Journal of Hydrologic Engineering

Hydraulic structures in rivers employ abutments of different shapes. An abutment causes a scour hole around the bottom of the structure in the river. The scour hole grows no further when the turbulence fluctuations and bed shear stress are unable to move sediment particles. Although there have been a multitude of studies on the rate of scour depth around abutments of different shapes, no research work seems to have been reported regarding the flow structure of local scour around semielliptical bridge abutments. This study experimentally investigated the flow field around semielliptical bridge abutments and compared the results with those around wing-wall and semicircular abutments. Results showed that the abutments with small dimensions (width and length) produced higher scour depth due to strong vortices and downflow than those with large dimensions. For the same aspect ratio, small and large abutments cause different patterns of the velocity field in streamwise and vertical directions. The abutment size (width and length) significantly influences flow separation and reattachment, manifesting larger separation and deeper scour hole in small size abutments. The equilibrium scour depth depends on the abutment size, with stronger circulation and vorticity in small abutments than in large ones. The internal boundary layer around small abutments plays a more significant role in the flow structure than that around large ones, generating a larger strong vortex trail with more scour in the vicinity of abutment and causing the failure of the structure. Unfavorable pressure gradients with nonlinear distributions of turbulence intensities and Reynolds stress at the scour hole upstream of the abutment are the primary factors causing the development of vortex trail and the scouring mechanism.

This study proposes a model using wavelet neural networks (WNNs) trained by a novel improved harmony search (IHS) algorithm to forecast daily groundwater level (GWL) in a shallow well and a deep well in Florida and Arkansas, respectively, for 1 year. Statistical characteristics of the GWL time series and the autocorrelation functions were determined first. Measured GWL series were then decomposed into several subseries using wavelet transform, and imposed as input patterns to the proposed model to forecast GWL in both wells. Efficiency of the IHS algorithm was affirmed by comparing its results with those of differential evolution (DE), harmony search (HS), and particle swarm optimization (PSO) training algorithms. Similarly, the efficiency of WNN was verified by comparing its results with those of the radial basis function (RBF) and multilayer perceptron (MLP) networks, all trained by the IHS algorithm. Predictive capability of the models was determined by Nash-Sutcliffe efficiency (NSE), Pearson correlation coefficients (PCC), normalized root-mean-squared error (nRMSE), and normalized mean absolute error (nMAE) indices. Results reflect that the wells have different statistical characteristics in their GWL series. The shallow well, in contrast to the deep one, has both homoscedasticity and stationarity properties in its observed noisy GWL fluctuations. The proposed model performed much better in both wells, with lower errors and higher PCC and NSE values compared to models with alternative training algorithms or network structures. All forecasts were more accurate for the deep well, as opposed to the shallow one, probably because of the highly noisy GWL fluctuations in the latter.

Acquiring discharge estimates is crucial in hydrological studies, extreme event analyses, and water resources management. This study demonstrates the improved fluvial acoustic tomography system (FATS), a promising approach for automatically measuring streamflows with high temporal resolution for a long period. Unlike past hydroacoustic systems, the FATS enables continuous capture of the depth- and range-averaged water velocity with a single transducer set. Streamflow measurements were performed on November 26, 2015, in a mountainous river (depth: 0.8 m under low-flow conditions, width: 115 m). This paper presents a 127-day record of the flow. The average stream velocities and streamflows are automatically estimated from the acoustic data and water level data. The real-time flow data can be monitored by the Internet access. The continuous FATS estimates indicate that the streamflow does not change smoothly with the river stage. Analyses reveal that the streamflow may not be estimated accurately from a smooth rating curve (RC). The relative biases of RC reach $\sim \pm 20\%$ even under higher flow conditions. The FATS captures short-wave irregularities in the streamflow time-series.

This paper establishes a novel approach to estimate monthly and annual direct runoff by combining the curve number method of the Natural Resources Conservation Service with an exponential distribution of rainfall depths. The approach was tested against observed rainfall and runoff for 544 watersheds throughout the contiguous United States. For more than half of the watersheds, the performance of the new approach is indistinguishable from the application of the method to daily rainfall when curve numbers are determined via calibration. For all watersheds, the uncertainty introduced by the approximation of the distribution of rainfall depths is far less than the uncertainty associated with the use of tabulated curve numbers based on soil and land-cover characteristics. The new approach does not appreciably increase the overall uncertainty associated with the application of the curve number method in ungauged watersheds. The approach provides reasonable estimates of monthly and annual direct runoff that can inform land-management decisions when daily rainfall records are unavailable.

