Journal of Hydrologic Engineering

Digital elevation models (DEM) are important inputs for topography in modeling floods for remote and inaccessible regions. DEMs often lack in accuracy near water bodies and rivers. The objective of this research is to present a DEM correction technique to improve the accuracy of flood simulation and inundation mapping. The key feature of this method is the variability in thalweg (deepest point along a cross section) locations and depth based on the river meandering, width, and side slope. The DEM correction technique is demonstrated by adjusting a national elevation dataset (NED) DEM along the Cumberland River near Nashville in Tennessee. The original (base DEM) and modified DEMs are used as main input of the 1D Hydrologic Engineering Center River Analysis System (HEC-RAS) model and corresponding performances were analyzed. The model using surveyed topography was calibrated for a high flood event (May 2010) and later validated for an intermediate flood event (2003), a high flood event (May 2010), and a low flood event (May 2013) using the modified DEM. It was found that the model with base DEM is capable of simulating at a very high stage but fails during low and intermediate stages. The applicability of base DEM is also limited for any event above 127 m and $\mathrm{3,000}{\mathrm{m}}^3/\mathrm{s}$ with specific biases. The model using modified DEM could be used for simulating large arrays of flow events. The root mean square error (RMSE) for simulated stage using modified DEM for 2003, 2010, and 2013 with the observed stage were 0.86, 0.23 and 0.52 m respectively. Comparison of simulated flood map for the May 2010 flood event using modified and base DEMs with observed flood extent showed errors of 2.66% (overestimate) and 13.38% (overestimate), respectively. The preliminary application of the DEM correction technique thus showed significant improvements in the quality of DEM data with corresponding increase of the HEC-RAS model accuracy.

This paper presents the most frequently used probability-density functions in groundwater hydrology and practical ways to apply them. The paper provides several examples of probability-density functions dealing with (1) their application to various types of groundwater phenomena, (2) the estimation of their parameters by the method of moments, and (3) the implementation of goodness-of-fit tests in probabilistic groundwater hydrology. The versatility of the log-gamma probability-density function to fit highly skewed groundwater data is demonstrated. Important univariate probability-density functions are covered, and the multivariate lognormal probability-density function’s applicability to goodness-of-fit testing and synthetic generation of random fields is elucidated.

Accurately forecasting the weekly seasonal streamflow of the Upper Blue Nile basin (UBNB) in Ethiopia is essential for managing large-scale water projects of Nile basinwide countries. A wavelet-based, artificial neural network calibrated by genetic algorithm (ANN–GA) model and a statistical disaggregation algorithm were integrated to forecast weekly streamflow of the UBNB. The July to October (JASO) streamflow of the El Diem station of UBNB shows strong interannual oscillations prior to the 1920s and after 1990s. Two ANN-GA models were developed to forecast the UBNB JASO streamflow, the first one using the February to May (FMAM) seasonal sea surface temperature (SST) of the global oceans as predictors to directly forecast JASO streamflow, while the second, a hybrid model, is developed to forecast JASO streamflow from two sets of predictors, which consist of FMAM SST and the July to September (JJAS) seasonal rainfall previously forecasted by the wavelet-based, ANN-GA also driven by FMAM SST as predictors. The forecasted JASO streamflow were then disaggregated to weekly total streamflow using the disaggregation model, Valencia and Schaake (VS). Results indicate that seasonal forecasts with up to 4 months lead time only based on SST as predictors achieved reasonable skill ($r^2=0.5$), while the hybrid model achieved a better performance ($r^2=0.8$). The disaggregated streamflow from the first model explained 69% of the streamflow variance, compared to 71% when the hybrid model was used. Based on the results, the proposed hybrid model that uses both SST and a forecasted JJAS seasonal rainfall as predictors achieves a marginally better forecasting skill of the UBNB weekly streamflow. This proposed method that directly forecasts the streamflow, rather than forcing a hydrological model with rainfall forecasts is useful for the management of river basins that lack observed hydrologic data.

