Journal of Hydrologic Engineering

Qanats have been considered as one of the significant and economical methods for groundwater abstraction, thus studying the flow into and along them through numerical modeling is of vital importance. In this study, based on a three-dimensional (3D) analysis of the flow in anisotropic heterogeneous aquifers, the submerged zone of a Qanat has been described using a numerical groundwater model. The 3D finite difference control volume approach has been used to discretize the flow equation in the aquifer. The flow from the aquifer into the Qanat has been described with conductance values, and the hydraulic conditions within the Qanat have been described by the Darcy-Weisbach equation. The boundary condition at the Qanat outlet was set so that both full and partially full channel flow can be considered. Designing a computer code for the proposed model, the 3D hydraulics of a Qanat-aquifer system have been simulated. Model verification has been done by simulating a horizontal well-aquifer system developed in previous studies, and by using the coefficient of variation as verifying criterion. Good agreement between results of the proposed solution and those of previous work has been obtained in the verification process (coefficient of $\mathrm{variation}<0.023$), and the small differences can be attributed to applying different pipe resistance coefficient values and convergence criteria. Furthermore, the model performance was evaluated using observed flow depth and discharge in the Gavgard branch of the Joopar Goharriz Qanat. With this modeling, the behavior of the system, spatially and temporally, can be evaluated for different assumptions and exploitation conditions, and, consequently, the most appropriate exploitation conditions can be defined.

Hydraulic conductivity ($K$), a proportionality constant in Darcy’s Law, is one of the most fundamental parameters in groundwater studies. A falling head permeameter (FHP) test is one way to determine $K$, and its results are computed assuming a zero specific storativity of the tested sample. This study closes this gap by analyzing the effect of specific storativity on the $K$ calculation in FHP tests. The authors develop a solution for flow in FHP tests considering a nonzero specific storativity in the Laplace domain and use the de Hoog algorithm to attain the inverse Laplace transform of this solution to yield solutions in real-time domain, then enter into this solution a wide range of values of hydraulic conductivity and specific storativity to examine the significance of storage effects when using a FHP to determine the hydraulic conductivity of a porous sample. The study confirms that the specific storativity has a nearly negligible effect, and the solution with a zero specific storativity assumption can be practically used for FHP data interpretation.

Characterizing pore structure and hydraulic properties of engineered media used in stormwater control measures (e.g., green roofs and bioretention) is critical for accurately predicting water and contaminant flow characteristics, water availability to plants, and aeration status. In this study, unsaturated hydraulic properties, namely, the water retention characteristic (WRC) and unsaturated hydraulic conductivity ($K$), were measured for 14 engineered media with varying compositions. Their WRCs were also used to investigate pore structures and estimate hydraulic conductivities. Results indicate that water retention dynamics in 10 pumice-based media involve complex interactions between interaggregate and intra-aggregate pores, for which the commonly used van Genuchten function is not suitable. The majority of the pores in the tested media have radii less than 1 mm, for which the Richards equation is applicable. A predictive $K$ function based on WRC and Mualem’s approach tends to underestimate $K$ at low water contents. More research is needed to study hydrological behavior of engineered media with complex pore systems under various flow conditions.

In situ bioremediation of groundwater has become one of the most widely used technologies for contaminated site treatment because of its relatively low cost, adaptability to site-specific conditions, and efficacy when properly implemented. According to many studies, enhanced bioremediation techniques can change the hydrogeological properties of the polluted aquifers, and the most notable change is biological clogging, resulting in reduction of porosity and hydraulic conductivity of the porous media. Because biodegradation kinetics in previous studies do not simulate microbial growth rate explicitly in the aquifer system, in this study biological clogging is accounted for during the cost optimization of in situ bioremediation system. A simulation-optimization approach based on extreme learning machine and particle swarm optimization (ELM–PSO) techniques is used to design an optimal in situ bioremediation system for a characteristic site. A two-dimensional finite-difference model is used to get the data for training and testing of ELM. Further, a single-objective function is considered to optimize pumping cost, facility capital cost, and well cleaning cost for a clogged well. The application of the ELM-PSO method to problems where clogging of wells occurs provides a more practical and realistic cost for a typical in situ bioremediation system.

Accurate hydrologic time-series prediction plays an important role in modern water resource planning, water supply management, environmental protection, and power system operation. In general, single-layer feedforward networks (SLFNs) can provide satisfactory forecasting results, but classical gradient-based learning algorithms are time consuming and easily trapped into local optimums. As a new training method for SLFNs, extreme learning machine (ELM) has faster training speed and stronger nonlinear mapping than gradient-based algorithms, and provides an effective technique for hydrologic time-series prediction. However, ELM may converge to local minimums in some cases due to the random determination of input weights and hidden biases. Thus, in order to overcome the shortcomings of ELM, this paper introduces a novel ELM–quantum-behaved particle swarm optimization (QPSO) model (ELM-QPSO) combining the advantages of ELM and QPSO. The proposed model adopts the QPSO algorithm to select the optimal input-hidden weights and hidden biases of ELM, and uses the Moore–Penrose generalized inverse to analytically determine the output weights. The proposed approach is assessed with daily runoff data of Xinfengjiang reservoir in China from January 1, 2000 to December 31, 2014. The results indicate that the ELM-QPSO can significantly improve the generalization performance of ELM for hydrologic time-series prediction, and that QPSO is an alternative training algorithm for ELM parameters selection.

