Date of Award:


Document Type:


Degree Name:

Master of Landscape Architecture (MLA)


Landscape Architecture and Environmental Planning

Committee Chair(s)

Shujuan Li


Shujuan Li


Carlos V. Licón


Joanna Endter-Wada


As residential urban development increases in the western United States, few studies have shown how different urban forms influence vegetation cover. The two studies in this thesis examine how to define and measure urban form in order to understand the relationship between urban form characteristics and vegetation cover.

In the first study, urban form was defined by using past methods of identifying and measuring urban sprawl. Past studies showed the most essential metrics that define residential urban form are building density, centrality, connectivity, land use mix, and parcel size. This study reviews these metrics and proposes revised unified definitions and measurement methods. It is recommended that consistent definitions and methods be used in further research of urban form.

Cache County, Utah was used as a study area to apply these methods as a representative community of the western U.S. Residential parcels were measured for each metric. A K-means cluster analysis assign each parcel to one of 50 groups based on metric measurement similarities. The 10 most widely used groups contained 87% of the residential parcels in Cache County. These groups represented urban forms with distinct identifiable characteristics such as Agricultural residential, urban sprawl, historic plat patterns, and evolved versions each.

The second study used the metrics and groups from the first study to check for correlations with vegetation cover. A normalized difference vegetation index (NDVI) of aerial imagery of Cache County was used to classify land cover into three groups: dense vegetation, light vegetation, and no vegetation. Measurements of vegetation cover were extracted for each parcel and each urban form group. Total vegetation cover (TVC) and dense vegetation cover (DVC) measured higher in urban areas than in the county as a whole. Agricultural residential groups had among the highest TVC, but had the lowest DVC. Non-agricultural residential groups had the highest DVC as a percentage of TVC with the exception of group 4 “Modern Suburban Sprawl” which had the lowest DVC as a percentage of TVC. Group 7 “Satellite Centers” had the highest TVC while group 3 “Dead End Semi-sprawl” had the lowest TVC. Both groups 3 and 7 had the highest DVC as a percentage of TVC.

A correlation analysis revealed that TVC had stronger correlations than DVC with urban form metrics. Building density had the strongest correlation with TVC (r = -0.62, p= 0). Correlations with TVC were also found with Parcel size (r = -0.23, p = <.001) and centrality (r = -0.21, p = <.001). Very weak relationships with TVC were found with
connectivity and land use mix.