Journal of Hydrologic Engineering

The presence of submerged vegetation divides flow velocity vertically into overflow and vegetation velocities connected by a strong shear. In addition, the flow structure exhibits periodic oscillations at the interface called Kelvin-Helmholtz vortices. The entire flow is divided into a combined vortex-overflow region that covers the overflow and upper part of the vegetation and a lower vegetation region below the vortices. A closed-form solution for longitudinal dispersion in this flow field is derived, which is an addition of the steady flow dispersion in the two regions and the dispersion caused by inefficient exchange between the two regions. The shear velocities in both the regions are modeled by using power law, whereas the vortices are modeled by the periodic components of the combined vortex-overflow stream function. The proposed dispersion model depicts shear-induced steady flow dispersion and vortex-induced dispersion caused by the inefficient exchange and shows reasonable agreement with past data.

A series of three-dimensional numerical simulations using a multidimensional hydrodynamic dispersion numerical model was performed to analyze various saltwater extraction schemes for mitigating seawater intrusion attributed to groundwater pumping in a coastal aquifer system. A steady-state numerical simulation was performed first to obtain initial (i.e., pregroundwater pumping) steady-state conditions before groundwater pumping, and then a transient-state numerical simulation was performed to obtain intermediate (i.e., postgroundwater pumping) steady-state conditions during groundwater pumping. In the subsequent series of transient-state numerical simulations as scenario and sensitivity analyses, four different saltwater extraction factors such as the amount of saltwater extraction, the number of extraction wells, the horizontal location of extraction wells, and the vertical interval of saltwater extraction were considered to determine an optimal saltwater extraction scheme for the coastal aquifer system threatened with seawater intrusion. The numerical simulation results show that seawater intrusion may be better mitigated when saltwater is extracted at 30% (up to 50%) of the groundwater pumping rate from a single extraction well, which is located horizontally midway between the pumping well and the coastline and is screened through the whole sand aquifer.

A unit hydrograph (UH) reflects the hydrologic response of watershed processes to a rainfall event. Most UH analyses use measured rainfall from a single rain gauge even though rainfall from a single gauge may not adequately reflect either the spatial or temporal variation during a storm event. The intent of this study was to better understand factors that influence UH accuracy, especially the spatial characteristics of rainfall. Analyses were made by using runoff and rainfall data, including radar rainfall measurements, from the densely monitored Walnut Gulch watershed, Arizona. Transmission loss (TL), as a physical process, is shown to be an important factor that influences UH shape. Multiple-storm, multiple rain gauge analyses show that rainfall from a single gauge is unlikely to yield a UH that reflects the watershed response. A comparison of UHs spatially averaged by using the temporal Thiessen method agree with UHs developed by using radar rainfall measurements. Intra-watershed analyses indicate that spatial variation of the physical processes, including channel processes that control transmission losses, can greatly influence the shape of a UH, thus suggesting that watershed subdivision is appropriate for analysis and design on even moderately-sized watersheds.

For evaluating the severity of drought in a certain month, two indexes of cumulative precipitation deficit (CPD) and maximum precipitation deficit (MPD) are used such that a regional frequency analysis has been carried out by L-moments. These measures of drought are strongly related to moisture deficits for the vegetation during the growing season. In this paper, 11 synoptic stations in Isfahan province, Iran, with a semiarid environment are used. The stations have a minimum of 10 years of data. The two drought indexes are calculated for all of them in whole months. The WeatherMan tool is used for completing missing data. The Food and Agriculture Organization of the United Nations Penman-Monteith (FAO-PM) method is used for calculating reference evapotranspiration [ ${\mathrm{ET}}_{(\mathrm{FAO}\mathrm{\text{-}}\mathrm{PM})}$ ]. According to previous studies, this method is suitable for this region. The Hosking homogeneity test is applied for identifying a homogeneous region. The Hosking goodness-of-fit test is performed to select the best regional distribution. The results show that the region is homogeneous in months 5 to 10, and in the rest of the months, the region becomes homogeneous after using clustering techniques and dividing the region into two smaller regions. Also, the best regional distribution is distinguished generalized logistic in the most months. Therefore, the severity of drought is estimated with various return periods by CPD and MPD. The maximum intensity of drought in the all stations happens in months 6 to 8, which are in late spring to midsummer. The severity of drought in Ardestan, Khoorbiabanak, and Naeen stations is greater than the remaining stations. Using the outcome of this study, the amount of required water can be estimated for agriculture each month, and the amount of water available is effectively managed in this province.

