Journal of Hydrologic Engineering

Accurate suspended sediment prediction is an integral component of sustainable water resources and environmental systems. This study considered artificial neural network (ANN), wavelet analysis and ANN combination (WANN), multilinear regression (MLR), and sediment rating curve (SRC) models for daily suspended sediment load ($S$) modeling in the Iowa River gauging station in the United States. In the proposed WANN model, discrete wavelet transform was linked to the ANN method. For this purpose, observed time series of river discharge ($Q$) and $S$ were decomposed into several subtime series at different scales by discrete wavelet transform. Then these subtime series were imposed as inputs to the ANN method to predict one-day-ahead $S$. The results showed that the WANN model was in good agreement with the observed $S$ values and that it performed better than the other models. The coefficient of efficiency was 0.81 for the WANN model and 0.67, 0.6, and 0.39 for the ANN, MLR, and SRC models, respectively. In addition, the WANN model presented relatively reasonable predictions for extreme $S$ values, acceptably simulated the hysteresis phenomenon, and satisfactorily estimated the cumulative suspended sediment load. Wavelet transforms provide useful decompositions of primary time series, so that wavelet-transformed data improve the ability of a predicting model by capturing useful information on various resolution levels. The proposed WANN model can be considered as a relatively new application of a combined wavelet and ANN model for suspended sediment prediction.

Air entrapment in soil is common in cases of farmland flood irrigation or intense rain. A simple, physically based model would be more useful than the complex two-phase (gaseous and liquid phase) flow model to describe water infiltration in layered soils with air entrapment. This study proposed a modified Green-Ampt model (MGAM) to simulate water infiltration in layered soils with consideration of entrapped air. A saturation coefficient $S_a$ was introduced in MGAM to account for the resistance effect of air entrapment on infiltration. $S_a$ had robust physical meaning, and was approximately equal to one minus the plus of the residual air and residual water saturation degree that could be determined from the soil water retention curve equation. In MGAM, the actual water content and hydraulic conductivity of the wetted zone were determined by multiplying $S_a$ with the saturated values. Infiltration experiments in a 300-cm-long five-layered soil column and a 280-cm-deep eight-layered field soil profile were conducted to test the applicability of MGAM. For comparison, the infiltration process was also simulated by the traditional Green-Ampt model (TGAM), in which the wetted zone was assumed to be fully saturated, and the Bouwer Green-Ampt model (BGAM), in which the hydraulic conductivity of the wetted zone was half that of the saturated hydraulic conductivity. The estimated $S_a$ values were very close to the measured saturation degree of soil layers at the termination of the experiment. The simulation results indicated that the TGAM overestimated the infiltration rate and cumulative infiltration, whereas the BGAM underestimated the infiltration rate and cumulative infiltration. Furthermore, the depths of the wetting fronts simulated by TGAM and BGAM were considerably smaller than those measured. The MGAM provided satisfactory simulation results and adequately described the infiltration process in both the laboratory soil column and the field soil profile.

