Journal of Hydraulic Engineering

Calculations are carried out first for steady fully developed laminar flow on a slope. Application of this result to field data published in a previous paper by another author suggests that many debris flows are laminar. Then singular perturbation techniques are used to obtain first‐order solutions for unsteady non‐uniform flow from a point source on a steep slope. Solutions are obtained for both laminar and turbulent flows, and it is found that the laminar flow model gives the best description of laboratory data obtained by a previous investigator. A number of differences are observed in the behavior of the outer (kinematic wave) solutions for laminar and turbulent flow, and it is suggested that it is probably not possible to describe all debris flows and avalanches with laminar flow models. It is also suggested that more comparisons need to be made between theoretical results and field data measurements.

Bed‐load sampling in sand‐bed rivers was studied. A mechanical sampler called the Delft Nile sampler was designed and operated in flume and field conditions. The flume data obtained for different bed material sizes in the range of 280–1,070 μm were used to determine the sampling efficiency factor of the bedload sampler. The sampling efficiency factor was found to be about 1.0 for bed material sizes larger than about 400 μm and about 1.5 for finer bed material sizes. Field measurements in the Nile river (Egypt) and in the Rhine river (The Netherlands) were taken to determine the proper number of samples to be taken in a cross section in relation to the errors involved. The bed‐load sampler was equipped with an underwater video camera to facilitate visual observations of the bed‐load sampling process. Based on the visual observations, five types of bed‐load samplings were distinguished and a computation method was proposed to determine the proper mean bed‐load transport rate from the collected samples at a certain sampling station. The error of the mean transport rate in a sampling station was determined as a function of the number of samples. The samples should be taken at different locations along a bed form (dune). Finally, the overall error of the cross‐section integrated bed‐load transport rate was determined as a function of the number of samples in a sampling station and the number of sampling stations in a cross section.

Point‐measurement methods for estimating the mean velocity of vertical distributions of streamwise velocity in ice‐covered channels are evaluated in this paper. The evaluation uses profiles generated numerically, based on a two‐power law description of vertical distribution of streamwise velocity. The two‐power law simplifies into the well‐known power‐law expression for open‐water velocity profiles. Its validity was verified using measured velocity profiles obtained from rivers and a laboratory flume. The profiles are representative of flows subject to various combinations of bed and ice‐cover conditions. Values of estimation bias (the percent error in estimating average velocity from a few point measurements instead of integrating over the entire vertical velocity profile) were determined for several point‐measurement methods, including those currently used by the United States Geological Survey (USGS) and the Water Survey of Canada, and proposed new methods. The systematic variation of bias values with ice‐cover and bed‐resistance conditions was also determined. The method with the least overall bias is the so‐called usual two‐point method, in which velocity measurements taken at 0.2 and 0.8 of flow depth are averaged to obtain an estimate of mean velocity. This method produces a remarkably small bias; its maximum value is close to 2%. Estimation accuracy is enhanced if a coefficient of 0.98 is applied to the two‐point method.

The dispersive‐flow energy dissipator is a new design that combines features of traditional hydraulic‐jump, slotted‐bucket, and ski‐jump dissipators. Dispersive‐flow boards, chute blocks, and an adverse‐current end sill are the main components of the dissipator. The new dissipator outperforms traditional dissipators, and allows shorter and shallower stilling basins with less tailwater depth requirement; thus decreasing construction costs. The dispersive‐flow energy dissipator is best suited for total heads of 50–80 m, and entering flow Froude numbers of 5.2–7.2.

Initial dilutions of four ocean outfalls (the Miami‐Central, Miami‐North, Hollywood, and Broward outfalls) on the east coast of South Florida were determined from dye and salinity studies. In the dye studies, continuous injections of the red dye Rhodamine‐WT into effluent were conducted; dye concentrations were measured using a deck‐mounted fluorometer with a ship‐towed sampler and from grab water samples. In the salinity studies, temperature and conductivity were measured using a towed conductivity‐temperature‐depth device (CTD); salinity deficit was taken as a tracer to determine initial dilution. Results show that initial dilutions determined from both methods are consistent. Data for initial dilution and for environmental and effluent parameters are interpreted using the dimensional analysis method. A comparison is made between the present data and data from previous studies. Data for Hollywood and Broward outfalls (single‐port discharges) are consistent with previous data. Data for Miami‐Central and Miami‐North outfalls (multiport diffuser discharges) are not consistent with data for single‐port discharges. A value of $C_1=0.15$ for the asymptotic solution for the buoyancy‐dominated nearfield is suggested.

The modeling of flows at hillslope scale over land surfaces that contain rills and interrill areas is presented. The flow in rills is treated as one‐dimensional “channel flow.” The flow at interrill areas is treated as two‐dimensional “sheet flow.” Sheet‐flow equations are averaged along the interrill area width (local averaging) to account for the interaction between interrill and rill sections. The flow depth at the interrill area outlet section is derived by assuming that the flow profile of the interrill area section has the profile of a sine function. To minimize computational efforts and economize on the number of model parameters, locally averaged rill and sheet‐flow equations are then averaged over the transects of a hillsope (large‐scale averaging). The expectations of the terms containing more than one variable are obtained through the Taylor series expansion around their mean values. Only the first two moments of the series expansion are considered. The results of the model are in good agreement with the field observations. The effects of average local slopes and rill occurrence probability on flow rates are shown to be quite pronounced. On a rilled surface, most of the flow occurs in rill sections. Comparisons of model results with observed results over experimental hillslopes are satisfactory.

The use of deterministic atmospheric general circulation models (GCMs) to understand potential global climate change under doubled ${\mathrm{CO}}_2$ forcing has prompted a need for better understanding of local hydrologic impacts. Incongruities in model resolutions do not allow for GCM output to be directly used as forcing in the smaller‐scale hydrologic models. In this work, daily spatial disaggregation techniques are applied to the upper Rio Grande basin in Colorado, simulating local temperature and precipitation regimes, and preserving spatial covariance structures at all spatial scales. Canadian Climate Centre GCM output is disaggregated to site‐specific locations within the study basin. The Precipitation Runoff Modeling System is then used to examine hydrologic sensitivity under the disaggregated climate forcing. The results from this sensitivity indicate that under spatially disaggregated, site‐specific, climatic forcing, significant snowpack‐accumulation decreases occur. This results in total annual runoff decreases of, on average, 17.7%. A seasonal shift toward earlier in the year is observed in peak runoff, soil moisture storage, and evapotranspiration.

A study was conducted to determine how fall distance, or drop height, beneath a culvert outlet affects the dimensions of localized scour. Eighteen experiments were performed in which flow discharged from a circular, 4‐in.‐diameter pipe onto a near‐horizontal bed for 316 min. The bed material was relatively uniform sand with median grain diameter, $d_{50}=1.86\mathrm{mm}$. The scour hole dimensions, depth, width, length, and volume, were documented for drop heights of 0, 1, 2, and 4 times the culvert diameter above the bed. A series of expressions were developed correlating the dimensionless scour hold parameters of depth, width, length, and volume to the modified discharge intensity, ${\mathrm{QA}}^{-1}{g}^{-0.5}{R_H}^{-0.5}$, and to the drop height. It was determined that culverts placed above the bed result in deeper, wider, and shorter scour holes than culverts placed adjacent to the bed. Criteria for adjusting the existing U.S. Department of Transportation scour hole prediction procedure for drop height is presented in this paper.