Here are a few impromtu notes I have received from people who have worled with me to learn MODFLOW

***********************************************************

Hi Eileen-

I just wanted to thank you for your class. I am currently working on a modeling project and after spending a day struggling with GMS, I fixed up all the files to run directly through MF2K in about an hour and now, the model runs without a hitch. Time to calibrate! I will say, GMS is an acceptable pre and post processor, but having a good knowledge of the text files makes the difference between having a working model and a lot of time billed for nothing.

Dan

***********************************************************

Dear Eileen,

I started my new job in Denver and after only one week of working there I wanted to thank you for so many things. First, thank you for making us learn how to use MODFLOW with only the text files (and no GUIs). When GUIs are making the project more difficult, I can rely on the raw text files and a little help from the MODFLOW manual to help solve most problems we run across in our modeling efforts. Second, thank you for teaching me so much about groundwater systems. It is true that the more you learn, the more you realize you don't know, but you have taught me many invaluable concepts about hydrology.

My sincerest thanks,
Paula

***********************************************************

Hi Eileen!

I, personally, use Ground Water Vistas at the office, but have also manipulated some Modflow files on my own and really understand why it is better to learn the long way first. I have had to teach myself most of the Modflow files so I am exited to actually have someone walk me through the steps from start to finish.

Gina

***********************************************************

Dear Eileen,

What we learned in Modeling class was right! NEVER TRUST A GUI.

1. Never trust a GUI. You need to be able to look at the data you think are in the files. I have found that strange paramater values can result using GUIs. Sometimes the GUI interpolation method creates some strange values. This is not necessarily an error in the GUI, but may result from the input data. For example, the GUI we used created a strange halo effect for VCONT. If we did not look at the data files we would not have known this halo was being produced. The halo was produced as a result of the data we imported
that was interpolated. We would have missed this error if we did not examine the resulting MODFLOW files.

2. Never trust a GUI II. Sometimes the GUI has difficulty with the imported shape files. For example, in our last project, we found errors in the MODFLOW stream input file. The errors resulted from some strange zigs and zags in the stream trace imported in the shape file. If we did not check the resulting stream file produced by the GUI, we would have missed some critical errors.

3. Never trust a GUI III. If you make input mistakes in the GUI, these mistakes show up in your files. For example, you may make mistakes entering stream information in GMS. GMS does some checking, but it doesn't know that you meant to input a stream width of 100 feet but you mistakingly entered a width of 1000 feet. Your conductance values will be 10 times higher than what you meant to input. You have to check your resulting stream file to see if the values are resonable. I have seen a river that suddenly rose 100 feet in elevation along its course because of a user input error. This model was being used without the users knowing they had this error.

4. GUIs may use defaults that the user may want to change. This is particularly true for defaults used for solver packages, output control, and the option to print input files to output files, and options for output formats. Many times, solver parameters need to be adjusted to reach convergence. The user needs to be able to examine these files in an editor, know what parameters are located in data fields, and change parameters in the files.

5. Generally, I believe that all files should be checked outside of a GUI. This can involve contouring data to see if parameters are distributed properly with reasonable values, or graphing data such as constant head elevations at a boundary to ensure constant heads are resonable, checking stream elevations relative to average land elevation in the grid cell, etc. I have written a code to check stream files for the stream routing package because I can't tell if the stream file created by the GUI is OK. This list goes on and on. Essentially the list includes ALL input data.

I could think of more and more reasons, but generally, modelers need the ability to examine the MF files and know what data fields they are looking at.

Judy

***********************************************************