Laserfiche WebLink
<br /><:7;) <br />'<:J 187 <br />"" <br />10 <br />~ value. The rate at which the standard deviation decreases, or the rate <br />00 <br /> <br />of convergence is seen to be low for N greater than 200. <br /> <br />Also, the 90% confidence interval for annual discharge is + 1% <br /> <br />for a simulation of 200 years. <br /> <br />Measurements of annual Powell dis- <br /> <br />charge are reported accurate to within ~ 5% (USGS, Water Supply Paper, <br /> <br />1973). <br /> <br />These results show that a simulation of 200 years yields a <br /> <br />value for average annual discharge whose reliability lies within that <br /> <br />of the model calibration data. Further refinement of the estimate of <br /> <br />average discharge through increasingly longer simulations is slight <br /> <br />for the reasons discussed above. <br /> <br />To investigate the sensitivity of mean Powell discharge to <br /> <br />different streamflow sequences, ten streamflow sequences of 200 years <br /> <br />were generated. Simulations at the lowest depletion level, DPL #1, and <br /> <br />at maximum allowed Powell storage values of SPM = 15 MAF and SPM = 27 <br />MAF were performed using each of the ten streamflow traces. <br /> <br />The variance in reservoir discharge decreases and the variance <br /> <br />of storage increases as the volume of the reservoir is increased. For <br /> <br />this reason the case with SPM = l5 MAF was used to examine the varia- <br /> <br />bility in mean discharge between simulations. The target discharge of <br />3' <br />Lake Powell was set to 8.23 MAF/yr (10.1 km /yr) in accordance with <br /> <br />the value specified in Section 5.5. <br /> <br />The output from the ten simulations indicates that differences <br /> <br />in estimates of average Lake Powell discharge could not be judged <br /> <br />statistically significant (Figure 6.2). The simulations were treated <br /> <br />as replications of a single experiment and their results combined. <br />