Journal of Hydrologic Engineering

The Watershed Environmental Hydrology model presents a new approach to the modeling of hydrologic processes in order to account for the effect of heterogeneity within natural watersheds. Toward this purpose, the point location–scale conservation equations for various hydrologic processes were upscaled in order to obtain their ensemble averaged forms at the scale of the computational grid areas. Over hillslopes these grid areas correspond to areas along a complete transect of a hillslope. The resulting upscaled conservation equations, although they are fundamentally one-dimensional, have the lateral source/sink terms that link them dynamically to other hydrologic component processes. In this manner, these upscaled equations possess the dynamic interaction feature of the standard point location–scale two-dimensional hydrologic conservation equations. A significant computational economy is achieved by the capability of the upscaled equations to compute hydrologic flows over large transactional grid areas versus the necessity of computing hydrologic flows over small grid areas by point location–scale equations in order to account for the effect of environmental heterogeneity on flows. The emerging parameters in the upscaled hydrologic conservation equations are areal averages and areal variances/covariances of the original point-scale parameters, thereby quantifying the spatial variation of the original point-scale parameters over a computational grid area, and, thus, the effect of land heterogeneity on hydrologic flows. Also, by requiring only the areal average and areal variance of parameter values over large grid areas, it is possible to achieve a very significant economy in parameter estimation.

The upscaled hydrologic conservation equations in the Watershed Environmental Hydrology (WEHY) model are capable of taking into account the effect of heterogeneity within natural watersheds. Model parameters that are capable of describing the heterogeneity of the flow domains also need to be estimated. The parameters of the WEHY model are related to the physical properties of the watershed, and they can be estimated from readily available information on topography, soils, and vegetation/land cover conditions. The fundamental assumption of the parameter estimation methodology in this study is that the heterogeneities of land characteristics over a model computational unit (MCU), representing an individual hillslope, are spatially stationary. Consequently, the parameters of the WEHY model that represent the statistical moments of point-location parameter values (such as the variance of log saturated hydraulic conductivity) do not vary over different transects of an MCU. The geomorphologic parameters that describe the rilled surface geometry of MCUs are derived directly from the digital elevation model map of the watershed by using Arc/Info tools or ArcView. The soil hydraulic parameters of the WEHY model require a soil texture classification database and/or field investigations. In this paper, a detailed methodology on the estimation of the geomorphologic and soil hydraulic parameters of the WEHY model is described. Applications of the methodology to the Upper Cosumnes River watershed in California and the Shiobara-Dam watershed in Japan are also presented in order to illustrate how to utilize the available geographic information system database and field survey data for the estimation of the parameters of the WEHY model.

The Watershed Environmental Hydrology (WEHY) model was applied to the Shiobara-Dam watershed for model validation for four historical events. Based on the analysis of the established geographic information system (GIS) database, which consists of the digital elevation model map, the land use map, the soils map, and the Landsat TM images, along with information on soil depths and soil properties, the modules of the WEHY model for the Shiobara-Dam watershed were selected and configured. A system of model computational units and stream reaches formed the basic building blocks for the watershed model. Using the model parameters, which were estimated from the established GIS database of the Shiobara-Dam watershed directly, good results were obtained when the WEHY model watershed runoff predictions were compared against their observed counterparts for several historical rainfall-runoff events. Interactions among the hydrologic flow processes at the watershed were examined in some detail.

The applicability and performance of the so-called M5 model tree machine learning technique is investigated in a flood forecasting problem for the upper reach of the Huai River in China. In one of configurations this technique is compared to multilayer perceptron artificial neural network (ANN). It is shown that model trees, being analogous to piecewise linear functions, have certain advantages compared to ANNs—they are more transparent and hence acceptable by decision makers, are very fast in training and always converge. The accuracy of M5 trees is similar to that of ANNs. The improved accuracy in predicting high floods was achieved by building a modular model (mixture of models); in it the flood samples with special hydrological characteristics are split into groups for which separate M5 and ANN models are built. The hybrid model combining model tree and ANN gives the best prediction result.

