Journal of Hydrologic Engineering

Watershed models are widely used in total maximum daily load (TMDL) studies to predict the impacts of pollutant discharges on the biochemical functioning and assimilative capacity of water bodies. The reliability of a TMDL is therefore tightly linked with the predictive capability of these models. While there has been an increasing availability and application of watershed models for TMDL studies, guidelines for model evaluation, including recommendations for an appropriate selection and implementation of calibration, validation, and uncertainty analysis strategies, remain at present limited. The ASCE Environmental and Water Resources Institute (EWRI) TMDL Analysis and Modeling Task Committee was established in part to identify existing and emerging challenges encountered by water resources professionals during any phase of a TMDL development and to produce documentation to address these challenges. This paper reviews existing approaches for model calibration, validation, and uncertainty analysis, including recommendations to establish baseline modeling practices to obtain a satisfactory watershed model.

Fissure seepage is the main factor that causes mine water disasters. Physical models of the fractured rock mass with various widths of fracture were designed to perform the seepage test, which demonstrated the seepage characteristics of the fractured rock mass. Furthermore, grouting tests were performed, with the network parallel electrical test conducted simultaneously, during the water and slurry seepage, respectively. The relationship between the seepage velocity and the hydraulic gradient during the seepage process was fitted to a quadratic function, in which the seepage flow resistance of the fractured water was a key factor affecting the seepage flow. The percolation velocity gradually decreased as the width of the percolation fissures increased under the relatively stable seepage pressure difference. The seepage velocity of the percolation medium was quantified by the change of the natural electric field along the path of the seepage. The change extent of the exciting current and the variation of apparent resistivity determined the location of the percolated medium and the filling effect on the space of the seepage channel.

Estimations of low-probability flood events are frequently used to plan infrastructure and to determine the dimensions of flood protection measures. Several well-established methods exist for estimating low-probability floods. However, a global assessment of the consistency of these methods is difficult to achieve because the “true value” of an extreme flood is not observable. A detailed comparison performed on a given case study brings useful information about the statistical and hydrological processes involved in different methods. In the present study, the following three methods of estimating low-probability floods are compared: a purely statistical method (ordinary extreme value statistics), a statistical method based on stochastic rainfall-runoff simulation (SCHADEX method), and a deterministic method (physically based estimation of the probable maximum flood, PMF). These methods are tested for two different Swiss catchments; the results show that the 10,000-year return level flood estimations exceed the PMF estimations by 3% and 18%. The analysis shows that the plausibility of an extreme flood estimation does not only depend on the applied method but also on its ability to represent flood-triggering processes, including precipitation input, spatio-temporal precipitation distribution, and runoff.

Extreme precipitation has posed a huge risk to society and the environment. It is crucial to be able to accurately analyze extreme precipitation in order to reduce its potential risk. This paper presents a validation assessment, exploration, and improvement framework and systematically studies three state-of-the-art downscaling methods applied to six different global climate model (GCM) results for extreme precipitation projection. For the purposes of illustration, the paper applies this framework to data collected from the Washington, DC, metropolitan area from 1950 to 1995. The assessment shows that existing downscaled GCMs do not adequately predict some extreme precipitation indices based on historical records, such as the annual maximum 2-day precipitation and number of days with precipitation more than 20 mm. To explore possible ways of improving the accuracy, marginal distribution and day-to-day serial dependency of extreme precipitation are studied for the downscaled GCMs and observed precipitation. The projection results are further improved by incorporating serial dependency from observed precipitation into downscaled GCM results by means of copulas. In conclusion, the proposed method provides a generic way to further improve downscaled GCMs for extreme precipitation. The analytical results validate the method and show that significant improvement can be achieved from this method.

Precipitation-based intensity–duration–frequency (PREC-IDF) curves are a standard tool used to derive design floods for hydraulic infrastructure worldwide. In snow-dominated regions where a large percentage of flood events are caused by snowmelt and rain-on-snow events, the PREC-IDF design approach can lead to substantial underestimation or overestimation of design floods and associated infrastructure. In this study, next-generation IDF (NG-IDF) curves, which characterize the actual water reaching the land surface, are introduced into the design process to improve hydrologic design. The authors compared peak design flood estimates from the National Resource Conservation Service TR-55 hydrologic model driven by NG-IDF and PREC-IDF curves at 399 Snowpack Telemetry (SNOTEL) stations across the western United States, all of which had at least 30 years of high-quality records. They found that about 70% of the stations in the western United States showed the potential for underdesign, for which the PREC-IDF curves underestimated peak design floods by as much as 324%. These results demonstrated the need to update the use of PREC-IDF curves to the use of NG-IDF curves for hydrologic design in snow-dominated regions.

