Laserfiche WebLink
<br />r <br /> <br />Next, the sub-cloud. wind field i.s used in concert with the upper air <br />observations to determine the proper location of the historic rainfall <br />field over the basin. Finally the vector wind difference in the <br />sub-cloud and cloud-layer winds is used to determine the degree of <br />skew in the historic rainfall pa,ttern over the watershed. <br /> <br />3. <br /> <br />Next the historical rainfall pattern is placed over the watershed and <br />the isohyetal pattern is planimetered to obtain subbasin aVE!rage <br />rainfalls. CSR model output is used to establish the tempDral distri- <br />bution of the rainfall in each subbasin. <br /> <br />The transposition of the historic rainfall patterns over the Boulder <br /> <br /> <br />Creek watershed is shown in Figures 6 tD Il for the Big Thompson flash flood, <br /> <br />the Cheyenne flash flood, and the Masonville flash flood, respectively. In <br /> <br /> <br />each case, the dynamic transposition is different than a direct application of <br /> <br /> <br />the historic pattern over the watershed. <br /> <br />In the case of the Big Thompson storm, only one of the two heavy rain <br />cells is apparent over the watershed due to the size of the storm. In the <br />case of the Cheyenne storm, the introduction of the topography of the Boulder <br />Creek watershed is considerably different than the rolling plains environment <br />near Cheyenne, Wyoming which results in a, more concentrated rainfall pattern. <br />The Masonville storm presented a more difficult problem in that historical <br />rainfall pattern information is severely lacking. Records suggest that the <br />accepted peak point rainfall was 8.1 inches in about 1 hour with about 7 <br />inches occurring in 30 minul;es. Output from the CSR model does not support <br />these high rainfall amounts and suggests a peak point total of 5.7 inches in <br />about 90 minutes, 5.4 inches in 60 minutes and 3.5 inches in 30 minutes. <br />Since historic data is miss:ng and the CSR model tends to be within 15 per <br />cent of the historic rainfall in most reconstitutions, the CSR model output <br />was used in this reconstitution. <br /> <br />While historic spatial rainfall patterns were available for the Big <br />Thompson and Cheyenne stormE:, no pattern was available for the Masollville <br />storm. A SUbjectively derived spatial pattern was constructed using CSR model <br /> <br />10 <br />