Laserfiche WebLink
depth to water measurements are typically estimated to be accurate withal plus or minus 0.1 <br />feet, while streannflows are estimated to be accurate within plus or nninus 10 percent. Weighting <br />factors also will be applied to account for factors such as clustering of observations in time or <br />space in addition to applying scaling factors and the measurement error-based weighting <br />factors. <br />A weighting factor will allocate the contribution of flow and head target types to the overall <br />objective function. Currently, tlne relative weighting for the different classes of target data are as <br />follows. <br />• Well measurements with survey elevations: 25 percent of total <br />• Wells that are not surveyed and more than 1 measurements in a calibration period: 20 <br />percent of total <br />• Wells that are not surveyed and one measurement iln a calibration period: 5 percent of <br />total <br />• Stream gage annual flows: 15 percent of total <br />• Stream gain/loss: 35 percent of total <br />5.0 Summary and Conclusions <br />CDM has completed SPDSS Task 48.2, the development of calibration targets and criteria for tlne <br />South Platte alluvial groundwater model. This TM describes the process used to identify the <br />target locations and time periods for which observed data will be compared to model <br />simulation results. Tlnis TM also discusses and presents the calibration criteria and the range in <br />acceptable differences between observed and simulated measurements. <br />Below are conclusions from completion of this task: <br />• Model calibration will include steady-state and transient time periods. <br />o The steady-state period will simulate average arunual conditions for 1991-1994, <br />wlnicln represents a period with relatively steady or consistent conditions. Tlne <br />results will be used as initial input to the transient (time-varying) calibration <br />followed by an initial transient period of 1995-1998 conditions. <br />o The transient period will simulate monthly conditions during the 1999-2005 <br />period. <br />o Following the calibration to these periods, a validation run will be completed <br />that simulates the entire study period of 1950 through 2006. <br />A combination of automated parameter estimation and manual techniques will be used <br />to assist in calibrating the model. This is considered to be the most effective and efficient <br />method to calibrate a model as large and complex as the South Platte River basin alluvial <br />grounndwater model and follows current modeling standards. Accord>ingly, an objective <br />function will be used to deternnine goodness of fit for the parameter estimation <br />simulations. <br />Usung automated techniques, tlne model will be calibrated on groundwater level <br />elevation data, changes in water levels fluctuations over tinne, streamflow, and stream <br />gain/loss. Additional criteria will be used to assist in evaluating model calibration <br />including distribution of residuals, tlne ability to match the target ground water budget, <br />P~ SPDSS T~8 2 Final TM 10-08-0t~.doc 18 <br />