Laserfiche WebLink
<br />of radius. Eq. 16 provides the changes in channel-bed elevation for a time <br />step due to transverse sediment movanent. These transverse changes, as well <br />as the longitudinal changes, are applied to the stream bed at each time <br />step. Bed-profile evolution is simulated by repeated iter ati on along <br />successi ve time steps. <br /> <br />XI. 'l'PSr AND CALIBRATI<>>I CF M~TICAL ftDEL <br /> <br />The accuracy of a mathematical model depends on the physical <br />foundation, numerical techniques, and physical relations for momentum, flow <br />resistance and sediment transport. Test and calibration are important steps <br />to be taken for more effective use of a model. Because of the difference in <br />sensitivity of simulated results to each relation or empirical coefficient, <br />more attention needs to be paid to those that generate sensitive results. <br />Major items that require calibration include the roughness coefficient, <br />sediment transport equation, bank erodibility factor, bed erodibility <br />factor, etc. <br /> <br />To determine the sensi ti vi ty of flow, sediment transport and channel <br />changes caused by the variation of each variable, different values of the <br />variable need to be used in simulation runs and the results so obtained are <br />ccmpared. Generally speaking, the rate of channel changes is more sensitive <br />to the sediment rate computed from a sediment equation but the equilibrium <br />channel configuration is less sensitive. For example, the constriction <br />scour at a bridge crossing, or the equilibrium local scour at a bridge pier, <br />is found to be more or less independent of the sediment equation, or <br />sediment size, since both inflow and outflow rates of sediment are affected <br />by about the same proportion. It may also be stated that the rate of <br />widening is sensitive to the bank erodibility factor but the equilibrium <br />width is not nearly as sensitive. <br /> <br />Field data are generally required for test and calibration of a model. <br />Generally, the channel configuration before and after the changes, flow <br />record, sediment characteristics are required. Data sets with more complete <br />infonnation are also more useful. Several data sets that are useful for <br />test and calibration are included in Appendix E. <br />22 <br />