It is held self-evident that a watershed is among the best instruments with which to measure the spatial–temporal distribution of precipitation during heavy rainfall periods. Therefore, this paper describes an approach to discriminate physics of options of precipitation, normally built into meteorological models, using integrated meteorology and hydrology modeling. A hydrology model, WASH123D, was employed to calibrate and validate typhoon-induced hydrographs in Lanyang River basin using measured rainfalls as the driving force in a previously published paper. The calibrated hydrological parameters of the watershed include Manning’s coefficients in rivers and land surface, river characteristics, infiltration capacity of subsurface media, and so on. For forecasting purposes, rainfalls generated using numerical weather prediction models are needed as these are not known a priori. The problem is which physics in meteorological models ought to be used to produce the spatial–temporal distribution of rainfalls. An integrated meteorological and hydrological model is employed to address this issue. To our knowledge, this is the first time that the discrimination of rainfall models is accomplished via hydrological modeling. First, rainfall amounts and distribution, simulated with different options of physics built into the weather research forecasting (WRF) model, are investigated and assessed. The assessments indicate that it is very difficult to decide which physics is applicable to special conditions in Taiwan. Then, the integrated WRF and WASH123D model is employed to model hydrographs. The contrast between the hydrographs, simulated using rainfalls predicted by WRF and the rain gauge observation, is discussed. The magnitude and time lag of flood peaks from the simulated flood hydrographs are compared with those observed. Model calibration of hydrographs using the integrated model makes it easier to determine which physics in WRF is applicable to Taiwan from our selected cases. It was found that the Betts–Miller–Janjic cumulus parameterization is superior to others in our simulations for the case when the typhoon moves directly into the study area from the Western Pacific Ocean.

Controversy has arisen as to whether the lack of appropriate consideration to surface water flood risk in urban spatial planning is reducing the capacity to manage urban flood risk. A screening tool is required which would allow spatial planners to identify potential surface water flood risks and explore their management opportunities. An urban water balance approach is presented. The hypothesis is that key hydrological characteristics, storage volume and location, flow paths, and surface water generation, capture the key processes responsible for surface water flooding. The model is assembled and run by using ESRI ArcGIS software. Surface sinks and their catchment areas are identified by using a Lidar DEM. Excess surface water is calculated by using a runoff coefficient that is applied to rainfall volumes, and no other losses are considered. A surface water accumulation module sums the excess surface water from the catchment area of each sink. A sensitivity analysis of model assumptions demonstrates that these are valid for a screening tool. An informal validation of the model with local authority data revealed that most of the known flood risk locations were highlighted by the model. The model is applied to Keighley and sample results illustrate how knowledge of sink storage can be interpreted to explore opportunities for flood risk management. The model is a useful tool for quickly assessing potential flood risk locations and basic management options.

From a hydrologic perspective, one limitation of pervious concrete pavement is the risk of clogging, defined as a reduction in hydraulic conductivity that reduces infiltration into the pavement or exfiltration into the subgrade. Accordingly, a laboratory study was performed to measure clogging by sand and clay (sodium montmorillonite) in a saturated pervious concrete pavement system, and the subsequent effect of surface cleaning by pressure washing. Both sand and clay caused measurable clogging that was not reversible by pressure washing. However, even after clogging, the infiltration and exfiltration rates were well above the average intensity of $66\mathrm{mm}/\mathrm{h}$ for the 100-year 1-h design storm for Denver. This result is encouraging, but should be interpreted with caution, because in these experiments the flow-limiting layer was never the pervious concrete, but rather the subgrade, which in this case was a thin layer of sand with a large hydraulic conductivity. Accordingly, this study suggests that pervious concrete would be effective for storm-water detention when provided with an engineered drainage system, but that its performance for storm-water infiltration will be limited by the rate of exfiltration to the subgrade.

