Box 3 – Running and Post-Processing the Model for Case Study 1
Supplemental material for running the Case Study 1 model is provided in CaseStudy1–Models.zip can be downloaded at the Groundwater Project. The material includes computer files for running the model described in the Case Study 1: Hypothetical Stream-Aquifer System. The files are designed to allow the reader to reproduce all the simulations and results related to the case study as presented in the Groundwater Project book “Groundwater Resource Development: Effects and Sustainability.”
The zip file contains a folder named “Case Study 1 – Models” with subfolders that include all the input and output files for the three scenarios presented in this book. The subfolder “MODFLOW-NWT.Model” contains a copy of the executable software for MODFLOW-NWT (version 1.1.4), which was used in the analyses of this problem, as described in this book. It also includes a copy of the model documentation report. The computer source code for MODFLOW_NWT and additional documentation can be obtained by clicking here.
The Case Study includes three different scenarios, and the files for each are contained in separate subfolders, labeled (1) Base Case (No Recharge and No ET), (2) Low ET Case, and (3) ET and Recharge Case. Instructions for running a simulation for each case are provided below. The Base Case represents a scenario in which there is no areal recharge from precipitation and no evapotranspiration (ET) losses. The Low ET Case includes mountain front recharge but no areally diffuse recharge. The third scenario (ET.and.Recharge Case) includes both ET discharge (at a higher rate than in the previous case) and areal recharge from precipitation.
The folder for each of the three scenarios includes two subfolders. One subfolder contains all the input files needed to run that simulation and the other subfolder contains all the output files. The input file folder also includes the ModelMuse project file used to generate the input files (you do not need to use ModelMuse to run the simulation, although it is possible to do that). ModelMuse is a USGS public domain model pre- and post-processor. The input folder also contains a batch file that can be used to run the simulation using those input files.
Running the Model:
There are a number of alternative ways that the input files for each scenario can be run with MODFLOW-NWT. We offer one straightforward way that is consistent for the three scenarios. Specifically, we have placed a batch file (“name.bat”) in each input folder (where “name” is the name of the scenario). Double-clicking on this batch file will cause it to execute a script contained within it. The scripts are written to link to the executable version of MODFLOW.NWT contained in the “Model” folder, start executing it, and provide it the name and location of the input files for each scenario. It will route all output files to the “Output” folder. Note that if you run (or re-run) the model with this folder and file configuration, the original output files will be overwritten and lost. If you want to save them for future comparisons, then you will first need to either rename the previous output files or move them to a separate new folder before running the simulation.
Creating head (water level) and drawdown contour maps using ModelMuse:
To start, if you have not already installed ModelMuse, download and install it by visiting the United States Geological Survey, ModelMuse web site: https://www.usgs.gov/software/modelmuse-a-graphical-user-interface-groundwater-models.
Once it is installed, go to the Case Study 1–Models folder, then down to the Base Case folder, then into the Input.Files folder and double click on the file “Base.Case.gpt” to open it in ModelMuse. You may want to stretch the ModelMuse window so you can see the entire model grid as shown in Figure Box 3-1. Model rows extend from left to right starting with row 1 at the top; model columns extend from top to bottom starting with column 1 on the left. The main window is a plan view of the grid, the lower window is the front-view (i.e., the cross section of the row highlighted in green on the plan view), and the window on the right is the side-view (i.e., the cross section of the column highlighted in blue on the plan view). The blue lines on the plan view indicate the model column shown on the right and the green lines indicate the model row shown at the bottom.

If you have not run the model directly from ModelMuse, then you need to import the data set you wish to contour. Under “File,” select “Import >Model Results” and navigate to the Case Study 1–Models folder, the Base Case folder and the Output.Files folder and select the file Base.Case.fhd and choose Open. A list of all the model simulation times for which the file contains head data will be shown and by default the last time (73051 days) will be selected. You can set “Display choice” to “Contour grid” and select OK and click on “Update the existing data sets with new values” to view a contour map of those heads, or you could scroll down to uncheck that and choose any other time. Figure Box 3-2 shows the head distribution for “Head: Period 1; Step: 1; Total Time: 1” which is a contour map of head for the undeveloped conditions. To change the default contours, open the Data Visualization dialog box and select Contour Data from the list on the left, then in the upper right enter a different contour interval and click apply and close. The drawdown distribution at a time that is 73,051 days after pumping began can be viewed by importing data from that time from the Base.Case.fdn file (Figure Box 3-3). Drawdown is the difference between the heads at two different times, typically predevelopment and a given time after pumping began.


To view smaller magnitudes of drawdown further from the well, we suggest adding two new contours by opening the Data Visualization dialog box and selecting Contour Data from the list on the left, then click Specify Contours in the upper right. Increase the number of rows from 6 to 8. Then scroll down and set the contour values for the two new rows to 0.01 and 0.05. Next click “OK.” and “Apply” and then “Close” in the Data Visualization dialog box. There are many options available for coloring and labeling the contours; feel free to experiment. Zooming in on the area around the pumping well shows that the drawdown at the well is on the order of 0.6 m and along the river is close to (but less than) 0.01 m.

