A GIS Assessment of Urban Sprawl in Richmond, Virginia

Abstract

In the United States, the urban sprawl debate has closely paralleled urban growth trends over the past few decades. Many studies indicate that it is the pattern, density, and rate of new urban growth that create the appearance of sprawl. Population dynamics are often cited as a driving force behind urban sprawl. This thesis uses Geographic Information Systems (GIS) mapping and land cover change analysis, neighborhood statistics, community surveying, key-informant interviews with planners and developers, and planning documents to measure sprawl. The study area includes the jurisdictions that comprise the Metropolitan Statistical Area (MSA) of Richmond (The City of Richmond, Chesterfield County, Hanover County, and Henrico County). Urban land cover increased by one-fourth, from approximately 559 square kilometers to approximately 746 square kilometers from 1992 to 2001. Over the nine year time period, population within The City of Richmond decreased from 203,056 in 1990 to 197,790 in 2000, while Chesterfield, Henrico, and Hanover Counties increased in population from 1990 to 2001. Until the early part of the 20th century, cities in the United States experienced dramatic densification as industry set up shop in the city and workers flocked there in large numbers to claim jobs. As population grew and technology became more advanced, many people left rural farming areas to settle near industrialized urban core areas. This allowed the population to earn more money and with the introduction of cheap transportation in the form of the automobile, suburbanization began to take place. With more cars came the need for more roads. With more roads came more opportunities for people and employment to move away from the city into rural areas, thereby setting up communities which some refer to as sprawling developments. GIS maps suburbanization in the form of urban land cover, transportation networks, and population densities within and outside core urban areas over any given time period in order to assess trends in urban growth. This study analyzes urban land cover data as well as interviews with local developers and planning documentation to understand development trends in Richmond from 1992 to 2001. These dates reflect the availability of National Land Cover Data (NLCD), which I reclassified in the GIS to show only those classes that represent urbanized land. I then compared the two years to show the level of urban growth over the nine year time period. Next, I analyze patterns of urban expansion by using mapping capabilities within the GIS and neighborhood statistics in order to show the density and connectivity of patches of new growth. Based on the density and connectivity of new growth areas, I classify patterns as one of three types of sprawl: linear along highways, cluster, and leapfrog. My threshold densities are; 0 to 400 30 meter pixels per square kilometer for low density, 401 to 700 for medium density, and 701 to 1200 for high density. I also interviewed local developers and planners to gauge their opinions on the issue of urban sprawl versus urban growth. Developers do not see themselves as contributors to sprawl while planners see their roles as buffers between unfettered growth and market forces. The results indicate that the Richmond MSA did experience an increase in urban land from 1992 to 2001 and that urban growth in the study area can be classified as urban sprawl with the use of GIS mapping, neighborhood statistics, and analysis of jurisdictional planning documentation coupled with interviews with developers, land owners, and local planners. The density of new development is greatest in Henrico and Chesterfield, but the pattern and character with which development has occurred in Hanover is synonymous with sprawl. Sprawl is also facilitated by inexpensive land with available infrastructure (water, sewer lines).

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 10 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.