Geoscience Reference
In-Depth Information
9.1 Introduction
Over decades, urban sprawl has been the dominant urbanization pattern throughout
much of the United States. Urban sprawl is the land-consumptive pattern of sub-
urban development characterized by increasing scattered, low-density residential
and commercial areas outside of city centers caused by population growth, rising
incomes, decreasing commuting costs, and dependent on extensive automobile use
(Wilson et al.
2003
; Interlandi and Crockett
2003
; Robinson et al.
2005
). Many stud-
ies have found that urban sprawl has an adverse impact on varied aspects of social
and natural environment, including community development, public health, forests,
and stream systems (Zhang
2001
; Ewing et al.
2003
; McGrath et al.
2004
; Wheeler
et al.
2005
). The main reason has been attributed to the low-density development
pattern of urban sprawl since per capita land use, energy consumption, and pollu-
tion generation are higher in low-density suburban areas than in high-density central
cities.
An increasing concern about the impact of urban sprawl is stream degradation
due to the increasing impervious surfaces and human activities in the process of
urbanization in watersheds. (Bowen and Valiela
2001
; Finkenbine et al.
2001
; Tong
and Chen
2002
; Interlandi and Crockett
2003
;Hattetal.
2004
; Aichele
2005
;Brett
et al.
2005
; Deacon et al.
2005
; Schoonover et al.
2005
). The impact is usually
studied through relationship analysis of urbanization and water quality indicators
using conventional statistical methods, such as ordinary least squares regression
(OLS) and Spearman's rank correlation analysis. Generally positive correlations
exist between concentrations of water pollutants and percentages of developed lands
(i.e. urban lands, as used in many studies) including commercial, residential, and
industrial lands. In other words, higher concentrations of water pollutants are related
to higher percentages of developed lands. For example, Tong and Chen (
2002
)
examined the relationships of land use and water quality on regional scale in the
watersheds of the Ohio State, USA. They found that TN (total nitrogen), TP (total
phosphorus), conductivity, and fecal coliform were significantly positively related
to commercial and residential lands. However, a relationship between an urbaniza-
tion indicator and a water quality parameter found in one study might not be true, or
even opposite in other studies. For instance, a study in the Ipswich River watershed,
Massachusetts, USA found that TN and TP had no significant correlations with
urban lands (Williams et al.
2005
). Therefore, the relationships between urbaniza-
tion and water quality indicators are not constant because watershed characteristics
and pollution sources are not the same in different watersheds. Especially for the
regions under rapid urban sprawl, the relationships between urbanization and water
quality indicators might vary substantially over space because great differences in
land development pattern exist between low-density suburbs and high-density cen-
tral cities. However, the potential different impact of urbanization on water quality
between suburban areas and central cities has been seldom studied and reported.
One of the reasons might be due to the fact that the conventional statistical methods
used in previous studies are global statistics, which analyze the average situation for