This paper investigated the variability of freshwater discharge ($Q$) in the Dez River Basin in Iran and its possible links to the North Atlantic Oscillation (NAO) and El Niño Southern Oscillation (ENSO). Nonparametric statistical tests were applied to analyze fluctuations of discharge and teleconnection indexes. The results indicated statistically significant monotonic decreases of the discharge and peak flow ($p$-value $<0.05$) across most of the subbasins. Furthermore, a downward step change was detected in discharge time series around the mid-1990s from 14.7% to 43.3% with respect to long-term values. Similarly, almost concurrent shifts detected in the Oceanic Niño Index (ONI) (from $+0.254$ to $-0.416$ in 1998) and the NAO (from $+0.196$ to $-0.155$ in 1995) that suggest changing conditions toward more cold phases of ENSO and negative phases of NAO coincided with discharge and monthly peak flow depletion. Comparing the amount of discharges during different phases showed that $Q_{(\mathrm{El}\mathrm{Ni}\tilde{\mathrm{n}}\mathrm{o})}>Q_{(\mathrm{neutral})}>Q_{(\mathrm{La}\mathrm{Ni}\tilde{\mathrm{n}}\mathrm{a})}$ and $Q_{(\mathrm{NAO}+)}>Q_{(\mathrm{NAO}-)}$. Furthermore, the Mann-Whitney U test verified statistically significant differences ($p$-value $<0.05$) between flows pertaining to different phases of both ENSO and NAO. Correlations between discharge and teleconnection indexes indicated that discharge is apparently affected by teleconnections considering the lag time.

Green infrastructure (GI) is being widely implemented in urban areas to capture and remove stormwater from the surface drainage system. Whereas most analyses have focused on diverted surface flow, here the authors demonstrate a method to quantify all components of a hydrologic budget at the site scale. The authors instrumented and applied mathematical modeling to a GI site consisting of a system of tree trenches and permeable pavement in Philadelphia, Pennsylvania. They utilized ParFlow.CLM version 743, a three-dimensional groundwater–surface water–land surface model, to quantify the water budget, including evapotranspiration, infiltration, and recharge to regional groundwater. They compared simulated and observed groundwater levels and analyzed the simulated monthly water balance for the site over 1 year. The authors found that snowmelt was an important source of recharge in the winter months of the 2016 simulation period. During the summer months when evapotranspiration exceeds precipitation, additional water captured by the GI contributing area enhances recharge to groundwater, altering water budget seasonality at the site scale. Simulation results indicate that the GI functioned as intended, converting runoff to recharge, with discharge to regional groundwater throughout the year.

This paper assesses the impact of climate change uncertainties on the Khorramabad River basin’s runoff in Lorestan Province, Iran. Five atmosphere-ocean general circulation models’ (AOGCMs) [Hadley Centre Coupled Model, version 3 (HadCM3), Center for Climate System Research-National Institute for Environmental Studies (CCSR-NIES), Commonwealth Scientific and Industrial Research Organization Mark 2 (CSIRO-MK2), Coupled Global Climate Model (CGCM2), and Geophysical Fluid Dynamics Laboratory (GFDL-R30)] projections of rainfall and surface temperature were applied to simulate climate in the periods 2040–2069 and 2070–2099 under the A2 and B2 greenhouse gases (GHGs) emissions scenarios. The AOGCM projections showed an increase in temperature and a decrease in rainfall over the future periods. The ranges of climate change scenarios were determined, and the models’ results were weighted for each month based on the $k$-nearest neighbors (KNN) method. Multiple time series of temperature and rainfall were generated with the Monte Carlo method based on their monthly probability distributions. The identification of unit hydrographs and component flows from the rainfall, evapotranspiration, and streamflow (IHACRES) hydrological model was calibrated and validated, and subsequently applied to simulate future river flow with downscaled climatic data from (1) five AOGCMs, and (2) a developed Monte Carlo model. The results showed that the average annual long-term runoff calculated with the developed Monte Carlo approach in the period 2040–2069 under the A2 and B2 scenarios decreased by 7.76% and 10.63%, respectively. The average annual runoff decreased by 13.06% and 29.49%, respectively, relative to the baseline period in the period 2070–2099 under the A2 and B2 scenarios. These results indicate a worsening availability of runoff through the remainder of the 21st century in the study region.

Spatial heterogeneity of a river basin increases the predicting uncertainty of streamflow using hydrological models, and for such river basins calibration and prediction become a challenge. For a large and low flow–producing river basin like the Pennar basin of India, single-site calibration may ignore spatial heterogeneity, which leads to the use of a multiple-site calibration approach. The present study used Soil and Water Assessment Tool (SWAT) model to develop a real-world numerical model for predicting streamflow in a large, low runoff–producing river basin and evaluated the performance of the model under single-site and multiple-site calibration approaches. Under multiple-site approach, the large basin was divided into smaller subbasins, and the calibrated parameters were sequentially applied to the entire basin, and so local conditions were incorporated very effectively into the calibration process. The results demonstrated the superiority of a multiple-site calibration over a single-site calibration approach in predicting streamflow. The overall improvement of model performance in predicting streamflow using multiple-site approach over single-site approach reached as high as 133% during calibration and 140% during validation, as observed at the basin outlet (Chennur station). It was also observed that, within the basin, precipitation contributed only 15.8% to surface runoff, 2.6% to total aquifer recharge, and 69.3% to evapotranspiration. Uncertainty analysis indicated that more than 64% of the observed streamflow was bracketed by a 95% prediction uncertainty (PPU) band under the multiple-site approach. Thus, the multiple-site calibration approach demonstrated its ability to improve model performance by capturing the heterogeneity of a low runoff–producing basin like Pennar.