To remove grid lines, under View, select “Show or hide 2-D Grid>Show Exterior.” To improve clarity of drawdown values and contours you can also select “Hide all objects” under the “Objects” pull-down menu. To export and save the image of the contour map, go to File>Export>Image (or click the camera icon). In the resulting dialog box, you have several options, but just click “Save image” to generate a file (select format or type) with the contour map. Click “Close.” If you prefer that the contours be a single color, you can change the color scheme to “Blue only” or “Black only” in the Data Visualization dialog box. The resulting image is shown with black contours in Figure Box 3-5.

Extracting and plotting MODFLOW budget data using GW_Chart software:
GW_Chart can be used to conveniently extract the budget data from the main MODFLOW or MODFLOW-NWT output file. If you have not yet installed GW_Chart, download and install it by visiting the United States Geological Survey, GW_Chart web site: https://www.usgs.gov/software/gwchart-program-creating-specialized-graphs-used-groundwater-studies.
Once it is installed, open “GW_Chart.” Under pull-down menu for “Chart Type/Convert,” select “Water Budgets.” Then in the lower right area, select “MODFLOW” under “Read Data From.” Next, under “File,” select “Open,” and then navigate to the output folder for the Base Case and select “Base.Case.lst” to open that file and allow GW_Chart to read all the water budget data. This will generate plots for all selected variables for either cumulative or rate budgets. However, these plots are not at a high resolution and do not provide the numbers.
Therefore, we want to Save/export data using the lower-right middle button (); name the output text file (e.g., “Base.Case.Budget.txt”) and select a destination folder. This will generate a text file that contains all the saved budget data for the model (as specified in the “output control” [“.oc”] input file).
Open the “Base.Case.Budget.txt” file with Notepad or other word/text processing software. The file lists “cumulative” data first, followed by a listing of “rate” data. For this exercise, we will work with annual rates. Select all lines in the “RATES” category (bottom half of the listing (lines 206 to 408), and then select “COPY.” Then open a blank Excel workbook, select the upper left cell (A1) and paste using the “Text Import Wizard.” Click “next” and then select delimiters that clearly and properly separate the data columns (specifically, select “tab” and deselect “space”; the latter step will assure that the headers in row 2 align correctly with the proper data columns below). Then click “next” and “finish” to complete the import process. You can give the worksheet a name (e.g., by changing default name of “Sheet1” to “Base.Case” in the lower left tab). Save the spreadsheet file to the Output.Files folder, giving it an appropriate name (e.g., “Rate.Budgets”).
Examine the headers in Row 2. If these column labels did not line up accurately with the data columns, adjust them manually for improved clarity in the spreadsheet. For improved clarity, select (highlight) all of Row 2 and then click on “Wrap Text” to see complete labels. You can also adjust column widths and number formats for data as desired. After completing these several steps, the first 11 rows (out of 203 Rows) and 17 columns (of 19) should look something like Figure Box 3-6.

To assess the sources of water to the well, and how these change with time, we need to complete a few more calculations, which are aided by the use of formulas in Excel. We will examine these changes over the 200-year simulation period, so it will be convenient to create a column for time in years (because the model units of time are days) so that variables can be plotted in terms of years. Therefore, in Column T, add a label in Row 2 (something like “Time, in years”). Then in Row 3 of Column T, add a formula to convert time in days (Column C) to time in years (assume one year equals 365.25 days). Copy and paste that formula (“=(C3/365.25)-0.00273785”) into the remaining cells of Column T. (Recall that the length of the initial steady-state stress period is arbitrary, and was set at 1 day, thus the subtraction of approximately 0.00273785 years) to account for the steady state portion of the simulation.
One source of water to the well is from a change in storage in the aquifer. In the MODFLOW budget terminology, “In: STORAGE” refers to the water that enters (flows into) the groundwater system by coming out of storage in the aquifer. We will use Column U to compute the net change in aquifer storage. So, in Row 2 of Column U, add a label “Net Change in Storage” or something similar. In Row 3 of Column U, insert a formula to compute net change in storage as the difference between the value in Column D and that in Column K (“=D3-K3”). Then copy that formula and paste it into every remaining cell in Column U. With this order of subtraction, the results with a positive sign will represent a reduction (or depletion) of storage.
Next, we want to calculate capture, which in this case can only include the capture of streamflow (1) by increasing the seepage losses from the stream (“In: STREAM LEAKAGE”), which equals recharge to the aquifer, and/or (2) by decreasing the discharge from the aquifer to the stream (“Out: STREAM LEAKAGE”). So, we need to calculate how these change with time. First, we compute the increase in recharge from the stream in Column V. Add a label in Row 2 of Column V (something like “Increased Seepage Loss from River”). The increase in seepage loss in any time step during the transient stress period equals the difference between the “In: STREAM LEAKAGE” during that time step and the respective value during the initial steady-state stress period when there was no pumpage from the well. Set up a formula for that column to compute these values (the formula should look something like: “=I4-I$3” in Row 4 of Column V). Follow a similar procedure for “Decreased Groundwater Discharge to River” by entering “=P4-P$3” in Row 4 of Column W). Then in Column X, calculate the Capture by adding the absolute values of Columns V and W, that is “=ABS(V4)+ABS(W4)” in Row 4 Column X. Finally, we want to compute the nondimensional fractions of the sources of water to the well. That is, we want to compute the storage depletion and capture fractions. Use Columns Y and Z to calculate the storage depletion fractions and capture fractions, respectively, by dividing Columns U and X by the well pumpage (Column M). Note that the sum of these two columns should always equal 1. Plot the change in the storage depletion and capture fractions over the 200-year simulation period (using Excel or your favorite graphic/plotting software package). The graph should look like Figure Box 3-7.