A neural network analysis of agrichemical occurrence in groundwater was conducted using data from a pilot study of 192 small-diameter drilled and driven wells and 115 dug and bored wells in Illinois, a regional reconnaissance network of 303 wells across 12 Midwestern states, and a study of 687 domestic wells across Iowa. Potential factors contributing to well contamination (e.g., depth to aquifer material, well depth, and distance to cropland) were investigated. These contributing factors were available in either numeric (actual or categorical) or descriptive (yes or no) format. A method was devised to use the numeric and descriptive values simultaneously. Training of the network was conducted using a standard backpropagation algorithm. Approximately 15% of the data was used for testing. Analysis indicated that training error was quite low for most data. Testing results indicated that it was possible to predict the contamination potential of a well with pesticides. However, predicting the actual level of contamination was more difficult. For pesticide occurrence in drilled and driven wells, the network predictions were good. The performance of the network was poorer for predicting nitrate occurrence in dug and bored wells. Although the data set for Iowa was large, the prediction ability of the trained network was poor, due to descriptive or categorical input parameters, compared with smaller data sets such as that for Illinois, which contained more numeric information.

This research utilizes Monte Carlo simulations of soil loss and runoff equations to assess impacts of the Jabiluka Uranium Mine site on Ngarradj, the downstream waterway. Ngarradj, a stream of high conservation value, debouches into wetlands of international importance within Kakadu National Park. The stream has a low background mean annual suspended-sediment concentration of $5.2\mathrm{mg}{L}^{-1}$ $(S=1.2\mathrm{mg}{L}^{-1})$ determined through monitoring. Erosion modeling indicated that a mean suspended-sediment increase of about $17.3\mathrm{mg}{L}^{-1}$ $(S=14.1\mathrm{mg}{L}^{-1})$ was expected from construction of the Jabiluka site’s waste rock dumps at Year 30 of the operation. The model indicated that the probability that the Jabiluka site activity would result in exceedence of the water-quality guideline for stream suspended sediment was 0.98 for that year. The evaluation was able to reach this conclusion based on 5 years of stream suspended-sediment concentration data and is highly conservative as a 1:1 sediment delivery ratio is assumed. Application of the simple methodology here not only enabled greater certainty concerning the risks of downstream sediment impact, but also identification of important management strategies.

In this paper, a method for streamflow forecasting in arid and semiarid regions is presented. The proposed methodology consists of three steps. In the first step, hydrologic seasons are defined to capture coupled effects of climatic and hydrologic characteristics of a basin. For this purpose, the correlation between average snow water equivalent (SWE), the similarity in basic statistical characteristics of the streamflows in different months, and the general characteristics of the study area are considered for defining hydrologic seasons. In the second step, a seasonal streamflow forecast time series is developed using autoregressive integrated moving average (ARIMA) models. In the third step, the fuzzy rules are developed to modify the statistical forecasts, utilizing average snow budget over a watershed, El Niño–Southern Oscillation (ENSO) climate signals, historical streamflow data, and time series of forecasted streamflows. Application of the proposed algorithm to the Salt River Basin in Arizona has shown improvement in the statistical seasonal forecasts. The most important application of the proposed methodology is to forecast streamflows six to nine months ahead in the summer of each year, when the ENSO signals are determined. This is of significant importance in water resources planning and management.

A model for estimating the watershed response is presented and used for assessing how the watershed size and spatial variability of the hydrodynamic parameters (i.e., wave celerity and dispersion coefficient) affect the comparative importance of advective processes with respect to hydrodynamic dispersive processes. A parameter $\Omega$ was defined to quantify this comparative importance. This parameter represents the fraction of the watershed response variance that is explained by advection. A series of simulations were performed for basins of different sizes and different spatial distributions of their hydrologic parameters. It was found that, for spatially uniform hydrodynamic parameters, the effect of hydrodynamic dispersion decreases compared to that of advection as the watershed size increases, and vice versa; and that, for nonuniform hydrodynamic parameters, the spatial distribution of the parameter values over the watershed, in conjunction with the watershed size, determines which process—advection or hydrodynamic dispersion—prevails.

In this technical note, a new and efficient control method based on fuzzy logic is proposed for the real-time operation of spillway gates of a reservoir during any flood of any magnitude up to the probable maximum flood. To demonstrate the performance of the proposed method, we simulate the control system using different probable overflow hydrographs. The results of the proposed control method are compared with the results of the conventional control methods. The results from the simulations indicate that the proposed method exhibits superior performance over conventional reservoir flood operation with much more reliable results.