Observed changes in midelevation snow cover and duration have raised concerns over future impacts of global warming on snowmelt-dependent water resources and ecosystems. However, predictions of future changes in snow hydrology and water supply from mountain basins are complicated by natural variability in climate and interactions among topography, vegetation structure, wind and radiation energy, and snow deposition. In this study, interannual variability in snow-cover development, snow melt, and runoff is assessed for a range of precipitation and temperature conditions typical of a mountain catchment, the Reynolds Mountain East (RME) basin, in Idaho. A spatially distributed energy and mass balance snow model, Isnobal, coupled with a windfield and snow redistribution model, is used to continuously simulate snow accumulation and melt for five individual snow seasons (1984, 1986, 1987, 2001, and 2006), representing the historic range of climatic variance. The modeling results compare well with the field measurements of snow water equivalent for all simulation snow seasons (Nash-Sutcliffe model efficiency coefficient of 0.81 to 0.97). The modeling scheme used in this study demonstrates spatial and temporal differences in snow-cover development and melt processes in complex mountain terrain. During all simulation snow seasons, the forested site was shown to hold more snow, generate more snowmelt, and melt later than the exposed site of the basin. Comparison of simulation results for wet and dry snow seasons showed that the snow melts earlier and a greater fraction of surface water input (SWI) is utilized to fill soil water storage during dry seasons than in wet snow seasons. This study demonstrates differences in snowmelt volume and timing during wet and dry seasons within a mountain basin. It will improve understanding of how interannual variability affects the delivery of water from the seasonal snowcover to the soil and streams of mountain basins, and can provide useful inputs to the decision-making process for reservoir and water management strategies for utilizing surplus or supplementing deficit water supplies during wet and dry years, respectively.

Derivation of travel time based on Saint-Venant equations may substantially improve rainfall-runoff routing, particularly in the Clark approach and GIS-based hydrological routing models. Kinematic wave (KW) and diffusive wave (DW) approximations are often applied in lieu of the dynamic wave (DYW) solution to study overland flow regime. In this study, for the first time, travel-time relationships are presented based on DW approximation. In terms of equilibrium time, the DW approximation improved results from 5% up to 15% over the KW approximation. After obtaining the semianalytical solution of the DW travel time, discharge hydrographs over a rectangular plane through the well-known time-area method were derived. Considering the DYW solution as the benchmark, it was found that the time-area approach by adopting the proposed DW wave travel time formulation, provided satisfactory results for $2.5<K_0{F}_0^2<25$ and $F_0>0.1$ . Using the travel time based on KW approach, the time-area method was found fairly accurate for $K_0{F}_0^2>25$ and $F_0>0.1$ . $K_0$ and $F_0$ are the well-known kinematic wave number and Froude number, respectively. Therefore, it is perceived that as $K_0{F}_0^2$ decreases, the difference between KW and DW approximations becomes greater.

Changes in climate were a growing concern during the last decade and will be even greater in the coming years. When investigating the impact from changes in the climate on urban drainage systems, two challenges are (1) what type of input rainfall data to use and (2) what parameters to use to measure the impacts. The overall objective of this study is to investigate the hydraulic performance of urban drainage systems related to changes in rainfall, and through these hydraulic parameters describe the impact of climate change. Input rainfall data represent today’s climate and three future time periods (2011–2040, 2041–2070, and 2071–2100). The hydraulic parameters used were water levels in nodes (e.g., as the number of floods, and frequency and duration of floods) and pipe flow ratio. For the study area, the number of flooded nodes and the geographical distribution of floods will increase in the future, as will both the flooding frequency and the duration of floods.

The natural hydrologic flow regime is altered by urbanization, which can be mitigated through best management practices (BMPs) or low impact development (LID). Typically, the effectiveness of different management scenarios is tested by comparing post- and predevelopment instantaneous peak flows. This approach, however, does not capture the extent of hydrologic change and the effect on downstream communities. A new hydrologic sustainability metric is presented here to quantify the impact of urbanization on downstream water bodies on the basis of the inundation dynamics of the flow regime. The hydrologic footprint residence (HFR) is designed to capture both temporal and spatial hydrological changes to an event-based flow regime by calculating the inundated areas and duration of a flood. The HFR is demonstrated for a hypothetical watershed and a watershed on the Texas A&M University Campus, located in College Station, Texas. For the campus watershed, three design storms (2-, 10-, and 100-year) and a set of historical events (during the period 1978–2009) are simulated for various management scenarios, representing predevelopment conditions, development on campus, BMP-based control, and LID-based control. The results indicate that the HFR can better capture alterations to the shape of the hydrograph compared with the use of the peak flow only.

