Journal of Hydraulic Engineering

TWO stochastic modeling approaches for multisite hydrologic forecasting are presented in this work. The models are of ARMAX class expressed in state‐space formulations. The first approach, the adaptive partitioning, models a large catchment by shifting the locations in their time‐lags and modeling the flows of the different subsystems one at a time. The adaptive partitioning approach reintroduces the partitioning proposed by Wood in 1981 in an adaptive and more flexible form. The second approach, the cascading, divides the large catchment into groups of locations and models the groups as subsystems. In both approaches, the models' parameters and noise statistics are updated on‐line in an adaptive manner along with the states. For this purpose, the work proposes an evaluation/ forecasting algorithm based on the three parallel filter theory‐, a state‐space, parameter‐space, and a noise‐space filter. The algorithm is a synthesis and development of two preceding studies by Hebson and Wood, in 1985, and Bergman and Delleur, in 1985. The proposed evaluation/forecasting algorithm introduces a more flexible and comprehensive algorithm for the adaptive parameter‐noise statistics estimation of stochastic models. The two multisite hydrologic forecasting approaches use this tool in modeling large‐scale systems. In this work, the proposed algorithm and the two modeling approaches have been applied to daily stream‐flow forecasting of the Fraser River, Canada.

An analysis of the effects of the passive plane porous bed of an open channel on the steady, uniform, fully developed, two‐dimensional turbulent shear flow is presented. The effect of the pervious boundary on the free flow is taken as a perturbation of an equivalent turbulent flow over an impervious bed of the same surface texture. The turbulence field is regarded as stationary and statistically homogeneous in planes parallel to the channel bottom. Expressions for the mean velocity, the Reynolds stress, and the induced stress distributions are derived and compared with available experimental data. The new concept of induced pressure is introduced, and its functional form is obtained. The presence of a porous bed will in general cause the friction factor to be directly proportional to the Reynolds number, which agrees with experimental findings. The validity of Phillips' formula for the computation of the induced stress is discussed and clarified. The presented analysis improves our understanding of the physics of the problem and offers new analytical results for further study.

This paper examines the oscillatory flow induced by progressive gravity waves near a sea bottom covered with three‐dimensional ripples. Previous works on the subject, in which only two‐dimensional bottom waviness were considered, are thus extended. By assuming small values of the ratio ε between ripple amplitude and the thickness of the bottom boundary layer, the solution is given analytically by means of a perturbation approach. Arbitrary values of the ratio r between fluid displacement and ripple wavelength are considered. For vanishing values of r an asymptotic solution is also determined as an independent check for the previous one. Even though the solution can provide the flow field over arbitrary bottom topographies, detailed results are presented only for the profile of brick‐pattern ripples, which is characterized by transverse crests joined by longitudinal bridges. Finally, relevance of the results to understanding the mechanism of three‐dimensional ripple formation (brick‐pattern ripples, tile ripples) is discussed.

The objective of this study is to review existing methods and propose alternative or additional devices for scour protection around bridge piers. Engineering methods used in dealing with pier‐scour problems include placing coarse riprap around the foundations. An inverted filter layer is normally placed beneath the riprap to overcome leaching problems. Practical difficulties in laying the filter according to specifications, especially in deep water, often cause failure of the riprap layer. The study examines the effect of using a slot in the pier, or a collar around it, as devices for controlling the depth of scour. Experiments conducted in clear‐water conditions show that the use of a slot alone (with a width equal to one‐fourth of the pier diameter) can reduce equilibrium scour depth by as much as 20%. A combination of slot and collar can further reduce the scour depth. These devices may be useful as additional or alternative devices for scour protection.

