Journal of Hydrologic Engineering

Terrain models created from triangulated irregular networks (TINs) can be used to delineate watershed and subbasin boundaries. However, an important criterion for accurately defining boundaries is that triangle edges of the TIN honor linear drainage features such as streams, canals, and roads. Typically, this requires a certain amount of labor-intensive, manual editing. A new algorithm has been developed for automatically creating TINs from a set of Geographic Information System (GIS) objects that correspond to the drainage features. The resulting TIN honors all of the drainage features and is well suited to automated basin delineation.

Constructing a lumped parameter hydrologic model of a watershed involves dividing the watershed into subbasins and determining their connectivity through the stream network. Subbasin boundaries and stream networks can be described by Geographic Information System (GIS) data layers derived from digital terrain analysis or digitized from maps. By intersecting the subbasin and stream network data layers, then interpreting the resulting geographic features, it is possible to construct automatically a node-arc network description of the watershed comprised of connected hydrologic elements, including subbasins, reaches, junctions, reservoirs, diversions, sources, and sinks of flow to and from the watershed. The procedure, called CRWR-PREPRO, has been automated in Arc/Info Arc Macro Language (AML) and ArcView Avenue programs, which produce an ASCII file readable by the Hydrologic Engineering Center's Hydrologic Modeling System (HEC-HMS). A step-by-step procedure for interpreting geographic data to identify hydrologic elements is presented using a modified version of the Tenkiller Reservoir watershed in Oklahoma as an example. A further example application is presented of the application of CRWR-PREPRO to the Upper Mississippi basin and part of the Missouri basin, which results in a hydrologic model containing more than one thousand hydrologic elements.

The presence of zeroes in a record of low flows usually has been interpreted as being indicative of a stream that goes dry during some periods. However, in some instances, zeroes may appear in a gauging record simply because the actual discharge was below a measurement threshold. For example, in some regions it is sometimes true that the lower limit of a stream gauge is above the true stream bottom, and any (censored) discharges not registered by the gauges are recorded as zeroes. This paper presents a comparison of alternative methods for treatment of zeroes in low-flow estimation procedures when tail models are employed for modeling purposes. Simulation experiments indicate that a Weibull tail model, as fitted using a likelihood-based method for inclusion of censored data, is the best estimator of the 10-year low-flow quantile when zeroes in a data set have arisen as a consequence of censoring. When zeroes in a data set exist because the stream is an ephemeral one, a lognormal mixed tail model fitted using maximum likelihood is the best performer. When one is not sure whether zeroes in a data set are real, it is recommended that the 10-year low-flow quantile be estimated by treating the zeroes as censored values and performing the estimation using the Weibull tail model fitted using maximum likelihood.

Laboratory experiments of runoff production in shallow soils were conducted on plots consisting of a bedrock layer, an intermediate soil layer, and a crushed stone cover layer. Artificial rainfall was supplied in varying amounts and time sequences; and surface runoff, soil layer interflow, and throughflow were measured. The purpose of the experiments was to investigate infiltration mechanisms in shallow soils and to characterize the effect of stone cover on the partitioning of rainfall at the soil surface. Two soil types were used: (1) A sandy loam; and (2) a silty clay loam. In the case of the sandy loam, the results showed that the majority of rainfall infiltrated readily into the soil, where it subsequently moved downward to the bedrock as throughflow. Strong throughflow response was attributed to preferential flow through large soil pores. Because these macropores formed naturally during the experiments, this work serves as a plot-scale demonstration of the role of rainfall as a factor of change of soil hydrology. For the silty clay loam, the majority of rainfall exited the plot as surface runoff from the upper surface layer, while the relative magnitude of interflow and throughflow components of storm response increased with rainfall duration. Increasing the slope of the soil layers resulted in increased surface runoff and throughflow and reduced interflow. The influence of surface storage as represented by the thickness of the stone cover was found to be minimal for both soil types. This suggests that runoff production in both cases is controlled by the infiltration capacity of the soil layer alone.

The possibility of making short-term prediction of rainfall is studied by investigating the existence of chaotic behavior in the rainfall data series. The minimum number of variables essential and the number of variables sufficient to model the dynamics of the rainfall process are identified. The behavior of rainfall over different record lengths is studied. The effects of the data size and the delay time on the correlation dimension estimate are also analyzed. Daily rainfall data of different record lengths from each of six stations in Singapore are analyzed. The correlation dimension method, the inverse approach of the nonlinear prediction method, and the method of surrogate data (to detect nonlinearity) are used in the analysis. The results indicate that the rainfall data exhibit nonlinear behavior and possibly low-dimensional chaos, which imply that short-term prediction based on nonlinear dynamics might be possible. The minimum number of variables essential is identified as 3 and the number of variables sufficient lies in the range between 11 and 18. The results also indicate that the attractor dimensions of data of longer record lengths are greater than that of data of shorter record lengths. The study suggests that a minimum of ∼1,500 data points is required for the computation of the correlation dimension. Recommendation on the selection of the delay time is also provided.