Dense reservoir networks, with thousand of small dams, can be increasingly found throughout the world, especially in water-scarce environments, such as the Brazilian northeastern region. Although the effect of individual small dams might be negligible, their joint effect has proved to be relevant on water and sediment connectivity. Literature, however, is scarce concerning the effect of such networks on water availability and/or sustainability of mesoscale or large-scale basins. This research intended both to assess the effect of the dense reservoir network of the semiarid Upper Jaguaribe Basin (UJB; $24,200{\mathrm{km}}^2$ , in Brazil) for a 45-year period (1961–2005) and to investigate a network arrangement that maximized its hydrologic sustainability. Imagery of the years 1970 and 2002 was analyzed to assess temporal evolution of the network. The Water Availability in Semiarid Environments (WASA) model, which proved valid for the Upper Jaguaribe Basin, was used to assess its sustainability for almost 100 different arrangements. The optimization process was performed using the Simplex-MSX algorithm. Results showed that the hydrologic sustainability of the basin increased throughout the analysis period. Nonetheless, since the late 1990s, the system showed clear signs of saturation. The optimization process identified that the best network arrangement should have residence time of approximately 3 years, with volumetric density of $0.2{\mathrm{hm}}^3·{\mathrm{km}}^{-2}$ , and the importance of the reservoir size continuity to obtain effective hydrologic performance.

Because the problem of prediction in ungauged basins (PUB) has become a central theme of hydrological research, there is a clear need to have an effective and efficient method that can help to transfer information from gauged basins to ungauged ones. Previous research relating the parameters of a hydrologic model to physical drainage basin characteristics has had limited success. Conventional regression procedures do not take into account multicollinearity among the basin characteristics, which have a great effect on the stability of regression equations. This study presents an unbiased ridge regression estimate (URR) that can provide robust regression equations in the presence of multicollinearity. The proposed method is applied to estimate parameters of the Xinanjiang model in 20 watersheds located in southeastern China. Compared with conventional estimations, such as ordinary least squares estimate and ordinary ridge regression estimate, the URR was reduced by 78% for root mean square error, 47% for variance, and 13.1% for Akaike’s Information Criterion. The coefficient of determination $R^2$ was increased by 49%.

Model subdivision is used to capture spatial heterogeneity in input parameters and it is well-established that spatial resolution (i.e., degree of aggregation) affects model output. However, a general consensus about the effect does not exist. The objective of this study was to investigate the effects of spatial resolution on model predictions in an urban catchment, and to understand the mechanism(s) responsible for the scale effect. The general approach is to develop models at various spatial resolutions, perform simulations, and compare the predictions of total outflow volume and peak flow. Models were developed on the basis of actual drainage networks, and artificial ones generated on the basis of a fractal algorithm by using the Artificial Network Generator (ANGel). Simulations were performed by using the EPA Storm Water Management Model (SWMM), and model output was compared for 50 storms. There was very little difference in the total annual outflow volumes predicted by the different resolutions. However, peak flows showed a dual scale effect. For the larger storms, model aggregation reduced peak flows, which can be explained by differences in infiltration. This effect was attributed primarily to the spatial distribution of the soil-saturated hydraulic conductivity and the length of overland flow. For the smaller storms, aggregation increased peak flows, which can be explained by the combined effects of overland flow and conduit routing. The results were consistent using actual and artificial networks. This study illustrates that a scale effect can be introduced by different processes, which can go in different directions (i.e. increase or decrease peak flows) and depend on the storm characteristics.

