Model application verification is important because it establishes the predictive capability of the model. In the verification process, the model conceptualization and parameter values determined during the calibration process are used to simulate independent conditions for which a set of field data exist. To fully assess the predictive capabilities of the model, several different hydrologic conditions which represent various time and space scales should be utilized during verification. This paper describes the verification of a groundwater flow model application in a heterogeneous, anisotropic aquifer. The computer code ARMOS was used because of its ability to simulate spatial heterogeneity, anisotropy and complex recovery systems configurations, which are present at the subject site.
In the modeling effort, water levels from three different time periods, water recovery from pumping well and trench systems, cumulative water recovery from extraction systems located in different areas, and infiltration tests at two different locations were used for model verification. The model successfully predicted all of the observed values within the range observed during calibration, except for one of the infiltration tests. Small-scale local heterogeneities, on the order of a few meters, may not be accurately simulated by the model. However, because of the comprehensive scope of the verification process, the model developed and calibrated for this site is demonstrated to be a useful and accurate tool for predicting groundwater flow at a scale of several meters or larger.