Journal of Hydrologic Engineering

Forum papers are thought-provoking opinion pieces or essays founded in fact, sometimes containing speculation, on a civil engineering topic of general interest and relevance to the readership of the journal. The views expressed in this Forum article do not necessarily reflect the views of ASCE or the Editorial Board of the journal.

Results of a simplified analysis of the hydrologic performance of undrained porous pavements, pavements without underdrains but that allow infiltration, are reported. The characterization of porous pavement behavior presented is appropriate for initial pavement design purposes. During any given storm, there are four distinct phases of porous pavement behavior: infiltration, accumulation, runoff, and drawdown. Analytical and graphical methods are presented for describing each of these phases from which a nonconvoluted approximate runoff hydrograph can be derived. This approach approximates peak runoff and time of that peak for a given storm depth, rainfall distribution, and pavement design. Second, a generalized method for calculating a porous pavement’s effective curve number (ECN) based on the National Resources Conservation Service (NRCS) [formerly Soil Conservation Service (SCS)] runoff curve number is presented. The analysis shows that for undrained porous pavements, the ECN is a function not only of the pavement and soil properties, but also of the storm depths and hyetographs used in the analysis. A standard calculation and presentation method is suggested that will allow local regions or municipalities to generate a local ECN design tool appropriate for their intensity-duration-frequency (IDF) data, which are only a function of the effective storage depth and the daily effective soil infiltration depth. A plot of the maximum ECN for all pavement depths and soil types is presented for each NRCS storm type as a reference for generating local ECN figures.

The return period of a rainfall event is estimated by the frequency analysis for a given rainfall duration. Thus, it is possible to derive different return periods with different rainfall durations for a given rainfall event. The longest derived return period is generally cited to represent the rainfall event. However, it is not clear if the longest derived return period is a representative measure of the given rainfall event. In this study, as a solution for this problem, a bivariate frequency analysis was introduced. As a first step, annual maximum rainfall events were selected by applying a bivariate exponential distribution. As an application, a total of 1,534 rainfall events observed in Seoul, Korea, over the last 46 years were analyzed. The annual maximum rainfall event series were then analyzed by applying a bivariate logistic model. The results were also compared with those from a conventional univariate frequency analysis. The findings of this study are summarized as follows: (1) the bivariate exponential distribution satisfactorily represented the duration and total rainfall depth data of all independent rainfall events, and the annually estimated parameters of the bivariate exponential distribution were more reasonable with respect to annual changes in the climatic conditions than those for the entire data period; (2) by using the bivariate logistic model, the return period was able to be assigned to each annual maximum rainfall event; and (3) rainfall quartiles of the univariate frequency analysis were bigger than those from the bivariate frequency analysis for rather short return periods of less than 30 years, but smaller for rather long return periods exceeding 100 years, primarily attributable to the smaller variance of the univariate annual maximum series.

This study analyzes variations in the model-projected changes in catchment runoff response after urbanization that stem from variations in the spatial distribution of impervious areas, interevent differences in temporal rainfall structure, and antecedent soil moisture (ASM). In this work, an ensemble of hypothetical imperviousness scenarios created for two small ($<1\mathrm{ha}$) watersheds were incorporated into the gridded surface subsurface hydrologic analysis (GSSHA) model, which was calibrated against 41 runoff events under natural conditions. Each event was resimulated for each imperviousness scenario. Variations in the model-projected changes in runoff were characterized and related to temporal rainfall dispersion, ASM, and two metrics: (1) proximity of imperviousness from the outlet, and (2) normalized number of downstream pervious elements. Key findings include the following: First, interscenario variations in the simulated runoff were relatively subdued on an event-mean basis but were much wider for individual events. For example, the coefficient of variation (CV) was less than 7.8% for runoff peak but was beyond 20% for certain events. Second, the rate of increase in simulated runoff peaks with elevated imperviousness tends to be lower for events with higher temporal rainfall dispersion and ASM, with one of the largest events exhibiting the slowest rate of increase. Third, both metrics were found to be negatively correlated with simulated runoff depth. These findings point to the possibility of refining the model projection by incorporating indicators of overall locations of impervious areas, rainfall dispersion, and soil moisture conditions.

