Monitoring Plant and Groundwater Interactions - Introduction to Phytoremediation of Contaminated Groundwater - page 231

Multispectral remote sensing is the collection of
reflected, emitted, or backscattered energy from objects
across the spectrum of electromagnetic energy (Jenson
2000). Hyperspectral sensing can be used to indicate areas
of groundwater discharge and possible consumption of this
groundwater by phreatophytes. This technique can detect
areas of constant wetness (Batelann et al. 2004).

the linear relation between ET and depth to water table
below the ET surface is replaced with a non-linear curve
that consists of segments of curves of different slopes
defined by the user. This flexibility allows the modeler to
more accurately simulate the higher zone of transpiration
that occurs as the water table is within the root zone
of phreatophytes, but it still assumes, like the original
MODFLOW EVT, that the ET rate will decrease as the
depth to water table increases below the ET surface. It also
assumes, like the EVT, that the ET rate will increase as the
depth to water table decreases. In fact, many plants will die
and transpiration will stop if the roots are flooded by a high
water table for a period of time such that oxygen levels are
depressed.
Baird et al. (2005) presented a simple solution to these
limitations. They developed two models that can take water-
level data output from MODFLOW as input for a Riparian
Evapotranspiration Package (RIP-ET); MODFLOW-2000
(Maddock and Baird 2003) and RIPGIS-NET (Ajami et al.
2011). The linear relation between ET and depth to water
below the ET surface is replaced, in their models, by a set of
plant-based physiological curves, or plant functional groups
(PFG) that are representative and unique for different plant
types, water tolerances, and root-depth ranges. This model
decouples the E and the T in ET . For example, rather than a
linear relation between ET and head decline, Baird et al.
(2005) use a Gaussian distribution, or multiple, nonlinear
segmented flux curves (Baird and Maddock 2005) such that
at depths near extinction and at higher water tables, the ET
rate is lower. As might be expected from the areas under
these curves, the MODFLOW EVT method will typically
overpredict ET relative to that estimated from the RIP-ET
package. Maximum ET occurs when the water table is within
the maximum root density, but with sufficient air supply to
support respiration. Also, these models account for the oppo-
site scenarios, where the water table increases and decreases
the availability of oxygen to support root respiration.
A novel development of how to simulate ET in numerical
models was reported by Shah et al. (2007). These authors not
only suggest changes to how ET rates and the depth to water
table are simulated but also recognize that plants will
remove water from the unsaturated zone. They state that
when the water table is within 1-2 ft (0.3-0.6 m) from the
land surface most of the simulated ET is derived from
groundwater. As the depth to water table increases, less ET
water is derived from groundwater and more is derived from
the unsaturated zone. Moreover, rather than using a linear
relation to describe this relation between ET and depth to
water table as is used in MODFLOW, an exponential decay
function is more appropriate.
The Simultaneous Heat and Water (SHAW; Flerchinger
1991) model can be a useful tool to investigate the water
budget of a site where plants will be installed. The SHAW

9.3.6 Tracers

The introductions of dyes, or organic compounds that appear
colored, have a long history of use in hydrogeologic studies.
Dyes such as rhodamine are used often in surface-water
studies to detect movement and dilution patterns. Dyes also
have been used in groundwater studies to investigate
groundwater flow in karst systems. Dyes also have been
used in plant studies to determine the source of water
being used by the plants. Robinson and Donaldson (1967)
added the fluorescent dye pontacyl brilliant pink to the soil
surrounding the roots of the woody phreatophytes willow
and wildrose during the summer. Fluorometric analysis of
the leaves indicated the presence of the dye in the leaves,
roots, stems, and in gas bags that were used to capture
transpiration. Moreover, the detection of the organic dye
molecule in the leaves provides early evidence that these
plants could take in a large organic molecule from the soil-
water solution and translocate it throughout the plant's vas-
cular system. Moreover, the use of radioactive isotopes, such
as tritium and 32 P, as tracers can also provide information on
the source of water used by trees.

9.3.7 ET and Groundwater Models

The effect of plants on groundwater levels can be evaluated
using numerical models. The USGS model MODFLOW
contains a module, called the ET Package (EVT), that
simulates the interaction between groundwater and vegeta-
tion. EVT simulates ET based on the simulated water level,
or head, in a model cell, where the head ranges from a
maximum elevation (ET surface) to a minimum elevation
(extinction depth). Maximum ET rates occur at or above the
ET surface; at the extinction depth, ET is zero. Between
these two extremes, the ET rate is variable and is linearly
dependent on the depth of the head below the ET surface.
These relations cannot be changed regardless of the
differences in groundwater use by different phreatophytes,
such as obligate or facultative, or wetland or transitional
plants.
There are alternatives to this approach of a linear
decrease in ET with increased depth of water table below
the ET surface. One is provided by Banta (2000), in which