Journal of Hydrologic Engineering

The theory of flow in porous media has been developed on the basis of the experimental Darcy's equation. Observations from engineering practice and laboratory studies suggest that in certain cases it is necessary to fundamentally reconsider this theory. This paper analyzes experimental data collected to date demonstrating that in certain fluid-porous media systems the flow may be described by a seepage law with a threshold gradient: no flow occurs until the hydraulic head gradient exceeds the threshold value, then the flow velocity follows Darcy's equation. Using this flow equation, the paper presents solutions to two practical problems that have not been previously reported. First, the paper studies the problem of refraction of flow lines at the interface of two homogeneous and isotropic saturated porous media with different conductivities and derives a generalized “tangent law.” Second, the problem of unsteady infiltration is studied with a numerical solution to the flow equation for the one-dimensional case. The predicted behavior differs from the known classical results for Richards' equation.

Hydrologic modeling and relatively simple monitoring were used to estimate the hydrologic balance for two geographically close and, in the undisturbed state, hydrologically similar, zero-order basins: one undeveloped forest and the other suburban. Continuous precipitation and streamflow were measured in each basin; the model was used to estimate time series of evapotranspiration and ground-water recharge over a 40-yr period. The suburban catchment was denuded of forest cover, soil thickness was reduced, and 30% of the area was covered with impervious surfaces. The amount of annual precipitation that becomes runoff ranged from 12 to 30% in the forested catchment and 44 to 48% in the suburban catchment where runoff from pervious areas accounts for 40–60% of the annual total. The peak flow rate per unit area for an approximate 24-h, 50-yr rainfall was more than 10 times higher from the pervious area at the suburban site than at the forested site. These findings emphasize the need to consider surface flow from all sources in the catchment when considering mitigation measures.

This paper presents the development and application of a monthly water balance model of the Nile River basin (WBNILE) for assessing potential climatic change impacts on Nile runoff. The WBNILE model consists of 12 Nile subcatchments, including characterization of the lakes region of equatorial Africa and the Sudd swamp. The model uses mean monthly climate variables of precipitation, temperature, relative humidity, and sunshine hours, and each basin is calibrated from observed monthly averaged discharge. Areally averaged temperature and precipitation changes from five global circulation models (GCMs) were imposed on each subbasin for assessing climate change impacts on runoff. Results showed the sensitivity of the basin to climate fluctuations, because four of five GCMs predicted significantly larger flows in equatorial Africa and the expansion of the Sudd swamps, whereas there was a range of results for the Ethiopian highlands of the Blue Nile and Atbara basins depending on the GCM scenario.

An approach for developing a simplified variable-parameter stage-hydrograph routing method from the Saint Venant equations is proposed for routing floods in any shape of prismatic channel and flow following a generalized friction law. This approach enables one to relate the parameters of the routing equation to the channel and flow characteristics, and also enables the development of a theoretically based procedure for varying these parameters at every routing time level. Further, it allows the simultaneous computation of the discharge hydrographs corresponding to given input-stage and routed-stage hydrographs.

The variable-parameter simplified stage-hydrograph routing method developed in Part I of this series is studied to determine its limitations, the criterion for its applicability, and its accuracy based on the assumptions used for its development. This method is evaluated by routing given hypothetical input-stage hydrographs through uniform rectangular cross-section channels and comparing the results with the corresponding numerical solutions of the Saint Venant equations. The discharge hydrographs as computed by the method are also compared with the corresponding Saint Venant solutions. The method closely reproduces the stage and discharge hydrographs obtained from the Saint Venant solutions subject to compliance with the assumptions of the method.

Simulation of catchment processes that influence the quantity and quality of storm water runoff are employed in many circumstances where it is proposed to modify the existing catchment management practices and it is infeasible to trial physically the proposed management practice. An essential and necessary component of any catchment simulation is the model whereby the rainfall over the catchment is estimated. Most of these models, such as Thiessen polygons, were developed prior to the development of the digital computer and the subsequent development of hydroinformatic tools. Presented here are the results of a study investigating the accuracy and reliability of hydroinformatic tools for modeling the spatial and temporal distribution of rainfall over a catchment. It was found that using spline surfaces within a geographic information system (GIS) produced robust and accurate estimates of rainfall and enabled real-time estimation of spatially distributed patterns. Furthermore, use of the GIS enabled estimation of alternative hyetographs for different locations within the catchment.

Melaleuca forests cover extensive areas of south Florida, and are widely assumed to have extraordinarily high evapotranspiration (ET) rates. This investigation indicates that the transpiration of the melaleuca forest is dominated by vapor deficits, with net radiation playing a secondary role. Evaporation of intercepted rainfall contributes between 12 and 23% to the monthly ET, with lower contributions during the dry season and higher contributions during the wet season. Annual ET from melaleuca forest is estimated to be approximately 1,600 mm, with 300 mm contributed by intercepted rainfall. Comparison of these results with predictions of other forest-ET models indicate that the Shuttleworth model overestimates the ET of melaleuca forest by about 5% in a typical month as a result of neglecting the canopy wetness effect. The Calder-Newson model, which equates forest transpiration to short-vegetation transpiration, produces annual ET estimates that are within 6% of the annual ET estimated in this study, however, the discrepancies in monthly ET range between −8 and +17%.

The knowledge of the behavior of a water supply system during hydrological drought events is useful to assess the risk of shortages and to undertake the necessary actions to reduce drought impacts. In the present paper relationships between performance indices of a water supply reservoir and severity of hydrological droughts identified using the concepts of runs have been studied. A linear relationship between water shortage index and mean intensity (or cumulative deficit) of droughts of the same length was derived analytically and verified by simulation. Simulation has been carried out during drought periods for two demand patterns by using either standard operating policy (SOP) or different hedging policies. Plots of performance indices versus drought characteristics show in some cases a definite pattern whereas in others the points present a relatively high scatter. When a hedging operating policy is applied, though the scatter of the points increases, some relationships still preserve a definite pattern. The slope of such patterns can be used to evaluate the ability of different hedging rules in reducing drought impacts.