Regional frequency analysis of extreme precipitation after a drought event based on the L-moment method has provided promising insights into drought risk management. The drought index of continuous noneffective rainy days was applied to analyze the droughts with the characteristics of increasing frequency, severity, and duration in the Heihe River Basin. Based on the typical drought records, the precipitation intensity after drought was shown to be greater than the long-term average values. Regional pooling of the precipitation after drought and under normal conditions for 1960–2010 from 14 sites across the Heihe River Basin was applied in a regional frequency analysis based on L-moments to estimate the quantiles for given return-period precipitation events in each of the three homogeneous regions for both the drought and normal series. From the study results, the following observations were made: (1) the Heihe River Basin (14 sites) can be classified into three homogeneous regions by the analysis of discordancy measure and heterogeneity measure. The results of goodness-of-fit measures showed that the Pearson type III (PE3) and generalized extreme value (GEV) distributions were suitable for the basin, but their respective performances for the drought and normal series were clearly different; and (2) the estimated quantiles at a given return period ($\mathrm{return}\mathrm{period}=1$, 2, 5, 10, 20, and 50 years) for the drought and normal series revealed that the high-intensity precipitation after drought was more than that under normal conditions in all three of the regions. In addition, the spatial variation of extreme precipitation in the drought and normal series was noticeable.

In the present paper, an analytical work for the description of the soil water balance and runoff production was adopted over a significant number of river basins belonging to a humid region of Southern Italy. The model is based on a stochastic differential equation, where the spatial heterogeneity of a basin is incorporated by a parabolic function describing the distribution of soil water storage capacity at the basin scale. The model provides an analytical description of the probability density function (PDF) of relative saturation of a basin as well as the PDF of daily runoff production. The proposed model includes five parameters that depend on climatic and soil characteristics. In particular, two parameters describe the rainfall process ($\alpha$ and $\lambda$), two characterize the distribution of soil water storage capacity ($w_{\max }$ and $b$), and the last is the soil water loss coefficient ($V$). Application of the model allowed the regionalization of model parameters based on physically consistent characteristics of the river basins. In particular, it was found that the soil water loss coefficient is strongly controlled by the fraction of forest cover of the river basin, while the parameter $b$, controlling the shape of the distribution of soil water storage capacity, is influenced by the basin topography.

Prediction of damage caused by floodgate maneuvers is treated by the determination of factors amplifying the magnitude of generated damage. The identification of vulnerable areas is obtained by analogy with the water-hammer phenomenon in pressurized pipes. The numerical model is based on the Boussinesq assumption for transient flow in open channels and solved by the Mac-Cormack scheme. Transient flow concerns a rectangular channel where the initial steady state is uniform. Transient regime is provoked by the movement of a floodgate at the upstream extremity of the channel producing sinusoidal fluctuations of depth. A condition of floodgate closure is imposed at the downstream extremity. The applied excitation has permitted analysis of the fluid structure interaction, examination of pulsations provoking the resonance of the free-surface waves, and study of the evolution of depths according to time along the channel. According to the frequency of the movement of the floodgate at the upstream end, distortion of depth profiles or oscillations over these profiles may appear. The frequency value also has a significant effect on the flow depth evolution and can provoke the flooding of the channel.

This study introduces the utility of the second-order neural network (SONN) method to model the reference evapotranspiration (${\mathrm{ET}}_0$) in different climatic zones of India. The daily climate data of minimum and maximum air temperatures, minimum and maximum relative humidity, wind speed, and solar radiation from 17 different locations in India were used as the inputs to the SONN models to estimate ${\mathrm{ET}}_0$ corresponding to the FAO-56 Penman-Monteith (FAO-56 PM) method. With the same inputs, for all 17 locations the first-order neural networks such as feed forward back propagation (FFBP-NN) models were also developed and compared with the SONN models. The developed SONN and FFBP-NN models were also compared with the estimates provided by the FAO-56 PM method. The performance criteria adopted for comparing the models were root-mean-squared error (RMSE), mean-absolute error (MAE), coefficient of determination ($R^2$), and the ratio of average output to average target ${\mathrm{ET}}_0$ values ($R_{\text{ratio}}$). Based on the comparisons, it is concluded that the SONN models applied successfully to model ${\mathrm{ET}}_0$ and performed better compared to the FFBP-NN models. This study also found that the SONN models yield better results using a fewer number of hidden neurons compared to FFBP-NN models. Better performance of SONN over FFBP-NN models suggest that SONN models can be used to estimate ${\mathrm{ET}}_0$ in different climatic zones of India.

