Journal of Hydrologic Engineering

A simulation of the transport of dibromochloropropane (DBCP) and ethylene dibromide (EDB) in the vadose zone and ground water was conducted in the Mililani area of central Oahu, Hawaii to revisit the ground water contamination predictions made by researchers at the University of Hawaii in 1988 for selected wells of the Honolulu Board of Water Supply. Examination of sampled DBCP results in these wells indicated a large discrepancy between the observed and predicted concentrations in 1988. Therefore, a postaudit study was initiated in the same area to assist in better defining projected concentrations of contaminants in well waters in the future. With the new data, different modeling approaches were undertaken for simulating flow and contaminant transport in both the vadose and saturated zones. The new simulations indicate travel times of $14–32\text{years}$ for contaminants to reach the water table, depending on the location. The approximate recovery time of wells in the study area is expected to occur after the year 2016 for DBCP. The predicted concentration of EDB in the wells is less than the current detection limit. Variabilities in model predictions due to parameter uncertainties and model limitations are also considered. Although the predicted concentrations from the present simulations match sampled data better than the concentrations from earlier modeling efforts, still a number of uncertainties remain.

This study deals with the estimation of sediment yields and sedimentographs of small watersheds. The proposed mathematical model is based on simultaneous solution of flow dynamics followed by dynamic erosions. The flow dynamics is based on the well accepted kinematic wave theory. As the storm rainfall proceeds, the process of generation of overland depth is a dependent function of the infiltration rates. This has been taken care of by the use of Phillip’s infiltration approach. Recent research has revealed that the flow resistance coefficients greatly influence the overland water depths. Therefore, the concept of depth dependent overland resistance coefficients has been incorporated in the overland flow computations to minimize the errors. The components of the soil erosion model have been modified for better prediction of sediment flow rates and sediment yields. The rainfall-runoff-sediment model as developed was applied to 34 storm events registered over six small watersheds located in different climatic regions of India and the United States. The model has satisfactorily reproduced the runoff hydrographs and the sedimentographs.

The aquifer system, which is the resource of water for the city of Milano (Italy), is a multilayered aquifer, characterized by sandy and sandy-gravel units, connected by discontinuous aquitards. The recharge area of the phreatic aquifer is close to the prealpine area, some tens of kilometers north of the city; nevertheless the local recharge given by rain infiltration through the soil is not negligible and contributes to the mass balance of the aquifer. A field campaign was carried out for $16\text{months}$ in the suburban area of the Lambro Park to evaluate the local recharge due to the rainfall infiltration through the unsaturated zone; the water table in this area lies about $15\mathrm{m}$ below the ground surface. Standard meteorological data (atmospheric pressure, rainfall, humidity, wind velocity and air temperature) were collected, along with incident and net radiation, to evaluate potential and actual evaporation from the bare soil using three methods (Bowen ratio, Penman, and Hamon equations). Simultaneously, the volumetric soil water content at different depths (down to $76\mathrm{cm}$ ) was measured with time domain reflectrometry probes. A finite-difference model, which solves the one-dimensional Richards equation in transient conditions, was developed to simulate the flow through the unsaturated zone, to evaluate the characteristic time of recharge and the mass balance. The water flow rate through the ground surface was assigned as the upper boundary condition and was evaluated from the mass and energy balance at the atmosphere–soil interface with different approaches based on the meteorological data and the actual soil water content. The lower boundary condition is given by the saturation condition at the water table. Characteristic retention curves were estimated in laboratory using Richards apparatus. In order to calibrate the model, the numerical results were compared with experimental data for two different periods: A dry period (July 2001) and a wet period (October 2001). Finally, the calibrated model was used to simulate the infiltration and the flow through the unsaturated zone for a $15\text{-}\text{year}$ -long period (1988–2003) and to estimate the total phreatic aquifer recharge. Moreover, the time lag between a variation of the infiltration/exfiltration rate and the corresponding variation of the recharge rate was evaluated; the values of the delay time of recharge are longer than those computed with models that approximate the transmission zone as a unique cell.

