Case Study:

SPARTACUS

Methodology

Land Use - Transportation Model

Structure

MEPLAN was used as the land use-transportation model in the SPARTACUS study. (Each of the three metropolitan areas in the SPARTACUS study had an existing, operational MEPLAN model.) The model has three interrelated modules:

The land use module produces a spatial allocation of activities (such as employment and population) and produces trade flows between zones;
The transport module assigns flow matrices to different transportation modes and routes and calculates the resultant transportation disutilities; and
The interface module converts land use trade matrices into transportation flow matrices and, conversely, converts transportation disutilities into trade disutilities.

MEPLAN is based on the principles of economic input-output models as well as random utility models from the transportation sector. Three key theories drive the model:

Interactions between activities are determined by input-output analysis (for example, coefficients state how much of an input factor is required by households in different socioeconomic groups);
Locational choices, mode choices, and route assignment are determined by random utility models; and
Capacity restrictions in the land and transportation markets alter prices and thus affect disutilities.

The three MEPLAN modules are applied dynamically, with land use affecting transportation and vice-versa; feedback typically occurs in five-year increments. The interface module is used in such a way that land use changes produce immediate changes in the demand for transport, whereas transport changes have only gradual effects upon the pattern of land uses and trades.

Data Requirements and Calibration

The MEPLAN model is custom-built for each area for which it is applied, based on the availability of information and on relationships which best reflect the area and the users' interests. Typical inputs for the land use submodel include population, employed persons, land and floorspace area, incomes, land prices, and goods vehicle travel by area. Employment must be separated into an exogenous (basic) and an endogenous (non-basic) element to drive the input-output mechanism.

The transportation submodel inputs, which describe the transportation network and trip patterns, are similar to those for typical four-step transportation modeling systems. The land use and transportation models may be built on different zonal structures; for example, economic data may be available at the jurisdictional level for the land use module, while transportation data may be provided at the traffic analysis zone (TAZ) level for the transportation module. The interface module translates between these different zone structures.

Additional information on the MEPLAN model is provided in the Sacramento case study.

Outputs

A large volume of information is generated by running the model. Typically, this includes:

Land use characteristics by zone, including households by socioeconomic group, employment by type, and land/floorspace by type;
Zonal cost information, including cost of living, income, and rental incomes;
Characteristics of transportation by mode, traveler type, and zone pair;
Screenline traffic counts and loaded networks;
Summary statistics such as vehicle-miles of travel, vehicle-hours of travel, energy consumption, and emissions; and
Total trips, average distance, average user cost, and average travel time for each travel type/user mode combination.

The model also offers an economic evaluation module that can be used to compare alternative policies or scenarios. This module is based on flows of money (e.g., transit fares) as well as other benefits such as time savings. Using principles of consumer surplus, the model provides a measure of the distribution of benefits among firms, households, the government, transport operators, and landowners/developers. In the SPARTACUS study, net economic benefits are divided by the population of the region to achieve an overall economic indicator value.

Comparison with Other Land Use/Transportation Models

MEPLAN is one of a number of packages available for modeling the interactions between land use and transportation. The overall level of effort as well as the sophistication of these packages varies considerably. Among models that have been applied in the U.S., DRAM/EMPAL and HLFM II+ require somewhat fewer data inputs and a lower level of modeling effort. MEPLAN, TRANUS, and UrbanSim are examples of models that require more data and effort to develop, but at the same time can model a broader range of relationships with greater sophistication.

Models such as MEPLAN, TRANUS, and UrbanSim have a strong economic foundation, and business and residential location decisions are modeled based on a range of factors. Land markets are key to the modeling framework, with prices driving development. This is in contrast to models such as DRAM-EMPAL that re-allocate land use only on the basis of transportation accessibility. UrbanSim advances the state of the practice by using the logit model structure, typically used to predict transportation mode choice, to model a broad range of decisions such as business and residential location.

At the same time, some of the economic data required to calibrate MEPLAN or other large-scale urban models may not be readily available. For example, in many urban areas, a single database on factors such as land prices. Data on goods movement - such as truck trip generation, truck flows, or goods flows - are also typically poor. Commonly available data sources may not provide a breakdown of exogenous versus endogenous employment at less than the county level. As a result, any region considering implementing a land use/transportation model should carefully consider the availability of data as it relates to modeling requirements, in addition to other factors such as staff resources.

[TOP]