Laserfiche WebLink
~ ~ <br />variation in the groundwater system is important, observation points with multiple <br />observations will be used to the maximum degree possible. Limited data are available from <br />early portions of the calibration period; thus, most calibration information will be from more <br />recent studies. The calibration period will extend from 1950 to the present, recognizing that <br />very few observations of groundwater levels are available prior to about 1980. <br />• Available data will be examined and weighting factors applied to overcome the effects of <br />clustering of observation locations, data outliers, varying well screened interval, and other <br />localized effects using engineering judgment. This process is recommended to develop a <br />representative distribution of calibration data throughout the model area and to avoid <br />excessive influence on the calibration from local areas with abundant information. This <br />process will be documented in descriptive fields in the modeling GIS repository and will be <br />discussed with the State throughout the activity. <br />• An approach to model verification will be presented in this task. Model verification is a <br />process in which a data set not used in the model calibration process is compared to the <br />calibrated model's estimates. Model verification provides an independent level of checking <br />that the model reasonably simulates observed conditions. Model verification will be <br />implemented in Task 48.3 by reserving some available data that will not be used directly in <br />the calibration. A balance between utilizing all available data and reserving some data to. <br />validate the model will be developed and presented to the State for review. It is anticipated <br />that the model verification data set will include at least 3 years. The types and locations of <br />data used in the model verification process will be finalized after receiving review comments <br />from the State and the PRC. After the model is calibrated to the target data set, the calibrated <br />model results will be compared to the verification period to assess the calibration. <br />• A model sensitivity analysis will be conducted in Task 48.3. The approach to the sensitivity <br />analysis will be presented as part of this task for concurrence by the State. This approach will <br />conform to ASTM D 5611-94 Standard Guide for Conducting a Sensitivity Analysis for a <br />Groundwater Flow Model Application. A selected set of model input parameters (likely to <br />include aquifer hydraulic conductivity, storage coefficient, streambed conductance, recharge, <br />evapotranspiration, boundary inflow, and well pumping) will be varied over defined ranges <br />(e.g., plus or minus 10 percent) in individual model runs, and the key calibration criteria will <br />be compiled. <br />Deliverables: <br />• ATM will be prepared documenting the calibration targets and the proposed approach to <br />calibration. It is anticipated this TM will be up to 10 pages in length, not including figures, <br />tables and appendices. A draft will be presented to the State for review and comment. After <br />resolution of comments, this report will be presented to the PRC for review and comment. <br />After resolution of comments, the report will be finalized as Appendix L of the final report as <br />described in Task 48.6 below. <br />SPDSS Phase4_gw_SOW_09-15-06.doc 12 <br />