7371 - Laserfiche WebLink

Home Browse Search

large upstream backwaters, The use of Manning"s equatIOn is''lTlore appropriate in <;leep streams with short back~!ater effects. The stage-discharge (regression) approach m<JY be used in all types of rivers. provided that the cross section is not subject to a variable backwater. Large rivers may be steep and vary rapidly. but most are low gradient and vary gradually; therefore the step-backwater model is the most commonly used approach to predict water surface elevations in large rivers. Table 1 lists several computer programs that are available to predict the water surface elevation at unmeasured streamflows. TABLE I. Computer programs for determining water surface ele- vations. Name Type Source IFG4 Regression U.S. Fish and Wildlife Service MANSQ Manning U.S. Fish and Wildlife Service R2~CROSS Manning U.S. Forest Service PSEUDO Step-backwater U.S. Bureau of Reclamation WSP" Step-backwater U.S. Fish and Wildlife Service WSP2 Step-backwater U.S. Soil Conservation Service HEC-2 Step-backwater U.S. Army Corps of Engineers HEC-6 Step-backwater U.S. Army Corps of Engineers "Modified and expanded from the PSEUDO program of the U.S. Bureau of Reclamation. Velocity - Another common attribute of habitat studies is the prediction of the velocity distribution at different water discharges. This feature distinguishes two- dimensional microhabitat analysis from most other types of river studies. Although it may be sufficient to determine the average cross-sectional velocity for certain types of studies, microhabitat analyses often require prediction ofthe lateral, longitudinal, and vertical velocity patterns in the river. There are basically two ways of making these predictions: (1) by empirical regression; and (2) by a more theoretical approach based on the concept of conveyance. The water column velocity can be measured at the same locations in the river at three or more discharges to develop log transformed regressions between the measured point velocities and total stream discharge. The concept of con- veyance uses the Manning equation to distribute velocities within the channel at different flows. Both techniques usually employ a mass balancing feature that ensures that discharges predicted by the simulation will equal the dis- charges originally inputJor stimulation. Either method can be used in the IFG4 program (Milhous et al. 1984), but the conveyance approach is preferred for two reasons: (1) it is mathematically stable when extrapolated over a wide range of flows; and (2) it often produces more accurate results with fewer data. At least three flows must be measured to develop an empirical velocity-discharge relation, whereas equivalent results can be obtained from only one flow mea- surement by using conveyance. The conveyance method is also superior when the current direction of the stream changes as a function of discharge (Fig. 2). In this example, the stream meanders more at low flow than at high flow, and the current streamlines change around obstacles as the discharge changes. Such streamline shifts cannot be satis- ~ -- -- ~/ ~ ~ Low Flow " , " ~ "--_/-' . ----""" ...- " / / ------ .4 ...... .... .... High Flow FIG, 2. Example of shifts in direction of current in a stream between low flow to high flow. factorily simulated by using log-linear regression, and may lead to some of the mathematical instabilities of this approach. Because each flow pattern is independently calibrated under the conveyance approach, streamline shifts can be simulated, although it does require measurements of the velocity distribution at more than one flow. Velocity at fish location - In small rivers, the water is often so shallow that measurements of the mean column velocity closely approximate the velocities experienced by the fish, conventionally termed the "nose velocity. " On the other hand, the mean column velocity in a large river may considerably exceed the near-bottom velocity faced by bottom-oriented species. The fish respond to changing dis- charges by changing their position in the water column - often by moving toward the bottom. Therefore, rather than the mean column velocity, the nose velocity must be used in habitat analyses in large rivers for determination of habi- tat suitability. If mean column velocity is used, suitability will be considerably underestimated. Annear and Condor (1984) concluded, for example, that output from the U$FWS microhabitat simulation model PHABSIM was biased toward low discharges when it was applied in large rivers. This conclusion may have been the result of using mean column velocity rather than nose velocity in the habitat computations. In conducting analyses involving nose velocities it is necessary to transform the mean column velocities from the ' hydraulic simulation model to the nose velocities used in the habit analysis. This transformation can be made empirically or from a theoretical distribution. To convert mean column velocities empirically, one must measure both the mean column and nose depth velocities during the collection of calibration data. These measurements are then used to cali- 17