chapter  3
20 Pages

Mapping Impervious Surfaces Using Classification and Regression Tree Algorithm

WithGeorge Xian

CONTENTS 3.1 Introduction ................................................................................................. 39 3.2 Selection of Study Areas............................................................................ 41

3.2.1 Seattle-Tacoma Area....................................................................... 41 3.2.2 Las Vegas Valley............................................................................. 43

3.3 Impervious Surface Estimations............................................................... 43 3.3.1 Data ................................................................................................... 44 3.3.2 Classifications of High-Resolution Images................................. 45 3.3.3 Landsat Imagery ............................................................................. 45 3.3.4 Regression Tree Models................................................................. 46 3.3.5 Imperviousness Estimates ............................................................. 47 3.3.6 Accuracy of ISA Estimations ........................................................ 50

3.4 Urban Land-Use Density and Percent Imperviousness ....................... 51 3.5 Discussion .................................................................................................... 54 3.6 Conclusion ................................................................................................... 56 Acknowledgments ............................................................................................... 56 References ............................................................................................................. 56

Urban development is usually associated with the conversion of land in rural areas to residential and commercial land use. As the extent of built-up land increases, further development generally raises concerns about the impacts of land use and land cover (LULC) change on urban and rural environmental conditions and on quality of life. Spatial distributions and patterns of LULC often affect socioeconomic (Douglass, 2000), environmental (Gillies et al., 2003), and regional climatic conditions (Arnfield, 2003; Kalnay and Cai, 2003; Voogt and Oke, 2003). The influences of urban environments on the

global population and the monitoring of spatial-temporal changes in large urban and suburban areas are both becoming increasingly important (Small, 2001). The ability to monitor urban LULC changes is highly desired by local communities and by urban management to help provide a more detailed picture of the human-influenced landscape (Carlson, 2003).