Journal of Hydraulic Engineering

A numerical simulation model for the dynamics of a doublediffusive system is presented. The model is an extension of an existing lake and reservoir model. In addition to adapting the parameterizations of the original model to suit a thermohaline system, the model accounts for the transport of salt and heat by double‐diffusive instability. The role of the sidewalls, both by virtue of their slope and by the differential heating that may occur there, are explicitly included in the treatment of the double‐diffusive fluxes. The use of the model as an aid in site selection for salt‐gradient solar ponds is described as an example of its application.

Analytical models are developed for the simulation of earthdam breach erosion. Using a reservoir water‐mass depletion equation, broad‐crested weir hydraulics and a breach‐erosion relation, solutions are derived for rectangular, triangular, and trapezoidal‐shaped breaches. Breach erosion is assumed to be either a linear or quadratic function of the outflow mean water velocity. Historical data are used to test the models. A sensitivity analysis is performed to determine the importance of the various parameters involved.

Experiments were performed in turbulent air flow in a trapezoidal duct with a mean aspect ratio of 3.5, corner angles of 64 and 116 degrees, and with both hydraulically smooth and partly roughened walls. Measurements in the fully developed flow region include wall shear stress and isocontours of the axial mean velocity, the normal and shearing turbulent stresses, and the turbulent kinetic energy. The formation of possibly symmetric secondary flows near smooth‐wall corners, and their suppression near smooth and rough‐wall corners, are documented. Finally, the inadequacy of gradient transport as a model of this flow is demonstrated.

Two trends in the adjustment of shape and slope of alluvial canals and rivers are identified. Using the concept of divergence from Lacey, these trends are quantified and reveal that the width, depth, and slope of alluvial channels are interdependent. The relationships evolved provide a simple procedure for the design of alluvial canals covering a wide range of bed material characteristics, sediment transport rate, and resistivity to bank erosion. Evaluation of the response of rivers to hydraulic structures and design of river link canals are also rendered easy with the aid of the two newly quantified trends.

The uncertainties implicit in suspended sediment transport curves are examined and the practice of using sediment load versus discharge is shown to be misleading, since the goodness of fit implied by this relation is spurious. The bases for equations to derive best estimates and confidence limits for suspended sediment concentration versus discharge are developed.

An integrated approach to the design of automatic control systems for canals using Linear Quadratic Regulator theory is developed. The one‐dimensional partial differential equations (PDE) describing open channel flow (Saint‐Venant equations) are linearized about equilibrium flow conditions and discretized spatially to provide a set of approximate ordinary differential equations (ODE) which describe the effects of gate openings on depth and flow rate. Standard linear quadratic techniques are used to design a regulator. The requirement to measure all states is obviated by the construction of an observer using only measurements of depth adjacent to control gates. Simulation results are presented which show dramatic improvements in transient response over the uncontrolled case. Use of these techniques in water conveyance canals facilitates more rapid changes in discharge, and by permitting longer periods of off‐peak pumping could greatly reduce pumping costs.

The Clark unit hydrograph is a three‐parameter synthetic unit hydrograph procedure that can be used in flood hydrology and other surface water runoff studies that require the development of a hydrograph or the reconstitution of a flood event. The technique is particularly valuable for unusually shaped watersheds, such as watersheds with large length‐to‐width ratios, or for application to watersheds containing several different physiographic areas, such as plateaus, escarpments, and valleys. The Clark unit hydrograph can be developed completely by a mathematical routing procedure that is computationally very efficient and lends itself to convenient computer applications. Although this unit hydrograph procedure has a strong theoretical basis and is very applicable to many watersheds, it has not gained wide application. Infrequent use by practicing engineers may be because of the difficulty in evaluating the storage coefficient R from recorded hydrographs, and the lack of empirical procedures to estimate R for ungaged watersheds. A procedure is presented to facilitate the evaluation of R from recorded hydrographs using the technique of hydrograph recession analysis, and an application is provided that demonstrates the development of an empirical, regional equation for estimating R for ungaged watersheds.

The principle of maximum entropy (POME) is employed to derive an alternative method of parameter estimation for the log‐Pearson type (LPT) III distribution. Historical flood data are used to evaluate this method and compare it with the methods of moments (MOM) and maximum likelihood estimation (MLE). The parameter estimates yielded by POME are comparable to those by MOM and MLE.