenarios. The model provides a very flexible system and scenarios are defined by
the user.
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-Processing Requirements
The output of the model can be spatially and graphically analyzed through Maplnfo network and
zone plots. Documentation indicates that post-processing may sometimes be required. The
type of post-processing needed was not specified.
NEXT STEPS FOR MODEL DEVELOPMENT
Planned future enhancements of MEPLAN include analysis of the sources of projection errors.
Additionally, topics for further model development are currently being explored.
B-29
-------
METROSIM
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
The METROSIM model covers an entire metropolitan area including its rural fringes. It can
include as many zones as necessary. The NYMTC-LUM version includes 3,500+ zones for the
New York/New Jersey metropolitan area. The zones can be as small as data permits. For
example, in Chicago applications, quartersections of 1/4 mile by 1/4 mile were used. The
structure of the model is such that it can be operated with many small zones or fewer large
zones.
Temporal Resolution and Extent
The METROSIM model can accommodate any number of time periods, but more than 30 time
periods are not recommended. Scenarios are run yearly, or some aggregation of years such as
2 years, 5 years, or decades.
METROSIM can produce a one-shot long run equilibrium forecast for transportation and land
use in a metropolitan area, or it can operate in annual increments and produce yearly changes
to transportation and land use from the existing situation until convergence to a steady state in
achieved.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
Transportation, land use, metropolitan growth inputs are used to produce outputs consistent
with economic equilibrium (market forces) subject to inputs, policies and land use restrictions to
be specified by planners or the users of the model. Cost-benefit measures are produced if
desired.
Input Pre-Processing Requirements
Normally, all preprocessing can be done by Alex Anas &Associates.
Model Assumptions
METROSIM does not incorporate any broad assumptions. Calibration procedure checks for
data inconsistencies, such as unavailability of sufficient land, etc.
Setting Parameters
METROSIM is calibrated using Census Tract or Traffic Analysis/Land Use Zone data. The key
data sets needed in a comprehensive application are:
B-30
-------
(1) The census transportation planning package (urban elements 1, 2 and 3);
(2) Data on the link-node destination of the metropolitan transportation networks by mode of
travel; and
(3) Data on real estate parcel characteristics and values (generally available from tax
assessors in many metropolitan areas).
When the third data set is not available, METROSIM can still be calibrated, although somewhat
more simply, using data sets 1 and 2.
Comparing Scenarios
METROSIM can address multiple scenarios of virtually any variety.
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-Processing Requirements
Normally, post-processing of model outputs by the user is not necessary. It is done by the
model. However, it may be required for some special purposes.
NEXT STEPS FOR MODEL DEVELOPMENT
Please contact the model developer, Alex Anas, at (716) 688-5816 or aanas@adelphia.net.
B-31
-------
SAM-IM
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
SAM-IM can address a regional level or a "microscopic" level (i.e., less than one acre). A
land use polygon is the basic unit of geography used in SAM-IM. It envelops an area of
homogeneous land use and density. In its most detailed scale, a land use polygon is a parcel.
In a more general application, it is a subdivision.
Temporal Resolution and Extent
SAM-IM forecasts future growth in a community using a user-defined future land use layers for
a scenario, as well as active and future developments. It will use this information, as well as
other factors defined by the user (e.g., undevelopable lands and redevelopment districts) to
perform land use allocations. SAM-IM can be used to create scenarios (i.e., conduct land use
allocations) for any historical period for which data are available.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
SAM-IM is primarily a growth model. It estimates growth and adds this estimation to
information on existing land uses to create a forecast of future land use. Either growth is
computed by comparing the forecast theme (i.e., with existing land use or growth) or is
computed by comparing the forecast with a base year DRAM/EMPAL (i.e., transportation
model) table.
SAM-IM is implemented as an integrated set of modules, each of which is designed to perform
a certain function. The modules are ArcView projects, or .apr files. The bullets provided below
explain the function of each of these modules and how they relate to one another.
• The Forecast module allows the user to specify and organize all of the socio-economic
forecasts.
• The Existing Land Use, Plan, Known Projects, and Other Lands modules all enable the
user to edit various types of land use polygons.
• The Scoring module allows the user to assign scores for vacant and redevelopable
lands that reflect the probability that the land will be allocated to a portion of the socio-
economic forecast.
• The Execute Forecast module allocates the forecast(s) to the future land use scenario
according to the rules developed in the other modules.
B-32
-------
• The Traffic Analysis Zone (TAZ) Dataset module produces new Travel Demand Model
(TDM) input datasets out of SAM-IM results to tie the land use planning process to the
transportation planning process.
Input Pre-Processing Requirements
SAM-IM requires some pre-processing of input data, as follows:
• All input land use information (e.g., existing land uses, planned land uses, active
development projects, etc.) must be converted from vector representations (i.e.,
polygons) to grid representation (i.e., raster cells).
• Active and planned development projects must be inserted into the plan that will
supercede what is shown in the general plan and generate a file containing a new "plan"
grid.
Model Assumptions
SAM-IM enables users to develop their own set of assumptions about land-use suitability;
projected growth demands (future population and employment trends and anticipated
development densities); and land use controls. The types of assumptions specific to each
module are identified below (Setting Parameters).
Setting Parameters
SAM-IM is "configurable" to the requirements of the local area, including land use coding
practices, forecast variables of interest, etc. SAM-IM automatically configures itself according
to project definitions provided in "configuration" files, which are prepared by the model
developer or the user.
The following parameters within each module of SAM-IM must be set or calibrated to operate
the model.
• Forecast Module
- Land use types
- Socioeconomic sectors (e.g., industrial employment, military, work at home, office
employment, etc.)
- Dwelling unit densities for each land use type
- Employment densities for each land use type
• Existing Land Use Module
- Existing land use information (i.e., a base land use condition)
- A previous scenario developed by SAM-IM that the user would like to select as the
base condition for generating a new scenario (i.e., land use forecast)
B-33
-------
• Plan Module
- Future land use layer for the scenario that represents a future vision of urban form
and defines how land in the future can be used and at what densities
• Known Projects Module
- Active developments
- Planned developments
• Other Lands Module
- Polygons that establish other "themes" for use in allocating land uses in a scenario,
such as undevelopable lands or redevelopment districts. They represent areas
where development will be prohibited from occurring based on whatever reasons
the user has in mind that are not otherwise reflected in the land use covers
themselves.
• Scoring Module
- Tools for generating GIS layers that reflect site potential based on the user's
standpoint -- and the availability of source data (e.g., flood plain information,
buffered highways, proximity to infrastructure, distance from urban lands, total
market size within distances of land etc.)
- Tools for creating, saving, and computing grids
- Tools for computing logit "choice" equations that reflect the probability that land will
be developed
• TAZ Dataset Module
- Grid cell size
- Forecast year to be developed
- TAZ and Land Use Sources
- Equations to use during the calculation of the trip generation variables
Comparing Scenarios
SAM-IM provides users with the ability to create project scenarios. A project scenario can
represent a new urban form alternative, or it can possibly represent different forecasting years
(e.g., 2000, 2005, 2010, 2020, etc.). The user interface for SAM-IM is organized around the
creation, modification and simulation of project scenarios.
The user may define scenarios by changing the information contained in each of the modules
contained in SAM-IM, such as existing land use, future land use, planned and active
developments, evaluation criteria, and regional forecasts.
B-34
-------
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-Processing Requirements
SAM-IM contains substantial capabilities for producing data sets for use by other urban
modeling systems, for example transportation models run by EMME/2. Land use allocations
produced by SAM-IM can be summarized for any unit of geography, whether it be traffic
analysis zones (such as used by transportation models), municipal boundaries, census tracts -
any statistical zone system that can be represented by polygon features. Further, SAM-IM
projections can be combined with other socioeconomic data represented by other geographic
features to produce data sets needed by other modeling systems. SAM-IM will format these
data sets according to the requirements of these other models, offering a range of output
format choices including ones that are completely user defined.
NEXT STEPS FOR MODEL DEVELOPMENT
A substantial enhancement program is planned for the next year. Among new features and
capabilities will be:
• Faster and more automated procedures
• Modeling scripts to redirect model procedures
• A sophisticated change in the data model to support mixed use developments
• Reallocation mechanisms for addressing allocation residuals
• Probabilistic allocations consistent with logit model formulations
• Development velocity curves to support known development projects already underway
• Simplifications to the "geographic calculator"
• Variable geographic resolutions to provide greater detail in downtown areas and less
detail in outlying areas
• Improved accessibility capabilities, including the capability for producing network travel
time matrices
• Direct support for DRAM/EM PAL
B-35
-------
SLEUTH
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
The spatial scale and extent of the SLEUTH model are user defined and, therefore, are
variable. Pixel sizes have ranged from 50 m to 1 km. Currently, national (48 contiguous states)
and local (Santa Barbara coast) data sets are being calibrated by the model.
Temporal Resolution and Extent
The SLEUTH model, using yearly intervals, can operate as far into the past or future as
available historical data will allow.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
The SLEUTH model allows users to see how urban areas grow over time by projecting urban
expansion. The model is a C program running under UNIX that uses the standard compiler
(cc). It contains a land cover deltatron model (LCD) that allows the user to conduct land cover
modeling. Users can customize the SLEUTH model for use with their own data sets. After
performing a three-phase calibration process, which allows the user to derive the best
parameters for historical modeling, the user runs the model to obtain projective information on
urban growth. The outputs provided by the SLEUTH model includes animations that illustrate
how urban areas grow over time and the impacts associated with growth.
This work is sponsored by the United States Geological Survey EROS Data Center and by the
National Science Foundation under the Urban Research Initiative. The SLEUTH model has
successfully projected urban expansion for communities on a regional level, including the San
Francisco Bay area and the Washington-Baltimore area. In the near future, the SLEUTH model
will project urbanization in other major metropolitan areas, such as Portland-Vancouver,
Chicago-Milwaukee, Philadelphia-Wilmington, and the New York metropolitan area.
Input Pre-Processing Requirements
Data layers must be on the same projection, same resolution, and map extent. All must be 8-
bit GIF files, with values set as data (not in the color table).
Model Assumptions
Assumes zoning and other policy-making does not alter overall coverage of urban growth. Only
"planning" factor is where roads are placed. The SLEUTH model can incorporate some zoning
alternatives using the excluded layer. The main assumption of the model is that the future can
be projected by the past, assuming historic growth trends continue.
