Laserfiche WebLink
<br /> <br />N <br />i-" <br />c.o <br />~ <br /> <br />The initial attempts at flow simulation presented difficulties which <br />were dealt with in the process of deriving a satisfactory set of flows, <br /> <br />Several attempts were made to adjust the data so that acceptable traces <br />could be generated. The results were judged to be satisfactory by <br />July of 1974. However, in early 1975 a number of programing errors <br />were noted which required additional modifications to the generating <br />program and data. It is expected that further modifications will also <br />improve the results beyond their present state. As stated before, <br />the results presently obtained are basically valid but do have some <br />recognized problems which necessitate further adjustments to the <br />generated data. <br /> <br />4.4 Comparison of Synthetic Flows and Historic or Present Modified Data <br /> <br />Several criteria and methods are available for illustrating the similari- <br />ties and differences between synthetic and historic flow sequences. <br />For this report, the comparisons made were chosen to illustrate <br />(a) the correlation structure among flow sequences, (b) the total <br />water supply at a station, (c) the periodic patterns of the traces, <br />and (d) probability distributions of long period average flows. <br />Although an unlimited number of ways of comparing traces are possible, <br />the ones applied here seemed most appealing to USBR personnel, <br /> <br />Comparisons made in this report are thQse made in the July 1974 report. <br />Although changes have been made it is felt that comparisons with the <br />more recent data would contribute essentially the same results. <br />Except where noted,comments within this section will refer to gener- <br />ated data obtained prior to July 1974. Also, since a repetition of <br />many of the supporting graphs and tables would serve no purpose, <br />they have been eliminated from this report. The interested reader <br />is referred to the July 1974 report. <br /> <br />4.4.1 Correlation Structure <br /> <br />As previously described, the method of data generation maintains <br />correlation among flows by preserving the periodic pattern of the <br />time series structure as w.ell as by the multiple regression analysis <br />. performed on the residual series. A check was made on the efficacy <br />of this procedure by inspecting the correlation matrix of historic <br />flows and that from one of the synthetic traces. Since the relation- <br />ship among as many stations as possible was desired, the period of <br />concurrent historic records from 1942-1956 was used for the analysis <br />in table 4. <br /> <br />As most of the values show, there is a very good representation of the <br />relationship between all stations, except the Virgin River. Since <br /> <br />28 <br />