Urban Growth Model

LAND USE DEVELOPMENT MODEL REVIEW 

INTRODUCTION

Urban development in the United States since the 1950s has been dominated by the movement of residential, commercial, and industrial land uses to the urban fringe and the conversion of rural areas near major metropolitan areas into low-density, predominantly single-family residential subdivisions and strip commercial developments. These development patterns, commonly referred to as “urban sprawl,” result in haphazard, low-density development patterns which consume large quantities of valuable agricultural land and generate excessive public costs of providing required community facilities and services. As a result, the issue of urban sprawl has attracted a great deal of public interest and academic attention (see, e.g., Audirac and Zifou 1989; Burchell, et al. 1998). 

The last forty years has also witnessed a continued effort to develop computer-based models for describing urban development patterns and determining the future impacts of public policy choices (Harris 1985; Batty 1994; Wilson 1998). These efforts have generated a voluminous literature (Klosterman 1994; Southworth 1995) but few operational models. However, this situation is changing rapidly as dramatic advances in computer technology and the availability of large quantities of spatially referenced data are stimulating a renewed interest in urban modeling in the United States and throughout the world (Wegener 1994).

This report will review and evaluate ten currently available, fully operational land use development models that can be used to address the environmental degradation and human health impacts of urban sprawl in the 15-county Northeast Ohio region. The review will consider academic and commercially available models to identify the model, or modeling approach, that is most appropriate for considering the implications of urban sprawl and alternate growth management strategies in Northeast Ohio. The review updates and extends similar reviews by Wegener (1994; 1995), the International Study Group on Land-use/Transport Interaction (Webster, et al. 1988; Webster and Pauley 1991), and Southworth (1995). 

The models to be considered in this review are: (1) METROPILUS, the latest version of Steven Putman's DRAM/EMPAL family of models; (2) the first California Urban Futures (CUF-1) model developed by John Landis and his colleagues in the early 1990s; (3) the second California Urban Futures (CUF-2) model developed by John Landis and his colleagues in the late 1990s; (4) the Portage County, Ohio, model developed by Jay Lee and his colleagues; (5) the What if? model developed by Richard Klosterman and his colleagues; (6) the SmartPlaces model distributed by the Electronic Power Research Institute;  (7) the TRANUS model developed by de la Barra and his colleagues at Modelistica, a Venezuelan company; (8) the UrbanSim model developed by Paul Waddell and his team at the University of Washington, and (9) the Medina County model developed by Chengri Ding and his colleagues at Cleveland State University. 

The report begins by briefly describing the ten models. It then evaluates the models with respect to the following considerations: (1) cost; (2) their ability to work with the data which are, or will become, available as part of this project; (3) the potential for viewing model results via the World Wide Web; (4) the understandability of the model assumptions and operations for non-technical experts; and (5) the theoretical soundness of the model results. It concludes by recommending one model to be used for sub-regional (i.e., county and sub-county) analyses and two closely related models to be used for the fifteen-county regional analysis.