B-36
-------
Setting Parameters
The automated calibration routines set initial parameters by calibrating the model from historical
data.
Five growth parameters must be set, -3 coefficients: 1) dispersiveness of growth, 2) growth at
new settlements (breeding), 3) outward expansion plus factors for, 4) the likelihood of
settlement up steep slopes, and 5) promoting new settlements near roads or transportation
networks. There are 10 additional system and "self-modification" (momentum) constants which
also must be set.
Comparing Scenarios
The SLEUTH model addresses multiple scenarios such as the default level of initial urbanized
areas, impacts of roads, and other growth factors.
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-Processing Requirements
None.
NEXT STEPS FOR MODEL DEVELOPMENT
The SLEUTH model's scale is being increased towards national and global coverage. Multiple
land use types are being introduced into the related "Deltatron" model. Predictive future horizon
is being extended as well.
B-37
-------
Smart Growth INDEX®
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
The size of the area evaluated by Smart Growth INDEX® is user defined; the user selects a
land-use cell ranging between 5 and 100 acres. Therefore, the model is suited to evaluate
small or large areas (e.g., community or region).
Temporal Resolution and Extent
Smart Growth INDEX® can operate in two different modes: it can provide grid-based forecasts
over time (yearly intervals up to 20 years) or it can provide a snapshot of a parcel at single point
in time.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
When operated in the "forecast" mode, Smart Growth INDEX® uses an exogenous population
forecast and user-selected policy constraints and incentives to determine where growth should
occur in the community. The results of this analysis are scored using a set of environmental
performance indicators. When operated in the "snapshot" mode, the model scores an
exogenous land-use plan or urban design using a similar set of environmental performance
indicators.
Model Assumptions
Smart Growth INDEX® assumes that population and employment growth are directly related to
a locale's accessibility to transportation and infrastructure services.
Input Pre-Processing Requirements
The following pre-processing of input data is required:
• All GIS inputs must be in ESRI shapefile format and must be provided at the smallest
available polygon levels. All coverages must have a common projection.
• Map units must be in feet.
• In cases in which multiple jurisdictions in a single region have multiple land-use
classification systems, they must be converted into a single land-use classification
before being imported into the model.
B-38
-------
Setting Parameters
The following parameters must be set or calibrated in order for the model to operate:
• Estimated resident, household, and employment growth for each interval year to a
maximum horizon of 20 years; commuteshed population; availability of rail transit; study
area urban character (CBD, urban, suburban, exurban); existing peak-hour and off-peak
levels of service on study area freeways and arterial; allowable number of lanes by
functional class; allowable densities of each land-use class; ratios of nonresidential
floor area to number of employees; ratios of residential to nonresidential uses in mixed-
use land-use classes; percent of infilling to be allowed on vacant lands; transportation
fuel consumption rates; building energy demand rates; transportation and building air
pollutant and greenhouse gas emission rates; and residential water consumption rates.
Comparing Scenarios
Smart Growth INDEX® addresses multiple scenarios by altering the parameters listed above
and modifying user-defined policy constraints and incentives that determine land availability for
development.
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-Processing Requirements
None.
NEXT STEPS FOR MODEL DEVELOPMENT
The ability to link to additional four-step travel demand models is in development.
B-39
-------
Smart Places
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
Smart Places is scalable and can be customized to meet study area needs of end user. Smart
Places allows the user to define the scope of the scenario to be considered for evaluation. For
example, the user can evaluate the entire design or interactively select land-use categories,
regions, or other boundaries within the design.
The spatial resolution of the model is based on the resolution of data provided by the end user.
Temporal Resolution and Extent
Smart Places does not include a temporal component.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
Smart Places looks at input to output relationships in "feature casts" each related to one or
more issues of interest. There are no bounds on either the feature or it's related linked
indicator definitions.
An example feature cast for solid waste is given below. Each feature like solid waste is
described by one or more indicators. Each indicator may be evaluated using one or more
measures. Each measure results in one or more linked indicators. Thus, the definition of each
set of feature relationships is given by its feature cast.
Example: Solid Waste Feature Cast
Solid Waste Indicators: Solid Waste Information Stock:
Residential Number of Structures by Type
Commercial Pounds of Solid Waste
Industrial Persons
Mixed Use Standard Industrial Classification
Public Land Area
Parks-Recreation Persons/Household
Utility Pounds Solid Waste/Process
Transportation Pounds Emissions/Ton of Solid Waste
Communications
B-40
-------
Solid Waste Link Indicators: Solid Waste Measures:
Combined Landuse Generation Rate/Landuse Type
Combined Energy Generation Rate/SIC
Combined Emissions Generation Rate/Capita
Combined Water Landfill Tons
Combined Wastewater Incineration Tons
Combined Solid Waste Compost Tons
Recycle Tons
Reuse Tons
Landfill Acres
Vehicle Miles/Ton
Vehicle Fuel/Mile
Input Pre-Processing Requirements
Certain data may require reformatting to accommodate the GIS.
Model Assumptions
Smart Places enables users to build their own set of assumptions about land use suitability,
land use controls, etc.
Setting Parameters
Smart Places requires the user to set various input parameters (i.e., attributes) to evaluate
alternative scenarios based on their goals and objectives. Required attributes will depend on
the evaluation models defined by the user and may include characteristics such as the number
of dwellings per acre, number of persons per unit, and kilowatt hours of electricity per unit per
month.
Comparing Scenarios
Smart Places allows the user to address multiple scenarios by changing various assumptions
and saving the results of alternative scenarios for comparison.
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-Processing Requirements
None.
NEXT STEPS FOR MODEL DEVELOPMENT
A partnership is being formed with Public Technology, Inc. to bring the model to cities in the
USA.
B-41
-------
TRANUS
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
The TRANUS model can be used for urban, regional, national or even international
applications.
The model is zone-based. Any number of zones may be defined at any degree of resolution.
For a metropolitan application the model is typically applied for disaggregations of Regional
Planning Districts, ranging from 60 to 250 zones. Model results may be further disaggregated
using a GIS-based procedure.
Even if the modeling system does not impose restrictions on the number of zones, there are
some obvious practical limitations. In the USA, very detailed data is available for population
and land use, but data on employment is less reliable at a finer scale and data on land rents is
usually coarse. The purpose of the study is also relevant in defining the spatial resolution of the
model. If the purpose of the study is to make strategic long-term projections, then a relatively
small number of zones may be sufficient. For short and medium-term projections looking into a
specific project in detail (e.g., a mass transit system), a finer degree of resolution may be
necessary. The amount of resources available for the study is also relevant. A quick strategic
exercise suggests a small number of zones, while a bigger and longer effort probably calls for a
large number of zones. Too many zones may become a nuisance, making it difficult to check
the results of the model and see any possible errors or inconsistencies.
Temporal Resolution and Extent
The temporal extent of the model is freely defined by the planner, there are no limitations.
There have been applications that consider 10 years into the past and up to 40 years into the
future. Others have focused on short-term projections such as 5 years into the future.
The time intervals for which scenarios are run are also freely defined by the planner, there are
no limitations. Five-year steps are the most common in past applications, with two and three for
short-term projections.
TRANUS deals with the temporal scale with a sophisticated 'scenario-tree' structure. Each data
item in the database has a unique time-scenario reference. After a base year scenario has
been defined, the first future scenario is connected to the previous one as a branch of a tree
and inherits all the data from it. There is no need to repeat or redefine information; only the
changes in each period-scenario have to be introduced. The same is true for subsequent
periods, which will be connected to the previous scenario in the tree. This arrangement is very
convenient when a project has many scenarios and time periods to deal with, minimizing the
possibility of errors.
B-42
-------
The scenario tree appears graphically in the user interface database. This structure is clear
and economic in terms of computer resources, because each scenario only has small amount
of data, since only the changes in relation to the previous scenario in the tree are stored. Any
number of periods, scenarios and branches may be defined.
The scenario tree appears in all windows of the interface. The model user may 'navigate'
between periods and scenarios to check for changes in the data, which are highlighted with
color codes. Data may be copied from one scenario to another. This may be done for a single
data item, for sets of data (e.g., all transit routes changes) or to a whole scenario. A network
view is available to show what elements have changed for a specific scenario.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
The TRANUS User Interface automatically generates all files needed by the model programs to
run, avoiding the annoying manual elaboration of input files with fixed formats and consequent
human errors. The database is object-oriented, which means that each data item is related to
all others in logical ways, guaranteeing consistency and minimizing the possibility of errors.
TUS also includes facilities for importing and exporting network data, a data validation
procedure and an unlimited 'undo' feature.
With the TRANUS User Interface (TUS) it is very easy to set a project for any person
familiarized with the Windows environment. Three menus: Project, Land Use and Transport
provide commands to create and edit the related entities. Each command opens an Edit Dialog
with a field for every data needed. A red color in a field indicates erroneous, inconsistent or
missing data. Additionally, a Validate command performs an overall validation of the database,
presenting a report of errors and direct access to correct them. TUS provides a context
sensitive help for all commands. Copy-and-paste facilities are provided to enable interaction
with spreadsheets. Linkjd and Zonejd codes are provided to facilitate interactions with GIS
systems.
Once the project has been defined in the database and the data for one scenario is complete,
input files are automatically generated and the user can run the model programs in a very
simple sequence. TUS presents assignment results numerically and graphically. A set of
complementary programs are provided to produce detailed reports in a variety of formats; the
reports may be opened by spreadsheets, where they can be organized at will, adding graphics
and prints.
Input Pre-Processing Requirements
The only pre-processing required, depending on the type of input, is arranging the fields in the
appropriate order in a worksheet and saving the file as comma delimited text.
Model Assumptions
There are no built-in assumptions about growth or possible changes of the variables into the
future. There are many ways in which growth may be specified for any variable in the system.
A distinction is made between exogenous and induced production. In urban and metropolitan
applications it is common to define some economic sectors, such as agriculture, manufacturing
and government, as exogenous to the model. For projections, the growth of these sectors may
be given by zone; alternatively, a study-wide growth may be specified (positive or negative),
together with a distribution function. A variety of function forms may be specified, such as
B-43
-------
linear, power or logit, with associated independent variables and parameters. Growth specified
through functions may be combined with zone-specific values.
In the case of endogenous variables, such as tertiary employment, households or land types,
the model uses the standard allocation procedure based on logit probabilities. The model user
may specify attractor functions to influence the location process in various ways, including
environmental quality indicators, neighborhood quality, crime rates, slope, floodplains and many
others.
