pictures/imod_logo.png

iMOD User Manual version 4.4 (html)


11.11Tutorial 11: Unsaturated Zone Package

This tutorial gives an introduction to a transient implementation of the Unsaturated Zone package (UZF), see Section 12.31.


Outline

This is what you will do:


Required Data

For this tutorial you need the following iMOD Data Files/folders:


Getting Started


Create a PRJ file

We will quickly generate a PRJ file using the auto fill option of the Project Manager.

Your window should look as follows:

pictures/tutorial13/config_top.png

Figure 11.148: Example of the Define Characteristics Automatically window.

In this manner we tell iMOD to add the given files for model layer 1,2 and 3 automatically. Let’s see how that works.

iMOD will pop-up a window with the files found in the Automatic Package Allocation window.

pictures/tutorial13/config_top_files.png

Figure 11.149: Example of the Automatic Package Allocation window.

Here we can inspect the results, or even modify this. We leave it like this as it is correct and continue.

Also constant values for a package or modules can be inserted in this manner, as the boundary conditions of our three-layered model are all equal one, we insert this information as follows:

That’s convenient, right? Let’s add the constant information for the other packages repeating the steps 12 up to 17.

Next thing is to add the confined storage coefficients (STO).

Almost done, we will add the Evaporation package (EVT) on the conventional way as it will be defined for our first stress period only.

Note: By defining a surface level (SUR) of 100.0 m+MSL and an extinction depth (EXD) of 8.0, there will be a linear reduction of evaporation with depth. At a depth of 8 m+MSL (2 meter above the bottom of model layer 1), the evaporation is 0.0.

Precipitation is the input parameter that varies per stress period. We will add that parameter with the automatic package allocation.

We have add now 459 definitions for precipitation in a single mouse click, that’s awesome isn’t it. Final thing to do is to add the configuration for the PCG solver.

Okay, we’re done, let’s save the modelling project.

Note: {installfolder} refers to the full path of the directory you installed iMOD in (e.g. D:\iMOD).


Simulate the model

With this project file we generate a standard MODFLOW2005 model, let’s do that.

pictures/tutorial13/simulation_config_layertypes.png

Figure 11.150: Example of the Layer Types window: assigning layer type ’Convertible (HNEW-BOT)’ to layer 1.

iMOD has found out what the transient content of your model is, it starts at 6\(^{\rm st}\) of October 2013 00:00:00 and ends at the 7\(^{\rm st}\) of January 2015 00:00:00. based on the entered precipitation (RCH) and evaporation (EVT) data.

pictures/tutorial13/simulation_config.png

Figure 11.151: Example of the iMOD Define Simulation Configuration window.

The model will generate results on a daily time step which is based on the occurrence of the input data, we can inspect this.

pictures/tutorial13/simulation_config_stresses.png

Figure 11.152: Example of the iMOD Time Discretization Manager for Simulation window.

We will leave the definition of the stress-periods for now.

iMOD will now create the necessary MODFLOW2005 file and runs the model, as the model is tiny, this will be finished rapidly. It will start the simulation directly thereafter. You’ll see that the model start in a separate DOS-command window and it will echo the simulation progress. As it is a transient simulation with 458 stress periods, it will consume probably 10 seconds to accomplish.


Inspect the result of simulation

Let’s inspect the hydraulic head of the first model layer and the computed recharge (equal to the input in fact) and generate time series.

iMOD will load all selected results files into the iMOD Manager and displays the result on the graphical canvas. Use your experience learned from the previous Tutorials to display the computed heads as time series as shown in the following figure.

pictures/tutorial13/results_rch_evt.png

Figure 11.153: Time Series of computed groundwater levels and precipitation.

Note: Modify the settings of the time series to assign the recharge from the BDGRCH on the second y-axes and use “BlockLines” as LineStyle.

As you may have noticed, the groundwater levels respond directly to the net-recharge. This is often, especially by these deep groundwater levels (> 2 meter depth) unrealistic. A delay of recharge through the unsaturated zone is a process that is taken care of by the UZF package.


Creating the UZF input

So, this tutorial is called UZF Package, but up to now we didn’t do anything with that. Well, most of the preparation we have done and running the model with the RCH and EVT packages, gives us a nice comparison of two different commonly used concepts. First we will delete the RCH and EVT packages.

Now we will add the input for the UZF package which is more-or-less a combination of the RCH and EVT package input.

Below is an example of your current window.

pictures/tutorial13/config_uzf.png

Figure 11.154: Example of the Define Characteristics Automatically window.

Let’s gather the appropriate files.

iMOD will combine the constant given values with the wild cards. At those location where the “inherent” is mentioned, iMOD will use the previous mentioned value/file.

pictures/tutorial13/config_uzf_files.png

Figure 11.155: Example of the Define Characteristics Automatically window.

We will leave it like it is.

It can take a little bit longer to refill the Project Manager. But once it is finished, we’re ready to start the model.