To facilitate estimation of streamflow characteristics at an ungauged site, hydrologists often define a region of influence containing gauged sites hydrologically similar to the estimation site. This region can be defined either in geographic space or in the space of the variables that are used to predict streamflow (predictor variables). These approaches are complementary, and a combination of the two may be superior to either. Here we propose a hybrid region-of-influence (HRoI) regression method that combines the two approaches. The new method was applied with streamflow records from 1,091 gauges in the southeastern United States to estimate the $50\text{-}\text{year}$ peak flow $\left(Q_{50}\right)$ . The HRoI approach yielded lower root-mean-square estimation errors and produced fewer extreme errors than either the predictor-variable or geographic region-of-influence approaches. It is concluded, for $Q_{50}$ in the study region, that similarity with respect to the basin characteristics considered (area, slope, and annual precipitation) is important, but incomplete, and that the consideration of geographic proximity of stations provides a useful surrogate for characteristics that are not included in the analysis.

In this paper we present algorithms based on the acceptance-rejection (A-R), and Markov chain Monte Carlo methods to draw samples from Halphen distributions. Quantiles computed with these generating techniques are compared to those given by the method of importance sampling. Results show that our choice of the instrumental distribution considered in the A-R, produce an efficient method to generate samples from Halphen distributions. The availability of such procedures makes it possible to approximate all integral quantities and to study statistical properties of parameters estimators in the case of small sample sizes usually encountered in hydrology.

Based on recent reviews of 11 physically based watershed models, the long-term continuous model soil and water assessment tool (SWAT) and the storm event dynamic watershed simulation model (DWSM) were selected to examine their hydrologic formulations, calibrate, and validate them on the $620{\mathrm{km}}^2$ watershed of the upper Little Wabash River at Effingham, Ill., and examine their compatibility and benefits of combining them into a more comprehensive and efficient model. Calibration and validation of the SWAT by comparing monthly simulated and observed flows and adjusting the model-assigned resulted in coefficients of determination and Nash–Sutcliffe coefficients for individual years and cumulatively for the calibration period (1995–1999) and for the entire simulation period (1995–2002) mostly above or near 0.50 with an exception of 0.05 and $-0.27$ , respectively, in 2001, relatively a dry year. Visual comparisons of the hydrographs showed SWAT’s weakness in predicting monthly peak flows (mostly underpredictions.) Therefore, SWAT needs enhancements in storm event simulations for improving its high and peak flow predictions. Calibration of DWSM was not necessary; its three physically based parameters were taken from SWAT. Validation of DWSM on three intense storms in May 1995, March 1995, and May 2002 resulted 1, $-29$ , and 16% errors in peak flows and 0, $-11$ , and 0% errors in times to peak flows, respectively. Comparisons of DWSM’s 15-min flow hydrographs with SWAT’s daily flow hydrographs along with the 15-min and daily observed flow hydrographs during the above three storms confirmed that DWSM predicted more accurate high and peak flows and precise arrival times than SWAT. DWSM’s robust routing scheme using analytical and approximate shock-fitting solutions of the kinematic wave equations was responsible for the better predictions, the addition of which along with its unique combination with the popular runoff curve number method for rainfall excess computation to SWAT would be a significant enhancement. Parameters and data of both the models are interchangeable and, therefore, are compatible and their combination will result in a more comprehensive and efficient model.

If two independent sets of data are available—the piezometric heads and the related source terms for two stationary flow situations, such that the hydraulic gradients are not parallel to each other—the aquifer transmissivity can be identified with the differential system method (DSM), provided that an initial value of transmissivity is given at least at one point. In field applications, it is difficult to collect two data sets that are independent throughout the aquifer, and the results depend on the location of the point where the initial value is assigned. These difficulties are reduced if several data sets are available and the differential system is obtained either by applying a least square technique to the whole ensemble of data sets, or by choosing the “best” couple of data at each point. Numerical tests on a synthetic aquifer show that both techniques yield good results and make the DSM more robust to noise on the piezometric heads than the standard DSM applied with two data sets only.

Treating the occurrence and severity of droughts as random, a hybrid model, combining a linear stochastic model and a nonlinear artificial neural network (ANN) model, is developed for drought forecasting. The hybrid model combines the advantages of both stochastic and ANN models. Using the Standardized Precipitation Index series, the hybrid model as well as the individual stochastic and ANN models were applied to forecast droughts in the Kansabati River basin in India, and their performances were compared. The hybrid model was found to forecast droughts with greater accuracy.