For projections the model uses time-periods (cross-sectional) with an explicit dynamic scheme
as shown in the figure below. The projection period is divided into time periods t,, t2, t3,... with
intervals of, say, 5 years. Changes in the location and interaction between activities
automatically generate changes in the demand for transport at the same time period. Any
changes in the transport system, such as new roads or improvements in the transit system, also
generate changes in demand for the same period. These are short-term dynamics. Changes
in the transport system imply changes in accessibility patterns that affect or feedback into the
activities and land use system. In this case the model assumes that such changes take time to
consolidate, as residents and firms change their travel patterns. For this reason, changes in
accessibility affect the land use system for the next time period ti+1.
Finally, the horizontal arrows that link the activity system from one period to the next represent
elements of inertia. The model user defines which elements of the land use system affect or
restrict changes occurring in the next period. The model user also defines the time steps.
Setting Parameters
All the equation coefficients and parameters must be set or calibrated in order for the model to
operate.
Comparing Scenarios
The TRANUS model system provides a large number of indicators to support the comparative
analysis of alternative scenarios. A special program is specifically designed to compare two
sets of transport results corresponding to two alternative scenarios; the program estimates the
consumers' surplus derived from the application of policies.
The range of policies and projects that may be compared and analyzed is very wide. These
may be broadly classified as land use policies, transport policies, and combined land use-
transport policies. Any number of projects and policies may be combined to form a scenario.
The list of possible policies that can be represented in the TRANUS system is very long. The
following is in no way exhaustive but serves as an indication.
• Urban development plans, smart growth initiatives, metropolitan plans, etc.
• Land use controls
• Impact of specific projects such as new industries, housing developments, commercial
centers, etc.
• Regional development plans
• Housing programs and policies
• Environmental protection programs, constraints to preserve special areas, etc.
B-44
-------
• New roads or improvements to existing ones
• Reorganization of the transit system (new routes, tariff structures, integrated fares, etc.)
• Bus-only lanes
• Mass transit systems
• Toll freeways, urban or regional
• HOV lanes programs and facilities
• Demand management programs, car bans or limitations, etc.
• Pricing policies, such as road-pricing, fuel taxes, parking surcharges, etc.
• Park-and-ride
• Selective road-pricing and congestion pricing
• Road upgrading and rehabilitation
• Road maintenance programs and policies
• Railway projects and improvements
• New port facilities, relocation or improvement of existing ones
• New or relocation of freight and passenger airports
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-Processing Requirements
None.
NEXT STEPS FOR MODEL DEVELOPMENT
Modelistica has a permanent and continuous research and development program for the
TRANUS model. Included in the effort is a to-do list of future developments and new features
for the TRANUS model, covering all areas. A new version is released at least once a year, with
several sub-versions approximately once a month. As of May 2000, the most recent version is
January 2000.
B-45
-------
UGrow
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
UGrow's spatial extent is variable depending on the size of area explored. Its spatial scale is
also variable depending on available GIS data.
Temporal Resolution and Extent
UGrow's temporal extent is between 1,950 and 2,100, depending strongly on the time frame of
interest, with yearly scenarios provided.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
UGrow is a system dynamics suite of models for urban policy design and testing. Numeric
(system dynamics), spatial (GIS - maps) and 3-Dimensional (fly through visualization) are tools
which are integrated to serve a community's needs. Most jurisdictions prefer a limited numeric
model and may substitute population forecasts by housing type (per decade) to serve as an
input to the spatial model. The spatial model uses this information to determine how much
land must be developed to meet future housing needs. The spatial model then places this
development in accordance to "growth rules" selected by the user groups. These rules are
essentially the specific zoning a community may apply to generate a variety of growth scenarios
and may read: permit growth of single family on 1 acre parcels only within 1 mile of paved
roads, and allow multifamily only within 2 miles of town centers, and prevent any development
within any designated open space or riparian area, etc. Different factions within a community
may use this capability to define growth as each may prefer it and the model will generate
maps, by decade, into the future showing these differing scenarios.
The UGrow developers can pull any spatial scenario into the 3D model and fly this "community
of the future" showing new development and its relationship to the currently built environment.
Input Pre-Processing Requirements
UGrow requires the setting of all initial conditions, through extensive model focusing and issue
generation, before any inputting of data.
Model Assumptions
The UGrow model's assumptions are adaptable to the ones chosen by the user.
Setting Parameters
Confidence is built in the model structure by comparing the UGrow model output against
historic data.
Comparing Scenarios
The UGrow model addresses multiple scenarios. The particular scenarios addressed depend
on the issues of interest to the user.
B-46
-------
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-Processing Requirements
None.
NEXT STEPS FOR MODEL DEVELOPMENT
UGrow developers are currently developing an "Event Model" which can move across proposed
community growth scenarios. The event may be a tornado, drought or storm depending on the
attributes assigned and will highlight the risks of development in certain areas. This work is
NOAA/DOD funded.
B-47
-------
UPLAN
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
UPLAN can be customized to meet the study area needs of the end user. Once the model has
been operated, the information is saved to a file. UPLAN can incorporate previous files with
each new run. Therefore, the program theoretically does not have an upper limit, although a
consistent land use data set must provide the basis for UPLAN and larger areas may be more
complex as multiple jurisdictions and General Plan land use maps are included.
Low-density residential is represented in a 10 acre parcel size (200 m cells), while all other land
uses are represented by 1/4 acre parcel size (50 m cells). The user determines all cell sizes.
UPLAN is appropriate and cost effective for areas that are experiencing or anticipating growth
and have a significant amount of open space potentially available for development. The model
develops growth in a specific pattern that aligns itself with the roadway system, slopes,
services, and general plan designations.
Temporal Resolution and Extent
UPLAN uses 1990, or any year, as its base year and can accommodate different projection
periods. The time interval for which scenarios are run is determined by the user.
Temporal scale depends on the UPLAN module being considered. The dates for the growth
and allocation modules are user defined.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
When operated in the "forecast" mode, UPLAN uses an exogenous population forecast and
user-selected policy constraints and incentives to determine where growth should occur in the
community. The results of this analysis are scored using a set of environmental performance
indicators.
Input Pre-Processing Requirements
The following pre-processing is required for the inputs listed in the UPLAN fact sheet:
• Growth projections must be summarized so they can be hand-entered into "UPLAN" to
provide a user-specific model
• The community must determine level of compliance with the General Plan
• The community must define future infrastructure plans. Standard GIS overlay functions
must be used to create the various "Grids" used to study area
• Community must define existing infrastructure plans
B-48
-------
Model Assumptions
The UPLAN system enables users to build their own set of assumptions about land use
suitability; projected growth demands (future population and employment trends, assumed
household characteristics, and anticipated development densities); and land use controls and
infrastructure. The types of assumptions are listed below under "Setting Parameters."
Setting Parameters
UPLAN requires the end user to set various input parameters called "Grids" to evaluate
alternative urban growth scenarios. The grids are overlaid with each other to develop a specific
scenario. The user does not need to input parameters to operate the model, UPLAN has a set
of default parameters that can be used to run the model. The required parameters are listed as
follows:
• Attraction Grids (areas development can occur in the future)
- Freeway Ramps: This grid represents the location of the on/off ramps of a freeway.
The logic being these areas are highly desirable locations for development.
- Minor Highways: This grid represents desirable locations for development since the
highway is considered to have multiple access points.
- Major Arterials: This grid is considered to have full access, therefore, it is
considered a desirable location for development.
- Minor Arterials: This grid is considered to provide intersections with other arterials.
- Cities and their Spheres of Influence: This grid represents the city and the area
where development will likely occur in the future.
- Light Rail Stations: This grid provides location of existing and future stations and is
considered an attraction for development.
- Industrial Allocation attractions: Airports, Ports, etc.
- Factor weights: numeric values indicating the relative importance of different factors
for determining suitability, with the user setting the number of buffers and their
distance from the attractor.
• Exclusion Grids (areas development should not occur)
- Regional General Plan: This grid is a generalized composite of all general plan
land use maps for the user's location.
- Rivers/Lakes: This grid represents naturally occurring waterways. A user- defined
buffer area can be inputted.
- Floodplains: This grid represents the 100-year floodplain elevation.
- Slope: This grid provides areas too steep for development and a weighing factor
for varying slopes.
- Public Lands: This grid provides all areas designated as publicly-owned.
- Fire Severity: This grid provides levels of potential fire severity.
- Existing Urban: This grid allows user to test redevelopment policies at a detailed
scale. It is best if this layer is in a fine grid, such as 25 m cells, so that vacant lands
are accurately depicted.
- Open Space: This grid provides location of "public open spaces" this includes
parks, etc.
- Farmlands: This grid provides agricultural locations.
- GAP Vertebrate: This grid provides location of important habits. The WHR tables
associate animals with the vegetation types.
B-49
-------
• Allocating Demographic Land Use
- Residential: (1) population projection; (2) demographic and land use characteriza-
tion the user wants to review; (3) persons per household; (4) percent of households
in each density class; and (5) average parcel size for each density class.
- Industrial/Commercial: (1) all the above from residential; (2) workers per household;
(3) percent of workers in each employment class; and (4) average floor space per
worker and floor area ratio for buildings.
• Existing Urban Land Use
- User indicates previously developed areas.
Comparing Scenarios
UPLAN offers four modules that enable the user to modify the allocated land use:
(1) Strict Compliance: Requires the model to adhere to the land use designations in the
user's General Plan,
(2) Limited Compliance: Allows the model to modify the land use designations in the
General Plan to incorporate the previous allocated land use,
(3) Industrial Compliance: Requires all industrial land uses be delegated to industrial land
uses, whereas all other land uses can go anywhere, and
(4) No Compliance: Allows the model to delegate all land uses to any other land use.
These four scenarios provide the user with alternatives in determining future land uses.
Scenarios can be further defined by developing new road layers or transit layers, to attract
development to different locations, and be defining environmental constraints, such as urban
growth boundaries or habitat preserves.
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-Processing Requirements
Data summaries, per user needs.
NEXT STEPS FOR MODEL DEVELOPMENT
The next step is developing a series of urban impact models that will run off of the land use
layer (costs from wildfires, costs from flooding, local service costs, habitat damage, agricultural
lands losses, surface water quality, threats to groundwater, etc.).
B-50
-------
UrbanSim
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
The UrbanSim model covers the metropolitan region and surrounding area. The coverage area
is user defined and can include surrounding watershed areas or other ecological areas.
