JAWRA article - Rain Gardens

• Project Type: General OWC
• Title: Performance Assessment on Rain Gardens
• Date: 12/15/2009
• Water Conservation - Doc Type: Reports

PERFORMANCE ASSESSMENT OF RAIN GARDENS
mottles (i.e., small areas of gray, red, yellow, brown,
or black that differ in color from the bulk soil), which
may indicate hydric soils (Richardson and Vepraskas,
2001) that would be associated with prolonged water
saturation. The soil was crumbled and repacked into
the core hole after soil inspection was complete.
Infiltration Rate Tests
equation for cumulative infiltration, i(t), using the
geometry of a spherical cap with a height of
R(t) + Lmax and soil with an initial and final moisture
content of 0 and 0 respectively, is:
i(t) = 3 ( — 0 0) ( 2 [R(t)] 3 + 3 [R(t)]�Lmax — Lmax — 4r0)
(1)
The MPD Infiltrometer (Figure 1) consists of a thin -
walled (2 mm thick) aluminum cylinder with a height
of 45 cm and an inner diameter of 10 cm. A transpar-
ent piezometer tube was attached to the outside of the
device beside a measurement tape for making water
level readings. After any mulch or detached/decaying
plant material was brushed aside the device was
pounded into the soil to a depth of 5 cm and then filled
with water to a height of 43 cm as to not overflow the
device. The water level over time was then recorded
either manually or automatically using an ultrasonic
sensor (1 reading /s averaged over 10 s) (MassaSonic,
M -5000; MASSA PRODUCTS CORPORATION, Hing-
ham, Massachusetts). MPD Infiltrometer measure-
ments were made at a number of locations throughout
each basin, based upon a restricted sampling grid that
avoided bushes, trees, and energy dissipation struc-
tures such as riprap or concrete, and did not otherwise
destroy plantings. The coordinates of each location
were determined using a Trimble ProXR GPS unit
(Trimble Navigation Limited, Sunnyvale, California),
which delivers submeter accuracy (accuracy varies
with proximity to base station) in the correct condi-
tions. Six MPD Infiltrometers were used at a time to
increase the rate of data collection.
The MPD Infiltrometer and the notations used in
the equations are illustrated in Figure 1. The original
Philip -Dunne permeameter technique involved placing
the device in a borehole. The device was therefore mod-
ified to incorporate surface infiltration and capture any
effects of sediment accumulation in the stormwater
BMP. Due to these modifications in the technique, the
methodology described by Philip (1993) for determining
saturated hydraulic conductivity needed to be altered
accordingly. This alteration included changing the
geometry of the source from a sphere to a hemisphere
and accounting for one - dimensional flow through the
soil contained within the bottom of the device.
The equations used in calculating the saturated
conductivity are only applied after R(t) (radius to the
sharp wetted front at time t) is greater than the dis-
tance ri + LmaX, where r is the radius of the cylin-
der and L a. , is the depth of insertion into the soil,
which can be determined from the volume of water
infiltrated and soil porosity. The mass conservation
The same analysis procedure described by Philip
(1993) was followed, which involves working through a
series of equations to solve for the two unknowns, Ksat,
the saturated hydraulic conductivity and � which is
the wetting front suction head for the unsaturated soil.
A computational spreadsheet procedure with the sol-
ver add -in and visual basic application was developed
to find solutions to the equations and automate the
computational process and obtain values ofKsat and 0.
For 57% of the tests, only three data points were
obtained manually with the times corresponding to
full, half - empty, and empty, as recommended by
Munoz - Carpena et al. (2002) for the Philip -Dunne
permeameter. The three -point method, however,
required additional data processing to meet the
requirements of the data fitting procedure used to
determine Ksat• An exponential fit of the three points
was used to generate the necessary water level vs.
time data. For the remainder of the tests, more data
points were obtained and the additional processing
was not required. A capacitance probe (Theta Probe
ML2x; Thermo Fisher Scientific Inc., Waltham, Mas-
sachusetts), which measures the dielectric constant of
the soil, was used to indirectly estimate the initial
and final soil moisture content of the top six cm of
soil in the vicinity of the infiltrometer. A soil specific
calibration using several gravimetric soil moisture
measurements was also conducted for each rain gar-
den. Bulk density measurements, required to convert
gravimetric water content to volumetric water con-
tent, were made using the core method (Klute, 1986).
Tests in which there was minimal change in water
level over —3 -h time period were terminated, suggest-
ing that the Ksat value was less than the smallest
measured Ksat value of 5.6 x 10 -7 cm/s. The calcu-
lated Ksat values for each measurement location were
entered into ArcView to provide a map showing the
spatial variability in Ksat for each rain garden.
The arithmetic mean, geometric mean, and the med-
ian were then calculated for each site. Graphs of the
cumulative distribution of the measured Ksat values
along with the theoretical normal and log- normal
distributions for the mean and standard deviation (SD)
of the data collected were plotted. Visual inspection of
the cumulative distribution plots along with the
JOURNAL OF THE AMERICAN WATER RESOURCES ASSOCIATION 1023 JAWRA