This paper describes the impacts of potential hydropower retrofits on downstream flow distributions at Lock and Dam No. 8 on the upper Mississippi River. The model used solves the complete Reynolds equations for two-dimensional free-surface flow in the horizontal plane using a finite element solution scheme. RMA-2 has been in continuing use and development at the Hydrologic Engineering Center and elsewhere for the past decade. Although designed primarily for the simulation of hydraulic conditions, RMA-2 may be used in conjunction with related numerical models to simulate sediment transport and water quality. In this study, velocity distributions were evaluated with regard to environmental, navigational and small-boat safety considerations. Aquatic habitat was defined by depth, substrate type and current velocity. Habitat types were quantified by measuring the areas between calculated contours of velocity magnitude (isotachs) for existing and project conditions. The capability for computing and displaying isotachs for the depth-average velocity, velocity one foot from the bottom and near the water surface was developed for this study. The product of this study effort is an application of the RMA-2 model that allows prediction of structural aquatic habitat in hydraulicaly complex locations. Elements of the instream flow group methodology could be incorporated to provide detailed predictions of impacts to habitat quality. Calibration of the numerical model to field measurements of velocity magnitude and direction is also described.

A two-dimensional horizontal finite element numerical model (RMA-2) was applied to a 15 mile (24 km) river channel-floodplain reach in West Germany. Previous applications of such models have been restricted to much smaller scales. The results indicate that finite element schemes may successfully estimate river stage in large scale floodplain applications. Computed stage hydrographs compared well with observed data using loss coefficients within expected ranges. Two-dimensional flow models have been applied to certain classes of river channel problems. Applications have included detailed analyses of flow patterns near structures such as bridges and floodplains. In all these problems the scale of interest has been small, e.g. reaches of river a few river widths long. Many estuary studies have been done that were of large scale; some of these utilized a hybrid (numerical plus physical) modeling technique. In a review of the application of finite element methods to river channels, Samuels reported that the river channel was resolved separately from the floodplain in only two studies. Missing in previous work is attention to large scale floodplain modeling. The work reported in this paper focuses on the feasibility and accuracy of applying a two-dimensional flow model to a large floodplain. Traditional floodplain studies have used semi-empirical flow routing with steady, one-dimensional computation of water surface elevations to define inundated areas. Keywords: Army Corps of Engines. (kr).

The Arkansas River between Pueblo, Colorado, and John Martin Dam, a distance of about 125 river miles, is an alluvial, sand-bed river. It meanders between bluffs in a flood plain about one mile in width. During geologic time the downstream (eastern) portion of this reach has been migrating southward due to heavy sediment loads from northern tributaries. A local flood control project is being planned for the town of La Junta, which is in the downstream one-third of this reach. A study was undertaken to evaluate the future performance of various flood control alternatives with regard to channel stability, sediment movement, and project maintenance. The alternatives considered were various channel and levee configurations. Evaluations were based on both long-term (100-year period) and short-term (single flood event) hydrologic scenarios. The primary tool used in this study was the movable boundary mathematical model HEC-6 entitled 'Scour and Deposition in Rivers and Reservoirs.' The hydrologic and sediment regimes of the study reach are complex due to four tributaries and eleven major irrigation diversions. This paper describes development of representative data for the long-term analysis, operation of the model, calibration and simulation strategies employed, interpretation of model results, and computational aspects of this application.