The Sacramento River is the largest river in California, and an important source of water for agricultural, municipal, and industrial users. Input to the Sacramento River comes from Shasta Lake and is controlled by operators of Shasta Dam, who are challenged with meeting the competing needs of these users while also maintaining a cold water habitat for Endangered Species Act (ESA) listed winter-run Chinook salmon. The cold water habitat goals are constrained by the volume of cold water storage in the lake, which operators attempt to selectively deploy throughout the critical late summer/fall window. To make informed decisions about the release of this limited cold water resource, skillful forecasts of downstream water temperature attributes at the seasonal time scale are crucial. To this end, we offer a generalized linear modeling (GLM) framework with a local polynomial method for function estimation, to provide predictions of a range of daily water temperature attributes (maximum daily water temperature, daily temperature range, number of hours of threshold exceedance, and probability of threshold exceedance/nonexceedance). These attributes are varied in nature (i.e., discrete, continuous, categorical, etc.), and the GLM provides a general framework to modeling all of them. A suite of predictors that impact water temperatures are considered, including current and prior day flow, water temperature of upstream releases, air temperature, and precipitation. A two-step model selection is proposed. First, an objective method based on Bayesian Information Criteria (BIC) is used in a global model to select the best set of predictors for each attribute; then the parameters of the local polynomial method for the selected best set of predictors are obtained using generalized cross validation (GCV). Daily weather ensembles from stochastic weather generators are coupled to the GLM models to provide ensembles of water temperature attributes and consequently, the probability distributions to obtain risk estimates. We demonstrate the utility of this approach by modeling water temperature attributes for a temperature compliance point on the Sacramento River below Shasta Dam. Regulations on the dam depress the water temperature forecasting skill; to show this, we present skillful results from applying the approach to an unregulated location in the Pacific Northwest. The proposed method is general, can be ported across sites, and can be used in climate change studies.

Hydrologic modeling of karst watersheds requires detailed information on geologic settings and hydrologic properties for efficient simulation of spring and streamflow hydrographs. In this paper, sinkholes located in the streambed of a karst watershed in Kentucky were conceptualized as orifices, and flow through these orifices was modeled as a function of sinkhole diameter. The karst flow components were incorporated in a soil and water assessment tool (SWAT), called KarstSWAT in this study. The KarstSWAT was able to reproduce the observed hydrographs better than the SWAT model in terms of both peak flow and hydrograph volume. Flow through a groundwater basin that is partly fed by the sinkholes and is discharged from a spring was estimated using a method called the successive summation routing algorithm (SSRA) in this study. The time series analysis of simulated spring flows showed that the quantity of water recharging the sinkholes was sufficient to reproduce the observed spring hydrographs. The proposed method could be adapted to karst watersheds dominated by sinkholes and to watersheds that drain water from sinkholes to springs.

Traditionally, levees are a popular measure widely adopted for flood control, accepted and trusted by populations living in floodplain areas. The presence of levees sometimes might even induce a false sense of safety in the population, influencing their decision to develop further in floodplains, because they feel safer. Thus, failures of levee systems are potentially devastating, as they might induce loss of human lives, damages to properties, and economic loss. This study proposes an innovative methodology to estimate the reliability of levee systems, accounting for different sources of uncertainty, and to divide and classify discrete levee reaches, according to different fragility classes. The reliability analysis is performed by evaluating the probability of failure, as a function of a certain failure mechanism, conditioned by a given hydraulic load. Fragility curves are determined using two different methods: Monte Carlo data generation and the so-called approximate, first-order reliability method. Thus, once having assessed fragility curves for each discrete levee reach, different fragility indexes are introduced to assess the reliability of the levee system. As a test case, this methodology is applied to determine the probability of failure due to piping along a branch of the Po River (Italy), where the probability of failure in case of a synthetic 100-year return period flood event is additionally estimated. The computed fragility curves are in agreement with historical observations of levee breaches along the River.