Although copula functions have been successively applied to flood frequency analysis, their application has usually been restricted to modeling bivariate dependence. This is because higher-dimensional expressions are only available for a few copula families, which may not have enough flexibility in modeling complex dependence structures. However, vine copulas, which have been recently introduced as an innovative method, can overcome such limitations. In this study, vine copulas are used for multivariate modeling of annual instantaneous maximum flows of three main tributaries located in the Euphrates River Basin, which is one of the most important sources of water for Turkey. The performance of vine copulas was compared with commonly used Archimedean (Clayton, Frank, and Gumbel-Hougaard) and elliptical (Gaussian and Student’s $t$) copulas. Statistical tests and tail-dependence assessments found the vine copulas to be most suitable for describing the dependence structure between variables. The developed vine copulas were used to obtain joint and conditional return periods of maximum flows, which can be useful for hydrologic design and management of water resources structures in the basin.

In the present technical note, two aspects commonly used to obtain the analytical solution of advection–diffusion equation (ADE) with time-dependent coefficients describing solute transport due to a continuous source in infinite and semi-infinite porous media, respectively, have been addressed. One is regarding describing a continuous source in an infinite medium and the other is in the context of the analytical solution of the ADE with time-dependent coefficients in a semi-infinite medium. Primarily, this note establishes that in an infinite medium, the correct concentration attenuation pattern from a continuous or instantaneous source may be obtained through the solution of the ADE only if the pollutant’s source is defined by a nonhomogeneous term of the ADE in the form of the Dirac delta function, whereas the solution of the ADE with a first-order decay term when the continuous source is expressed by means of an initial condition in the form of the Heaviside function does not exhibit the expected attenuation pattern. In the second part, it is shown that the analytical solutions of the ADE with temporally dependent coefficients in a semi-infinite medium may be obtained, contrary to it being held in the literature that it could not be obtained.

The hydrological simulation program Fortran (HSPF) [Hydrological Simulation Program Fortran version 12.2 (Computer software). USEPA, Washington, DC] and the precipitation runoff modeling system (PRMS) [Precipitation Runoff Modeling System version 4.0 (Computer software). USGS, Reston, VA] models are semidistributed, deterministic hydrological tools for simulating the impacts of precipitation, land use, and climate on basin hydrology and streamflow. Both models have been applied independently to many watersheds across the United States. This paper reports the statistical results assessing various temporal (daily, monthly, and annual) and spatial (small versus large watershed) scale biases in HSPF and PRMS simulations using two watersheds in the Black Hills, South Dakota. The Nash-Sutcliffe efficiency (NSE), Pearson correlation coefficient ($r$), and coefficient of determination ($R^2$) statistics for the daily, monthly, and annual flows were used to evaluate the models’ performance. Results from the HSPF models showed that the HSPF consistently simulated the annual flows for both large and small basins better than the monthly and daily flows, and the simulated flows for the small watershed better than flows for the large watershed. In comparison, the PRMS model results show that the PRMS simulated the monthly flows for both the large and small watersheds better than the daily and annual flows, and the range of statistical error in the PRMS models was greater than that in the HSPF models. Moreover, it can be concluded that the statistical error in the HSPF and the PRMS daily, monthly, and annual flow estimates for watersheds in the Black Hills was influenced by both temporal and spatial scale variability.

This study focuses on the detection of temporal changes in annual maximum gauge height (GH) across the continental United States and their relationship to changes in short- and long-term precipitation. Analyses are based on 1,805 U.S. Geological Survey records over the 1985–2015 period and are performed using quantile regression. Trends were significant only at a limited number of sites, with a higher number of detections at the tails of the distribution. Overall, there is only weak evidence that the annual maximum GH records have been changing over the continental United States during the past 30 years, possibly due to a weak signal of change, large variability, and limited record length. In addition to trend detection, the extent to which these changes can be attributed to storm total rainfall and long-term precipitation was also assessed. The findings of this study indicate that temporal changes in GH maxima are largely driven by storm total rainfall across large areas of the continental United States (east of the 100th meridian, the West Coast). Long-term precipitation accumulation, on the other hand, is a strong flood predictor in regions where snowmelt is an important flood-generating mechanism (e.g., northern Great Plains, Rocky Mountains), and is overall a relatively less important predictor of extreme flood events.

To study the effect of aggregate size and total porosity of soil on the hydraulic parameters of the Green-Ampt (GA) infiltration model, eight laboratory constant-head infiltration tests were conducted on a clay soil with different values of aggregate size and porosity. Hydraulic conductivity, wetting front suction head, and air entrapment coefficient behind the wetting front of the GA model were measured for the conducted tests. With the use of double-ring infiltrometers, falling-head infiltration tests were also conducted in a field with soil of a clay texture, and a new theoretically based method was proposed for simultaneous estimation of the hydraulic parameters of GA model from falling-head infiltration test data. The results indicated that the estimated hydraulic conductivity of the GA model for laboratory soil columns followed the Kozeny-Carman equation. The air entrapment coefficient is nearly independent from soil aggregate size, but it increases as soil porosity increases. Furthermore, a perfect agreement between the measured infiltration field data and the falling-head approach of the GA model validated the capability of the proposed method to correctly estimate the hydraulic parameters of the GA model.