Characterizing urban subpixel composition using spectral mixture analysis - Urban Remote Sensing: Monitoring, Synthesis and Modeling in the Urban Environment - page 114

(a) (b)

(c) (d)

20 km

0 2 3 4

FIGURE 8.5 Maps of model complexity for the settlement of Seringueiras, Rondonia, for the years (a) 1986, (b) 1990, (c)
1995, (d) 2000. Water, burned, and unmodeled pixels are black.

8.3.2 Discriminating urban tree and
lawn cover in a western US city

contribute to urban environmental quality by mitigating air
quality, sequestering carbon, and lowering building energy use
through shading, evapotranspiration, and wind speed reduction
(Small and Lu, 2006). While an additional benefit of turf grass
is enhanced carbon uptake by soil, lawns are also associated
with a number of negative effects, including non-point source
pollution from over-use of chemicals and fertilizer, as well as
increased summer water use, especially in Western cities (Milesi,
et al ., 2005). For example, fifty-four percent of annual house-
hold water use in Denver, Colorado, is dedicated to landscaping
(Denver Water, 2009).
The objective of this study was to inventory urban lawn and
tree cover in Denver to provide city resource managers with a
rapid assessment of the distribution and total area covered by each
vegetation type (Fig. 8.1b). The challenge of such an assessment
is the spectral similarity between lawn and tree vegetation; one

Knowledge of the distribution and characteristics of urban veg-
etation is necessary to quantify and model urban environmental
conditions at various spatial scales (Small and Lu, 2006; Franke
et al ., 2009). The distribution, type, and quality of urban vegeta-
tion strongly influence urban environmental processes, such as
the cycling of water, nutrients, energy, and carbon (Small, 2001;
Lee and Lathrop, 2006; Petaki et al ., 2006), as well as substantially
impact the social and economic characteristics of a neighbor-
hood (Nowak et al ., 2001). In most US cities, the dominant
vegetation types are trees and turf grass (i.e., lawn); both con-
tribute to moderating the urban climate, reducing storm water
runoff, and providing habitat for wildlife. Trees, in particular,