Vegetative filter strips (VFS) are an effective approach to reducing edge-of-field discharge of sediment and pesticide. The web-based Vegetative Filter Strip MODel (VFSMOD) system, developed to assess the trapping efficiency of VFS, has a limited capability in regards to considering the effects of shallow groundwater table, which significantly affects estimation of infiltration and runoff in the VFS. In addition, the system does not simulate the transport of pesticides that can be implemented in agricultural areas. Thus, the web-based VFSMOD system was enhanced in this study by incorporating the groundwater table in simulating soil moisture dynamics and by adding a new water quality module for evaluating the VFS performances effectively. Results show that the trapping efficiencies of sediment (61%) and a pesticide (57%) were simulated with the VFS width of 10 m for a single storm event, improving the VFSMOD performance with a shallow groundwater table. The optimum VFS widths for multiple storm events were found for a given reduction goal.

In a groundwater outburst through a fault, the velocity is generally so high that the relationship between the velocity and the hydraulic gradient deviates from the linear Darcy regime. This study uses the Forchheimer equation to model the flow in both the aquifer and the fault zone. The model is used to investigate the effects of non-Darcy flow on a specific inrush episode at the Zhongguan iron mine by varying the fault permeability. It shows that the flow behavior in the aquifer was not always Darcy flow or non-Darcy flow, but changed depending on the fault permeability. The region in the aquifer near the fault is most likely affected by non-Darcy flow, and the non-Darcy-effect region expands away from the fault with the increase of the fault permeability. Considering a 10% non-Darcy effect for engineering, the Forchheimer equation can be approximately used to describe the flow mechanism from Darcy to non-Darcy. Furthermore, the Forchheimer equation is also acceptable for predicting the groundwater inflow during groundwater outburst through fault.

The regional climate models (RCMs) used in the analysis of the impact of climate variables on the hydrology of river basins needs appropriate preprocessing (bias correction) to represent and reproduce future climate with a fair degree of accuracy. The performance of bias corrections methods was assessed in this investigation on the basis of their ability to minimize error on climate variables and streamflow. This work compares the performance of five bias correction methods applied for precipitation and four methods for temperature in modeling the hydrology of the river catchments of the Western Ghats of India. The Western Ghats are a mountainous forest range along the entire west coast of India that plays a major role in the distribution of Indian monsoon rains. Simulations were used to evaluate the performance of the bias correction methods. Using raw RCM, bias corrected precipitation and temperature time series, streamflows were estimated by the soil and water assessment tool (SWAT) hydrological model. The results indicated that the raw RCM-simulated precipitation was biased by 42% and the temperature was biased by 12% across the catchments investigated. Subsequently, a bias of 65% was found in the streamflow. The performance of the delta change correction method was consistently better for precipitation (with Nash-Sutcliffe efficiency, $\mathrm{NSE}>0.75$ for 5 catchments) and temperature ($\mathrm{NSE}=1$) compared with other methods. Good performance was observed between the observed and bias corrected streamflow (daily time scale) for the catchments Purna ($\mathrm{NSE}=0.97$), Ulhas ($\mathrm{NSE}=0.64$), Aghanashini ($\mathrm{NSE}=0.82$), Netravathi ($\mathrm{NSE}=0.89$), and Chaliyar ($\mathrm{NSE}=0.90$); low performance with an NSE of 0.3 was observed for the catchments Kajvi and Vamanapuram. The methods failed for Malaprabha and Tunga catchments. The results indicate that the delta change correction method performed best in analyzing the hydrological impact of climate variables on the windward side of Western Ghats of India.

By coupling the standardized precipitation index (SPI), frequently used as a meteorological drought index, with the Tennant concept, popularly used in environmental flow (EF) assessment, this paper attempts to predict the EF condition of a catchment using easily available rainfall data only, which is useful for ungauged catchments. Its application is demonstrated using the data of three catchments of the Mahanadi Basin (Ghatora, Kurubhata, and Salebhata), two catchments of the Brahmani-Baitarini Basin (Anandpur and Jaraikela), two catchments of the Godavari Basin (Hivra and Nandgaon), and four catchments of the Narmada Basin (Mohegaon, Hridaynagar, Manot, and Sher), which fall in subtropical and subhumid climatic regions of India. A coefficient of determination of more than 0.7 for most of the catchments showed the existence of a relationship between SPI and % average annual flow (AAF) (used to describe the EF condition). Therefore, the EF condition of the previously named catchments can be ascertained using SPI derived from easily available rainfall data.