Evapotranspiration has a major role in agricultural and forest meteorology researches, hydrological cycle, irrigation scheduling, and water resources management. There are many methods to estimate the potential evapotranspiration including mass transfer, radiation, temperature, and pan evaporation-based methods. The present study aims to compare radiation-based methods to determine the best method under different weather conditions. The results discussed in this paper are from the data collected in the study area, but the method can be used in other similar regions. For this purpose, weather data was collected from 181 synoptic stations in 31 provinces of Iran. The potential evapotranspiration was estimated using 22 radiation-based methods and compared with the Food and Agriculture Organization of the United Nations (FAO) Penman-Monteith method. The results show that the Stephens method estimates the potential evapotranspiration better than other methods for provinces of Iran. However, the values of $R^2$ varied from 0.93 to 0.98 for 15 provinces of Iran. Therefore, the methods were calibrated and precision of estimation was increased (the values of $R^2$ were less than 0.99 for 4 provinces in the modified methods). The radiation-based methods estimated the potential evapotranspiration in the central provinces of Iran (solar radiation between 24.0 and $25.0\mathrm{MJ}·{\mathrm{m}}^{-2}·{\mathrm{day}}^{-1}$, annual relative humidity less than 50%, and sunshine more than $250\mathrm{h}·{\mathrm{month}}^{-1}$) better than other provinces. The most precise methods were the Berengena-Gavilan method for Esfahan (ES) (before calibration), the Stephens method for Zanjan (ZA) and Lorestan (LO), and the Stephens-Stewart method for Semnan (SE) (after calibration). Finally, a list of the best performances of each method is presented to use in other region studies according to mean, maximum, and minimum temperature, relative humidity, solar radiation, elevation, sunshine, and wind speed. The best temperatures to use radiation-based equations are 10–26°C, 16.5–24.0°C (with the exception of Jones-Ritchie), and 5–13°C for mean, maximum, and minimum temperature, respectively. The results are also useful for selecting the best model when radiation-based models must be applied based on available data.

Human activities are known to increase interference with runoff. Conversely, human activities related to utilizing and managing water resources are primarily determined by annual runoff processes dominated by precipitation distribution. With this view, the soil and water assessment tool was used to quantify human-induced annual runoff changes at different periods and under different patterns of precipitation in seven catchments in Northeast China. The conclusions are as follows. First, although human activities have distinct regional characteristics, an increase in reduced runoff is found for the catchments under investigation; human-induced runoff changes are more significant in the catchment where water resources are limited. Second, the interannual runoff distribution is significantly disturbed in the catchment with large reservoirs. Third, human-induced runoff changes are similar under all patterns of precipitation in the catchment where annual precipitation is less than 500 mm and intense human activities play a dominant role in runoff. Fourth, in general catchments, runoff changes more significantly during relatively drier years or years with uneven precipitation distribution. Finally, human-induced runoff change is related to both annual precipitation characteristics and operations of the reservoirs for catchments in which reservoirs played a significant role.

A permeable pavement system can capture stormwater to reduce runoff volume and flow rate, improve onsite groundwater recharge, and enhance pollutant controls within the site. A new unit process model for evaluating the hydrologic processes of a permeable pavement system has been developed in this study. The developed model can continuously simulate infiltration through the permeable pavement surface, exfiltration from the storage to the surrounding in situ soils, and clogging impacts on infiltration/exfiltration capacity at the pavement surface and the bottom of the subsurface storage unit. The exfiltration modeling component simulates vertical and horizontal exfiltration independently based on Darcy’s formula with elaborating Green-Ampt approximation. The developed model can be arranged with physically-based modeling parameters, such as hydraulic conductivity, Manning’s friction flow parameters, saturated and field capacity volumetric water contents, porosity, and density. The developed model was calibrated using high-frequency observed data. The modeled water depths are well matched with the observed values ($R^2=0.89$; $\mathrm{Nash}\text{-}\mathrm{Sutcliffe}\mathrm{coeff.}=0.90$). The modeling results show that horizontal exfiltration through the side walls of the subsurface storage unit is a prevailing factor in determining the hydrologic performance of the system, especially where the storage unit is developed in a long, narrow shape; or with a high risk of bottom compaction and clogging.

