GLY 5828, Chemical Hydrogeology

Michael Sukop, Florida International University

 

Assignment 8: Inverse modeling of conservative Bromide and reactive Phosphorous

 

  1. Use inverse modeling to fit the CDE to the Bromide tracer test data given at http://faculty.fiu.edu/~sukopm/GLY5828/W31BR_OBS.DAT. Values in the files are time in days and concentration in mg/L. Assume a one dimensional system. The pulse duration was 7 days. The pulse concentration was 20.8 mg/L and the background concentration was 0.5 mg/L. Base your initial guesses for v and D on the following:

 

 

Explain how you arrive at your initial guesses. Compare your guesses with the fitted estimates. Plot the results as a dense line for the CDE model and show the observations on the same graph using open symbols.

 

  1. Use the parameters estimated by fitting the Br- data above as fixed inputs and fit the well 31 Phosphorous data given at http://faculty.fiu.edu/~sukopm/GLY5828/W31P_OBS.DAT. Use a two site conceptualization and give the final estimates for R, Kd, beta, omega, alpha, and f. Use the following time-varying boundary condition (multiple pulses).

 

Pulse       Start Time       End Time     Concentration

     1            .0000            395.0000            .2670

     2         395.0000         456.0000           1.9900

     3         456.0000         487.0000            .1610

     4         487.0000         579.0000           3.1200

     5         579.0000         671.0000           1.0400

     6         671.0000         822.0000           3.9200

     7         822.0000         853.0000           6.3000

     8         853.0000         884.0000           2.1500

     9         884.0000        1157.0000            .2850

    10        1157.000        4627.0000            .4340

Comment on the meaning of the fitted parameters and their values. Plot the results as a dense line for the CDE model and show the observations on the same graph using open symbols.

 

  1. Project the future Phosphorous concentrations 5000 feet from the source out to 200 years from time zero. Use the same time-varying BC, with the last pulse extended beyond the end of the simulation. Plot the prediction.