Laserfiche WebLink
<br />5.4 Model Performance and Calibration <br /> <br />An important part of any numerical modeling process is to determine the <br />sensitivity of the results to variations in certain parameters. As stated in <br />section 5.2.2, the river bed profile at the end of a flood event did not <br />change drastically due to changes in the inflowing sediment load. Changing <br />the Kanning n value did have a greater effect on computed results. <br /> <br />The discussion in Section 4.4 of the long-term simulation explains the <br />role and procedures for calibrating movable boundary models. After performing <br />sensitivity analyses for different Manning n values, the same n values that <br />were used for the long-term simulation were found to be appropriate for use in <br />the single-event simulation. For the main channel, n=0.04; for the overbank <br />areas, n=0.05. <br /> <br />To maintain numerical stability in the HEC-6 computation scheme, a minimum <br />time step of two hours was used for the peak f low periods. The most stable <br />HEC-6 hydraulic parameter weighting factors (see (9), p. 3.13) were used for <br />all of the simulations. <br /> <br />Ideally, a simulation would be most sensitive to the key parameters such <br />as channel geometry and the shape and duration of the flood hydrograph. Based <br />on the above sensitivity analysis, it was decided that this was the case for <br />the single-event analyses and extra efforts were made to accurately depict <br />these parameters. <br /> <br />5.5 Simulation Results <br /> <br />The results of the single-event analyses using HEC-6 for existing and <br />project conditions are presented here using the following three formats. <br /> <br />1) Graphs of calculated stream bed thalweg profiles at the peak flow and <br />at the end of each event (50 year, 100 year and standard project floods) are <br />presented in Appendix E. <br /> <br />2) Graphs of calculated changes in thalweg elevation at selected cross <br />sections vs. time are presented in Appendix F. This was done for the standard <br />project flood since it was considered to be the most severe event. <br /> <br />3) Tables 5.1 through 5.3 summarize the important data for each of the <br />three flood magnitudes for existing and project conditions. <br /> <br />5.5.1 ExistinK Conditions <br /> <br />The anticipated sediment response of the river for existing conditions is <br />shown in Appendix E on pages E-l, E-2, E-7, E-8, E-13 and E-14 (Pages E-l, <br />E-7, and E-13 are at peak flow and E-2, E-8, and E-14 are at the end of the <br />event). These figures correspond to the 50 year, 100 year and standard <br />project floods respectively. General trends indicated deposition near river <br />mile 1199.3 and 1197.9, scour near river mile 1199.7 and 1196.3. The maximum <br />scour and deposition at the areas listed above were on the order of one to one <br />and a half feet. It must be understood that since HEC-6 is a one-dimensional <br />model, material that is deposited or removed is distributed uniformly over the <br />entire movable bed portion of each cross section. Since the movable bed width <br />for existing conditions was on the order of 5,000 feet, a bed elevation change <br /> <br />56 <br />