Laserfiche WebLink
<br />Streamflow Sequences <br />We had recommended in our Denver meeting that we base the ,criteria and <br /> <br /> <br />the test studies only on the 31-year period 1970 through 2000. Our test <br /> <br /> <br />studies with the one exception previously mentioned are based on this <br /> <br /> <br />concept. <br /> <br />Initially we selected for study 13 - 31-year continuous sequences or <br />those which began with each fifth year of the 1906-1968 period starting <br />in 1906, and continuing through 1966. For the studies, we assumed 1906 <br />followed 1968 in the streamflow sequences. Two studies (Nos. 14 and 15) <br />were completed with all 13 sequences, but realizing the magnitude of the <br />number of studies which would be necessary using 13 sequences in combina- <br /> <br />tion with many other variations in other parameters, we reduced the flow <br /> <br />sequences to 4 for the subsequent studies. The four selected sequences <br /> <br />are: #3 (1916-1946), #6 (1931-1961), #8 (1941-1908), #12 (1961-1928).' <br />These are shown on the graph and provide a representative cross section of <br />sequences which we might reasonably expect during the next 31 years. The <br />first sequence (sequence #3) is about 103% of the 1906-1968 average with <br />higher than average years occurring early in the sequence. The second <br />(sequence #6) is about 86% of average and is the lowest 31-year sequence of <br />record. The third (sequence #8) is about 93% of average or about lower quartile <br />and contains the lowest 4 and ~2-year sequences of record. The fourth is 111% <br />of average with below average years occurring early in the sequence. 'Nearly <br /> <br />all studies were run with all four sequences. However, we did run a few <br /> <br /> <br />studies using the 1953-1920 sequence which is 101% of average and includes <br /> <br /> <br />the most critical 4 and 12-year sequences at the beginning of the study. The <br /> <br />6 <br />