This paper conducts a distributed parallel computation associated with domain decomposition with overlapping for the simulation of ground-water solute transport in the regional areas. Contaminant transport model based on the two-dimensional finite analytic numerical solution, an analytical-based numerical method, is taken as a discretization tool in this study. All the computation is carried out on the IBM/SP1 parallel computer with multiple processors, and a parallel virtual machine is utilized as a parallel computation tool for data communication and message passing among processors. The performance of the parallel computation is assessed through the numerical simulation in the illustrated example. The parallel implementation demonstrates that the simulation improves on accuracy, greatly reduces computational time speeding up to 5.5 times for using eight processors as compared with that of a unit processor, and the optimal efficiency 0.91 is found by using four processors. The extension to a three-dimensional parallel computation for the finite analytic method and the future aspects to be explored further also are addressed.

This paper emphasizes the supportive role of geostatistics in applying ground-water models. Field data of 1994 ground-water level, bedrock, and saltwater-freshwater interface elevations in south-central Kansas were collected and analyzed using the geostatistical approach. Ordinary kriging was adopted to estimate initial conditions for ground-water levels and topography of the Permian bedrock at the nodes of a finite difference grid used in a three-dimensional numerical model. Cokriging was used to estimate initial conditions for the saltwater-freshwater interface. An assessment of uncertainties in the estimated data is presented. The kriged and cokriged estimation variances were analyzed to evaluate the adequacy of data employed in the modeling. Although water levels and bedrock elevations are well described by spherical semivariogram models, additional data are required for better cokriging estimation of the interface data. The geostatistically analyzed data were employed in a numerical model of the Siefkes site in the project area. Results indicate that the computed chloride concentrations and ground-water drawdowns reproduced the observed data satisfactorily.

The design of infiltration basins used to dispose of treated wastewater or for aquifer recharge often requires estimation of ground-water mounding beneath the basin. However, the effect that the unsaturated zone has on water-table response to basin infiltration often has been overlooked in this estimation. A comparison was made between two methods used to estimate ground-water mounding—an analytical approach that is limited to the saturated zone and a numerical approach that incorporates both the saturated and the unsaturated zones. Results indicate that the error that is introduced by a method that ignores the effects of the unsaturated zone on ground-water mounding increases as the basin-loading period is shortened; as the depth to the water table increases, with increasing subsurface anisotropy; and with the inclusion of fine-textured strata. Additionally, such a method cannot accommodate the dynamic nature of basin infiltration, the finite transmission time of the infiltration front to the water table, or the interception of the basin floor by the capillary fringe.

A site was established by the U.S. Department of Agriculture-Agricultural Research Service on the Reynolds Creek Experimental Watershed in southwest Idaho in the fall of 1987 and operated through the spring of 1994, to compare precipitation catch between nine precipitation-measuring systems. This site was established as a part of the World Meteorological Organization's program to compare current national methods of measuring solid precipitation (snow), so the primary emphasis of this study was the measurement of snowfall. Over seven seasons, four of the systems measured snowfall and total catch, which included snow, mixed snow and rain, and rain events, within 4% of the Wyoming shielded gauge, which had the greatest total catch. These measuring systems were the Alter shielded gauge and the dual-gauge system from the United States, the double-fence shielded gauge from Russia, and the Nipher shielded gauge from Canada. The unshielded universal recording gauge that was mounted with its orifice at 3.05 m had the least catch in all precipitation categories, which amounted to 24% less snow, 18% less mixed snow and rain, and 10% less rain than was measured by the Wyoming shielded gauge.

This technical note presents a method for calculating the flow from springs and may be useful in the management of aquifers where the spring flow is a critical factor. The note examines the conjunctive operation of a groundwater-fed spring with pumping and artificial recharge directly from and to the same aquifer. Rorabaugh's (1963) solution for a one-dimensional homogeneous aquifer of a finite length is used, so a simple solution for the aquifer outflow due to a series of instantaneous discrete recharges or to pumping uniformily distributed along the aquifer is obtained. Only a few components of the infinite terms of the analytical solution are used; these are the first Fourier components of the sinusoidal eigenfunctions that solve the problem. As the eigenvalues of the solution are (2n + 1)²α, the factor to be applied to each component is exp[−(2n + 1)²αt], which decays very fast with time for high values of n. Thus, the structure of the solution can be perceived as an extension, with several cells, of the simple approach of the unicellular model.

At present, various methods are available to analyze storm runoff data. Among these, application of the Z-transform is comparatively new, and the technique has been used to identify rainfall from analysis of a single-storm runoff. The technique works under the premise that the rainfall-runoff process behaves as a linear system for which the Z-transform of the direct runoff equals the product of the Z-transforms of the transfer function and the rainfall. In the hydrologic literature, application aspects of this method to the rainfall-runoff process are lacking. As a result, some of the conclusions reported in the literature are questionable. The present study provides a rational basis for the application of the Z-transform technique to the analysis of storm-runoff data. Using data from a previously reported study, the study also explains the selection aspects of the unit hydrograph and the rainfall and lays the necessary background for the correct use of this method by hydrologists.

Reduction in deviation between predicted and observed runoff ordinates, and accurate estimation of loss rate parameters, have received considerable attention from many researchers. Linear programming is an optimization tool that is being used to determine the optimal unit hydrograph for a watershed from observed rainfall-runoff data, but requires the rainfall losses a priori. Nonlinear programming models overcome this problem but require good initial values of decision variables. The method presented herein determines the optimal loss rate parameters and unit hydrograph using infiltration theory, unit hydrograph theory, and linear programming. A comparison of the results with previous methods shows the better predictive capability of the model.