Groundwater is considered as an important source of freshwater for a variety of purposes including drinking, domestic, industrial, and irrigation uses. Because of increasing population and life standards, there is a growing need for the optimum utilization of groundwater resources. In this paper, a multiobjective particle swarm optimization model with a new evolutionary strategy based on the compromise solution of the Pareto-front optimal solutions is presented. The advantage of this proposed model stems from using a unique Pareto-compromise solution to drive the fitness calculations of the evolutionary process. The new evolutionary strategy is verified on a variety of multiobjective standard test problems with either connected or disconnected Pareto fronts. The proposed multiobjective evolutionary strategy is reminiscent of single-objective optimization, in that its fitness assignment and convergence criteria are both based on tracking a single evolving solution over the search history. Details of the model development and implementation are described and an example application related to groundwater management is presented to demonstrate the capabilities of the proposed model. The proposed model showed its ability to drive the Pareto-optimal solution for the example application and consequently its ability to be applied in real-life groundwater management problems.

A nonlinear relationship between peak discharge and runoff volume (or depth = volume per unit area or rainfall amount), reported in the literature, was derived based on the standardized peak discharge distribution (SPDD) with regression analysis. However, the SPDD regression-based runoff model may only predict the mean behavior of peak discharge for a given runoff depth (i.e., the conditional expectation of peak discharge for a given runoff depth). This study proposes the application of entropy theory to derive the joint frequency distribution of peak discharge and runoff depth and the distribution of peak discharge conditioned on runoff depth. The conditional expectation of peak discharge (i.e., predicted peak discharge) for a given runoff depth is then compared with that obtained from the second-order SPDD regression-based runoff model. The entropy-based method is validated using data from 27 watersheds of different areas located in different climate regions in the United States. The results show that (1) with properly defined constraints, the entropy-based method may properly model the joint distribution of runoff depth and peak discharge and conditional distribution of peak discharge given runoff depth, and (2) the proposed method performs better than the regression-based method in terms of representation of extreme values that otherwise may be considered outliers.

Nonstationary behavior of recent climate increases concerns among hydrologists about the currently used design rainfall estimates. Therefore, it is necessary to perform an analysis to confirm stationarity or detect nonstationarity of extreme rainfall data to derive accurate design rainfall estimates for infrastructure projects and flood mitigation works. An extreme rainfall nonstationarity analysis of the storm durations from 6 min to 72 h was conducted in this study using data from the Melbourne Regional Office station in Melbourne, Australia, for the period of 1925–2010. Stationary generalized extreme value (GEV) models were constructed to obtain intensity–frequency–duration relationships for these storm durations using data from two time periods, 1925–1966 and 1967–2010, after identifying the year 1967 as the change-point year. Design rainfall estimates of the stationary models for the two periods were compared to identify the possible changes. Nonstationary GEV models, which were developed for storm durations that showed statistically significant extreme rainfall trends, did not show an advantage over stationary GEV models. There was no evidence of nonstationarity according to stationarity tests, despite the presence of statistically significant extreme rainfall trends. The developed methodology consisting of trend and nonstationarity tests, change point analysis, and stationary and nonstationary GEV models was demonstrated successfully using the data from the selected station.

In this work, a conceptual lumped model was developed to simulate runoff and analyze hydrological processes with the goal of incorporating the morphological information into a probability-distributed model (PDM). PDMs usually describe the process of runoff generation as the result of soil saturation excess caused by precipitation with soil storage capacity represented by a spatially distributed quantity and described by a probability distribution. The proposed model, called topography-based probability distributed model (TOPDM), based on a simple water balance whose components are basin soil moisture storage, precipitation, drainage to groundwater, evapotranspiration, and Dunnian and Hortonian surface runoff, is the result of the combination and integration of the topographic index within a PDM. A TOPDM was applied to the Baron Fork basin in Oklahoma in this research, and simulation results show that it provides a reasonably good estimation of runoff and a realistic representation of physical process in the catchment confirmed by the comparison with soil moisture data and with the results provided by a more complex, physically based distributed hydrological model.

A new method is proposed to compute prediction errors so as to not penalize models for good predictions with a temporal error. Traditional goodness of fit measures are based on model error calculated by pair-wise comparison of observed and simulated values at the same time, (which the authors term isotemporal predictions). However, even good models could have a random temporal error. Relatively small time shifts in predictions could produce large individual errors, especially if there are rapid changes in the system. This could lead to misleadingly poor goodness-of-fit statistics. The authors propose a modification to the calculation of model error that pairs each data point with a prediction that is closest in a Euclidean sense, (which the authors term proxitemporal prediction), instead of the same time. A normalization criterion is proposed to make the prediction and time scales commensurate based on the slope of the model predictions. The method is tested using stochastic simulation for a sinusoidal model and applied to an uncalibrated Escherichia coli water quality model. The stochastic simulation showed that the ability of the new method to accurately capture the true goodness of fit depends on the true prediction error and the true temporal error.

