Journal of Hydraulic Engineering

Two standard modes of laboratory sediment transport flume operation are sediment‐feed and recirculating. In the case of uniform sediment, the final equilibriums achieved in the two modes are equivalent and independent of initial conditions, even though the specified flow and sediment constraints differ. When the sediment consists of a mixture of sizes, however, this symmetry breaks down. In a sediment‐feed flume, both the feed rate and size distribution are specified by the user, and must be identical to those of the bed load at final equilibrium. The final state thus becomes independent of initial conditions. In a recirculating flume, however, the equilibrium size distribution of neither the bed load nor the bed surface material can be specified in advance. As a result, the final state depends on the initial conditions and can include the possibility of partial transport. The equilibriums themselves are equivalent in that they obey the same laws of flow and sediment transport. Their a‐priori predictability, however, differs due to the difference in governing constraints. When fractional transport is scaled by the bulk size distribution, the two flume modes may produce very different transport relations, even for the same sediment. A consistent approach requires that the transport be scaled by the size distribution of the bed surface.

To obtain the condition of stable meandering of a water rivulet on an inclined smooth plate, the bend theory including the effect of the surface‐tension force is studied. The governing equations (momentum equations and equation of continuity) are St. Venant equations of shallow water flow along the flow direction, for which the curvilinear coordinate system is employed. Applications of the bend theory of meandering rivers and the surface‐tension‐force effect lead to the bend equation of meandering water rivulets. The linear stability analysis of this bend equation for infinitesimally small disturbance derives the criterion that determines the relationship between the meander wavelength of water rivulet and flow properties such as water depth, flow velocities, and so on. The derived criterion equation contains one unknown parameter. This parameter shows the deviation of contact angles of meandering water rivulet from the straight one. This value is obtained by comparing the theoretical wave number of meandering with observed ones. As a result, the theoretical computation is found to be in good agreement with the observation.

An experimental investigation of the flow structure at a 90° open‐channel diversion is presented. Velocity data obtained in a laboratory flume show that the flow at an open‐channel diversion is three‐dimensional, exhibiting similar characteristics to river bend flows. The behavior of three‐dimensional features, including the dividing stream‐plane and the strength of a secondary circulation, were found to depend upon the roughness of the main channel bed and the ratio of diversion flow velocity to main flow velocity. Results of this study indicate that a description of sediment‐transport behavior at diversions requires the understanding of the three‐dimensional flow structure and the application of sophisticated modeling techniques. The results also suggest a strong analogy between diversion flows and bend flows. The similarity between diversion and bend flows may justify the application of relatively simple bend‐flow models to predict the three‐dimensional flow features at open‐channel diversions.

The flow in the upper Niagara River is affected by weed growth in the summer and by ice formation and accumulation in the winter. Changes in flow resistance tend to occur abruptly with time and space during ice‐affected flows and make it difficult to forecast flows for power generation. Since 1988, the New York Power Authority and Ontario Hydro have collaborated on the development of a computer‐based system for forecasting flows and optimizing power generation from the Niagara River. A hydraulic routing model has been developed for forecasting the flow in the upper Niagara River under uncertainties due to ice consisting of: (1) A steady‐flow model; (2) an unsteady‐flow routing model; (3) a linear, time‐varying reservoir model for forecasting the flow into the Grass Island Pool; (4) a time‐varying, stochastic model for the stage‐discharge relation at Fort Erie, the upstream boundary of the reach; and (5) Kalman filtering schemes to be coupled with these components models, to reduce uncertainties in parameters and variables to be estimated. This paper presents components (2)–(5) and related concepts and modeling techniques.

Functional relationships to describe surface wind mixing, vertical turbulent diffusion, convective heat transfer, and radiation penetration based on data from lakes in Minnesota have been developed. These relationships have been introduced by regressing model parameters found either by analysis of field data or by calibration (minimizing the difference between measured and predicted temperatures) in simulations on individual lakes, against gross lake properties such as surface area or Secchi depth. Results of the deterministic lake water temperature stratification model using the functional relationships are not much different than results using the individual calibrations on a great variety of lake surface areas, depths, and transparencies. The model also requires no on‐lake weather but uses input from existing off‐lake weather stations. First order uncertainty analysis showed moderate sensitivity of simulated lake water temperatures to model coefficients. The numerical model which can be used without calibration has an average 1.1°C root mean square error, and 93% of measured lake water temperatures variability is explained by the numerical simulations, over wide ranges of lake morphometries, trophic levels, and meteorological conditions.

This study illustrates the development of the finite‐analytic (FA) numerical solutions associated with an optimal (variable) time‐weighting factor for water flow in unsaturated porous media with irregular boundaries. This method entails the nine‐point and five‐point elements such that it can simulate quite well the irregular flow domain without using boundary‐fitted transformation. For an element near the irregular boundaries, a five‐point FA algebraic solution is derived; for a regular element, the nine‐point FA algebraic equation is used. In execution of numerical computation, the irregular five‐point boundary element is recasted into the nine‐point element, allowing a regular structured grid system to be used to solve for the water flow in an irregular domain. Two problems are solved in this study. The first is a transient, two‐dimensional flow of water under a strip source infiltration of constant flux with regular boundaries. Uniform clay loam, uniform coarse sand, and layered system of soil are used in the numerical simulation. It is shown that the finite‐analytic numerical solutions conform to accuracy and numerical stability very well. Mass conservation for the FA method also is illustrated. The second problem simulates unsaturated flow of leachate from landfill with irregular boundaries. The water pressure head and content in the unsaturated zone are predicted as a function of time and space.

A new point estimate method (PEM) suitable for estimating the statistical moments of the response of water resources models is illustrated. The importance of using the correct approach for modeling spatial variability in water resources modeling is also demonstrated using the PEM and the simulation of a backwater profile in an open channel. Random field models are conceptually more realistic than single random variable methods currently used for determining spatial variability. Although the mean model response using the single random variable model or the random field model is similar, there is significant variance reduction with the use of the random field model. The proposed point estimate method used to estimate the moments of the response of the backwater model proved to be efficient. It does not require knowledge of the underlying distribution of the random variables, and is an alternative method suitable in water resource modeling.