Laserfiche WebLink
<br /> <br /> 01. <br />Table 3. Error in simulated peak flow in relation to observed peak flow [All values in percent]. , <br />~ <br />Stonn CUHP/ CUHP DR3M HEC HSPF PSRM SWMM 1:; <br />SWMM TR20 ~t <br />Harvard Gulch Watershed <br />1981/518 2.8 3.8 3.7 220 -30 -24 22 -62 <br /> ~ <br />1983/6/5 -13 -5.0 -2.3 120 -46 -52 25 -98 <br />1985/1/15 -18 43. -16 260 -15 -50 83 -100 ~ <br />1985/7/19 -4.5 -6.0 6.0 210 -21 -12 40 -64 , <br />1987/6/9 -38 -15 -41 40 -52 -84 -10 -100 .. <br />1987/6/30 11 22 -5.2 50 0~74 -17 46 -98 , <br />Mean .. -10 7.0 -9.2 150 -27 -40 34 -87 <br />RMSt 19 21 19 171 33 47 44 89 <br />S.D" 17 22 18 92 20 27 31 18 <br />Surrey Downs Watershed <br />1980/4/18 48 62 200 -58 -26 89 -81 <br />1980/8/17 39 -23 190 46 -24 89 -11 <br />1981/6/30 2.6 27 38 -25 -54 26 -89 <br />1981/9/20 -26. 44 26 -64 -67 3.7 -100 <br />1981/12/1 3 57 200 60 -30 100 86 <br />1981/1/15 68 -8.7 200 -16 17 -23 <br />Mean 28 26 142 15 -36 54 .36 <br />RMS 42 42 162 52 40 67 74 <br />S~D. 34 35 85 56 20 43 70 <br />-,Meancrror <br /> -- MEAN = Z;c <br /> . where ,. z (Predictl!d-Observed)'IOO <br />t. Root Mean Square error Observed <br /> RMS' E n ,.. Numb,." of Storms <br />t. Standard Deviation <br /> <br />Simulated storm volumes differed from observed storm volumes (table 4) by -100 to 190 percent at Harvard Gulch <br />and by -100 to 240 percent at Surrey Downs. The RMS storm volume error ranged from 17 to 101 at Harvard Gulch <br />and from 15 to 142 at Surrey Downs. The average RMS storm volume error is 56 at Harvard Gulch and 54 at Surrey <br />Downs. The standard deviation of the errorin the predicted storm volume ranged from 19 to 70 at Harvard Gulch and <br />from 16 to 98 a~ey Downs. The average standard deviation in storm volume error is also about 40 percent greater <br />at Surrey Downs (62) then 'at Harvard Gulch (38). In general, sinJu1ated storm volumes tended to be overpredicted for <br />larger storms and underpredicted for smaller storms at both sites, except for the largest storm at Surrey Downs which <br />was underpredicted. <br /> <br />Models based on the SCS curve number (HEC-I and TRZO) for generating runoff generally had the poorest fit. <br />HEC-1 simulations substantially overpredicted peak flows, and TR20 simulations substantially underpredicted peak <br />flows; this may indicate the sensitivity of the simulations to user judgment of the SCS curve number. HEC-1 results <br />presented represented simulations based on 6 subbasins; results also were provided for a simplified model that used <br />only 3 subbasins~ The simplified model produced slightly smaller peak flow error, but about the same runoff volume <br />error, An additional HEC-I analysis of the May 1981 storm at Harvard Gulch, performed by the CASC2D modelers, <br />had only a -22 percent peak flow error which underscores the sensitivity of the model to the SCS curve number. A <br />comparison of runoff simulation-techniques in west-central Florida indicated somewhat less, but comparable error, in <br />simulated peak-flows and storm volumes for TR20 and HEC-1 simulations (Trammer and others, 1996). In that <br />study, average unca1ibrated-model peak-flow and storm-volume error averaged 45 and 43 percent. respectively, for <br />TR20 simulations and 105 and 27 percent, respectively, for HEC-l sinJulations. <br /> <br />7-168 <br />