Citilabs - Introducing Cube Land: Bringing Land Use to Your Models

Introducing Cube Land: Bringing Land Use to Your Models

Background
Citilabs couples a proven land use model with the power of Cube Base and ArcGIS to deliver a tightly integrated transportation and land use modeling framework.

In January 2009, Citilabs introduced Cube Land—the latest addition to its Cube transportation planning software suite. The Cube Land module allocates households and employment in a study area according to basic economic principles of real estate market equilibrium. In the process, it generates forecasts of commercial and residential units built by type and zone and, according to bid-rent theory, Cube Land determines real estate value based upon the amount the highest bidder would be expected to pay in an auction. Operationally, this willingness to pay is a function of location externalities and transportation accessibility. These characteristics make Cube Land an especially useful tool for integrated land use and transportation planning.

Cube Land is based on Dr. Francisco Martinez’s MUSSA II model—a highly regarded and innovative framework already described in academic literature1. Through a series of pioneering studies in Santiago, Chile, Dr. Martinez and his laboratory developed a comprehensive, robust theory of land use and real estate market dynamics, and devised techniques for describing these complex relationships. Citilabs has adopted this proven and widely-accepted model and coupled it with the unparalleled power of Cube Base’s modeling platform and ArcGIS, providing a streamlined interface for the development of integrated transport and land use models.

Building Cube Land Models
Cube Land models are highly sophisticated and can be used to construct very detailed models or sketch level models that require much less data.

Although the Santiago case study involved the collection and manipulation of highly detailed data sets, Cube Land can be calibrated using relatively little data. The basic data file relates the characteristics of a zone to the households or firms located therein. Digitized parcel data can also be paired with census data or any of several commercially available inventories of businesses and residents to provide data for estimation and calibration of bid function coefficients. In the United States, for example, the Public Use Microdata Sample (PUMS), American Housing Survey, Journey-to-Work tabulations, and Longitudinal Employer-Household Dynamics data sets may all provide valuable starting points for building a simplified Cube Land model. With richer data, it is possible to estimate and calibrate bid function terms that represent more complex interaction effects and location externalities.

Once calibrated, Cube Land requires even less data for developing scenarios. Because Cube Land operates at a zone-based level of geography, it is not necessary to maintain base year land use or socio-economic information at grid or parcel level detail. Rather, the zoning system can be explicitly defined to correspond to the level of detail available. Furthermore, for future years, household totals, firms, and real estate by category are only needed at the regional level. These figures are typically available from macro-economic models or projections provided by state demographers, and can be modified to analyze how shifting conditions in the economy at large affect urban growth and decline.

Cube Land’s real estate supply model can run in a variety of modes reflecting fixed, bounded, and constrained supply assumptions. Fixed supply scenarios may be used to constrain the level of development to base year conditions, or to analyze demand response to a hypothetical, exogenously defined scenario, such as a guided plan or a major new development. Bounded supply scenarios can be used to limit the amount of change in supply from a base condition, representing restrictions on redevelopment or an incremental model. Constraints on the kinds of development may occur in different locations may also be imposed. Taxes and financial incentives can also be represented.

Integrating Cube Land
Cube Land is easily integrated with existing transportation models and can be implemented using a variety of approaches, data formats, or travel models.

In practice, Cube Land can be added to any existing Cube application as a new program step in Cube Base. Cube Base provides the graphical user interface for managing the model and an embedded ArcGIS interface for defining inputs, mapping, and reporting. Cube Base also handles any conversion steps needed to re-format data inputs (or outputs) to fit within the model structure.

With an existing Cube travel demand model, transportation accessibility and attractiveness measures are extracted and applied to the Land model automatically. Furthermore, since Cube Base is an open modeling platform, any travel demand model that can be launched from a command line shell can be directly integrated with Cube Land. Conversely, Cube Land can be executed using text format control and input files from a command line shell, or called from a third-party product without the requirement to install any other Cube software.

Implementation Approaches
Cube Land supports several approaches to represent the dynamics of transportation and land use.

Cube Land can be configured to implement any of a number of widely-known approaches to integrated transportation/land use models. For example, Cube Land can be added as pre-generation or post-assignment step in a typical four-step or activity-based travel model. Cube Land can also be integrated as an urban modeling system to solve multi-level equilibrium problems. The multi-level approach uses supply and demand components of land use and transportation which alternate with feedback between the two subsystems. As is common practice with other land use models, a lagged stepwise approach may be implemented, wherein feedback between the land use and transportation subsystems occurs over successive time intervals, rather than within a single year. Finally, the Cube Land model may be treated as a separate application within Cube’s Application Manager, then run manually when needed to generate new input assumptions for transportation scenarios.

A Step Toward Better Planning
Cube Land promises to improve modeling by allowing planners to represent and understand how market conditions and transportation decisions influence growth.

Clearly, adding Cube Land to the transportation planning process will make transportation models more dynamic, realistic, and effective. Beyond being a proven approach to land use modeling, Cube Land offers several options for implementation and a sophisticated number of alternatives for integrating transport variables. Cube Land promises to improve modeling by allowing planners to represent and understand how market conditions and transportation decisions effect growth. By representing and understanding the relationship between land development and transportation investments, planners and their stakeholders are better equipped to establish policy, and prioritize transportation investments and development initiatives.

Notes:
1 Martinez, F. “Towards a Land use and Transport Interaction Framework” in Hensher and Button, ed.s Handbook of Transport Modelling, 2008. San Francisco: Elsevier