A one-dimensional (1D) hydrologic-hydraulic model for simulating dual drainage in urban areas is presented. It consists of four modules: (1) rainfall-runoff transformation, (2) 1D flow routing on a street network, (3) flow interception at street inlets, and (4) flow interaction between surface water on the streets and the underground storm-water system by interfacing with the EPA-SWMM5 engine (U.S. Environmental Protection Agency-Storm Water Management Model). The hydrologic model (first module) transforms rainfall to runoff using the kinematic wave approximation and simulates the infiltration process with the Green-Ampt method. The street network model (second module) is based on a finite-volume shock-capturing scheme that solves the fully conservative Saint-Venant equations and can be used to model both subcritical and supercritical flows. The inlet model (third module) computes the amount of water intercepted by inlets. The formulation of boundary conditions at the street crossings is generalized and can be used for any number of streets, any combination of inflowing and outflowing streets, and flow regime (e.g., subcritical and supercritical flows). Flow interaction between surface water on the streets and underground storm-water system is achieved by interfacing the proposed model with EPA-SWMM5. This interaction allows flow to enter from streets to the underground storm-water system and vice versa. The proposed model has several potential applications such as the identification of critical zones for flooding (e.g., zones with high water depths and flow velocities) in urban developments and can be used to take appropriate measures for drainage control (e.g., to increase number and/or size of inlets), to determine the consequences of different degrees of inlet clogging, and to assess flooding hazards through the application of suitable hazard criteria. A summary of criteria used for storm-water hazard assessment is presented. To demonstrate the dual-drainage model’s potential, an application is performed in a catchment of the metropolitan area of Chicago, Illinois. The results obtained are promising and show that the model can be a useful tool for storm-water management and flooding hazard assessment in urban areas.

When applying hydrological models, different sources of uncertainty are present, and evaluations of model performances should take these into account to assess model outcomes correctly. Furthermore, uncertainty in the discharge observations complicates the model identification, both in terms of model structure and parameterization. In this paper, the authors compare two different lumped model structures (PDM and NAM) considering uncertainty coming from the rating curve. Limits of acceptability for the model simulations were determined based on derived uncertainty bounds of the discharge observations. The authors applied the DYNamic Identifiability Approach (DYNIA) to identify structural failure of both models and to evaluate the configuration of their structures. In general, similar model performances are observed. However, the model structures tend to behave differently in the course of time, as revealed by the DYNIA approach. Based on the analyses performed, the probability based soil storage representation of the PDM model outperforms the NAM structure. The incorporation of the observation error did not prevent the DYNIA analysis to identify potential model structural deficiencies that are limiting the representation of the seasonal variation, primarily indicated by shifting regions of parameter identifiability. As such, the proposed approach is able to indicate where deficiencies are found and model improvement is needed.

Theoretical considerations and empirical evidence indicate a linear relationship between the land surface temperature ($T_s$) and the corresponding evapotranspiration (ET) rate under spatially constant wind and net energy conditions at a homogeneous vegetated surface. Such a relationship lies at the core of the popular surface energy balance algorithm (SEBAL); the satellite-based energy balance approach for mapping evapotranspiration with internalized calibration (METRIC); and the lesser known calibration-free evapotranspiration mapping (CREMAP) method, just to name a few. The present findings are based on analytical solutions of the coupled turbulent heat and vapor transport equations and further corroborated by monthly reanalysis data of $T_s$, ET, and sensible heat transfer rates over extensive areas in North America and Europe, where the CREMAP method has previously been applied.