UrbanSim currently operates at level of a spatial grid of user-defined resolution. Current
applications of the model have used 150 meter resolution.
Temporal Resolution and Extent
UrbanSim operates on annual simulation steps, and can be run as far into the future as the
user requires, though accuracy clearly diminishes with longer time horizons. Capacity to
represent model uncertainty over time is being developed. The annual time steps of the model
are not a fixed feature of the model. Time steps could be varied to be longer or shorter.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
The use of the modeling system consists of three phases. The first phase, data preparation,
entails assembling and integrating the necessary data to load the database for the model. The
second phase, calibration, entails specifying the model components or using the existing
specifications, using tools provided in the model system to generate special data extracts
formatted to facilitate use in econometric software such as Limdep, and estimating the
coefficients of the model equations using the local database. Further calibration of model
coefficients is completed once all the model equations are estimated individually. The third
phase of the model use, application, involves the construction of policy scenarios that reflect
land use and transportation system plans, pricing and other regulatory policies towards
development, such as environmentally sensitive lands, or the use of Urban Growth Boundaries.
The model is run on individual scenarios, and the results of each scenario are exported on an
annual basis. These results can be visualized using the UrbanView component, or exported to
external database and GIS tools for further manipulation and analysis. Further work is planned
to develop policy performance indicators that provide summary measures to facilitate
comparison of scenarios.
Input Pre-Processing Requirements
UrbanSim requires some pre-processing of input data, as follows:
Creation of a spatial grid for the study area, at a user-defined resolution
Integration with environmental and policy constraints, using polygon overlay on the grid
Integration of parcel data, using polygon overlay to intersect with the grid
B-51
-------
Integration of business establishment data, using point in polygon assignment to link to
the grid
Integration of household travel survey data, using point in polygon assignment to link to
the grid
Processing of census STF3A and Public Use Microdata, to prepare inputs for household
synthesis.
Model Assumptions
The model assumes that external policy assumptions (such as a land use plan) will be binding,
and that underlying behavioral preferences calibrated in the model are invariant over time and
capture the most significant aspects of the relevant behavior of the real estate market demand
and supply interactions with travel. Work is being done to introduce a capacity to test the
degree to which land use plan policies are binding, based on historical analysis of development
and land use plan designations. These effects can be calibrated in the developer model using
local data. The model is designed to reflect probable outcomes of policy changes and general
trends in metropolitan development. It cannot be expected to predict with tremendous accuracy
at the level of the grid cell for two or more decades into the future, and users are likely to
generally use and distribute more summary forms of the model predictions. The grid detail
provides flexibility in aggregating results using different zonal configurations.
Setting Parameters
The following parameters must be set in order for UrbanSim to operate:
Land price model coefficients
Developer model coefficients
Residential location model coefficients
Employment location model coefficients
Mobility rates
Comparing Scenarios
The user interface for the model is organized around the creation, modification, and simulation
of policy scenarios. The user may define scenarios combining transportation infrastructure and
pricing, land use plans, density constraints, urban growth boundaries, development impact fees,
and policies regarding development of environmentally sensitive lands such as wetlands, high
slopes, or floodplains.
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-Processing Requirements
None. The output data can be loaded into external software for further analysis, as needed by
the user.
NEXT STEPS FOR MODEL DEVELOPMENT
The principal areas of current and future development of the model system are the following:
Model components to simulate land cover, water demand and nutrient load.
Model components to implement activity-based travel modeling within the system.
B-52
-------
Refinements to the UrbanView visualization component.
Development of a refined user-interface for the model and visualization system.
Development of Web-access tools for model visualization and potentially for model
operation.
Development of tools to facilitate user-specified evaluation indicators for the model and
visualization.
Development of calibration tools to internalize the capacity to estimate all model
parameters.
Development of a modeling language or building blocks to facilitate development and
modification of model components with minimal software coding.
Development of data mining and robust data cleaning tools to facilitate data integration
for the model system.
Development of approaches to integrating the model system with ongoing monitoring
and assessment of planning objectives and short-term planning and capital
improvement decisions.
Development of microsimulation of real estate market interactions, with prices emerging
from the interaction of consumers and suppliers.
The UrbanSim project is a long-term research project designed to provide tools to assist in
planning sustainable and equitable approaches to metropolitan development. Using an Open-
Source approach to software development, the project seeks collaboration with academic
researchers and with practitioners in the further development and application of these tools.
Those interested in potential collaboration should contact the project team at the University of
Washington.
B-53
-------
What if?
Additional Technical Information
GEOGRAPHIC AND TEMPORAL SCALE OF MODEL
Spatial Resolution and Extent
What if? may be customized to meet user needs, although it is best suited for sizes larger than
a single parcel of land. Because it incorporates user data, the resolution is that of the data
provided.
Temporal Resolution and Extent
Since What if? is highly customizable to meet user needs, the temporal resolution is user
defined, based on available information and study interests. The model can accommodate up
to four different projection periods.
APPROACH TO MODELING INPUTS FOR OUTPUT GENERATION
General Description
Each module of What if? applies user-defined decision criteria (e.g., identifying and weighting
land-use suitability factors) to base data (e.g., current land uses) to derive study-area
conditions. These decision criteria are applied to land-use information stored in geographic
information system (GIS) data bases to create maps and reports showing 1) the relative
suitability of different locations for various land uses under alternative development scenarios
(suitability analysis maps) and 2) where future development may occur (allocation maps). Each
module of What if? performs a specific function:
• The suitability module applies standard "weighting and rating" procedures based on
users' inputs for various land-suitability criteria and their relative importance to create
maps and summary reports showing the relative suitability of different locations for
different land uses.
• The growth module converts the five main categories of land use demand (i.e.,
residential, industrial, commercial, preservation, and locally oriented uses) into
equivalent future land-use demands based on projected growth for the study area. The
system user defines projected growth for each land use type (e.g., total number of
households, total regional employment by industry and commercial category) and the
system calculates the projected demand in each projection year for each land use type.
What if? can account for up to four future projection periods.
• The allocation module produces maps and reports indicating where projected growth
may occur given user-specified information on land suitability, land-use demand,
infrastructure provision, and land-use plans and controls.
Input Pre-Processing Requirements
The following pre-processing of input data is required:
• GIS Inputs: Standard GIS overlay functions must be used to create Uniform Analysis
Zones (UAZs) for the study area. UAZs are map layers of information on natural
features, infrastructure plans, existing land-use patterns, and approved comprehensive
B-54
-------
plans or zoning ordinances that are combined to create homogeneous land units. GIS
"UNION" commands are used to combine these layers into a single layer made up of
UAZs containing information from each of the constituent layers. UAZs are stored as
ESRI shapefiles for incorporation into What if?.
• Growth Projections: Growth projections must be prepared so they can be hand entered
into What if? as the following: 1) number of residential households per year; 2)
percentage of residential households per scenario; 3) housing density for each housing
type; 4) average household size for each housing type; 5) residential vacancy rates; 6)
proportion of existing housing units that will be lost to demolition, fire, and so on; 7) total
regional employment; 8) density of employees per industrial/commercial type; 9)
average square footage of floor space per employee; 10) industrial/commercial floor
area (FAR); 11) vacancy rates; 12) amount of land that should be preserved - total
amount, percent of study area; and 13) acres needed to meet various local demands
(e.g., recreational infrastructure) per 1,000 new people.
• Alternative Development Scenarios: The community must define its own suitability,
growth, and allocation scenarios.
• Land-Use Classifications: Community must define its desired land-use classifications.
• Infrastructure Plans: Community must define its future infrastructure plans.
Model Assumptions
The What if? system enables users to build their own set of assumptions about land-use
suitability; projected growth demands (future population and employment trends, assumed
household characteristics, and anticipated development densities); and land-use controls and
infrastructure. The types of assumptions specific to each module are identified below (Setting
Parameters).
Setting Parameters
The following parameters must be set or calibrated in order for the model to operate:
• Suitability Module
- Suitability factors such as slopes, soils, 100-year flood plain, flooding potential,
endangered species, stream buffers, and other factors that the user has data on
and wants to consider. These must be identified for each land use under
consideration.
- Factor weights: numeric values indicating the relative importance of different factors
for determining suitability.
- Factor ratings: numeric values indicating the relative suitability within a particular
factor type.
- Permissible land use conversions: identification of which land uses may be
converted from their current use.
• Growth Module
- Residential: 1) total number of households in the region; 2) the study area's share
of regional households; 3) breakdown of housing type for new residential
construction; 4) housing density for each housing type; 5) average household size
for each housing type; 6) residential vacancy rates; 7) proportion of existing housing
units that will be lost to demolition, fire, and so on.
B-55
-------
- Industrial/Commercial: 1) total regional employment; 2) density of employees per
industry/commercial type; 3) average square footage of floor space per employee;
4) industrial/commercial floor area; 5) vacancy rate.
- Preservation: amount of land that should be preserved in total acres and as
percent of study area.
- Local demands: estimates of acreage needed per 1,000 new people for
neighborhood retail, public/semipublic areas, parks and recreation, or other locally
determined need.
• Allocating Demand Module
- Allocation priority: order in which projected land-use demands are to be allocated.
- Infrastructure controls: identification of previously defined infrastructure plans,
indication of the type of infrastructure required for each land use type (e.g., user
indicates whether it is "not affected," "required," or "excluded").
- Land-use controls: user can indicate previously defined land-use plans and zoning
ordinances.
Comparing Scenarios
The model addresses multiple scenarios as follows: What if? offers three modules that enable
the user to: 1) determine the relative suitability of different locations for different land uses and
to specify the relative importance of alternative suitability factors and ratings for the different
factor types, 2) determine the projected demand for different land uses in each projection year,
and 3) allocate the projected demand for each land use to locations on the basis of their
suitability for that use and user-specified land use controls and infrastructure expansion plans.
Within each of these modules, different scenarios are possible. The scenarios are user defined
and used by What If? to determine future land-use patterns.
ADDITIONAL INFORMATION ON MODEL OUTPUTS
Output Post-Processing Requirements
None.
NEXT STEPS FOR MODEL DEVELOPMENT
Planned future enhancements of What if? include: 1) a version that can be operated via the
World Wide Web, 2) an interface with traffic demand models, and 3) a fiscal impact module.
