Environmental Engineering Reference
In-Depth Information
(if available). Incoming precipitation can be
intercepted by vegetation, contribute to the
snowpack, fall on impervious material, or infil-
trate the soil. Evapotranspiration may occur
from water in each of these four reservoirs.
Evapotranspiration is calculated on the basis
of water availability and potential evapotrans-
piration as determined by the Jensen-Haise
method (Equation (
2.31
)) or by two alternative
methods.
Precipitation falling on impervious areas
becomes direct runoff. For other areas, direct
runoff is calculated by an empirical method,
similar to the SCS curve number method
(Equation (
2.37
)). Accretion or depletion of the
snowpack is governed by air temperature as
well as precipitation. Infiltration is determined
as the difference between water (either pre-
cipitation, snowmelt, or surface run on from
adjacent areas) arriving at the soil surface and
runoff. (An alternative approach taken in other
watershed models is to first calculate infiltra-
tion with the Green-Ampt equation (Singh,
1995
), for example, and to then set runoff equal
to the difference between water arriving at the
soil surface and infiltration.) Drainage from
the soil-zone reservoir occurs when its storage
capacity is exceeded. A portion (determined by
an empirical factor) of that excess water goes
to the subsurface reservoir; the remainder goes
directly to the groundwater reservoir. Water
can also drain from the subsurface reservoir to
the groundwater reservoir. There is no lag time
in the exchange of water among these reser-
voirs; drainage is assumed to be instantaneous.
Recharge for each HRU consists of all drainage
to the groundwater reservoir. Total recharge
for the watershed is the sum of recharge for all
HRUs. Water in the groundwater reservoir can
discharge to a stream (base flow) or flow out of
the watershed as groundwater flow; ground-
water discharge rates vary linearly with ground-
water reservoir storage.
Since its initial release (Leavesley
et al
.,
1983
) PRMS has undergone multiple revisions
that have added alternative options for calcu-
lating fluxes and have made the model more
user friendly. Improvements include incorp-
oration of a modular modeling system to link
available databases (Leavesley
et al
.,
1996
) and
a GIS tool for delineating HRUs on the basis
of elevation, topography, climate and vegeta-
tion zones, and depth to water table (Viger and
Leavesley,
2007
).
Example: Puget Sound Lowland
Bauer and Mastin (
1997
) used the Deep
Percolation Model, DPM (Bauer and Vaccaro,
1987
; Vaccaro,
2007
), to construct detailed water
budgets for three small watersheds in glacial-till
mantled terrains in the southern Puget Sound
L of w l a n d of f Wa s h i n g t of n . T h e g of a l w a s t of e s t i m a t e
the amount of recharge that penetrated the till.
The DPM is a distributed parameter, physically
based watershed model developed specifically
for estimation of groundwater recharge. Input
requirements and water exchange calculations
are similar to those of PRMS.
The surface of most of the Puget Sound
Lowland is characterized by rolling, hilly
glacial-till mantled areas and level glacial-
outwash bench land. At the three study sites,
the aquifer in the glacial outwash deposits
is overlain by 3 to 20 m of till. From October
1991 through September 1993, discharge was
measured at the mouth of each catchment
using a flume or culvert and a stage recorder.
Piezometers were installed within each catch-
ment to monitor groundwater levels. Soil-water
content was measured at three depths with
time-domain reflectometry probes at the pie-
zometer locations. Precipitation was measured
in each basin; daily maximum and minimum
air temperatures were obtained from nearby
National Weather Service stations. Solar radia-
tion was measured at one location and assumed
to be uniform for all watersheds.
Most precipitation and recharge occurred in
winter at the three sites; little rain fell during
summer (
Figure 3.7
). Average annual recharge
calculated by the DPM ranged from 37 to
172 mm (
Table 3.2
). Thicker till is the appar-
ent reason for the relatively small amount of
recharge at the Clover site. A unique aspect of
this study was the use of groundwater-level and
water-content data in model calibration; simu-
lated and measured soil-water storage were
in good agreement (
Figure 3.8
). Independent
estimates of recharge were obtained by ana-
lyzing soil cores for tritium concentrations and
