FLOOD04478 - Laserfiche WebLink

Home Browse Search

96 'I F.L Ogdm ~t Ill./ JO"TnDI of H.~drology 228 (2000} 82-100 300 FL Ogd~" d al./ Journal of Hydrology 228 (2000J 82-Joo 97 " o 1 ~ & ~ E g , ~ " t o ~ - 1la ~.E : " ~ ~ ~ N . ~ .. ~ ~ , . ~ '" ~ " H s :; ~o :;; "E '. & ~.E , z ~ , E' ..g ~ ~& ~l !.5 E ~ " . ~ .. ~ ~ " . i 8[ E 1.5 ] .c g ~ ~~ ~ .~'~::i 3 VI ~; ~ ~ ~ 1 g gl h ] !~ ] ':;; : rf .~ M ~ .< :> 5 <::0 N ~ o o o ~_~ i , . . ; "E ] ~ ]~ ~, ] ; ~ . . . o ] ! ~ ~ ~ " '2 & . ~ " f . .~ " '" ~ ~ " ~ ~ ::i N -, R . o ~ . ~ ~:? '6 ~E 11= ~~ N ~ , ~ N . ~ , N ~ ~ ;e ~ o ;;; ~ o o . ~ ;; ~~- o E 5;;; "~ ~ 200 ..~ . ~ . ~ o 100 ~ ;e ~ ~ , " , o 16:00:00 21 :00:00 00:00:00 03;00:00 Time, July 28129, 1997 (MOl) :0'1. :,:~. :;~;; .....-... K halved --- ~dotJbled - Relerence Run \ \ \ \ \ '~', Fig. ]0. Effect of systematic :t]00'I> e!TOrs in soilsaluralt'd hydraulic conductivity estimllle5011 predicted runoff at stalion VI. ~ was evaluated by performing two tests with fully spatially-varied inputs. In the first test, all calibrated K. values from the reference run were multiplied by 2.0. In the second test, they were divided by 2.0. The resulting hydrographs from these two simulations are plotted together with the reference hydrograph at station Ul on Fig. 10, and are quite revealing. The simulated hydrograph with KI reduced by a factor of twO is very similar to the reference simulation, with a slight 5% increase in peak discharge. When K, is doubled, the net effect is a more significant 13% reduction in peak discharge. These two tests reveal that for extreme rainfall in an urban catchment, factor of 2 errors in soil K. have a much smaller effect on runoff predictions. In the limit, as the rainfall rales increase relative to K., the watershed behaves more like an impervious surface. This is affirmed by the high runoff production efficiency (81 %) determined from the reference simulation. e ~ ~ ~ ~ ~ ~ ~ ~ IS, Discussion and conclusions ] ~- ~~E .~ -; .~ ] .0 0> . .g gi e';: ~o '" . c.!! . 0 .IS '" " The high quality data set collected in the aftermath of the unfortunate "ood that impacted Fort Collins, Colorado, on 28 July 1997, provides an opportunity to study the hydrology and hydraulics of an urban flash llood as a result of an extreme (> 500 year aver- age recurrence interval) convective storm. This study ;f&. ~~ provides useful quantitative information on the propa. gation of rainfall estimation errors from a number of sources through extreme event runoff predictions. This study also explores the iml?act of watershed char- acteristic data uncertainty and errors on simulations of extreme ftash "oods. Watershed characteristics, particularly stormwater control structures, had a very significant impact on the flood discharge downstream from the large detention basin fanned by the railroad embankment. For this reason, our analysis focused on inflows to the deten- tion basin at station U 1 (see Fig. I). It is interesting to note that Horsetooth reservoir (see Fig. 2) provided significant relief to Fort Collins during this flood. Had Horsetooth reservoir not existed, an additional 2600000 m3 of runoff volume would have passed through Fort Collins. Peak discharges and flow dura- tions would certainly have been higher along all of Spring Creek. This. is an excellent example of a water supply reservoir providing flood control protection. From the analysis it is clear that the polarimetric radar-rainfall estimates are of high quality, as shown in Fig. 3. Underestimation by the WSR-88D might not be the only problem when using single-polarization radar derived rainfall as inputs to hydrologic models. It is clear that the WSR-88D rainfall field is different from the reference field as seen in the model output even when the stann total rainfall volume is adjusted, as shown in Fig. 7. However, the WSR-88D rainfall