B-56
-------
Appendix C
Current Trends in Community Growth and Planning
-------
-------
Current Trends in Community Growth and Planning
Lying at the heart of growth-management initiatives is the desire to control the rate of sprawl. Narrowly
defined, sprawl is the "unplanned, uncontrolled, and uncoordinated single-use development that does not
provide for an attractive and functional mix of uses and/or is not functionally related to surrounding land
uses..." (National Research Council, 1998). See Exhibit C-l for some of the characteristics that have
been used to define sprawl.
Many communities associate sprawl with traffic
congestion, environmental degradation, a loss of
heritage landscapes, a deteriorating sense of
community, central city and inner suburb decay, and
the accompanying isolation of disadvantaged
populations. These communities are witnessing the
loss of precious landscapes at an alarming pace, are
often subject to increased property taxes despite the
strong economy, and are spending more and more time
on congested roadways.
Communities around the country are trying new
planning tools and techniques in an attempt to curtail
some of the negative effects of sprawl. Through
innovation and the use of land-use change models,
they are finding ways to preserve community
character, revitalize downtowns, preserve open space,
and obtain an improved quality of life. In Maryland, a
Smart Growth initiative aimed at restoring downtown
economies, a sense of community, and the
environment, with a focus on placing funding
priorities on infrastructure investment, successfully
passed the ballot in 1997. Portland, Oregon, continues
to attract national visibility through the successful implementation of its growth boundary and its
expansion of Portland's light rail system. In Michigan, a statewide coalition of transportation experts
and grassroots activists launched an initiative to develop technically feasible alternatives to congestion.
By doing so, they were able to address the loss of farmland and concerns over the increases in traffic
congestion. More initiatives like these are occurring across the country.
Successful implementation of the above techniques, however, requires community members and planners
to have a better understanding of the links between land use, transportation, housing, and employment
location decisions. An understanding of land-use models that bring together these links will help
communities determine how to use such models to assess the implications of growth, project the
outcomes of different planning choices, and strengthen decision makers' abilities to effectively manage
growth in the interest of the community.
1.0 Suburbanization: A Snapshot
While sprawl has become a hot topic over the past few years, suburbanization, as a settlement form, has
had a long history in the United States. In the late 1860s, Riverside, Illinois, became a model for
suburban development-a model that subsequently has been copied in virtually every major city across
Exhibit C-1. Characteristics
Defining Sprawl
/* >
*- Low-density development
*• Development requiring dependence on
the automobile
*• Segregated land uses (e.g., com-
mercial, industrial, or residential)
*• Large distances and poor access
between housing, jobs, and schools
*• Consumption of land occurring at a
faster rate than population growth
*• Consumption of agricultural land/or
environmentally sensitive land
Source: U.S. General Accounting Office. April
1999. Community Development: Extent of
Federal Influence on "Urban Sprawl" Is Unclear.
C-l
-------
the country. Characterized by curving tree-lined streets and expansive lawns, this model still dominates
concepts of land subdivision for single-family detached units today.
By the 1920s, with rising affluence and the growth in automobile ownership, middle-income families
introduced the first massive residential migration to the suburbs. A second mass migration followed
the Great Depression and World War II in the 1940s. Perhaps the most famous suburban housing
development of this era was Levittown, New York. Levittown and other similar large suburban
developments were made possible by government-backed financing and mortgage insurance programs.
The 1956 Interstate Highway Defense Act further facilitated residential and industry relocation outside
central cities with the largest public works program ever completed in the United States and 60 billion
dollars' worth of investment toward the construction of an initial 40,000 miles of limited-access
highways. By the 1950s, most postwar development was occurring in suburban areas; population
decline and physical deterioration became some of the problems central cities had to face.
Traditional zoning separating urban land by residential, commercial, industrial, and agricultural uses has,
in recent years, been faulted for contributing to unsustainable development patterns and lowering our
quality of life through increased reliance on the private automobile and greater traffic congestion.
Ironically, traditional land-use zoning, despite its recognized shortfalls today, was first introduced in the
1920s as a means of protecting the health and safety of inhabitants of major cities. During the first
decades of the 20th century, urban residents living in dense tenements void of direct sunlight or air were
at great risk of fire and illness. Residents of smaller, but growing, cities often found themselves next
door to pollution-churning factories. Zoning, through the separation of incompatible uses and
specification of the density of those uses and bulk of building on the land, enabled cities such as New
York to better protect the health and welfare of its citizens. Today, with our economy based on cleaner
high-technology and service industries, there is less need for universal application of traditional zoning
laws. Nevertheless, there is a continual push toward the suburban form of development, encouraged in
large part by interstate and intrastate highways providing greater accessibility to the outlying areas.
Citizens still desire to own homes on quarter-acre lots away from the congested roadways and dense
development that are characteristic of urban centers. Unfortunately, along with this desire, come longer
commutes and negative impacts on the natural environment.
2.0 Changing Trends
Across the country, sprawl is providing benefits, such as increased home and business ownership due to
lower property costs, but it is also taking its toll on environmental and fiscal resources and is putting
great strains on our quality of life. The costs of delivering public services to newly developed areas, for
instance, is going up rapidly. Not only does sprawl consume open space as it moves outward from a city,
but it is doing so at a rate much faster than population growth. For example, in the Philadelphia
metropolitan area, population grew by a mere 3.8 percent between 1970 and 1990, while the amount of
land in the region used for urban purposes grew by 36 percent (Katz, 1997). On a national scale,
between 1950 and 1990, urbanized land expansion grew at three times the rate of population growth
(Rusk, 1999).
The American Farmland Trust examined the U.S. Department of Agriculture National Resources
Inventory data spanning the decade from 1982 to 1992, and concluded that almost 14 million acres of
America's farmland were lost to development, 31 percent of which was prime farmland (Sorenson et al,
1997). In the Washington, D.C., metropolitan area, more than 200,000 acres of farmland, barren land,
forests, and wetlands were lost during the building boom decade of the 1980s; this area is nearly five
times the size of the District of Columbia (National Center for Resource Innovations, 1997).
C-2
-------
Residents in many metropolitan areas rank traffic congestion as one of the most serious local problems.
The Texas Transportation Institute estimated that, in the country's 70 largest urban areas, the total cost of
traffic congestion amounts to almost $74 billion per year with 88 percent of that cost attributed to delay
(Schrank and Lomax, 1998). Annual costs per individual driver are as high as $1,290 in the Washington,
B.C., area (Schrank and Lomax, 1998). On a national level, drivers stuck in traffic consumed more than
6 billion gallons of fuel in 1996-enough to fill 670,000 gasoline tank trucks. Drivers from the most
congested cities use five times more fuel than those in the least congested cities (Schrank and Lomax,
1998).
While staggering, these numbers are not entirely surprising given that the increase in total vehicle miles
traveled was double that of population growth between 1970 and 1996 (Bureau of Transportation
Statistics, 1998). This increase in travel, combined with larger vehicles, has largely offset gains made by
catalytic converters and improved engine technology. Although mandated by the Clean Air Act, 130
U.S. cities still fail to meet National Ambient Air Quality Standards for ozone, carbon monoxide, sulfur
dioxide, or particulate matter pollutants (USDOT/USEPA). Nationwide, 25 percent of all air pollution is
the result of vehicular travel (USDOT/FHWA, 1999). Each year, the average car emits over 600 pounds
of air pollution (Washington State Department of Ecology, 1997).
Air quality is not the only natural resource concern of communities. Water quality is inversely related to
the land area devoted to impervious surfaces (parking lots, rooftops, and roads). As little as 10 percent
impervious surface area can lead to degraded streams and rivers in a watershed (American Rivers
Association, 1997). The increased flow of runoff and accompanying chemical constituents not only
affects aquatic plants and animals, but also hastens erosion, alters the shape of stream beds, and changes
the temperature of the stream, taxing the survival of certain aquatic species.
The same forces that have shaped suburbanization have led to disinvestment of our central cities and
isolation of our most vulnerable population-the urban poor. As more jobs move out into the suburbs,
people who cannot afford cars have difficulty reaching employment areas on transit systems that do not
adequately service low-density suburban areas. Public dollars that could otherwise be spent on
maintaining central city infrastructure and addressing the needs of urban citizens are absorbed by newly
developing areas. Too often, low income and minority populations of these central city communities are
further disenfranchised by the siting of hazardous waste incinerators, regional transportation arteries, and
other undesirable infrastructure.
Hand-in-hand with these discouraging effects of sprawl and suburbanization are dramatic fiscal impacts.
Increasingly, localities are realizing that certain types of growth do not always pay their way. New jobs
mean new residents. New residents need new schools, water and sewer lines to connect to their new
homes, and new fire and police stations for public safety. The Maine State Planning Office calculated a
60-percent increase in local government spending per household (approximately $1,700) from 1980-81
to 1990-91 that can be attributed to rapid low-density growth. More than $300 million was spent on new
school construction in fast-growing areas between 1975 and 1995 despite a drop in student enrollment
(PAS, 1999).
3.0 Initiatives
An increasing awareness of the potential costs of sprawl has prompted citizens, planners, and local
government officials to reassess where we are and where we want to be. A desire for more sustainable-
development patterns and a high quality of life have instigated a number of new initiatives that are
addressing sprawl-related problems. These solutions have risen from the local, state, regional, and
C-3
-------
federal levels of government. Because each initiative is unique to a particular sprawl situation,
communities should recognize that their individual planning strategies must be based on their own legal
authorities, political environments, and cultures. The state, local, regional, and national initiatives
discussed below are a few of the sustainable actions currently being implemented across the country.
3.1 State and Local Initiatives
By structure of the U.S. Constitution, land-use planning and zoning are functions of state and local
governments. Thus, significant changes in urban development patterns are highly dependent on local
initiatives. The states and local governments deal with the problems attributable to sprawl in different
ways, from regional growth boundaries and zoning to fiscal and tax-base measures. The discussion
below summarizes two of the more prominent programs: 1) land-use planning and zoning, and
2) adequate public facilities tests.
Land-Use Planning and Zoning
Communities have long attempted to shape development through zoning and other planning tools. The
backbone of planning in the United States is the general or comprehensive plan supported by the zoning
ordinance. This type of plan provides a long-range view of a community's projected population growth
and land use, transportation, housing, utility, and recreational resource and expansion needs. Most plans
include a statement of goals and policies that articulate a community's vision for its future. While
general in nature, the comprehensive plan should guide a jurisdiction's zoning, subdivision regulation,
budget, and capital improvements plans.
