JAWRA - Stormwater Runoff

• Project Type: General OWC
• Title: Stormwater Runoff Quality and Quantity from Traditional & Low Impact Development Watersheds
• Date: 12/15/2009
• Water Conservation - Doc Type: Reports

SEDAN AND CLAUSEN
the postconstruction period (Table 4). It was expected
that pervious concrete pavement would reduce peak
discharge compared with traditional stormwater con-
veyance (USEPA, 1993a; Brattebo and Booth, 2003).
A goal of LID is to maintain peak flow and flow vol-
ume at predevelopment conditions. That goal was
achieved in this study.
Nutrients - Nitrogen
Traditional Watershed. During the postcon-
struction period, no significant change was observed
in NO + NO -N (p = 0.590) and NH -N (p = 0.103)
concentrations in runoff from the traditional
watershed when compared with values predicted by
the calibration equation. However, NO + NO -N and
NH -N mass exports significantly increased
(p < 0.001) based on the difference in regression
intercepts (Table 3).
Total Kjeldahl nitrogen concentrations in runoff
from the traditional watershed significantly
decreased (p < 0.001) by 76% in the postconstruction
period. The decrease was likely due to less input of
organic nitrogen. Runoff during the calibration period
was collected from a lawn, while runoff during the
postconstruction period was primarily from the
asphalt road. Conversely, the mass export of TKN
significantly increased (p < 0.001) by 40 times during
postconstruction (Table 3).
Other studies have reported nitrogen concentra-
tions in runoff comparable to these from the tradi-
tional watershed in the postconstruction period
(Table 3). For example, the Nationwide Urban Runoff
Program (NURP) reported similar concentrations of
NO3 + NO -N (0.74 mg/1) and TKN (1.9 mg/1) as
event mean concentrations (EMCs) from residential
runoff (USEPA, 1983b). The nitrogen concentrations
in runoff from the traditional watershed were also
similar to residential runoff observations found else-
where in Connecticut (Dietz and Clausen, 2004).
Urbanization has increased storm flow and subse-
quently increased pollutant loadings (Myers et al.,
1985), which was evident in this study. Increased
mass exports of NO + NO -N, N11 -N, and TKN
were primarily due to increased storm flow from the
traditional watershed, since the concentrations had
decreased or not changed at all (Table 3). The NURP
study (1983b) reported exports of NO + NO -N
(5.8 kg/ha/year) and TKN (5.8 kg/ha/year) which
were slightly higher compared with the control and
traditional watersheds (Table 3).
Low Impact Development Watershed. Mean
weekly concentrations of NO + NO -N in runoff from
the LID watershed increased significantly (p = 0.014)
by two times during the postconstruction period
(Table 4). NH -N concentrations in runoff signifi-
cantly decreased (p = 0.012) by 50% when compared
with values predicted by the calibration equation
(Table 4). TK 4 concentrations in runoff significantly
increased during the postconstruction period
(p = 0.008) by 44% (Table 4).
Unlike results from Dietz and Clausen (2004),
homeowner education of BMP lawn practices was not
effective in this study. A survey of homeowners in
the study watersheds saw no difference in lawn care
practices, such as whether they fertilized and the
number of applications per year between the tradi-
tional and LII) watersheds (Table 5). However, more
LID residents applied fertilizers themselves than in
the control and traditional watersheds, which had a
higher percentage of professional service applicators.
The increase in TKN concentrations in runoff was
likely due to organic nitrogen from the grass swales,
since NH -N concentrations decreased. Grass clippings
and other detritus were expected sources of organic
nitrogen. Rushton (2001) also found TKN concentra-
tions were higher in runoff from parking lots with
grassed swales as compared with lots with no swales.
During the postconstruction period, no significant
change (p > 0.277) was observed in NO + NO -N
mass export in runoff from the LID watershed. How-
ever, the mass exports of NH -N and TKN in the LID
watershed significantly decreased by 71% (p < 0.001)
and 33% (p = 0.05), respectively, when compared with
values predicted by the calibration equation (Table 4).
Mass export reductions in NH -N and TKN from
the LID watershed were attributed to the decrease in
storm flow in the postconstruction period. Reduced
exports of nitrogen from the LID watershed (Table 4)
were somewhat similar to observations by Rushton
(2001) who found that reduced runoff from pervious
pavement with grassed swales had reduced NH load-
ings by up to 85 %.
TABLE 5. Household Survey Results for Lawn
Maintenance by Watershed (2003- 2004).
Control
Question (n = 40)
Traditional
(n = 15)
LID
(n = 16)
x2
Fertilize lawn? ( %)
Yes 85
100
88
2.557
No 15
0
13
Fertilize times /year? ( %)
1 -2 48
20
29
5.861
3 -4 45
67
71
>4 7
13
0
Lawn care source? ( %)
Self 75
67
100
5.987*
Professional 25
33
0
Note: N.S. = not significant
*p < 0.05
JAWRA 1004 JOURNAL OF THE AMERICAN WATER RESOURCES ASSOCIATION