Laserfiche WebLink
Historical periods of previous studies: Of the twenty-five reports reviewed, study periods <br />ranged from 1906 to 1993. Twelve of the twenty-three studies extended back to approximately <br />1950 (range from 1947 to 1953), two studies had periods starting later than 1953 (1961 & <br />1974), four began their study periods between 1938 and 1943, three began on or around 1930, <br />one in 1911, one in 1906, and the final study used stochastically generated data. <br />• Desired period of record: Of those interviewed, at least four people specifically wanted the <br />extended data set to include the 1950s. The others wanted the extended data set to include as <br />many years as existing partial records would allow. Extension of data back to 1929 would <br />allow StateMod simulations to be compared to ongoing studies on the San Juan and Yampa <br />basins. <br />• Limits on data extension: The greatest limitation on extension of records will be the <br />determination of natural streamflows from gaged data. <br />• Stochastic model: The stochastic models have only been used on limited studies. Only two of <br />the studies reviewed utilized stochastic modeling for the generation of streamflows. <br />• Interviews indicated that the use of stochastically generated flows would require education of <br />the users as to the uses and limitations of synthetic streamflows for planning purposes. <br />• Other methods: The additional two methods identified for the generation of streamflows were <br />the use of tree ring data and the indexed sequential method. Of the twenty-four studies <br />reviewed, only one utilized the indexed sequential method and one utilized tree ring data. <br />Recommendations <br />The following recommendations should be considered in the scope of work for Tasks 11 and 12, <br />based on the above summary. <br />• The data extension effort should focus, at a minimum, on the 1950s period to the extent that <br />data is available. Comments received from persons interviewed suggested that, if possible, data <br />should be extended to 1929 to be consistent with ongoing studies in the Yampa and San Juan <br />basins. <br />• If developed, the stochastic data generation will require education of users on the benefits and <br />limitations of stochastic modeling. <br />• The use of other methods, such as the indexed sequential method or tree ring data, was thought <br />to have some value, but was not overwhelmingly supported. <br />Appendix E E-5 <br />