Environmental Engineering Reference
In-Depth Information
Spatial and temporal variability are import-
ant both in constructing a conceptual model
and in selecting and applying methods. Diffuse
recharge can occur somewhat uniformly across
the area of interest at specific times during the
year, but patterns of land use or soil permeabil-
ity can result in a nonuniform distribution of
recharge. Focused recharge occurs at specific
points in a watershed; a stream can be a source
of focused recharge during part of the year but
a sink for groundwater discharge at other times
of the year.
Space and time considerations must also be
tied to study objectives. What types of recharge
estimates are desired? Long-term averages?
Annual or monthly values? Historical rates of
recharge for periods prior to human develop-
ment? Is an average estimate required for an
entire aquifer or watershed? Or is information
on spatial variability needed for an aquifer
vulnerability study? Answers to these ques-
tions will help guide selection of appropriate
methods.
Each estimation method is associated with
specific space and time scales. The spatial and
temporal scales of interest in a study need to
be matched with those of the selected estima-
tion techniques. If one is interested in recharge
estimates for a large watershed (say in the
order of hundreds of square kilometers), then
methods that integrate over large areas (e.g.
water-budget methods, streamflow hydrograph
analysis, groundwater-flow modeling) might
be preferable to methods that provide an esti-
mate at a point in space (e.g. the zero-flux plane
method). This is not to say that point estimates
are not useful in large basins. Integration of
multiple point estimates within a watershed
can produce a meaningful watershed-wide esti-
mate if sufficient data points are available. New
modeling tools, such as combined watershed/
groundwater flow models (
Section 3.6
), can pro-
vide estimates of recharge at discrete locations
and times in a watershed, while also providing
an average estimate for the entire watershed.
In term of time scales, some methods provide
recharge estimates for individual precipitation
events. Other methods produce estimates that
might be averaged over a single year, multiple
years, several decades, or even centuries and
millennia. If a study site is in an area of shallow
water table where recharge occurs consistently
on a seasonal basis, then methods based on trac-
ers in the unsaturated zone may be of only lim-
ited use. Bomb pulses of tritium and chlorine-36
most likely would have been flushed out of the
unsaturated zone in these areas. If variability
in interannual recharge was also a concern at
this site, then natural environmental and his-
torical groundwater tracers would likewise be
of little use, because these tracer methods pro-
vide an average recharge rate that is integrated
over several years.
Time constraints of the study are also
important. If recharge estimates need to be
developed in a short time (months), then tech-
niques based on long-term monitoring (several
years) are clearly inappropriate if the requisite
data are not available. Tracer techniques may
be suitable for a short-term study because they
usually require only a one-time sampling and
provide an estimate of recharge averaged over
multiple years.
As with time constraints, financial con-
straints often are a controlling factor in select-
ing methods. Analyses of unsaturated-zone
cores for tritium and chlorine-36 can be very
useful in some arid regions. However, the cost
of obtaining the samples (especially from thick
unsaturated zones) and analyzing the samples
in the laboratory may be prohibitively expen-
sive for some studies. One point should be clear,
though. The cost of applying a method does
not necessarily correlate with the accuracy or
appropriateness of the method. The fact that one
method costs considerably more than a second
method should not imply that the results of the
first method are any better than those of the
second. Similarly, just because a method is very
inexpensive to apply does not mean that there
is no value in using the method. Any method
that is affordable and easy to apply merits
consideration.
The anticipated accuracy of estimates
obtained with any method should be evalu-
ated as thoroughly as possible before applying
a method. This is no simple task. It is difficult
to apply formal error analysis because there are
