Committee:

John Peponis, Ph.D., Professor, School of Architecture, Georgia Institute of Technology, Advisor

Sonit Bafna, Ph.D., Associate Professor,  School of Architecture, Georgia Institute of Technology

Ellen Dunham Jones, Professor, College of Architecture, Georgia Institute of Technology

Steve French , Ph.D.,  Dean, College of Architecture, Georgia Institute of Technology

Saif Haq, Ph.D., Associate Dean for Research, Texas Tech

Thesis title: “Assessing the measures of street connectivity” a comparative study of the largest American cities.

Abstract:

The thesis offers a systematic characterization of the morphology of street networks in Metropolitan Areas in the United States in order to help the assessment of current conditions. In the face of the prevailing variability of street network types, the thesis also examines which measures of street connectivity best capture significant differences on the ground and, by implication, might be most fruitfully used in assessing street connectivity in the context of policy development and planning for possible urban futures.

Street segment length, block size and street connectivity are measured for all segments within each of the 24 largest Metropolitan Statistical Areas in the US. In order to characterize local areas, values are also studied for a stratified sample of more than 4500 buffer zones, each of 2 mile diameter, chosen at variable distance ranges from the city-centers. As a point of comparison, a smaller sample of 96 areas is chosen to illustrate street network types identified and discussed in the existing literature on urban morphology and street network types and connectivity. The typological sample provides benchmarks for the characterization of the random sample of buffers.  

The thesis profiles the variation of Metropolitan Areas, and typical local areas within them. In addition, the thesis concludes that the measure of “metric reach” which captures the street network length accessible within a specified network distance from any given point, used on its own, is a more sensitive descriptor of street connectivity than some of the measures traditionally used, such as the distance between, or the density of, intersections. Metric reach also provides the most consistent characterization of trends such as the reduction of street network density with increasing distances from the historic city-center, or the identification of multiple centers at varying distances from the historic city center, whether these emerge as recently grown edge cities, or survive as centers of old towns absorbed into the expanding urban fabric.