To study the impact of human activity on river basin sediment erosion and sediment transport in the upper Yangtze River, variations in the sediment delivery modulus and the incoming sediment coefficient during the period from 1956 to 2015 were investigated with a cumulative curve method and the Mann-Kendall test. Then, the contribution of human activity on sediment load reductions was calculated quantitatively with a hydrologic method. The results showed that the values of the sediment delivery modulus and the incoming sediment coefficient decreased with time, which indicated a decrease in basin sediment yield and an increase in sediment deficit in recent decades. Moreover, human activities such as reservoir construction, soil conservation, and sand excavation were the main reasons for sediment load reductions, whereas there were no significant trend changes in rainfall, which had a minimal effect on the sediment yield changes in the upper Yangtze River. Of the total reductions in sediment load, sediment trapped by reservoirs contributed close to 88%, whereas soil conservation and other factors only contributed 12%.

A micropool is a small wet pool placed in front of the outlet structure installed at a stormwater extended detention basin. An extended detention basin is sized with a user-specified drain time in order to meet the stormwater quality criteria. The outlet structure is often composed of a vertical perforated plate and a top horizontal grate to release the loaded basin over a long period of time. When the perforated plate becomes clogged, the siphon system housed in the micropool becomes surcharged and can continually drain the stored water volume to meet the design criteria on the basin’s drain time. This study presents a new design method to size a micropool. It is recommended that a micropool be sized by the debris particle’s float velocity, water quality capture volume, and basin’s drain time. The depth of a micropool shall take into consideration evaporation loss and sediment dead storage volume. The debris samples collected in the field represent typical urban floating particle sizes observed in stormwater basins. It is recommended that a micropool be sized to capture up to 80%–90% of floating particles accumulated in a detention basin. This new method provides a reasonable surface area and pool depth to achieve the goals set forth for micropool designs.

The curve number (CN) rainfall-runoff method is described, with inferred extension to daily time-step models, which are the bases for most total maximum daily load (TMDL) assessments. Although originally developed for engineering design for extreme rainfall events, its flexibility, its basis in soils and land use, and its history of accepted use have encouraged adaptation and adjustment to the continuous daily time-step models needed for TMDL evaluations. The original event-based CN model itself, background assumptions, improvements, problems, and evolved usage are described. Included are the roles of the initial abstraction ratio ($I_a:S$), antecedent conditions (ARC), the limits of application with different land types, model sensitivity, effects of land slope, and seasonality. Inconsistencies and recent improvements are stressed, including (1) the adjustment of the abstraction coefficient $I_a:S$ to 0.05 from the traditional value of 0.20; (2) awareness of the nonapplicability of the CN method in all cases and the preferred use of distributed/weighted runoffs from an array of CN source areas over the use of runoffs from averages CNs; and (3) use of local data and on-site inspections for calibration and perspective. The application of CN-based watershed hydrology models for TMDL studies should be conducted cautiously because it can produce a biased estimation of hydrograph components in certain soils and landscapes. Model parameter calibration in the application of CN method for continuous simulations of TMDLs is strongly suggested.

Intensity-duration-frequency (IDF) relations are essential for estimating extreme rainfalls for the design of various hydraulic structures. However, in current engineering practice, the construction of these relations represents a challenging and tedious task since it involves the uncertainty analysis of different probability models and the frequency analyses of a large amount of extreme rainfall data for different durations at a given site or over many different locations. In particular, the selection of the best probability model for extreme rainfalls is the most difficult decision since it requires two main challenging tasks: (1) a detailed evaluation of the descriptive and predictive abilities of each selected distribution as well as the analysis of its uncertainty and (2) a systematic comparison of the accuracy and robustness of all candidate models based on a number of graphical and numerical performance criteria. Hence, this paper proposes a decision-support tool, herein referred to as SMExRain, that can readily be used to identify, in an objective and systematic manner, the most suitable distributions for the accurate and robust estimation of design rainfalls. Therefore, SMExRain represents an efficient practical tool for establishing reliable IDF relations for a given site or for many sites of interest. The tool is based on the theoretical procedures developed by the authors and others. An illustrative application of SMExRain is presented using extreme data for the Ontario region (Canada) to demonstrate the efficiency and usefulness of the proposed tool in engineering practice.