Zoning is the predominant land-use control employed by local governments. Local governments track
their land resources through an inventory of land area by designated use. One popular land-use
classification system adapted by many localities around the country is the Anderson system, in which
more general land-use classes, such as urban and nonurban, are further broken down through a hierarchy
of increasingly detailed levels (Kaiser et al., 1995). Exhibit C-2 illustrates how the Anderson
classification can be adapted for a county or small city.
Exhibit C-2. Adaptation of Anderson Land-Use Classification System
for County or Small City
Urban
Nonurban
Residential
Commercial
Industrial
Agriculture
Forest
Low-density
Medium density
High density
Mixed-use (commercial/residential)
General retail (central business district, shopping center)
Professional services and office
Light industry/warehousing
Heavy industry
Industrial/research park
Cropland
Orchard
Deciduous
Evergreen
Source: Kaiser etal., 1995
C-4
-------
Traditional zoning separates land uses into four primary categories or classifications: residential,
commercial, industrial, and agricultural. While it continues to be universally implemented across the
country, traditional zoning has in recent years come under scrutiny as a contributor to sprawl. This form
of zoning separates homes from places of work, leading to longer commutes and empty downtowns after
the evening rush hour. New zoning tools such as planned-unit and transit-oriented development, overlay,
and floating zoning are being used to encourage new development patterns, especially mixed-use
development with greater flexibility in residential and commercial densities.
Beyond comprehensive planning and zoning, various communities and states have adopted additional
tools to manage growth. One such tool is urban growth boundaries. Urban growth boundaries
concentrate development inside a mapped boundary line and preserve agricultural land outside. The
best-known example of a successful urban growth boundary is Portland, Oregon. In 1973, the state of
Oregon passed legislation requiring that all urban areas designate growth boundaries to keep sprawl in
check. Between 1980 and 1989, more than 90 percent of Oregon's new residents were located inside
Urban Growth Boundaries (Weitz and Moore, 1998). Portland complements its land-use planning with a
nationally recognized light rail transit system and encouragement of higher-density mixed-use
developments around transit stations. Between 1990 and 1995, annual transit trips per capita increased
4.4 percent during a period when ridership on similar transit systems elsewhere dropped (USDOT, 1997).
Adequate Public Facilities Tests
Some communities have introduced tests of adequate public facilities and focused investment planning as
a means of addressing a community's capacity to service additional development. In Montgomery
County, Maryland, subdivisions will not be approved unless developers demonstrate that adequate
facilities exist to service the proposed development. These facilities include roads and public
transportation facilities, sewer and water service, schools, police stations, firehouses, and health clinics.
A growth policy report prepared annually guides the County planning board's implementation of the
adequate public facilities ordinance. Maryland's recently adopted Smart Growth legislation focuses state
investments on county-identified priority funding areas as a means to encourage development where
infrastructure is already in place and in locations consistent with the Smart Growth and Neighborhood
Conservation Act.
3.2 Regional Initiatives
While politically difficult to implement, regional initiatives are particularly important because local
governments are not empowered or motivated to influence regional land-use patterns in accordance with
metropolitan-wide growth management objectives. The Minneapolis-St. Paul metropolitan area has had
regional property tax-base sharing in place since the mid-1970s. Under this arrangement, localities pool
40 percent of their tax increase over 1991 assessment for commercial and industrial properties. These
combined revenues are redistributed among the localities according to the population and overall tax base
of each. Since tax-base sharing began, Minneapolis has moved from a beneficiary to a contributor,
reflecting its economic revitalization success, while St. Paul has reduced its share of the recipient pool.
Overall, this revenue-sharing has ameliorated the per-capita disparity between the area's richest and
poorest communities fivefold (Nelson and Duncan, 1995). Despite complaints from suburban
jurisdictions, local officials have stated that, without it, no meaningful regional land-use and
development controls could be imposed. The Minneapolis-St. Paul area has recently strengthened
regional planning by placing all sewer, transit, and land-use planning under one operational authority,
transforming the Met Council from a $40-million-a-year planning agency to a $600-million-a-year
regional government (Orfield, 1997).
C-5
-------
3.3 National Initiatives
The intent of the following national initiatives is not to replace state and local land-use planning control,
but to lend support to local Smart Growth initiatives, demonstrate new alternatives, and articulate the
potential cost of allowing current development patterns to grow uncontrolled.
President's Council on Sustainable Development
In 1993, President Clinton created the President's Council on Sustainable Development (PCSD). Until
its expiration in June 1999, the PCSD advised the president on new approaches to integrate economic,
environmental, and equity issues. The Council comprised a partnership of leaders from business,
multiple levels of government, and community, environmental, labor, and civil rights organizations. The
guiding mission of the PCSD was to accomplish the following:
• Forge consensus on the identification and development of innovative economic, environmental,
and social policies and strategies
• Demonstrate how policy can be translated into actions that foster sustainable development
• Increase the visibility of sustainable development
• Evaluate and report on progress by recommending national, community, and project-level
frameworks for tracking sustainable development
Under the umbrella of the PCSD, the Metropolitan and Rural Strategies Task Force addressed the central
components of sustainable development-land use, ecosystems, transportation, public safety, and
affordable housing-by encouraging local and regional collaboration among federal, state, and local
government agencies; public interest and community groups; and businesses in both metropolitan and
rural communities.
The Clinton-Gore Livable Communities Initiative
In January 1991, Vice-President Gore initiated a comprehensive Livability Agenda to strengthen the
federal role in support of state and local community-building efforts aimed at ensuring a high quality of
life and sustainable economic growth. As part of this initiative, the Clinton Administration has proposed
that billions of dollars be directed toward investments that support livability programs. Better America
Bonds, increased mass transit funding, the promotion of regional collaboration, and school planning are
among the major funding areas. More than a million dollars would be dedicated to the purchase of open
space, in addition to tax incentives rewarding farmland and parkland preservation. At the same time, the
Administration has promised to modify those federal subsidies, such as highway funding, that encourage
sprawl. The Livability Agenda integrates the commitments of more than a dozen federal agencies.
Transportation Equity Act for the 21st Century
Some of the alternative transportation investment strategies outlined by the Livability Agenda are funded
through the Transportation Equity Act for the 21st Century (TEA-21) that was signed into law in June,
1998. For instance, TEA-21 provides $41 billion over the 6-year authorization for transit programs. This
is $10 billion more than that provided under the Intermodal Surface Transportation Efficiency Act
(ISTEA), the previous surface transportation bill. The legislation strengthens environmental protection
efforts through the continued funding of key ISTEA programs. The Congestion Mitigation and Air
Quality Improvement program provides a flexible funding source to state and local governments for
transportation projects that help them meet requirements of the Clean Air Act, whether they be transit
projects, travel-demand management strategies, traffic-flow improvements, or public fleet conversions to
cleaner fuels. Expanded provisions of TEA-21 enable more funds to be used for bicycle facilities and
C-6
-------
pedestrian walkways, as well as educational programs that address pedestrian and bicycle safety
considerations. Other changes in the law ensure the consideration of bicyclists and pedestrians in the
planning process. TEA-21 authorizes more than a million dollars for technical assistance and grants to
states for the purposes of developing scenic byway programs. The Transportation and Community and
System Preservation Pilot program is a comprehensive initiative of research and grants to investigate the
relationship between the efficiency of the transportation system and community and environmental goals.
Beyond these transportation investments, TEA-21 reaffirms earlier surface transportation policy through
its requirement of a nexus between transportation and land-use planning.
Federal Environmental Regulations
While not directly meant to affect development and growth, environmental laws are in place that
increasingly necessitate growth management so communities can remain compliant with the law.
Presented below is a brief overview of several of these:
• Clean Water Act (CWA). This law focuses on controlling discharges of pollutants to water
bodies, including oceans, lakes, rivers, and streams, as well as wetlands. Through this regulation
and associated regulatory programs, communities must consider how activities associated with
different land uses, particularly commercial, residential, and industrial land uses, will affect their
local water resources. Many of the programs implemented under the authority of the CWA, such
as the storm water and nonpoint source programs, advocate the preservation of open spaces
within communities as an effective means to control the amount of pollutants washed from the
land into the water.
• Coastal Zone Management Act/Coastal Zone Act Reauthorization Amendments
(CZMA/CZARA). The U.S. coastline historically has attracted settlements, resulting in a wide
range of uses from industrial and port developments to small fishing villages and high rise
condominiums. Preservation and protection of coastal resources, such as beaches, dunes, and
wetlands, are the foremost objectives of the CZMA. Communities within states that have a
comprehensive coastal zone management plan are encouraged to implement management
measures that will prevent the degradation of coastal waters and surrounding ecosystems. The
1980 amendments to the CZMA encourage "special management area planning" to identify in
advance those critical ecological resources that need protection. Those designated areas benefit
from special protection and programs that serve to manage conflicts between development and
resource conservation (Kaiser et al., 1995). An example of this form of critical area analysis is
Maryland's designation of all land within a thousand feet of the Chesapeake Bay as critical areas
and subject to special protection.
• Safe Drinking Water Act (SOWA). A strong connection exists between land development and
the quality of drinking water. The SDWA acknowledges this connection by promoting
watershed protection as a means to protect ground and surface water sources of drinking water.
Recent amendments to the SDWA require states to devise and implement source water
assessment plans that examine potential land uses that may degrade drinking water sources.
Communities use this information collected through the assessment to improve protection of
their drinking water supplies by developing a source water protection program. Through
improved understanding about the location of and potential impacts to drinking water supplies,
communities can use land-use restrictions, such as zoning, to control development within a
watershed and potentially reduce the cost of drinking water treatment.
• Clean Air Act (CAA). Good air quality is an important consideration in determining a
community's quality of life. Clean air affects not only human health, but also the aesthetics of a
community. The Clean Air Act regulates the amount of air pollutants released by a wide range
C-7
-------
of sources, including motor vehicles, factories, and small businesses. Each state has goals it
must reach for different air pollutants. When these goals are not met, a state must develop and
implement a plan to reduce the levels of air pollutants. Communities play an important role in
achieving the air quality goals contained in state implementation plans. Efforts to reduce air
emissions often cause communities to consider alternatives to traditional development for their
air quality benefits, such as investment in public transportation systems versus the creation of
more or wider streets.
