Selecting Calibration Values and Formulating Calibration Targets for Ground-Water Flow Simulations

Simulation of ground-water flow is often a desired or promised outcome of many field investigations. Frequently, the study design, and data collection and analysis are performed without regard to the desired accuracy of the modeling results. Rarely are calibration values and targets set prior to simulation or even considered when designing the study. This neglect of fundamental field information needed in model calibration leads to unrealistic expectations from modeling efforts. We suggest a systematic approach to assessing the level of error associated with sample information selected for calibration assessment. It is proposed that this procedure occur prior to the initiation of ground-water flow modeling. Our procedure involves: 1) selecting calibration values, typically heads, fluxes and a water balance; 2) setting calibration targets based on error analysis of calibration values; 3) assessing the results of calibration efforts based on the agreement between simulated values and calibration targets. Application of the procedure is demonstrated using results from a flow model of an unconfined intermontane basin aquifer.