Base flow recession analysis is required to estimate the long-term, reliable component of the hydrograph and water cycle, and for drought management, inflow design and analysis, and contaminant and nutrient transport. Operationally, recursive digital filtering (RDF) techniques are commonly applied; however, questions have been raised about the reliability and parameterization of these methods. A comparison of three base flow separation methods, including two popular RDFs and one hydrologic model, is performed in two base flow-dominant subbasins of the Grand River Basin in southern Ontario, Canada. The Eramosa River and Whiteman’s Creek subbasins are similar in size (236 and $383{\mathrm{km}}^2$, respectively) and have similar annual runoff distributions yet significantly different physiography governing runoff generation. Subbasin physiographic characteristics are found to result in significant differences in annual base flow distributions, which are well captured by the hydrologic model but not by the RDFs. Stable water isotopes (SWIs) are applied for verification of base flow separations using two-component mixing model separations and show better visual agreement with a base flow derived from the hydrologic model over the RDF methods. Limitations include insufficient isotope data to quantify a statistically significant model fit and results that are tied to subbasin physiography and basin-specific processes. Results support the application of SWIs for regional hydrograph separation and highlight the need for more efficient, yet physically based base flow separation in regions with complex physiography.

This study addresses the sensitivity of short-term flow forecasting in the Seine River basin ($\mathrm{43,800}{\mathrm{km}}^2$, France) to the spatial distribution using a semidistributed model [Transfer with the Génie Rural model (TGR)]. The basin was decomposed into intermediate basins depending on the gauging stations selected for this study. A lumped hydrological model was applied on each intermediate basin and a routing model was used to propagate the discharge through the river network. Discharge data at the gauging stations were assimilated using a Kalman filter and tests for flow forecasting were performed with a lead time of up to 72 h. Several spatial configurations, defined by a selection of one or several gauging stations, were tested and the performances were compared with a reference lumped model currently used operationally by the regional flood forecasting center. Results showed that the forecasting performance improves with an increase in the degree of spatialization. Nevertheless, this improvement was not systematic and the integration of some particular gauging stations degraded the model performance. In addition, it was shown that integrating some other stations (generally the most upstream) led to a negligible improvement. This suggests that in an operational context, where the model has to be robust and computationally efficient, some efforts should focus on finding the optimal spatial distribution, which is not necessarily the one using all the available stations.

Spatially variable channel geometry in natural rivers produces nonuniform flow and spatial gradients in the shear stress field. The travel distance required for the flow to acquire the capacity bedload concentration and attain a new equilibrium bedload transport rate upon encountering a region of higher or lower shear stress is defined as the bedload adaptation length ($L_b$). Estimates of $L_b$ are used by some numerical morphodynamic models to account for nonequilibrium bedload transport in the computation of local transport rates. However, current methods for estimating this parameter are uncertain and often crude. The authors therefore conducted experiments designed to measure $L_b$ for a uniform sediment mixture in a laboratory flume. Instantaneous bedload transport rates were determined by counting passing sediment particles on digital imagery collected at variable distances downstream from a zero-transport boundary in a small flume. The flume was operated at three bed slopes in order to assess $L_b$ over a range of hydraulic conditions. Bedload adaptation length was found to be about $30\pm 8$ particle diameters at a relatively low excess dimensionless shear stress ($\theta -{\theta }_c=0.018$, where $\theta$ is the dimensionless shear stress and ${\theta }_c=0.0436$ is the critical dimensionless shear stress) and about $100\pm 30$ particle diameters at a moderate level of excess dimensionless shear stress ($\theta -{\theta }_c=0.032$). The experiments failed to resolve $L_b$ at higher shear stresses. These results support physically based models that cast $L_b$ as an increasing function of excess shear stress. They also suggest that $L_b$ may be small relative to the resolution of the numerical mesh used in many modeling applications. In such cases, model performance may be insensitive to the choice of any arbitrary small value of $L_b$.