On-site treatment of residential wastewater is important for rural and remote regions where centralized wastewater treatment may not be feasible. Sand filtration is a proven method for secondary treatment of septic tank effluent (STE) and can be an economical option where soil conditions do not favor the installation of a typical disposal field. This study investigates the hydraulics and treatment performance of eight lateral flow sand filters (LFSFs) receiving domestic STE at the Bio-Environmental Engineering Centre (BEEC) in Truro, Nova Scotia, Canada. The main objectives of this study were to evaluate the long-term treatment performance associated with the LFSFs and to investigate the effects of filter length and wastewater loading on performance. Previous research conducted at the BEEC indicated that current technical guidelines used in Nova Scotia for the design of LFSFs may be conservative and that the design of these systems could be optimized. The hydraulics and treatment performance associated with: (1) two shortened length filters (5.5 m versus 8 m standard length), and (2) six standard length filters receiving double the recommended hydraulic loading rate were evaluated. Results were compared to the previously documented performance of standard length filters operating at the recommended loading rate. Filters included fine (${\mathrm{d}}_{10}=0.15\mathrm{mm}$), medium (${\mathrm{d}}_{10}=0.17\mathrm{mm}$), and coarse (${\mathrm{d}}_{10}=0.30\mathrm{mm}$)-grained sands with hydraulic conductivities ranging from $1.5\times {10}^{-4}$ to $1\times {10}^{-3}\mathrm{m}{\mathrm{s}}^{-1}$ at slopes of 5 and 30%. Hydraulic residence times (HRT) were determined for each filter several times during the study period using a conservative tracer. Average removal efficiencies ranged from 82.4 to 96.9% for total suspended solids (TSS), 96.7 to 98.4% for five-day biochemical oxygen demand (${\mathrm{BOD}}_5$), 4.3 to 5.2 log reduction for Escherichia coli (E. coli), 43.7 to 93.4% for total phosphorous (TP), and 40.9 to 57.2% for total nitrogen (TN). Water quality analysis indicated that both the shortened filters and the filters receiving higher hydraulic loading rates provided a similar level of treatment for most parameters. This suggests that sand filtration is not only a physical treatment process but can be attributed mainly to Fbiological degradation of contaminants that occurs within the biomat. Filter length and loading rate adjustments did not appear to affect the HRT of the filters, suggesting that the biomat also plays an important role in the hydraulic functioning of the filters. Based on performance and tracer study results, LFSFs appear to maintain their hydraulic function and provide adequate contaminant removal at linear loading rates ranging from $66\mathrm{L}{\mathrm{m}}^{-1}{\mathrm{d}}^{-1}$ to $132\mathrm{L}{\mathrm{m}}^{-1}{\mathrm{d}}^{-1}$ of domestic STE. Long-term performance monitoring must continue in order to adequately assess the effects of filter length and wastewater loading on treatment performance and to further evaluate current design guidelines.

Various methods have been proposed to downscale the coarse resolution general circulation model (GCM) climatological variables to the fine-scale regional variables; however, fewer studies have been focused on the selection of GCM predictors. Additionally, the results obtained from one downscaling technique may not be robust and the uncertainties related to the downscaling scheme are not realized. To address these issues, the writers employed independent component analysis (ICA) for predictor selection that determines spatially independent GCM variables. Cross-validation of the independent components is employed to find the predictor combination that describes the regional precipitation over the upper Willamette basin with minimum error. These climate variables, along with the observed precipitation, are used to calibrate three downscaling models: multilinear regression (MLR), support vector machine (SVM), and adaptive-network-based fuzzy inference system (ANFIS). The presented method incorporates several GCM grids in the downscaling process that allows considering more predictors in the model calibration and removes the predictors correlation and dependence by ICA. Also, the study uses several downscaling techniques to develop an ensemble of precipitation time series that can be used in hydrologic climate impact assessment. The performance assessment of the results indicates that the procedure is successful in choosing the predictors for downscaling the GCM data both in monthly and seasonal timescales. The study shows that by choosing proper predictors the MLR model is an efficient method for precipitation downscaling.

Rain gauge monitoring networks are highly important for precipitation and erosion estimation. In general, measurement accuracy depends on the precipitation as well as on the network size and design. This paper proposes a method for assessing the optimal location of new monitoring stations within an existing rain gauge network. It takes account of precipitation as well as the prediction accuracy of rainfall erosivity. A well known geostatistical variance-reduction method is applied jointly with simulated annealing as an algorithm for objective function minimization. With respect to the first objective, the kriging variance of intensity estimation is minimized considering a reference duration of one hour. The erosion-related objective, meanwhile, is focused on the minimization of the kriging variance of estimation of the logarithm of the erosivity factor. The spatial variability of rainfall observed during the extreme rainfall event recorded in March 1973, which was the heaviest rainfall event in the period 1973–2003, is selected as the basis for developing the geostatistical approach. Scenarios in which the size of the initial network is increased by 25, 50, and 100% are tested. Another application is shown for increasing the network to meet the minimum requirement of the World Meteorological Organization relative to the network spatial density. Optimal single objective networks are compared for both the intensity and erosivity factor objectives. With regard to the new station locations, it is found that the optimal single objective networks differ by just one or two stations. Multiobjective optimization techniques are then applied to help in the network augmentation process. If both the spatial distribution of the rainfall intensity and the spatial distribution of the erosivity index are captured, this helps with the selection of the most relevant new stations. Results are identical to examples of single objectives when equal weighting factors are adopted for both objectives for the smallest augmentation scenario (only three new stations to be implemented). Different optimal rain gauge locations are obtained if one aspect is emphasized more than the other (rainfall versus erosion and vice versa) when a greater number of new stations is to be added (50 and 100% augmentation).