Soil texture characteristics of a small coastal granite catchment in the Pearl River Basin in southern China were analyzed, revealing loamy sand and sandy loam as the main soil types in this catchment. Rainfall-runoff processes were then analyzed on the basis of the observation of precipitation, runoff yield, and soil water content at two experimental plots in this catchment. The two study plots represented forest and shrub cover, respectively, and both were underlain with highly weathered granite. Average runoff yield coefficients for the forest and shrub plots were 0.113 and 0.042, respectively. A relatively large proportion of interflow was generated only when rainfall was sufficiently large in amount or intensity. A logarithmic relationship existed between precipitation and the maximum infiltration depth for plot A. The main direction of rainwater movement at plot scale was vertical. The runoff yield was composed primarily of infiltration excess overland flow and interflow at the plot scale, whereas overland flow (saturation excess overland flow generated near the stream), interflow and bedrock fissure flow were the main storm flow components of this granite catchment.

The calibration of hydrological models using multiobjective algorithms generates several competitive solutions, usually referred to as a Pareto-optimal set. Pareto-optimal solutions are nondominated (i.e., accurate in different ways) and give users several decision scenarios and alternative trade-offs between model evaluation objectives. For decision-making purposes, a single solution is often chosen to represent the properties of the problem under consideration. From a practical standpoint, users are interested to know if selected solutions will continue to be nondominated when evaluated for future periods. In this paper demonstrates an evaluation framework to compare four autoselection methods commonly applied to select a solution from the Pareto-optimal set. The Pareto-optimal sets were generated by using the nondominated sorting genetic algorithm-II (NSGA-II) to calibrate the soil and water assessment tool (SWAT) for simulations of streamflow in the Fairchild Creek watershed in southern Ontario, Canada. The analysis was conducted for 15 calibration outputs in different periods, and each output was evaluated for another 15 different validation periods, resulting in a total of 225 evaluations for each autoselection method. Only a subset of nondominated solutions during the calibration phase remain equally accurate when evaluated at a future time. The results showed that a selection criteria based on a compromise between a representative pathway in parameter space and a dominant variability in objective space is important to finding solutions that remain nondominated for several validation periods. That is, the most suitable solutions are those that have commonalities in parameter space and whose responses at the watershed outlet are similar.

Discriminating between competitive statistical models is an important problem in hydrological frequency analysis. The present study deals with discrimination between the two-parameter lognormal ( $LN2$ ) and the two-parameter log-logistic ( $LLOG2$ ) distributions, or, equivalently, between the normal ( $N$ ) and the logistic ( $LOG$ ) distributions. Previous work using the likelihood ratio (LR) statistic suggested that discrimination between these distributions is difficult, and that criteria other than LR need to be studied in hope of finding criteria with better discriminating power. In the present study, several criteria other than LR are considered, and their ability to discriminate between the $LN2$ and $LLOG2$ distributions is assessed using Monte Carlo simulation. The results confirm previous findings that discrimination between the two distributions is difficult with small samples, but a criterion based on the Shapiro-Wilk statistic appears to be the most appropriate for sample sizes typically encountered in hydrology. Two hydrological examples are presented to illustrate how obtained results can be implemented in practice.

A stochastic kinematic wave model for open-channel flow under uncertain channel properties is developed. The Fokker-Planck equation (FPE) of the kinematic open-channel flow process under uncertain channel properties is developed by using the method of characteristics. Every stochastic partial differential equation has a one-to-one relationship with a nonlocal Lagrangian-Eulerian FPE (LEFPE). As such, one can develop an LEFPE for the governing equation of any hydrologic or hydraulic process as the physically based stochastic model of the particular process. A linear, deterministic, differential equation in Eulerian-Lagrangian form, LEFPE provides a quantitative description of the evolution of the probability density functions of the state variables of the process at one shot to describe the ensemble behavior of the process instead of the commonly used many Monte Carlo simulations to quantify the same ensemble behavior. Under certain assumptions, the nonlocal LEFPE reduces to the classical local FPE, which is more convenient in practical applications. The numerical solutions of the resulting FPE are validated by Monte Carlo simulations under varying channel conditions. The validation results demonstrated that the developed FPE can express the ensemble behavior of the kinematic wave process under uncertain channel properties adequately.