At any location, flood magnitudes can result from multiple causes (e.g., riverine flooding, tidal surges, snowmelt runoff, pluvial flooding, etc.) that occur simultaneously or independently. Analyzing flood data that resulted from simultaneous occurrences but treating the individual magnitudes as independent factors can lead to inaccurate estimates of flood depths, probability, and risk. While the assumption of independence is computationally convenient, it can be hydrologically incorrect. A method of joint flood frequency analysis that is based on copulas is proposed to address cases where the assumption of independence does not apply. For example, flooding at a coastal site can be the result of high riverine flows that simultaneously happen with high storm surges. The individual flood records would not adequately describe flood depths where both flood types occurred. The proposed method can be applied with two or more causes of flooding. The method is demonstrated using data from a site in coastal Florida, with the results showing that differences in flood estimates can be significant when independence is assumed for cases when a dependency exists.

This study considers the estimation of aquifer parameters for a spatially heterogenous aquifer from pumping test data. An approach is proposed that is based on modeling the unknown parameters as smooth Markov random fields. The associated inverse problem is formulated using the Bayesian framework and the posterior probability distribution of parameters is explored using Markov chain Monte Carlo. The method is evaluated by a numerical simulation in which measurements are taken in four observation wells. Even such a minimalist example presents significant computational challenges. Therefore, to obtain a computationally feasible solution, a model reduction is carried out and the estimation problem is reduced from over 1,000 parameters to 40 parameters. The approximate posterior distribution is then sampled using an adaptive Markov chain Monte Carlo sampler in order to quantify parameter uncertainty. This paper compares the parameter with predictive uncertainty and discusses the consequences of the model reduction.

Common probabilistic scores are used to evaluate the performance and reliability of interpolated rainfall scenarios based on geostatistics. The kriging variance is used jointly with the best linear unbiased estimator at the interpolation sites to render an ensemble of rainfall depth at each time step, providing dynamic information about the uncertainty of the interpolated field. Tests are conducted with rainfall data, extending from June 1, 2000, to October 31, 2005, from 50 stations located in the Saint Lawrence Valley, Canada. Three categories of variograms are examined: automated variograms that change at each time step; variable variograms that associate an average variogram to a specific interval of mean areal rainfall; and unique variograms, which are averages over the whole series. Simple, ordinary, and universal kriging are also examined. Rainfall scenarios constructed from automated variograms, simple kriging, and 30 neighboring stations presented acceptable sharpness and overall performance, including reliable exceedance probability and confidence intervals. An optimal choice of neighborhood is critical to performance and reliability.

Growing concerns have been given to parameter uncertainty in hydrological modeling because of the associated effects on water resource management. In this study two uncertainty analysis methods, the sequential uncertainty fitting algorithm (SUFI-2) and the generalized likelihood uncertainty estimation (GLUE), were employed to analyze the uncertainty of surface flow and sediment yield modeling results by the soil and water assessment tool (SWAT) model. The two methods were also compared in terms of the capability in quantifying parameter and prediction uncertainties and the computational efficiency. The results showed that the SUFI-2 method was capable of examining the uncertainty by using the Latin hypercube sampling scheme. It also demonstrated the advantages in analyzing the effects of uncertainties for distributed hydrological models with complex structure and high computational demands. The GLUE method is specialized in reflecting parameter correlations and uncertainties associated with parameters and predictants.

General circulation models (GCMs) use transient climate simulations to predict climate conditions in the future. Coarse-grid resolutions and process uncertainties necessitate the use of downscaling models to simulate precipitation. However, in the downscaling models, with multiple GCMs now available, selecting an atmospheric variable from a particular model which is representative of the ensemble mean becomes an important consideration. The variable convergence score (VCS) provides a simple yet meaningful approach to address this issue, providing a mechanism to evaluate variables against each other with respect to the stability they exhibit in future climate simulations. In this study, VCS methodology is applied to 10 atmospheric variables of particular interest in downscaling precipitation over India and also on a regional basis. The nested bias-correction methodology is used to remove the systematic biases in the GCMs simulations, and a single VCS curve is developed for the entire country. The generated VCS curve is expected to assist in quantifying the variable performance across different GCMs, thus reducing the uncertainty in climate impact–assessment studies. The results indicate higher consistency across GCMs for pressure and temperature, and lower consistency for precipitation and related variables. Regional assessments, while broadly consistent with the overall results, indicate low convergence in atmospheric attributes for the Northeastern parts of India.

