Laserfiche WebLink
CRWCD: <br />Mike Gross <br />Dave Kanzer <br />DWD: <br />Steve Schimitzer <br />Harris Water Engineering, Inc. <br />Steve Harris <br />The interviews discussed each person's interest in the CRDSS and experience with using extended <br />data sets and stochastic modeling. A series of questions were asked to each person concerning data <br />extension. Generally, there was support for the need to extend the data set. The period generally <br />desired included the 1950s, while some also wanted to see the late 1920s and 1930s. The major <br />reasons were for consistency with other studies and because the 1950s drought has been historically <br />used as a measuring stick for system operations. <br />A need for obtaining daily flows back into the 1920s for the Colorado and San Juan Rivers was also <br />expressed in the interviews. Although studies on these rivers may require such information, the <br />feasibility of extending daily flows back that far is low due to its complexity and the declining <br />accuracy of estimated data the further it is extended. Amore likely approach maybe to develop <br />daily flows from one period of good record, such as the 1950s, and use that for analysis. Although <br />there are methods by which daily flow data can be obtained, its development would probably not <br />serve the overall long-term needs of the CRDSS. Therefore, it will not be further addressed in the <br />data extension feasibility study. <br />The greatest concern with data extension was the need for accurate diversion data. The USBR has <br />developed historical natural flow data, but did not segregate the individual diversions and <br />depletions (diversions minus return flows) in developing their data set. DWD only removed <br />diversion data that they envisioned changing in operation studies, thus avoiding dealing with <br />individual historical diversions. The use of demand based upon average, wet, and dry years or <br />some other relationship was thought to generally be acceptable where actual diversion data did not <br />exist. The overall opinion was that extension of the data series would be useful, but that <br />assumptions for the development of demand data and the methods of data filling should be well <br />documented. <br />A series of questions were also asked about the needs for stochastic modeling. In general, the <br />people interviewed were not very familiar with the use of stochastic modeling and its benefits. The <br />interviews indicated that if a stochastic modeling component is developed, significant education of <br />the users will be required in order to help them utilize this tool. The use of stochastic modeling was <br />identified primarily for larger scale (basin-wide) planning level studies. It would probably not be as <br />applicable for detailed analysis of specific smaller reaches or for detailed daily or weekly flows, <br />such as studies done on the 15 Mile Reach. <br />Finally, suggestions were taken regarding other data extension methods to consider. The only two <br />methods mentioned were the indexed sequential method and the use of tree ring data. The latter <br />was not thought to be worth pursuing by most persons interviewed. <br />Summary <br />Review of the literature and interviews with interested individuals have led to the following <br />conclusions: <br />Appendix E E-4 <br />