FLOOD09697 - Laserfiche WebLink

Home Browse Search

I o ~ ( !! ,~ STORM OF MAY 22-23, 1968' "! \ ~ -' \ ~ / ,.. -t> s/,. sw 3 WASATCH NATIONAL FOREST " Swl % \ \~ \ Figure 3.10. lsohyetallines for the event of May 22-23, 1968. ure 2.7) published as "Hourly Precipitation Data" by the U. S. Deparlment of Commerce. By interpola- tion from the hourly data estimates were made of the 30-minute precipitation quantities of Ihe recording gage. The daily information from the non-recording gages was distributed in time on the same basis as the observed data from the recording gage. The 30-min- ute precipitalion thus computed al each gage location was spatially distributed in accordance with the 'Thiessen network of Figure 2.7. Some questions might be raised on the use of the Thiessen procedure rather than isohyelallines. For illuslrative purposes, Figure 3.10 shows isohyetal1ines and the precipita- tion stalion totals for a single slorm event. This pro- cedure of spatially distributing poinl precipilation measurements is generally regarded as more accurate, but it is also more difficult to implement in a com- puter. Some isohyetal charts for specific events were developed, and significant differences were nol de- tected between the results of the loohyelal and the 'Thiessen weighing methods. Because il is readily im- plemented on the computer,lhe Thiessen technique was adopted for lhis study. Calibration of the urban model was based on prololype data from three slerms. Model oulput was compared to measured flow by using the sum of the squared devialions as Ihe objective function for the pattern search procedure described previously. The three slerms gave varying values for the five paramet- ers, and an average value was selected to provide the c10sesl agreement between predicted and observed hydro graphs for the Ihree slonus. In order 10 vary watershed parameter values with urbanization it was necessary to relale the para- 37