Remote sensing and GIS for urban environmental modeling, monitoring and visualization.

摘要

In recent decades, use of satellite images for urban applications in the remote sensing community has increased. However, there also has been a shift from exclusive land use mapping, in early studies, toward recent land cover and land use mapping. Understanding land cover composition is essential for urban environmental analysis. The V-I-S (Vegetation-Impervious surface-Soil) model provided an effective method for distinguishing land cover composition in urban and periurban areas.; Based on the V-I-S model and as a step further, a supervised classifier for TM images had been successfully developed, in previous research, to estimate six ground component percentages on urban areas. These six ground components are Vgg: green grass vegetation, Vts: tree/shrub vegetation, Ibr: bright impervious surface, Imd: medium impervious surface, Idk: dark impervious surface, and Sdv: soil/dry vegetation. This research is to extend the capacity of the supervised classifier to ETM+ images and MSS images. Moreover, these classifiers are applied to remotely sensed images covering Salt Lake City areas to derive ground component percentages over the past 28 years. By comparing the ground component percentages from different years, urban growth can be analyzed qualitatively, as well as quantitatively. A scientific visualization method is developed to demonstrate the rapid urban area expansion in the format of computer animation.