Resource Conservation and Recovery Act (RCRA). The focus of this regulation is the proper
management of solid and hazardous wastes. RCRA requires thorough documentation of these
materials from "cradle to grave." Through RCRA, a community is able to determine the location
of firms producing hazardous materials and where haulers dispose of them. This information,
such as the location of improperly disposed residues from chemical production, can help
communities make decisions about emergency planning, infrastructure investments, and future
land-use controls to reduce the risk associated with hazardous and solid waste materials.
Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA).
This regulation, often referred to as Superfund, addresses the problem of improperly disposed
wastes and the contaminated lands left behind. Superfund comprises many components,
including determining who is responsible for creating the waste site, conducting an assessment of
the hazardous substances on a site, and implementing cleanup of the site. Communities with
Superfund sites listed on the National Priorities List often struggle with the issue of how to use
these vacant lands. The Brownfields program initiated under Superfund helps communities turn
these vacant or underutilized sites into productive properties. Safe redevelopment of former
Superfund sites allows a community to infill existing developed areas and prevent the pressures
of growth from consuming undeveloped lands or greenfields.
National Environmental Policy Act (NEPA). Mitigating the environmental impacts of federal
actions, such as the siting of federal facilities and the issuance of permits and licenses, is NEPA's
focus. By requiring the preparation of environmental impact statements (EISs), the federal
government examines the likely consequences of proposed actions and evaluates several feasible
alternatives. The review process of an EIS allows the public to voice objections to a proposed
federal action. Such objections can lead to a modification or cancellation of proposed projects,
or a significant delay, thus influencing community development decisions.
Endangered Species Act (ESA). Compliance with this regulation often requires land-use
planning and regulatory mechanisms to conserve the ecosystems on which the species depends.
Through the development of Habitat Conservation Plans (HCPs), parties involved in proposed
land cover conversion must determine the impacts on the ecosystem of concern, identify steps to
minimize those impacts, and describe alternatives to the proposed changes. The process of
negotiating and implementing HCPs can alleviate development pressures through the creation of
habitat preserves.
Executive Order 12898 on Environmental Justice. President Clinton issued Executive Order
12898 on February 11, 1994, to establish environmental justice as a national priority and ensure
that "all communities and persons across this Nation should live in a safe and healthful
environment." This was the first presidential effort to direct all federal agencies with a public
health or environmental mission to make environmental justice an integral part of their policies
and activities. The Order focuses federal attention on the environmental and human health
conditions of minority and low income populations-those most susceptible to industrial,
hazardous waste, and large transportation infrastructure sitings. The U.S. Department of
-------
Transportation is now focusing more attention on understanding and addressing the needs of
these more transit-dependent populations who often find it difficult to access growing
employment opportunities in suburban areas.
4.0 The Future of Community Planning
Communities are discovering innovative ways to manage growth so as to enhance its benefits and
minimize its costs. In 1999 alone, approximately 1,000 state land-use reform bills were introduced in
legislatures across the country (American Planning Association, 1999). Approximately 200 of these bills
have been enacted into law, tightening land-use laws, authorizing more innovative and flexible land-use
controls, and reforming "business-as-usual" processes. Supported by other local, state, regional and
national initiatives, this modernization of planning will provide new opportunities for Smart Growth.
In addition, an increasing number of communities are investing in geographic information systems (GISs)
to inventory their land uses, infrastructure, natural and cultural resources, and to automate their mapping
needs. A lesser number of these communities are taking full advantage of GIS as an integrated land-use
planning and information management tool. Much of the land-use modeling being done today is
performed by regional metropolitan planning organizations that largely emphasize transportation system
design. Nonetheless, with technology rapidly expanding in scope and availability, more communities
will turn to more sophisticated GIS analysis and land-use modeling to meet their analytical needs of
today and tomorrow. Through the use of these models, communities can begin to assess the implications
of growth, project the outcomes of various planning options, and, ultimately, manage growth in a smarter
way.
C-9
-------
C-10
-------
Appendix D
Key Terms and Definitions
-------
-------
Key Terms and Definitions
The descriptions of the land-use change models summarized in this guide contain terms commonly used
in professions such as computer modeling, land-use planning, socioeconomics, transportation planning,
and environmental protection and restoration. This appendix provides "plain English" definitions of
terms used throughout the guide.
Aggregate Modeling
Algorithm
Attribute
Basic Industry
C
C++
Calibration
Causal Models
Census Tract
An approach to travel demand modeling that employs large population
aggregates, defined in geographic, social, or economic terms, as the
fundamental unit of analysis. In a typical application (such as the
regional network-based travel models that rely on coarse-grained zone
systems), the variation in key characteristics (such as income and
household size) between population aggregates is less than the internal
variation subsumed within population aggregates.
A step-by-step approach for computing a solution to a mathematical
problem. Solutions to some mathematical problems may be computed
by applying any one of several alternative algorithms; the solutions will
not necessarily be identical. For example, comparison of traffic
assignments computed with different algorithms generally will reveal
different numbers of vehicles assigned to a link
A piece of information describing a map feature. The attributes of a Zip
Code, for example, might include its area, population, and average per
capita income. Attribute data is one of the two main types of data in a
GIS, the other being spatial data.
The export sector that sells its goods and services to consumers outside
the metropolitan area.
A computer programming language.
An enhancement of C that supports object-oriented programming.
Testing and tuning of a model to a set of field data not used in the
development of the model. Also includes minimization of deviations
between measured field conditions and output of a model by selecting
appropriate model coefficients.
Attempt to define relationships among system elements. As with time
series analysis, past data are important to causal models. A causal model
is the most sophisticated kind of forecasting tool. It expresses
mathematically the relevant causal relationships. It is, in essence, a
mathematical description of the underlying process. Hence, the purpose
is not merely to project, but also to explain the process. Regression and
econometric models are examples of causal models.
A small, relatively permanent statistical subdivision of a county. Census
tract boundaries normally follow visible features, but could follow
governmental unit boundaries or other nonvisible features as well. A
census tract may contain anywhere between 2,500 and 8,000 people.
D-l
-------
Census Summary Tape U.S. Bureau of the Census' 1990 Census Data on population items such
Files la (STFla) as age, race, sex, marital status, Hispanic origin, household type, and
household relationship. Population items are cross-tabulated by age,
race, Hispanic origin, or sex. Housing items include occupancy/vacancy
status, tenure, units in structure, contract rent, meals included in rent,
value, and number of rooms in housing unit.
Census Summary Tape
Files 3a (STF3a)
Correlation
Demographics
Deterministic
Disaggregate Models
Econometric Model
Empirical
Expert System
U.S. Bureau of the Census' 1990 Census Data on population and
housing items, including age, mobility limitation status, occupation,
class of worker, place of work, educational attainment, poverty status,
employment status, private vehicle occupancy, family type, race, farm
and nonfarm population, residence in 1985, school enrollment, group
quarters, Hispanic origin, sex, household type and relations, travel time
to work, income in 1989, urban and rural population, industry,
veteran/military status, means of transportation to work, workers in
family in 1989, value of housing unit, mortgage status, vehicles
available, occupancy status, year householder moved into unit, year
structure built, and rent.
A statistical measure of the extent to which two variables behave alike or
are related.
The statistical characteristics of a population (e.g., income, education,
race, and home ownership).
An approach to problem solving that quantifies the results, using an
exact model, also referred to as an exact mathematical model. For
instance, estimating the time it takes a car traveling at 50 miles/hour to
travel 50 miles can be solved using a deterministic approach:
The exact mathematical model is: time (hours) = distance (miles) /
velocity (miles/hour) and the exact (deterministic) solution is one hour.
The same problem can be solved using a stochastic approach.
In common usage, models developed to represent the behavior of
individual decision makers (e.g., persons, households, firms).
A more sophisticated regression approach often involving a system of
interdependent regression equations that describe some sector of
economic sales or profit activity. Although expensive, it better
expresses the causalities involved than an ordinary regression equation
and, hence, should project more accurately.
To rely or base something on observation (data). An empirical approach
uses existing observation/data to develop relationships to solve a
problem (i.e., there is no hard science involved).
A modeling approach that incorporates human judgment and expertise,
both quantitative and qualitative, in a decision-oriented framework.
D-2
-------
Feature
Geographic
Information System
(CIS)
Growth Controls
Hedonics
Impact Fees
A map representation of a geographic object. Store sites, customer
locations, streets, census tracts, and Zip Codes are examples of map
features. Features are drawn as points, lines, and polygons in ArcView
GIS.
A configuration of computer hardware and software that stores, displays,
and analyzes geographic data. See Section 3.2 of this guide for a more
detailed explanation and illustration.
Tools used to manage growth through limits placed on land use, bulk,
and density that include down-zoning, open space acquisition, annual
permit limits, and urban growth boundaries.
Of, relating to, or marked by pleasure. With respect to land-use models,
hedonics refers to quality-of-life indicators.
A mechanism used by local governments to offset the impacts a new
development will cause. Impact fees are charged to developers seeking a
building permit.
Input-Output Analysis A method for analyzing the uses of capital and labor, the disposition of
goods, and the flows of money in an economy within a given spatial
setting, and for obtaining a picture of the distribution of economic
activity within a region and with respect to a system of regions, through
a matrix of coefficients which relate inputs to outputs. Essentially
derived from a revenue expenditure accounting system.
Land Suitability
Analysis
Layer
Linear Regression
Analysis
Determining where it is appropriate to locate new development based on
an array of factors (e.g., soil type, slopes, desired preservation areas)
often determined by the community.
A set of related map features and attributes stored as a unique file in a
geographic database. A GIS can display multiple layers (e.g., counties,
roads, and hamburger stands) at the same time. See Exhibit 3-1 for an
illustration of regional GIS layers.
A type of regression analysis in which the functional relationship
between two or more variables is described by a straight line, as opposed
to a curve. Linear regression using the least squares method (defined at
regression) is a procedure sometimes used to arrive at trip production
and trip attraction rates as a function of land use or household
characteristics.
D-3
-------
Logit
Lowry Model
Markov Process
Metadata
Multinomial Logit
Methods
Nested Logit
Non-Basis Industry
Nonpoint Source
Pollution
Nutrient Loading
Origin-Destination
Pairs
A choice model formulation based on the principle that individuals
maximize utility in choosing among available alternatives. The logit
formulation involves specifying a utility function for each individual,
with a deterministic component (that is, one which depends on
characteristics of the individual and of the alternatives) and a stochastic
disturbance (or error term). The form of the logit model follows from the
assumption that the error terms are independent and share the same
probability distribution. This assumption under certain conditions may
produce erroneous results, which can be overcome by using nested logit
formulations.
