Laserfiche WebLink
less in the cooler, wetter spring and fall. The monthly regressions shown in Exhibits 5, 6, and 7 tend to <br />support these observations. <br />Based upon the preceding discussion, four options exist in handling the missing diversion data: <br />1. The first option is to bypass filling the missing diversion data, regressing gaged and natural flows for <br />the existing period of record (1975-1991) to generate natural flows for the periods of missing data. <br />This would assume that diversions are relatively constant and the gaged flows are only linearly <br />shifted from natural flows. <br />2. The second method is to characterize diversion data by average, wet, and dry year conditions (or <br />availability of water according to snowpack), and assume that diversions are constant for each type <br />of year. Since diversions do not seem to vary much according to streamflow for the senior decrees, <br />this would not have much effect on their estimates. But for more junior decrees it would provide an <br />allowance for variation of yields based upon priority. This method will work if synthetic flows are <br />generated and provide a consistent systematic method for estimating diversions with synthesized <br />data. <br />3. The third method is to utilize consumptive use data to estimate an average efficiency of irrigation <br />and then calculate probable agricultural diversions. Municipal and industrial diversions would have <br />to be estimated separately using another approach and combined with the irrigation diversions to <br />arrive at a total. Although this method would be deterministically based, the historical data does not <br />indicate that efficiencies are constant. In fact, the data suggests that the demand is generally <br />constant indicating that the diversions are made rather consistently with little regard to the actual <br />crop irrigation requirement. Using this method would also require generation of climatic data where <br />streamflows are synthesized. <br />4. The final method is to digitize existing diversion data. This data exists to a questionable extent and <br />varying detail. Therefore, it would still be necessary to synthesize some data in order to fill periods <br />where not enough information is available. <br />There are trade-offs between making selected assumptions and the cost and complexity required in <br />filling data gaps. The first method listed above, consisting of ignoring diversions and regressing natural <br />flows to gage flows, is relatively simple and the least complicated approach. The accuracy of this <br />method is limited by changes in diversion patterns over time and by the changes in diversions resulting <br />from various hydrologic conditions. It also makes simulation of various operational scenarios difficult <br />because estimates of diversions, return flows, and reservoir operations will not be available to simulate <br />the changes in operation. <br />The second method simulates diversion and return flows, but does not account for changes in diversions <br />that may have occurred over time. This method will provide a consistent record of diversions for use in <br />the simulation of other system operations. Although diversions will be based upon snowpack, or winter <br />climate, they cannot be adjusted based on climatological variances during the irrigation season. They <br />will simply be based on the availability of water. This simplification seems to be justified, however, <br />through the low correlation between streamflow and diversions found in the analysis of selected senior <br />decrees. <br />The third method, using estimates of consumptive use to determine diversions, will require <br />climatological data for the period and would assume that the historical diversions have varied directly <br />Appendix E E-17 <br />