This study aims to perform a comprehensive trend analysis of precipitation and streamflow at the watershed scale. Multiple techniques were employed to examine the temporal trends in annual, seasonal, and monthly precipitation as well as streamflow in the Xiangxi River watershed, China. Results showed that heterogeneous trends existed in precipitation and streamflow. To be specific, statistically insignificant trends in annual and seasonal precipitation were discovered, but varying monthly precipitation exhibited complex changes, especially for the decreasing precipitation in September. Annual and autumn average streamflow in addition to monthly streamflow in April, May, September, October, and November were found to experience a downward trend. Also, negative trends were detected in all low flow statistics, including annual minimum instantaneous streamflow; $Q10$, $Q30$, monthly, seasonal-minimum, and annual-minimum daily-mean streamflow; and 7-, 14-, and 30-day consecutive minimum streamflow. For high flow statistics, monthly maximum streamflow showed a slight upward tendency in dry months (December, January, and February), while its values in April, May, September, and October indicated a decreasing trend. In addition, $Q70$ and $Q90$ were found to have a statistically significant negative trend. Sequential Mann-Kendall (MK) test revealed an earlier beginning of the trends in low flow statistics than the other stream flows. Change point analysis suggested that changes in streamflow statistics had occurred since 1989, and accordingly, a new relation between precipitation and streamflow was established after 1988. Anthropogenic activities, especially small hydropower development, were found to be mainly responsible for changes in streamflow regime during the last 50 years. Thus, the obtained findings would help to better implement hydrologic engineering design and water resources planning and management over the watershed under changing environment.

Pump and treat (PAT) is one of the commonly used techniques for groundwater remediation in which the contaminated groundwater is pumped, treated, and put back to the aquifer system or other sources. For an effective PAT system design, simulation-optimization (S/O) models are very useful; the simulation model helps in predicting the spatial and temporal variation of the contamination plume, and optimization models are used to minimize the cost of pumping. For large-scale groundwater flow and transport simulation, grid-based or mesh-based models such as the finite difference method (FDM) and the FEM have been found to be cumbersome and time consuming. Meshfree (MFree)-based numerical models that use a set of nodes scattered within the problem domain and on the boundaries of the domain regardless of the connectivity information between them are found to be very effective for large-scale field problems. In this study, the MFree point collocation method (PCM) is used for the groundwater flow and transport simulation, and an evolutionary algorithm based on particle swarm optimization (PSO) is proposed for optimization. The S/O model based on PCM and PSO with multiobjective (PCM-PSO-PAT-MO) strategies are developed and applied for remediation of an unconfined aquifer polluted by total dissolved solids (TDS) using PAT. The PCM-PSO–based models are found to be simple and effective in groundwater remediation design using PAT.

A large number of dams were built in Italy in the first decades of the last century to supply the needs of industrial, electric power, agricultural, and drinking purposes. Among them, the Montedoglio dam, an important reservoir on the Tiber River located in central Italy, with a drainage area of $276{\mathrm{km}}^2$ and a maximum storage volume of about $153{\mathrm{Mm}}^3$. The dam is an earth-fill structure with overfall spillway partly controlled by two sluice gates. On December 29, 2010, due to the partial sudden collapse of the spillway, a huge volume of water flooded the valley below the dam, severely damaging the territory, but luckily without causalities. Considering that scarce data are available for this type of event worldwide, this paper aims to illustrate the collected data of the studied event in terms of reservoir levels, using a discharge hydrograph observed at downstream gauged river sites and the flooded valley area. This study aims also to simulate the breach evolution and the downstream propagation of the outflow. Considering the Nash-Sutcliffe performance measure in simulation of the recorded reservoir levels, the time of breach formation and the corresponding discharge coefficient are obtained through an optimization procedure, yielding values of 0.02 h and 0.15, respectively. Finally, the Manning roughness coefficients for the downstream main channel and the floodplain areas are detected through a methodology based on one-dimensional hydraulic modeling by considering different flooding scenarios of valley.