Incorporates the spatial distribution of population, employment,
retailing, and land use within a compact, iterative procedure. Models
such as DRAM/EMPAL are successors to the Lowry Model.
A stochastic process that assumes that in a series of random events the
probability of an occurrence of each event depends only on the
immediately preceding outcome.
Information on the content, quality, condition, and other characteristics
of data. Useful in understanding and locating data.
A logit model of choice among more than two alternatives.
A representation of the structure of relationships between travel choices
an individual makes based on empirical data that provide the basis for
projecting the number of trips that will be made on each mode. A nested
logit is used for travel choices that have important similarities (e.g., bus
and light rail transit). Conceptually, nested logit analysis involves the
grouping of similar alternatives into one or more "secondary" logit
models, with a "primary" choice among the bundles of similar
alternatives. They can be any number of levels and branches in a nested
logit hierarchy, limited only by models through methodical estimation of
each standard logit model in the hierarchy.
The local sector; sells its goods and services within the metropolitan
area.
Pollution that originates from multiple sources over a relatively large
area, as opposed to that released though pipes. Nonpoint sources can be
divided into source activities related to either land or water use,
including failing septic tanks, improper animal-keeping practices,
agricultural and forestry practices, and urban and rural runoff from roads
and parking lots, etc.
The total amount of pollutants in the form of excess nutrients (usually
nitrogen or phosphorous) entering a water body.
Transportation network information about the number of area trips going
to and from origins and destinations.
D-4
-------
Object-Oriented
Programming
Object-Oriented
Database
Open Space
Preservation
Overlay Districts
Regression Model
Regression
Sedimentation
Semi-Empirical
A programming style that rigorously integrates data and actions that can
be taken on those data into single components called objects (DHS). One
example is Arc View's Avenue Script.
Each data item in a database is related to all others in logical ways,
guaranteeing consistency and minimizing the possibility of errors.
A tool used to preserve scenic, natural, and historic resources, including
prime agricultural and forestal lands, environmentally sensitive areas
such as flood plains, and steep slopes.
A land-use tool that is superimposed over existing zoning districts on a
particular zone, creating an additional set of requirements to be met
when those special resources are affected by proposed development.
Relates the variable being forecasted (i.e., the dependent variable) to
other economic, competitive, internal variables (i.e., independent
variables) and estimates a regression equation using the method of
ordinary least squares. Relationships are primarily analyzed statistically,
although any relationship should be selected for testing on a rational
basis.
A mathematical technique for exploring relationships between sets of
observations on two or more variables. A functional relationship
between the variables is postulated, and a line or curve fit between the
plotted observations so as to minimize some function (usually the
square) of the deviations between the plotted points and the line or
curve. The result is the equation of the best-fit line or curve describing
the dependent variable in terms of the other variables, which is often
used for projective purposes according to the goodness-of-fit. Several
types of regression analyses exist, including time-series, time interval,
time step, and temporal resolution.
Process of deposition of waterborne or windblown sediment or other
material. One example would be sediment deposited in a stream as the
result of runoff from a construction project. Also refers to the infilling of
bottom substrate in a water body by sediment (siltation).
Semi-empirical combines empirical techniques with some science
concepts; for example, estimating the dynamic flow of cars at a certain
point in a highway can use historical data only on the numbers of cars
with time (empirical). When the empirical approach includes some
science (e.g., statistical theory), then the approach becomes a semi-
empirical one.
D-5
-------
Smart Growth
Spatial Data
Spatial Analysis
Stochastic
Travel Demand
Management (TDM)
Uniform Analysis Zone
(UAZ)
Volume to Capacity
Ration
Validation
Vehicle Miles Traveled
(VMT)
Watershed
Smart growth recognizes connections between development and quality
of life. It leverages new growth to improve the community. The features
that distinguish smart growth in a community vary from place to place.
In general, smart growth invests time, attention, and resources in
restoring community and vitality to center cities and older suburbs. New
smart growth is more town-centered, is transit and pedestrian oriented,
and has a greater mix of housing, commercial and retail uses. It also
preserves open space and many other environmental amenities. But there
is no "one-size-fits-all" solution. Successful communities do tend to
have one thing in common-a vision of where they want to go and of
what things they value in their community-and their plans for
development reflect these values. (Source: ICMA and the Smart Growth
Network, 1998)
Spatial data represents the shape, location, or appearance of geographic
objects. It can be in vector, raster, or image format. One of the two main
types of data in a GIS (the other being attribute data).
The process of modeling space and examining and interpreting the
results. Spatial analysis is useful for evaluating suitability and capability,
for estimating and projecting, and for interpreting and understanding.
There are four traditional types of spatial analysis: topological overlay
and contiguity analysis, surface analysis, linear analysis, and raster
analysis.
A statistically based approach characterized by randomness. Similarly,
the results of the stochastic approach are based on probabilities.
Any step that can be taken to reduce the amount of travel in an area or to
and from a particular activity center. TDMs include carpooling, transit,
and congestion pricing.
Homogenous land units.
A ratio that refers to number of vehicles on a roadway versus the
capacity of the roadway.
Subsequent testing of a precalibrated model to additional field data,
usually under different external conditions, to further examine the
model's ability to project future conditions.
A unit of measurement used to estimate the impacts related to
transportation, such as traffic congestion and pollutant emissions from
mobile sources.
A drainage basin. The area bounded by a topographic divide in which
all water flows into the same water body.
D-6
-------
Zone The basic geographical unit for conventional travel demand analysis. A
study area is divided into zones, the number and size of which depend on
the size and land use patterns of the area, the geometry of the roadway
network, the nature of the problem, the computing resources available,
census boundaries, and political boundaries. Zone boundaries are
defined so that land uses and activities within are homogenous, to the
extent practicable.
Zoning The basic means of land use control employed by local governments in
the United States today. Zoning divides the community into districts
(zones) and imposes different land use controls on each district,
specifying the allowed uses of land and buildings, the intensity of such
uses, and the bulk of the buildings on the land.
D-7
-------
D-8
-------
Appendix E
References
-------
-------
References
American Planning Association. 1999. Planning Communities for the 21st Century. A special report of
the American Planning Association's Growing Smart Project.
American Rivers Association. 1997. Combating Sprawl to Save Rivers and Communities.
Beimborn, E., R. Kennedy, and W. Schaefer. 1996. Inside the Black Box: Making Transportation
Models Work for Liveable Communities. Citizens for a Better Environment and the
Environmental Defense Fund. Washington, B.C.
Bureau of Transportation Statistics. 1998 Pocket Guide to Transportation.
Chang, R., and P. K. Kelly. 1995. Step-By-Step Problem Solving. Richard Chang Associates, Inc.
Irvine, CA.
Deakin, E. 1995. Land Use Model Conference Keynote Address. In Travel Model Improvement
Program Land Use Modeling Conference Proceedings. Travel Model Improvement Program.
DOT-T-96-09. U.S. Department of Transportation, U.S. Environmental Protection Agency, and
U.S. Department of Energy.
Foote, K., and M. Lynch. 1997. The Geographer's Craft Project. University of Texas at Austin.
Department of Geography.
International City/County Management Association (ICMA) and the Smart Growth Network. 1998. Why
Smart Growth: A Primer.
Kaiser, E. J., D. Godschalk, and F. S. Chapin, Jr. 1995. Urban Land Use Planning. 4th Edition.
Chicago: University of Illinois Press.
Katz, B. June 1997. Curb the Sprawl of the Suburbs. Hartford Courant and Philadelphia Enquirer.
Miller, E., D. Kriger, and J. Hunt. 1999. Integrated Urban Models for Simulation of Transit and Land-
Use Policies. Transit Cooperative Research Program Report 48. Washington, D.C.: National
Academy Press.
Nelson, A. C., and J. B. Duncan. 1995. Growth Management Principles and Practice. American
Planning Association. Chicago: Planners Press.
National Center for Resource Innovations. 1997. A study conducted for the Washington Post by Frankel
and Fehr.
National Research Council. 1998. Report 39: The Costs of Sprawl-Revisited. Transportation Research
Board. Transit Cooperative Research Program. Washington, D.C.: National Academy Press.
Orfield, M. 1997. Metropolitics: A Regional Agenda for Community and Stability. Washington, D.C.:
Brookings Institution Press.
PAS (Planning Advisory Service). 1999. Public Investment (memo).
E-l
-------
Parsons Brinckerhoff Quade and Douglas, Inc. 1999. Land Use Impacts of Transportation: A
Guidebook. National Cooperative Highway Research Program Report 423A Washington B.C.:
National Academy Press.
Rusk, D. 1999. Inside Game Outside Game: Winning Strategies for Saving Urban America.
Washington, B.C.: Brookings Institution Press.
Schrank, D., and T. Lomax. 1998. Urban Roadway Congestion. Texas Transportation Institute. Texas
A&M University.
Sorenson, A. A., R. P. Greene, and K. Russ. March 1997. Farming on the Edge. American Farmland
Trust and Center for Agriculture and the Environment, Northern Illinois University, DeKalb,
Illinois.
South-worth, F. 1995. A Technical Review of Urban Land Use-Transportation Models as Tools for
Evaluating Vehicle Travel Reduction Strategies. Oak Ridge National Laboratory Report 6881,
U.S. Department of Energy.
Sumner, S. \992.ManagementInformationSystems: The Manager's View.
Taha, H. A. 1976. Operations Research: An Introduction.
USDOT. 1997. Transit Trends Over Time: Population Oregon: A Comparison with 20 Cities of Similar
Transit Service District Population Size, 1990-1995. National Transit Database (metro
publication).
USDOT/FHWA. 1999. Transportation Air Quality: Selected Facts. FHWA-PD-99-015.
USDOT/USEPA (U.S. Department of Transportation/U.S. Environmental Protection Agency).
Alternative Choice Print Ad "Mass Transit" for It All Adds up to Cleaner Air campaign.
U.S. General Accounting Office. April 1999. Community Development: Extent of Federal Influence on
"Urban Sprawl" Is Unclear. Report to Congressional Requesters.
Washington State Department of Ecology. 1997. Breathing Easier (fact sheet).
Weitz, J., and T. Moore. 1998. Development Inside Urban Growth Boundaries: Oregon's Empirical
Evidence of Contiguous Urban Form. Journal of American Planning Association, 64(4): 424.
August 1, 2000
E-2
------- |