Document Type

Thesis - University Access Only

Award Date


Degree Name

Master of Science (MS)

Department / School



The population of Sioux Falls, South Dakota has been steadily increasing from 72,468 in 1970 to 112,500 in 1995. This has resulted in a considerable amount of rural-to- urban land conversions occurring in the southwest and southeast regions of the city. Change detection methods have been developed to monitor these land cover changes and determine their extent. Change vector analysis is one of these methods being developed to study land cover change and to utilize the historical collection of multiresolution satellite data available. The change vector analysis procedure involves: (1) co-registering multiple dates of Landsat multispectral scanner (MSS) data, (2) formulating a signal-to-noise model, (3) radiometrically rectifying the images, (4) transforming the images to scene-based measures of brightness and greenness, (5) calculating change vectors from brightness and greenness, and (6) visualizing the change vectors. Change vector images were calculated for two time periods, 1973 to 1984 and 1984 to 1990. The change vector images were validated using 50 random samples collected for each change image. Of the 100 validation samples, 37 percent were classified as land use/land cover transitions and 63 percent were changes in the condition of an unchanging land use or land cover. Sixty percent of the land use/land cover transitions were rural-to-urban transitions. The change vector images were able to detect and characterize urban changes in terms of brightness and greenness.

Library of Congress Subject Headings

Landscape changes -- South Dakota -- Sioux Falls -- Remote sensing
Urbanization -- South Dakota -- Sioux Falls -- Remote sensing
Vector analysis
Landsat satellites



Number of Pages



South Dakota State University