Laserfiche WebLink
Once the predictive simulations were completed, it was necessary to carry out a <br />sensitivity analysis on the model for two primary reasons. First, the sensitivity <br />analyses would aid in determining which aquifer parameter or model input had <br />the greatest impact on the predictive results. This would also assist in identifying <br />any weakness in the data used in the model. Second, the sensitivity analyses <br />would allow for determination of the reasonableness of the assumptions applied in <br />the model. The procedure involves, for one selected parameter at a time, varying <br />the input values over a likely reasonable range and then running the model. The <br />model predictions are then compared to those generated by the originally utilized <br />set of values to see if any significant changes in the drawdown configuration <br />results. For this study, sensitivity analyses were run on aquifer hydraulic <br />conductivity and stream conductance, which involves an inter-related set of <br />values for stream width, stream head, and stream bottom conductance and <br />configuration. It was found that the model results are fairly insensitive to the <br />aquifer hydraulic conductivity, assumedly because the reasonable range of values <br />expected for materials typical for such aquifers is generally quite high. Thus <br />neither the higher values or the lower values reasonably expected really vary to a <br />great magnitude. Similarly, the stream conductance values exhibited a fairly low <br />sensitivity. One likely explanation for this is the very close proximity of the <br />stream to the project area, the high transmissivity of the aquifer (this tends to <br />focalize the drawdown effects, thus limiting the reach of the river over which <br />impacts will be felt), and the generally large surface area that the stream covers in <br />12 <br />Martin and Wood Water Consultants, Inc. <br />