Cohesive sediment beds occur in a variety of natural and man‐made environments. A parameter that has a very significant influence on the sediment behavior is the critical shear stress. This is the shear stress at the sediment surface above which erosion will take place. The critical shear stress can be evaluated in the laboratory under conditions closely resembling those found in the field by use of an annular flume. The main purpose of the work described in this paper is to determine experimentally and numerically the velocity and surface shear stress fields generated by the flow of water in an annular flume. Very good agreement is found between measured circumferential velocity profiles, obtained by laser doppler anemometry, and profiles predicted by the computational fluid dynamics program HARWELL‐FLOW3D. Measurements of the bed and wall shear stresses are made using hot‐film probes and these also agree reasonably well with numerical predictions. Also of interest is the dependence of the sediment shear strength on sediment depth. By comparing measured eroded sediment contours with predicted shear stress profiles a simple method emerges for describing this dependence.

The incidence method, graph‐theoretically formulated using incidence matrices, is useful as a rigid water‐column model for analyzing slow transients in complex pipeline systems. This incidence method is enhanced by state‐space notation and combined with the control theory. It leads to a more comprehensive analysis of transients than do numerical computations of initial‐ and boundary‐value problems. The number of variables required in the incidence method is reduced to a minimum, i.e., state variables for describing behavior of the system. To constitute state variables, linearly dependent discharges of pipes are explicitly expressed by linearly independent ones, using one of the incidence submatrices. To determine usefulness of the state‐space method, adequate valve operations for smoothly controlled transients are predicted by associating the theory of discrete optimal regulator with the actual valve characteristics, such as the speed of valves and the adequate range of valve opening.

Laboratory data, showing the effect of simulated floating debris accumulation at bridge piers on the local scour depth, are presented. The experiments are conducted using idealized impermeable cylindrical debris shapes attached concentrically to a cylindrical pier. The maximum scour depth recorded is 3.6 times the pier diameter. The design method proposed by Melville and Sutherland in 1988 is shown to predict the maximum possible local scour for a pier with simulated floating debris accumulation provided that an effective pier width is substituted for the actual pier width. The effective pier width is calculated as a weighted average of lengths (vertical dimension) of the debris and exposed pier.

This paper investigates the effect of a spoiler on local scour around submarine pipelines. Experiments conducted in a unidirectional flow show that the spoiler is capable of increasing both the scour depth beneath (tunnel erosion) and downstream of (lee erosion) the pipeline. The increase is related to the location of the spoiler compared with the direction of the incoming flow. The spoiler attached on top of the pipeline is effective in increasing the blockage to the incoming flow, which deflects a large proportion of the flow downward beneath the pipe, resulting in an increased scour depth. The increase in lee erosion, due to the protrusion of the spoiler above the pipeline, is important in the stimulation of the self-burial potential of the submarine pipeline. Spoilers attached to the lower half of the pipeline effected the formation of a wake zone beneath the pipe, which acts to deflect the flow downward, causing the formation of a deeper tunnel scour hole.

In this paper, a brief review is made to describe the situation and main problems of distorted hydraulic models. The concept is reaffirmed that the vertical distribution of velocity in a distorted hydraulic model is dissimilar to the prototype. For this reason, the measuring point of the vertical average velocity that is usually used cannot be directly employed. A similarity law for surface velocity is proposed, the factors affecting it are discussed theoretically, and the coefficients in the similarity scale are determined from field and laboratory data. An example of practical application is given. After using the proposed similarity law, the discrepancy of surface velocity of model to prototype decreased essentially. It is advantageous for hydraulic engineers using surface velocity to replace the vertical average value in physical models. It has been shown to be effective and satisfactory in our laboratory for hydraulic model studies with a large distortion. The similarity law proposed in this paper is considered to be used commonly.

Manning was looking for a dimensionally homogeneous velocity equation but did not achieve it. Later the coefficient in one of his equations was related to Kutter's n. The dimension of n is reviewed here. If n is indeed a boundary roughness geometry measure, its dimension should be length to the one-sixth power. Subsequently, a new coefficient n_g = (✓g/K_n) n is defined such that an improved, dimensionally homogeneous Manning formula is introduced. A short table of values of n_g is provided. Finally, an equation to evaluate the error in using Manning's formula for fully-developed turbulent, steady, uniform flow in rigid boundaries is given.