Producing realistic surface water flow patterns can be difficult for hydrologic models when there is insufficient grid resolution as a result of computational constraints or when available digital elevation model (DEM) data are relatively coarse. This technical note describes an algorithm that allows for more realistic overland flow by incorporating subgrid-scale stream location data without increasing grid resolution. The algorithm takes stream location data from the National Hydrography Dataset (NHD), maps it to the hydrologic model horizontal grid coordinates, and produces a list of ordered points along the stream. Because the algorithm uses indexes to efficiently order points along the stream, large-scale meanders require special treatment, whereas small (grid) scale meanders are explicitly included in the algorithm logic. Slopes are ensured to be continuous along the stream’s path as defined on the model grid, distinguishing this approach from traditional “stream burning” algorithms. Stream coordinates on the model grid are calculated along with corresponding elevation and slope values so that the stream can then be integrated into the DEM if desired. The algorithm’s flow routing capabilities are demonstrated by using an integrated surface water-groundwater model, ParFlow, under rain and recession conditions. This case study is performed by using the real topography of Owens Valley, California, and NHD flowline data for the Owens River. An approach using a DEM that has undergone standard processing to fill sinks and a typical “stream burning” approach [similar to those often used in geographic information system (GIS) applications] fail to route flow out of the flood plain in ParFlow’s overland flow model that partitions and routes topography-driven flow along adjacent cells in one of four cardinal directions. In contrast, this approach, with the river elevations and continuous slopes integrated into the DEM, routes water to the river and out of the catchment, creating more realistic surface flow patterns in the region. Although the method is applied here to address problems associated with a flat flood plain, it may also be applied to any area in which stream flow is discontinuous because of insufficient resolution of topography on a model grid.

Although phenomenon-mimicking algorithms, such as genetic algorithms, particle swarm optimization, and harmony search, have overcome the disadvantages of mathematical algorithms, such as the nonlinear least-squares method, segmented least-squares method, Lagrange multiplier method, a hybrid of pattern search and local search, and the Broyden-Fletcher-Goldfarb-Shanno technique, the algorithms have an inherent shortcoming. They require a tedious and skillful parameter-setting process for the algorithm parameters, such as the crossover rate, mutation rate, acceleration coefficients, harmony memory considering rate, and pitch-adjusting rate. Thus, this study proposes a novel parameter-setting-free technique interfaced with a harmony search algorithm and applies it to the parameter estimation of the nonlinear Muskingum model, which is an optimization problem with continuous decision variables. Results show that the proposed technique found good model parameter values while outperforming a classical harmony search algorithm with fixed algorithm parameter values.

The transient storage model with space-variable coefficients is solved by using the generalized integral transform technique (GITT) coupled with Laplace transform method. The spatially dependent flow velocity and solute dispersion are considered in the governing equations of the transient storage model. The correctness and accuracy of the derived analytical solution is verified by comparison with the analytical solution for constant flow velocity and the dispersion coefficient. Some illustrative examples are presented to demonstrate the application of the derived analytical solution. The increasing flow velocity and solute dispersion cause the faster transport of solute. Moreover, the faster the flow velocity and solute dispersion increase, the faster the solute transports.