Variable selection is a process of reducing data dimensionality by eliminating irrelevant or redundant variables. This is an essential step in any application that involves a large number of input variables and is a challenging task for model developers. Enumerative variable selection is usually impractical for large numbers of variables, hence efficient and effective variable selection tools are needed for real world problems. This paper describes a novel application of the gamma test with both forward and backward selection methods. These methods are tested with a flood regionalization problem. It has been found that the gamma test is able to provide useful input variable combinations and its best result outperforms the conventional forward cross-validation method. However, the gamma test results still have undesirable uncertainties. This is reflected by the difference between the gamma test backward selection and forward selection, as well as the uncertain Gamma statistic. More explorations are needed to improve or falsify the gamma test.

Computer models have become vital decision-making tools in many areas of science and engineering including water resources. However, models should be properly evaluated before use to improve the likelihood of making sound decisions based on their results. The model evaluation technique practiced today in hydrology assumes that model parameters are season insensitive and attempts to identify “optimal” values that would describe watershed behavior during dry and wet seasons. This assumption could compromise accuracy of model predictions. This study demonstrates performance improvement that would be achieved when a season-based model evaluation approach is pursued. A global sensitivity analysis (SA) model has been used to investigate seasonal sensitivity of streamflow parameters of a watershed simulation model on the headwaters of the Little River Watershed, one of the United States Department of Agriculture’s experimental watersheds. Two separate analyses have been performed: the conventional approach in which model parameters are assumed to be season insensitive; and a season-based evaluation in which the influential parameters may vary for months with a low runoff coefficient and months with a high runoff coefficient. The sensitivity analysis helped to identify dominant model and watershed behaviors for the conventional annual approach and for the wet and dry seasons. The SA results show that the influential parameters exhibited modest seasonal sensitivity for the experimental watershed. Model calibration was then performed by using the dynamically dimensioned search (DDS) algorithm for the conventional and season-based approaches using the principal parameters identified by the global SA model. Performance of the calibration attempts have been verified with the traditional split-sampling technique and also by assessing effectiveness of the model in predicting internal watershed behaviors through comparison of simulated streamflow with observations at multiple internal sites not used for calibration. Several efficiency measures have been used to test goodness of the model simulations. The season-based model evaluation technique showed superior performance compared with the traditional method of assuming constant model parameters across seasons. The watershed simulation model exhibited reasonable accuracy in simulating streamflow at the internal sites and for the verification periods when parameter values are allowed to vary from dry to wet season. The “optimal” parameter values identified by the calibration attempts showed significant seasonal sensitivity.

This case study presents a risk analysis reassessment for the oldest dam in the Tennessee River basin—the Wilson Dam—based on postdam flow data. The hydrologic risk of old Wilson Dam was computed from historical flow data (spanning pre and postdam periods) and reservoir volume at the dam site. Additional flow data not previously used in the design phase of the dam helped to update more robustly the probability of flood occurrence that exceeded a particular return period during the life of a dam. The generalized extreme value (GEV) distribution was fitted to historical peak flow (at the dam site) and annual maximum reservoir volume using the $L$ -moment method. This reassessment approach has wide application in reservoir water and safety management for aging dams. The study underscores the need for a review of risk analysis for aging dams that have extensive postdam flow data, particularly in the United States. Furthermore, this case study also demonstrates the unique value of the $L$ -moment method in incorporating postdam flow data for more robust risk analysis.

The selection of suitable site characteristics and the number of clusters play an important role for finding homogeneous regions in regional flood frequency analysis. The present study investigates the partition of the Mahanadi basin into homogeneous regions by applying different clustering techniques by using fewer but influential variables. As such, the entire basin is not hydrometeorologically homogeneous. Principal component analysis has been initiated in finding appropriate site characteristics (variables) as per priority. Out of seven variables, four variables are selected on priority. Possible numbers of cluster are found by applying the Kohonen self-organization map and Andrews plot. Other clustering techniques, such as hierarchical clustering fuzzy C-mean (FCM) and $K$ -mean, are applied on prioritized variables to verify the result of clustering. The intercomparison of clustering techniques gives the optimum number of sites to be placed in a particular cluster. The sites clustered as per FCM give more homogenous results. The entire basin is divided into two homogeneous clusters. The regional L-moment algorithm is used to test the homogeneity and to identify an appropriate underlying frequency distribution. An index flood is also stated relating to catchment characteristics by using the multiple linear regression approach. The results are compared with the previous studies of flood frequency on this basin. The study faces the limitation of less data availability to predict longer return period values ( $Q_T$ ).