Laserfiche WebLink
<br />390 Computational Fluid Dynamics <br /> <br />Influence of chan1U!/ 5hape on deposilion volume. Results in region A in the Above . <br />Lava Chuar reach are consistent with the conceptual model of Melis (1997) for the <br />relation between sboreline types and fan-eddy attributes and the potential for sand <br />storage. Melis (1997) suggested that tbe greatest potential for sand storage would <br />occur in reaches with the greatest density of debris faDS and downstream from the <br />tightest channel constrictions, and that the least potential for sand storage would <br />occur in reaches with Sleep shorelines and few debris fans. The model results for <br />these four reaches support the conclusion that recirculation zones are the most <br />effective sites for storing large volumes of sand that are likely to endure on a time <br />scale of months to years. The parts of these deposits that are beneath the water <br />surface during normal dam operation, however, are still subject to erosion at lower <br />discharges by scour, although at a much slower rate than occurs in the main channel. <br />These inundated deposits may also erode rapidly during routine dam operation <br />(Cluer, 1995). Deposits along the channel sides that are perched above the 708 mJls <br />stage, albeit smaller than the recirculation deposits, may bave greater potential for <br />preservation. Little is currently known about tbe rates at whicb erosive processes, <br />such as aeolian transport, operate. <br /> <br />Potential negalive impaclJ of high flows. The cases studied so far have led to the <br />generalizations listed above. Particular sites may respond differently. Schmidt el aI. <br />(1999) and Hazel elal. (1997, 1999) documented considerahle variation in response <br />to the 1996 test flow within reaches with similar morphOlogy. Even at a given site, <br />periodic mass failure of rapidly accumulated sand deposits can lead to a temporal <br />variability in sand deposit volume (Andrews el al., 1999). Widening of the main <br />channel flow at higher stages may diminish deposits formed at lower discharges in <br />some reaches, as shown in A in the Above Lava Chuar reach and as was suggested by <br />Melis (1997). This is especially likely if high releases are necessary to lower Lake . <br />Powell, as bappened during 1983-1984 when sustained high flows peaking at <br />2720mJ/s caused erosion at some archaeological sites (US Department of Interior, <br />1995). Although results so far have shown deposition under most conditions in , <br />sheltered areas, such as recirculation zones, under some conditions erosion can occur ., <br />within recirculation zones (Joseph Hazel, Northern Arizona University, oral com- j <br />munication, 1999). Erosion in recirculation zones is most likely to occur with deeper .' <br />flows, caused by some combination of increased stage and initial low elevation of the <br />sand deposit, if the channel morphology is conducive to high flow velocities and iI <br />combined with low sand supply in the main stem. Erosion and deposition in reaches , <br />with gradually varying channel width are more likely to show greater sensitivity to <br />variations in water and sand discharge. <br /> <br />14.8 CooclusioD <br /> <br />The model used in this srudy provides a pbysically based, predictive method fOf, <br />examining the effects of dam releases and sand supplies on sand deposits. PredictiODl. <br />of changes in sand bars can be made at specific sites and general trends can be infenec!, <br />from modelling results over a range of conditions and at multiple sites. The model_ <br />constructed with sufficient complexily to represent processes with an accuracy~' <br />menSllrate with the purpose of the study using available computational power. ' , <br /> <br /> <br />. <br />