Computing streamflow changes in Excel
To determine how streamflow varies in a downstream direction, first use a text editor to open the main listing (output) file “Base.Case.lst” from the Base Case simulation (folder: “Case Study 1–Models\Base Case\Output.Files”, file “Base.Case.lst”). Find the stream listing data at 200 years (time step 200 in stress period 2) by searching for “STREAM LISTING PERIOD 2 STEP 200”, which will be just above the last budget print out at the bottom of the file. Then copy the label lines and data (a total of 83 lines) and then paste it into cell A1 of an Excel spreadsheet. Next click on Data>Text to Columns, and choose “fixed width”, then “next”, then “finished.” Delete all data except the stream reach number and the flow out of the reach (these are Columns E and H of the pasted data). Now, the reach number should be in column A in sequentially increasing order. Insert two new columns after column A to compute the distance downstream in meters and kilometers, respectively. Use formulas based on the knowledge that each reach covers a distance of 804.67 m. Column D should then include the streamflow out of each reach (in m3/d) after 200 years of pumping. Next use Excel or other plotting or graphic software to plot the results, which should look like Figure Box 3-8.

To examine the stream losses and gains, add Column E with the label “Stream Loss(-) Gain(+)”. In cell E4 enter the formula “=D4-20000” because there is a stream inflow of 20,000 m3/d at the upstream end of the stream. In cell E5 calculate the difference between outflow in the reach and outflow in the upstream reach “=D5-D4” and copy the formula down the column. Because we are subtracting the flow at the upstream end of the reach from the downstream end, a negative value indicates stream loss in the reach and a positive value indicates gain. The result is shown in Figure Box 3-9.

To determine the impact that 200 years of pumping had on the streamflow as well as the gains and losses, we want to add the predevelopment data to the previous plots. Follow the above steps for the predevelopment conditions by searching in the same “Base.Case.lst” file for “STREAM LISTING PERIOD 1 STEP 1”, which will be close to the top of the file (for us it is line 1675 of the Base.Case.lst file.) As done before for 200 years, copy the data into the spreadsheet. Make similar calculations as done for the 200-year time and add the data to the graphs. They should look like Figure Box 3-10.

Plotting well hydrographs from MODFLOW output using GW_Chart:
GW_Chart can be used to extract and plot water-level changes over time at individual nodes (cells) of the MODFLOW grid. This produces a well hydrograph showing either water levels (heads) or drawdown (changes in water level from an initial value), as selected by the user.
To do this, first if you have not yet installed GW_Chart, download and install it by visiting the United States Geological Survey, GW_Chart web site: https://www.usgs.gov/software/gwchart-program-creating-specialized-graphs-used-groundwater-studies.
Once it is installed, open “GW_Chart.” Then, under the pull-down menu for “Chart Type/Convert,” select “Hydrographs.” In the Data box, make sure that “MODFLOW head or drawdown file” is selected. Next, set Column-Row-Layer to the cell location for the pumping well (30,40,1 for the Base Case). Then under “File,” select “Read Heads or Drawdown,” and then navigate to the output folder for the Base Case and select either the formatted head file or the formatted drawdown file. We chose to view drawdown.
To get a better-quality graph with a time scale in years, we need to place the data into a spreadsheet or plotting package. You can select (highlight) all the data in the two columns of data listed in the upper middle box of GW_Chart, and then copy it (Ctrl-c). Next, paste it into cell A1 of a blank Excel worksheet. Alternatively, under the File pull-down menu, you can select “Save Heads or Drawdowns” to save the two columns of data to a separate new file. The first time shown is “1”, which represents the arbitrary length of the initial steady-state stress period. There is no drawdown for this stress period, but MODFLOW uses the first period as the reference time and the final converged head is not known at the start of the calculations, so it calculates the change in head from the initial conditions to the converged steady state conditions. Therefore, change the drawdown value for the first time “1” to a value of 0 in Column B. Add a header (or label) to column C of the spreadsheet “Time, in years” and then define the values in that new column by a formula to convert time in days (Column A) to time in years (for example, the formula for C3 would be “=(A3-1)/365.25”). Copy and paste that formula into rows 2 through 202 of Column C. Then plot Column C on the horizontal axis and Column B on the vertical axis to produce the hydrograph as shown in Figure Box 3-11.

Streamflow changes with time are recorded in the Gage output file “Base.Case.sfrg1”. In this case, we defined a stream gage as being located at the most downstream cell of the river (reach 80), representing surface water outflow from the system being simulated. Streamflow at the gage for the Base Case is illustrated in Figure 3 Box 3-12.
