This chapter evaluates model performance based on the accuracy of the numerical solution. Two major types of tests contribute to this process: verification and validation. Verification compares analytical and numerical solutions. For this process to be legitimate, the analytical and numerical solutions must start with the same assumptions, such as initial and boundary conditions. Validation compares numerical solutions with data from real world systems such as laboratory column and field studies.

Few simple analytical solutions describing solute transport coupled with reactions exist. Poulsen (1991) used an analytical solution of the advective-dispersive equation coupled with first order degradation. Fry et al. (1993) developed an analytical solution to the solute transport equation with rate-limited desorption and first order decay. However, like most analytical solutions incorporating kinetic equations of higher order, the method of solution ultimately requires numerical methods. This chapter only compares numerical solutions with simple analytical solutions.

Because of these limitations described above, verification tests will separate the advective-dispersive processes from the reaction processes and evaluate each individually. An exception is one-dimensional solute transport with first order decay. It is the only coupled case possessing a simple analytical solution. The following list presents the verification tests included in this chapter and Appendix F:

Conservative Solute Tests
One-Dimensional Continuous Source
One-Dimensional Pulse Source
Two-Dimensional Pulse Source

Non-Conservative Solute Tests
One-Dimensional Continuous Source with First-Order Decay

Kinetic Reaction Tests
Linear Sorption (Appendix F)
Langmuir Sorption (Appendix F)
Freundlich Sorption (Appendix F)
First Order
Single Monod (Appendix F)
Double Monod (Appendix F)
Competitive Monod

In addition to verification tests, this chapter includes two validation-type tests. This chapter tests model performance in simulating conservative solute transport in stratified porous media (Sudicky et al., 1985). It also uses the model to reproduce biostimulation modeling studies conducted by Semprini et al. (1991).

Criteria defining acceptable and unacceptable error quantities do not exist. This is a function of model application. Some simulations may require high accuracy, while others may not. This chapter does not make decisions on the acceptability of performance. It simply presents the sensitivity of errors to various input parameters.