Environmental Engineering Reference
In-Depth Information
runoff) is derived entirely from groundwater
discharge, several other potential sources
require consideration. These sources include:
stream bank storage; snow and ice packs; slowly
draining natural features, such as wetlands,
which serve as surface-water reservoirs; and
human-developed diversions such as dams,
flood-storage basins, and water withdrawal and
return structures. The effect on streamflow of
slow, steady release of water from these sources
may be indistinguishable from that of ground-
water discharge (Halford and Mayer,
2000
).
Hence, methods of estimating base flow could
incorrectly interpret these releases as ground-
water discharge. It is difficult to assess the
magnitude of these release rates from stream-
flow record alone. Comparison of stream stage
and groundwater elevations over time and use
of local-scale groundwater flow models (
Section
3.5
) can aid in quantifying the rate at which
water is released from bank storage. Tracers,
such as specific conductance or temperature,
could possibly be used to identify the different
sources of streamflow (
Section 4.6
). But per-
haps the safest course of action is to identify
periods when the contribution to streamflow
from these storage reservoirs may be import-
ant and to simply remove those periods from
the analysis.
Measured or estimated base flow is often
divided by the surface drainage area at the
measurement point to derive an average rate in
units of length per time, such as millimeters
per year (mm/yr). This procedure assumes that
aquifer boundaries coincide with watershed
boundaries and that the area of the aquifer that
contributes to groundwater discharge is iden-
tical to the surface drainage area (Kuniansky,
1989
; Rutledge,
1998
,
2007
). In fact, ground-
water basin and watershed boundaries can dif-
fer (Tiedeman
et al
.,
1997
). Miscalculation of the
aquifer contributing area will lead to a propor-
tional error in recharge estimates.
The base-flow index, BFI, is the ratio of mean
annual base flow to mean annual streamflow.
The BFI offers a convenient means for compar-
ing the relative contribution of groundwater
to streamflow at different locales. Typically,
the BFI ranges between 0.4 and 0.8 (Stricker,
1983
; Kuniansky,
1989
). Streams that are not
connected to an aquifer could have a BFI of 0.
Winter
et al
. (
1998
) determined BFIs in excess
of 0.9 for streams in parts of Michigan and
Nebraska.
4.2 Stream water-budget methods
The water-budget equation for a reach of stream
can be written:
Q
sw
= +− −
+++ +
Q
sw
P
ET
sw
∆
S
sw
out
in
off
(4.2)
RQ Q
Q
trib
inter
seep
where
Q
sw
out
is streamflow (discharge) at the
downstream end of the reach,
Q
sw
in
is streamflow
at the upstream end of the reach,
P
is precipita-
tion falling on the stream,
ET
sw
is evaporation
from the stream, Δ
S
sw
is change in stream stor-
age,
R
off
is surface runoff to the stream,
Q
trib
is
flow from tributaries (or to diversions) within
the reach,
Q
inter
is i inter (flow (flow f from t he u nsat-
urated zone; this term is sometimes lumped in
with
R
off
), and
Q
seep
represents exchange of water
between the stream and the subsurface.
Q
seep
may be positive, indicating groundwater dis-
charge or base flow (
Q
bf
) to the stream, or nega-
tive, indicating stream loss to the subsurface.
In either case,
Q
seep
may represent recharge.
Base flow is indicative of diffuse recharge (as
described in the previous section) and can be
expressed as a volumetric flow rate (L3/T)
3
/T) or as a
volumetric flow rate per contributing unit area
(L/T). Stream loss, on the other hand, indicates
focused recharge, recharge from a surface-wa-
ter body; such recharge can be expressed as a
volumetric rate or as a rate per unit length of
stream (L
2
/T).
Q
seep
is often determined as the residual in
Equation (
4.2
); all terms in the equation (except
Q
seep
) are independently measured or estimated.
Generally, the surface-flow terms dominate
Equation (
4.2
), and measurement times are often
selected so that precipitation, runoff, and inter-
flow are zero.
ET
sw
and Δ
S
sw
can be estimated
independently, but for practical applications
on naturally flowing streams the magnitude
of these terms is generally quite small relative