The paper describes the results of a study based on the integration of remote sensing and geographical information system techniques to evaluate a distributed unit hydrograph model linked to an excess rainfall model for estimating the streamflow response at the outlet of a watershed. Travel time computation, based on the definition of a distributed unit hydrograph, has been performed, implementing a procedure using (1) a cell-to-cell flow path through the landscape determined from a digital elevation model (DEM); and (2) roughness parameters obtained from remote sensing data. This procedure allows the taking into account of the differences, in terms of velocity, between the hillslopes and the stream system. The proposed procedure has been applied to two watersheds in Sicily, in order to establish the level of agreement between the estimated and recorded hydrographs, using as a tool to calculate the excess rainfall a simplified version of the probability distributed model.

This study aims at deriving intensity-duration-frequency (IDF) curves from bivariate rainfall frequency analysis using the copula method, which does not assume the rainfall variables to be independent or normal or have the same type of marginal distributions. The bivariate distributions are then employed to determine conditional return periods and IDF curves needed for urban drainage design. The IDF curves are tested using rainfall data from catchments in Louisiana and are also compared with those of Technical Paper 40 (TP-40) of the U.S. National Weather Service as well as those published in the literature.

Differentiation of upper and lower regions of the unsaturated (vadose) zone is important for modeling evapotranspiration, infiltration, and water table recharge. This study developed a methodology to describe relative moisture conditions in both zones for modeling soil hydrologic response. A comprehensive distributed parameter model was developed based on deterministic and physically based soil and vegetative properties. The Integrated Hydrologic Model (IHM) uses a unique relative soil moisture approach for land segment integration and is intended to simulate the complex interaction between surface-water and groundwater systems. No prior rigorous investigation or validation of the upper and lower soil response has been made. In this study, field soil moisture observations and soil characterization data were used to formulate a new basis for the upper and lower zones in the IHM. Several tests were performed to illustrate how the model reduces field-scale variability in soil moisture behavior and enhances representation of antecedent conditions. Results show that the two zones commonly exhibit different relative moisture conditions. Evidence is presented to document the existence of prolonged (many days) air entrapment and excess pore pressure, which affect soil water storage and observed water table levels.

This technical note presents a combination of isohyetal map, kinematic wave, and rational methods for estimating flash flood synthetic hydrographs in arid regions. They are combined in a logical sequence for the calculations of time of concentration and synthetic flood hydrograph. This empirical approach relates flood peaks to the average temporal and areal intensities of storm rainfalls. The application of the methodology is presented for two arid region wadis in the central western part of the Arabian Peninsula. Hydrograph peaks yield practically reliable estimations of the observed flood magnitudes from individual storms.

This note simplifies the widely used parameter estimator methods, namely, method of moments (MOM) and probability weighted moment (PWM) for the three-parameter generalized extreme value (GEV) distribution for improved parameter estimation. The important shape parameter $\left(\kappa \right)$ of the GEV distribution used in annual maximum series and partial duration series models is expressed as a function of skewness $\left(\gamma \right)$ . Validity of the approximate PWM estimates of shape $\left(\kappa \right)$ and, in turn, scale $\left(\alpha \right)$ and location $\left(\xi \right)$ parameters is checked using the ratio of variance and bias with simulated data, whereas the MOM-approximated estimates are tested using standard error of quantile estimates. The resulting expressions exhibit better accuracy when applied to field data of five watersheds.

This study develops a quantitative relation defining the extent of influence that a dam imposes on annual peak streamflow, and employs the developed relation to retain the use of some USGS gauges in the state of Maryland for regional flood frequency analysis (FFA). Stage-storage–discharge curves for 34 dams in Maryland were provided by the Maryland Department of Environment Division of Dam Safety. Each watershed/dam system was analyzed using HEC-1. The results were used to form two linear equations that calculate the influence distance, or the downstream reach length affected by each dam, for a 5 and 10% level of regulation. The level of regulation is the measured amount, in percent, that the peak flow in a stream is influenced by the presence of the structure. When the influence distance predictor equations were applied throughout the state of Maryland at the 10% regulation level, we found that 32 of the $91\text{gauges}$ that lie downstream of dams could be retained for use in regional FFA.