pictures/imod_logo.png

iMOD User Manual version 5.2 (html)


12.13PWT Perched water table package

A perched water table (or perched aquifer) is a (temporary) water table that occurs above the regional groundwater table in the unsaturated zone. This occurs when there is a (relatively) impermeable layer above the regional groundwater table in the unsaturated zone. With the PWT-package a perched water table can be schematized in iMOD, the perched water table concept is given in figure 12.4.

pictures/runfile/pwt_figure_1.png

Figure 12.4: Conceptual schematization of a perched water table.

In the following pages the concept of the perched water table package is described and illustrated by several hydrologic situations. Hereby, the following figure (figure 12.5) is used which that represents the perched water table in terms of model parameters. Important to understand is that there can be only a single perched water table in each vertical column. Once a perched water table exists, both the horizontal and vertical flow component will be reduced up to zero when the pressure head above the perched water table drops below the top of the aquitard that creates the perched water table.

pictures/runfile/pwt_figure_2.png

Figure 12.5: Conceptual schematization of a perched water table in a groundwater model.

The PWT package is applied using the following assumptions, these are described in the following table in more detail.

I.

There is no storage in the PWT aquitard and the driving force for vertical flow equals the pressure head of layer x minus the top of the PWT-aquitard.

There are two situations to distinguish

  • No Perched Water table
    This situation is depicted on the left figure, the perched water table is below the top of the aquitard yielding a zero flux through the aquitard

  • Thickness of a perched water table
    This situation is show on the right figure, in this particular case the vertical flux through the aquitard is computed as:

    \begin{equation} \begin {array}{l} \text {d}H/(H_i-H_{i+1}) \; \text {whereby} \\ \text {d}H = H_i-T \; \text {: thickness of the perched water table} \\ T : \text {top of the aquitard} \\ H_i : \text {pressure head of modellayer }i \\ \end {array} \end{equation}

pictures/runfile/pwt_figure_3.png
Schematization of vertical fluxes using the PWT-Package

II.

The model cells with the PWT-package are considered to be the top most layer with saturated groundwater.

If the PWT cells are not within the first model layer the transmissivity above the PWT cells is recalculated. On the figure left, the transmissivities are \(10\) and \(10\) for the first two modellayer above the PWT. Since the PWT package will compute the transmissivity only for the first modellayer above the PWT layer, iMODFLOW will redistribute the transmissivities such that they all are lumped in the first modellayer above the PWT layer. This is shown in the figure right. Now, the transmissivitiy of the first modellayer is \(0.01\) (actually this is equal to the parameter MINKD in the runfile) and the first modellayer above the PWT layer has \(20\).
pictures/runfile/pwt_figure_4.png
Schematization of transmissivity when the PWT-Package is used in second model layer

III.

The model cells with the PWT-package are considered to be unconfined and thus also have a phreatic storage coefficient.

In order to compute the effective transmissivity \(T_\text {e}\), the permeability is computed initially by \(k=T/(TOP_\text {aquifer} - TOP_\text {aquitard})\). This permeability is used to compute the transmissivity \(T_\text {e}\) as function of the pressure head as \(T_\text {e}(h)=k(h-TOP_\text {aquitard})\).

IV.

The model cells in layer \(i+1\) have a phreatic storage coefficient, unless the pressure Head of layer \(i+1\) is greater than the bottom of the PWT aquitard. In this case an elastic storage coefficient is used.

In case the underlying aquifer becomes unconfined, and therefore the elastic storage coefficient is used, this is illustrated in the figures below. pictures/runfile/pwt_figure_5.png
Schematization of transmissivity when the PWT-Package is used in second model layer

The numerical implementation is such that the horizontal conductances and the vertical resistances are calculated on the heads of timestep \(t-1\). This is in order to avoid numeric instability.