Laserfiche WebLink
<br />',~1 <br /> <br />RESULTS OF MODEL CALIBRATION <br /> <br />Calibration Using Rainfall-Simulator Data <br /> <br />Model calibrations proceeded through a sequence of steps or phases, <br />more or less predicated on data availability. Initially, available data <br />from simulator runs of July, August, and October 1981 were analyzed on a <br />trial-and-error basis. Adjustment of Green-Ampt infiltration parameters, <br />surface-retention capacity, and surface-roughness coefficient were made <br />to gain insight into the sensitivity of parameters and to best reproduce <br />individual runoff events. These adjustments and their influence on runoff <br />volume and hydrograph shape readily can be determined by using both the <br />speed and graphics capability of the desk-top computer. <br /> <br />Results of the trial-and-error approach of fitting runoff for the <br />summer and fall runs of 1981 for the upland ponderosa area (plots 1 and 2) <br />are shown in figures 5 and 6. Simulation runs shown in figures SA, sB, and <br />6A were intended to represent application of a uniform rainfall intensity <br />of about 2 in/h. Simulation runs shown in figures 5C, 6B, and 6C were <br />intended to represent variable (step-function) rates of rainfall application. <br />Tipping-bucket rain gages were used to approximate the time-varying rainfall <br />rates. Observed discharge is shown by the "+" symbol and simulated response <br />by the continuous solid line. End of rainfall is shown by the vertical <br />dashed line rising from the time-axis. The time- and rate-axes are scaled <br />to a common magnitude for ease in comparing relative magnitude and response <br />characteristics. <br /> <br />In these fitting attempts, as in the majority of all fitting attempts, <br />the hydraulic conductivity term, KSAT, grossly controlled the goodness of fit. <br />The other parameters exerted a second-order influence affecting the shape and <br />timing of the rising hydrograph. Surface-retention capacity, SURF, generally <br />can be adjusted for good agreement between simulated and observed time of rise. <br />However, the range in value and absolute magnitude of this parameter is small, <br />from about 0.05 to 0.10 in. The pressure-head parameter, P, and surface rough- <br />ness, FRIC, affect the shape of the rising limb of simulated hydrographs. Mag- <br />nitude of P is very small (0.1 in of water) and approximately the same order <br />of magnitude as the depth of flow, Ho' Values of P commonly reported are in <br />range of 5 to 50 in of water. Roughness values are in general agreement with <br />values for shortgrass prairie, reported by Woolhiser and others (1970). <br /> <br />Results of sensitivity analyses of parameters involved in the infiltration <br />equation are shown in figure 7. These data were obtained by varying the value <br />of each parameter from 30 percent less to 30 percent more than the value for <br />the parameter obtained in the best-fit simulation, while holding the other <br />parameters at their best-fit value. Resulting runoff for each simulation <br />then was recorded. Sensitivity of calculated runoff to applied rainfall is <br />also shown in figure 7. These data were obtained by holding all parameters <br />at their best-fit values and varying applied rainfall from 20 percent less to <br />20 percent more than the measured rainfall for that run. Of the parameters <br />used in fitting simulated runoff to observed runoff, KSAT is by far the most <br /> <br />12 <br />