Laserfiche WebLink
and observed values, called the residuals. These are defined prior to beginning fl1e calibration <br />process and are considered as goals based on the modeling objectives and measurement or <br />other errors inherent in the field measurements. Calibration criteria are defined for each type of <br />calibration target. The criteria for each calibration target are discussed in Section 4.1. In <br />addition to the evaluation of residuals, the model calibration process will also include maps, <br />1listograms, and scatterplots of residuals to help identify spatial trends. <br />The head and flow criteria are compared to the results provided uz the automated model <br />calibration package (PEST; Doherty, 2004) through its parameter estimation objective function, <br />discussed in Section 4.2. T11e head and flow data used for calibration have varying degrees of <br />uncertainty. This uncertainty in the observed values is factored into the calibration process by <br />assign>11g weighting factors. Weighting factors that will be assigned to each calibration criterion <br />are discussed in Section 4.3. <br />4.1 Target Criteria <br />Calibration targets have been developed for use with the automated parameter estimation <br />objective function ('numeric' targets, described in Sections 3.2 through 3.4) and for the manual <br />evaluation of model calibration using other targets ('non-numeric' targets, described in Section <br />3.5). The model calibration will first rely on the numeric targets and their criteria, with <br />simulations meeting these criteria in fl1e objective function being further evaluated with the <br />non-numeric targets and their criteria to determine if additional calibration is needed. Eac11 set <br />of targets and criteria are presented in the sections below. <br />4.1.1 Criteria for Head and Flow Targets <br />The calibration criteria are listed below. They are considered to be reasonable calibration goals <br />based on review of the data and eng>1leering judgment but may need to be revised based on the <br />>1litial model calibration efforts. There are no generally accepted standards for an acceptable <br />agreement between modeled and observed heads and flows in guidance documents from <br />organizations such as the ASTM or various state agencies. <br />• T11e absolute value of t11e mean difference of head residuals for wells with surveyed <br />elevations across all time periods will be less fllan 5 feet. <br />• The absolute value of fl1e difference between observed 1leads at surveyed wells across all <br />time periods will be less fllan 5 feet for 75 percent of the observations. <br />• Observed and calculated average annual streaniflows will be withi1110 percent for all <br />flows greater than 25,000 acre-feet/year. <br />• Observed and calculated average annual stream gain/loss within a given reach will be <br />wifllin 10 percent for all flows greater than 25,000 acre-feet/year. <br />4.1.2 Criteria for Other Non-numeric Targets <br />• Residuals for head and flows should be randomly distributed on a geographic and <br />temporal basis. <br />• T11e model volumetric water balance calculated by the model will be within 1 percent for <br />fl1e steady-state calibration simulation and within 5 percent at the end of fl1e transient <br />simulations. <br />• The residual for each of the components of the groundwater budget will be less than ten <br />percent for 75 percent of the budget components. <br />P~ SPDSS T~8 2 Final TM 10-08-0t~.doc 16 <br />