Model Simulation

iMOD will now create the necessary MODFLOW2005 file and runs the model, it will consume probably a little bit longer than our previous model to accomplish.


Inspect the result of simulation

Let’s inspect the hydraulic head of the first model layer, the computed groundwater recharge, the precipitation and evaporation and generate time series.

iMOD will load all selected results files into the iMOD Manager and displays the result on the graphical canvas.

pictures/tutorial13/results_uzf.png

Figure 11.156: Time Series of computed groundwater levels with the RCH and EVT and the UZF package.


Modify the parameters of the UZF package

As you may have noticed, the UZF package generates some ground water recharge whenever multiply rainfall events appear on a short notice (dark green line). Short, heavy rainfall events, may not even feed the groundwater due to the strong evaporation or instant surface runoff. The evaporation has been modelled with an extinction depth of 8 meter, this may be a bit unrealistic and deplete the groundwater more than it should; let us change some parameters to see the effect of this.

First we should know what the Brooks-Corey Exponent (\(\epsilon \)) represents. The Brooks-Corey Exponent is used to compute the unsaturated hydraulic conductivity \(K(\theta )\) as a function of the moisture content \(\theta \).

\begin{equation} K \left ( \theta \right ) = K_s \left [ \frac {\theta - \theta _r}{\theta _s - \theta _r} \right ]^\epsilon , \end{equation}

whereby \(K_s\) is the saturated hydraulic conductivity; \(\theta _r\) is the residual water content; \(\theta _s\) is the saturated water content; and \(\epsilon \) is the Brooks-Corey exponent. This function describes how the hydraulic conductivity approaches the saturated conductivity as the water content \(\theta \) increases. This can be a linear relation ship (\(\epsilon =1.0\)), or a non-linear whereby the hydraulic conductivity increases more slow than the water content increases (\(\epsilon > 1.0\)) or faster than the water content increases (\(\epsilon < 1.0\)). The next figure shows the relation ship for different values of \(\epsilon \) and the effect on the hydraulic conductivity \(K(\epsilon )\).

pictures/tutorial13/brooks_corey.png

Figure 11.157: Empirical relation between water content (\(\theta \)) and hydraulic conductivity \(K(\theta )\) for different values for the Brooks-Corey Exponent (\(\epsilon \)).

In our case, it seems that our model has too less amount of recharge. In order to increase the recharge we need to decrease the Brooks-Corey Exponent (\(\epsilon \)). The hydraulic conductivity will increase more rapid for a slight increase of the water content, allowing precipitation seeps through the subsoil more quicker.

The easiest way to modify the parameters for the UZF package is to re-read the input, let’s do that:

If you have successfully carried out the above mentioned steps, the time series should like the figure below.

pictures/tutorial13/results_uzf2.png

Figure 11.158: Time Series of computed groundwater levels for the combination RCH-EVT and the two variants with the UZF package.

With a Brooks-Corey (\(\epsilon \)=2.0) the computed hydraulic head is significantly different and almost align (though more smooth) with the RCH-EVT combination. If you examine the water balance in the output files (*.LST-files), the following tables can be derived to clearly show the influence of the UZF package on the net recharge of the aquifer.

Table 11.9: Summary of water balance for the different model configurations for the unsaturated zone (\(uz\)) and saturated zone (\(sz\)).

Parameter RCH-EVT UZF(\(\epsilon \)=4) UZF(\(\epsilon \)=2)
Precipitation\(_{uz}\) - 2365470 2365470
Evaporation\(_{uz}\) - -2053611 -285329
Seepage\(_{uz}\) - -403764 -2311417
Net Storage\(_{uz}\) - -86867 201259
Total\(_{uz}\) - -5038 -954
Recharge\(_{sz}\) 2365470 403764 2311417
Evaporation\(_{sz}\) 2075648 -878899 -2107888
Net Storage\(_{sz}\) 290137 475197 203549
Total\(_{sz}\) -315 62 -20

The UZF package is able to reduce the net recharge significantly, the Brooks-Corey Exponent (\(\epsilon \)) is very sensitive in the amount of ground water recharge that seeps through the unsaturated zone and influences the behaviour of the groundwater level. Feel free to experiment more with the UZF parameters and observe how the influence the model outcomes. If you decrease the (SWC) Saturated Water Content of Unsat. Zone (IDF), the yielding hydraulic heads will be almost similar to the RCH-EVT combination. It is also interesting to start with a very wet subsoil (IWC) Initial Water Content (IDF), and see how the model responds on that.

Note: It is possible to examine per cell the moisture content in more detail, especially during the model building phase this could be desirable. This functionality is not steered via iMOD, but can be easily added to the MF2005 file directly, please visit the website of the USGS for this adjustments.

Note: It is possible to examine per cell the moisture content in more detail, especially during the model building phase this could be desirable. This functionality is not steered via iMOD, but can be easily added to the MF2005 file directly, please visit the website of the USGS for this adjustments.