Laserfiche WebLink
<br />III. ANALYSIS OF HISTORIC AND "PRESENT MODIFIED" FLOW DATA <br /> <br />3.1 Data Available <br /> <br />N <br />..... <br />-.J <br />GO <br /> <br />The information available to develop the stochastic data base for <br />Colorado;River Basin consists of the recorded historic data in the <br />basin, v~rious earlier estimates of historic and modified data, and <br />informat~on on depletions such as the recent Comprehensive Framework <br />Studies.i/ All of these sources of information were utilized in <br />developi~g the stochastic data base. Table 1 lists the historic <br />and "the, 1968 level present modified flow" data and the quality data <br />available at selected stations within the Basin, <br /> <br />Initially it was planned that the data base for the model would include <br />the 17 water quality stations reported upon in the "Progress Reports <br />on Quality of Water, Colorado River Basin" and the additional loca- <br />tions netessary for operation of the pertinent reservoirs in the <br />Basin. the Colorado River at Glenwood Springs, Colorado, was sub- <br />sequently deleted from this list; stations were added for the Animas <br />River, White River, and Yampa River to facilitate the Western U.S. <br />Water Plan studies, <br /> <br />3.2 Scope of the Analysis <br /> <br />A primary criterion for stochastic data generation techniques is that <br />the basiq data are homogeneous or can be analyzed to quantify the <br />nonhomog4neity, The meaning of homogeneous here is that there are no <br />man-induded time trends in the data and that the difference between months <br />and yearS results from the random inputs of nature. A major problem <br />with hyd~ologic streamflow data is'that essentially all the records <br />contain man-induced time trends, and that these trends are reflected <br />in the data through changes in the deterministic and random components. <br />A correct approach for developing a stochastic data base would be to <br />analyze ~ll climatic and streamflow data within the basin. Assuming <br />there wo~ld be no significant man-induced changes in precipitation and <br />other climatic data, these data would be used to detect and remove <br />the nonhqmogeneity in the streamflow data. In general, this process <br />could be ;applied from the headwater areas down through the remainder <br />of the basin. Each streamflow station as analyzed and modified would <br />be used to test and remove the nonhomogeneity in downstream and <br />subsequent stations. <br /> <br />Time and ~ersonnel limitations precluded a full analysis of all <br />data within the Basin. A shortened approach was therefore used. In <br />the approkch, all the available historic data for selected stations <br />were analyzed in the data analysis submodel, The means from this <br />analysis ~ere then adjusted to reflect a selected level of depletion <br />within th~ Basin, The data in the Upper Basin and above Hoover Dam <br />were adju~ted to reflect a mean flow for the 52-year period from 1914 <br /> <br />1/ See references, page <br /> <br />15 <br /> <br />