Vietnam is one of the countries severely impacted by climate change. The Saigon-Dongnai River basin is one of the largest river basins and the economic center in the south of Vietnam. Trian watershed is located in the upper Saigon-Dongnai River basin and is one of the largest subbasins of this river. However, no study has been conducted to assess the impact of climate change on this region. It is, hence, necessary to evaluate the potential future impact of climate change on this watershed, particularly on flood frequency, because flood events negatively impact economic and social aspects. The downscaled atmospheric data, from the simulations of five Regional Climate Models (RCMs), are used as input data for a physically-based hydrological model to simulate future streamflow data. The changes in the frequency of flood peak extracted by the peak over threshold (POT) approach is compared between historical and future time periods. The results indicate that there is a significant increase in flood magnitude under climate change for the Trian catchment. Specifically, the 100-year return level of Trian reservoir is increasing up to 32.34% in one of the future scenarios. Moreover, results of this study also indicate that directly using the asymptotic distribution to model the POT data set sometimes provides wrong insights.

Impervious surfaces have long been recognized as one of the critical factors influencing the rainfall-runoff relationship in urban areas. Published guidance on how to treat urban imperviousness in hydrologic simulations varies, and few studies exist for the southwestern United States. In recent years, small-storm hydrology has become increasingly important to support water quality evaluations. Methods developed for large, infrequent storms are often applied to simulate runoff from smaller rain events, with questionable results. To test the impact of imperviousness on small-storm runoff, impervious cover in a $1.5\text{-}{\mathrm{km}}^2$ urban basin in central New Mexico was mapped and differentiated into directly connected impervious areas (DCIAs) and unconnected impervious areas (UIAs) that are separated from the drainage system by a pervious buffer. Measured runoff from 25 small-magnitude storms was compared to model scenarios with varying treatment of imperviousness. Simulation results revealed that during small rain events, all runoff originated from DCIAs, while UIAs contributed no flow. For this particular basin with 28% DCIAs and 16% UIAs, losses from UIAs approached a constant value of approximately 4 mm. Findings were applied to a $142\text{-}{\mathrm{km}}^2$ watershed to assess model performance for larger storms. The impact of losses from UIAs on the runoff response decreased with increasing storm magnitude. This study presents a modeling approach for DCIAs and UIAs that increased accuracy for small runoff events in urban areas without negatively affecting model performance for larger storms. Use of this method can improve model predictions, particularly when hydrologic simulations built with flood control in mind are used to assess the impact of more frequent storm events.

The U.S. EPA Storm Water Management Model (SWMM) was used to simulate upland runoff production and the subsequent performance of a downstream, ponded infiltration basin installed adjacent to a highway in suburban Maryland. The SWMM’s performance was evaluated with a unique, rich suite of in situ flow and water quality observations. The availability of these in situ observations creates a novel opportunity to explore the performance of SWMM across small scales in space and time. In order to systematically explore the ability of SWMM to leverage these observations, an automatic Monte Carlo–based calibration framework was developed and a multiparameter sensitivity analysis was conducted. As expected, the calibrated model showed better skill in terms of reproducing water quantity observations relative to water quality observations. An uncertainty analysis showed model predictions (flow and water quality) were consistent in the sense that the model was able to encapsulate the observations between 5 and 95% confidence intervals. Example code for use by other researchers to employ the techniques discussed in this paper is made publicly available with this manuscript.

Streamflow information is collected worldwide and used in the areas of engineering design, water supply planning, and hydrologic and water quality analysis, among others. Available data from intermittent and ephemeral streams in arid regions are sparse because streamflow is highly variable and infrequent. Due to a lower perceived priority to monitor intermittent and ephemeral streams, it is more difficult to justify the economic expense of traditional monitoring equipment. This study demonstrates that time-lapse photography can provide a viable, low-cost, low-tech option for measuring stage (water depth) at stream gauging stations. A time-lapse camera was set up to record images of a channel and staff gauge in an ephemeral urban catchment in central New Mexico. During flow events, stage can be read from the time-lapse images and converted to a discharge time series. In the course of a 2-year test period, 33 runoff events were recorded. With approximately $200 in materials, this method costs only a fraction of conventional gauging stations.