Journal of Irrigation and Drainage Engineering

An evaluation was made of the capabilities of watershed models when applied in the absence of site calibration data and limited validation data to predict the effects of major land‐use changes upon hydrology. Members of the task committee were surveyed for representative models and asked to assess the accuracy to be expected from each. A total of 28 surface hydrology models were considered and most were judged capable of providing “good” accuracy. The reasons for this confidence were explored and appear to be based upon personal experience, possibly tempered by belief in the model originators. In view of the limited number of model comparison studies conducted and the less than encouraging results often obtained, it appears that when models are applied at sites without hydrologic data, considerable care must be taken if reasonable results are to be expected. In addition, there appears to be little justification for using a model more complicated than necessary under these conditions.

Flumes having a local diminution of channel width are widely used as discharge measurement structures. The present investigation considers the modified version of a cylinder positioned axially in a nearly horizontal, prismatic channel. This particular configuration has several advantages over the usual one: determination of discharge at arbitrary, well‐defined channel cross sections needs no supplementary installations, and the device is simple and inexpensive. Moreover, the device enables a mobile registration of discharge at selected locations of a sewer or channel system. Discharge‐head relations are determined using both the traditional hydraulic approach and more sophisticated models that account for the streamline slope and curvature. Considerations include rectangular, trapezoidal and U‐shaped channels. Experiments in rectangular channels corroborate the proposed discharge‐head relation and reveal the internal flow mechanisms in the vicinity of the cylinder. Examples show the computational procedure, and graphs enable a simple application of results to problems.

A graphical method for the design of stable alluvial canals connected in a distributary system is presented. Under a given plan for water distribution, this method may be used to determine the stable width, depth and slope of each canal in the system. It has the advantages of relative simplicity and handiness as a design tool. Because of the sediment problems in systems design, the geometries and slopes of all canals in the same system must be interrelated in order to maintain sediment equilibrium. Without the use of structural sediment controls, sediment equilibrium is maintained by designing canals with an equal sediment concentration. Under such a system design, the power expenditure per unit weight of water stays approximately the same throughout the system; and the channel slope is inversely proportional to the one‐sixth power of the discharge. Canals with greater discharges have higher velocities but flatter slopes.

The method of characteristics (MOC) was revised to solve convective‐dispersion equations in a two‐dimensional, integrated saturated‐unsaturated porous medium. Instead of usual bilinear interpolation schemes, the revised procedure utilizes a three‐way linear interpolation scheme to assign seepage velocities to moving points in a two‐dimensional grid system. The coordinates and concentrations of moving points reintroduced into the two‐dimensional system are assigned using a random distribution. The accuracy of the numerical simulator was tested by comparison with available analytical solutions, numerical solutions and field data. Comparisons provided excellent agreement in almost all cases. Considering the nonhomogeneous field conditions encountered in the experimental plot, the simulated results are considered to be very good. The MOC eliminated numerical dispersion in simulation of miscible displacement in saturated, as well as unsaturated soils. A typical two‐dimensional drainage problem was solved in a nonhomogeneous, integrated saturated‐unsaturated porous media using the revised MOC.

A theoretical analysis of the free flow discharge characteristics of the throatless flume is presented. A pressure correction factor which accounts for the non‐hydrostatic pressure distribution at the control (throat) section is determined experimentally and incorporated in the free flow discharge equations developed from energy principles. The equations are valid for geometrically similar flumes in which the depth of flow in the approach channel is less than twice the channel width. Tests were conducted in three geometrically similar flumes. Both the present data and previously published data are used to validate the expressions developed empirically.

Flood storage in reservoirs controlled by free‐overflowing, rectangular spillways is investigated by accounting for arbitrary reservoir bathymetry and single‐peaked inflow hydrographs. The general solution of the storage equation depends only on a storage factor. It is found that adequate substitution of the effective hydrograph by a model hydrograph has only a secondary effect on the maximum reservoir outflow when using a proper shape factor. The proposed method allows an immediate application to realistic cases. In particular, the maximum reservoir outflow and the corresponding maximum reservoir level are represented graphically. Of further interest are cases in which the spillway crest length is varied for cost optimization purposes. Computations are explained in detail by examples, one of which includes a comparison of the present solution with the result of the numerically integrated basic equation.

A gravity drainage problem for a horizontal soil stratum underlain by a highly pervious material is analyzed. A non-linear relation between velocity and gradient is assumed. Capillary rise is also taken into account. The general differential equation (non-dimensionalized) for the problem is developed and analytically solved. The degree of drainage as a function of time and the real time of complete drainage are calculated and presented for typical values of the non-linear factor. It is shown that the analyzed non-linear drainage occurs faster compared to linear flow.

Low-cost electronic digital ``stick-on'' clocks were modified and used with a water sensor to determine irrigation rate-of-advance. The clock-sensors, spaced intermittently in a test furrow or in a border or basin, were used to determine the time that the advancing water front reached a given point. It was demonstrated that obtaining field data with a limited number of personnel can be greatly facilitated by using the clock-sensors.