HW#14 - Monte Carlo Simulation

Note: You may work in pairs on this assignment.

In this exercise we will take the MODFLOW model we calibrated for an earlier assignment and perform a Monte Carlo analysis. Before you begin, download and unzip the following archive:


Then open the texmod.gpr project inside GMS. You should see the following model:

This project represents a calibrated MODFLOW model with three parameters:

  1. K for the lower layer
  2. K for the upper layer
  3. Recharge

Furthermore, the optimal parameter values have already been imported to the parameters dialog to the starting values column.

We will randomize each of the three parameters and perform a Monte Carlo simulation. We will then do a probabilistic capture zone analysis. Do the following:

1) Go into the MODFLOW Global Options and change the run option to Stochastic.

2) Go to the Parameters dialog and turn on the randomize option for each of the three parameters.

3) Change the Stochastic option to Latin Hypercube. Use four segments for each of the K parameters and three segments for the recharge parameter.

4) Enter a standard deviation of 0.15 for the two K parameters and 0.075 for the recharge parameter.

5) Save the project to a new name and run the model. You may notice that some of your model instances will not converge. This is most likely caused by one of the wells going dry in some cases.

6) Read in the results when the solution is complete. Click on several different solutions to view the results.

7) Perform a probabilistic capture zone analysis.

a) Right-click on the folder containing the stochastic solutions and select the Risk Analysis option.

b) Select the Probabilistic capture zone analysis option and click Next.

c) Change the tracking duration to Specified duration and enter a time of 20000 days.

d) Change the particle placement option so that there are four particles per cell.

e) Leave everything else at the default values and select Finish

When the particle tracking process finishes, select the two new data sets at the bottom of the folder containing the stochastic solution to view the results. There will be one data set per well.

Note that the capture zones are quite uniform. This is because we only have three parameters and they cover large regions of the model. It is also due to the fact that the standard deviations we used were quite small.

Before zipping your solution for uploading, you will need to delete the text output. Go to the subfolder containing your MODFLOW solutions and sort the files by size. You should see a bunch of files with a *.out suffix (or they may have a *.glo suffix). Select and delete all of these files (they are just bulky text files generated by MODFLOW). Then zip up your entire project, minus the original texmod portion.

Submittal Instructions

Save the GMS project with the completed solution. Zip up all files associated with the project.

Upload your zip archive via Learning Suite.