Journal of Hydrologic Engineering

Detention ponds are known as an efficient low-impact development (LID) strategy for reducing the negative impacts of urban floods. Their performance in flood peak attenuation is directly related to the size, location, and type of connections to the main drainage system, as well as the shape of their inlet/outlet structures and operational procedures. In this paper, a methodology is developed to design multiple detention ponds in urban drainage systems considering rainfall uncertainties. Harmony search (HS), a popular metaheuristic algorithm, is coupled with USEPA’s Storm Water Management Model version 5.1 (SWMM 5.1) and run under a significant number of synthetic rainfalls generated by copulas, a kind of joint probability distribution, to find the optimal place and size of online/offline detention ponds in the network. The approach is tested on the east main stormwater drainage network of Tehran, the capital city of Iran. Results reveal that offline construction of ponds has generally better performance than online construction. An optimal design and arrangement of ponds efficiently decreases flooding in the whole network up to 12.5% and outperforms the previous design proposed by consultant engineers for the rehabilitation of the studied network.

Total maximum daily load (TMDL) reports archived in the USEPA database can provide useful guidance for the development of new TMDLs and watershed management plans by aiding the selection process for the most appropriate modeling tool. The database contains more than 70,000 individual documents; therefore, a rapid screening tool is needed to elicit information about previous modeling studies that might help guide stakeholders and regulators in dealing with the TMDL application at hand, save time, and lead to a more cost-effective regulatory outcome. The paper introduces a smart web-based software tool for TMDL report selection based on different water management criteria. The tool uses an automated search method based on frequency of common water body impairments and models to categorize and select TMDL reports. Additionally, this tool provides better insight on the relationship between the modeling tools used and the impairments they address. This tool has proven useful in reviewing the state of integrated modeling (IM), applications of remote sensing (RS), application of basic versus mechanistic modeling, margin of safety (MOS) assessment, and the state of practice regarding relationships among impairments, models, and regions where TMDLs for various pollutants are being developed. Despite limitations on direct access to all TMDLs developed and reported to the EPA by the user, the tool can be improved over time to derive a better understanding of the relationships between these impairments, data, and the TMDL development process. Although the MOS is not directly quantified in the current version of the TRS tool, this feature may be incorporated in future updates.

Stochastic weather generators are generally used to produce scenarios of climate variability on a daily timescale for hydrological modeling and water resource planning applications. Most of the available weather generators assume extreme rainfall series as stationary series. However, it is currently perceived that global climate change is increasing the intensity and frequency of extreme rainfall events and creating a nonstationary component in extreme rainfall time series. Consequently, the realistic modeling of rainfall extremes in a nonstationary context is indispensable. In this study, we propose a modified version of a $k$-nearest neighbor (KNN) weather generator that incorporates nonstationarity in the extreme rainfall series. The proposed algorithm first models the nonlinear trend in the extreme rainfall series that exceeds the defined threshold $u$ and perturbs the original-KNN-simulated extreme rainfall using the knowledge available in the nonstationary model. The proposed algorithm is demonstrated with three case studies, and the performance of the proposed algorithm is validated using various extreme precipitation indices. The results of the three case studies indicate that extreme rainfall characteristics are consistently well simulated with the proposed algorithm. Particularly, based on the results of the three case studies, the proposed algorithm decreases the root-mean-square error (RMSE) in rainfall simulation with respect to the original KNN algorithm by at least 40%.

Reservoirs, as the largest water systems for controlling water and supply demands, have been particularly important due to their operational requirements. Applying practical tools to evaluate the performance of reservoir systems plays a key role in their effective operational management. This study focused on developing a simulation–optimization tool for the design and operation of multiattribute reservoir systems (SOMAR), which includes four parts: input data, simulation, optimization, and output results. Its different components can be selectively activated for each problem. SOMAR can use both standard operation policy or given rule curves for reservoir system simulation in accordance with the user’s preference. It applies a comprehensive evolutionary algorithm for optimization. In this study, SOMAR was validated in seven types of reservoir system problems, including simulations based on (1) standard operation policy, (2) given rule curves, (3) design optimization, (4) long-term operation optimization, (5) design and long-term operation optimization, (6) rule curve optimization, and (7) design and rule curve optimization. The validations indicated the capabilities of SOMAR in analyzing reservoir systems with high reliability of the final results by considering different aspects of these systems.

The discretization of the generalized Nash model (GNM) has been made in such a way that it can be applied easily to the sample-data system for flood routing. First, the variable $S_n$-curve was introduced to simplify the GNM. Under a commonly used assumption that the streamflow changes linearly between any two measurements, the simplified GNM was further discretized to a linear expression of the inflows and outflows. Then, the discrete generalized Nash model (DGNM), as well as its multistep-ahead form, that is, $m$-DGNM, was obtained, by which the outflow was produced by the old water stored in the river reach and new water from the upstream inflows. The weight coefficients of the old and new water were expressed in explicit functions of the $S_n$-curve, which makes the DGNM and $m$-DGNM easy to compute and more intuitive than the GNM. To test the predictive ability of the DGNM, a comparison with the widely used Muskingum method and dynamic wave model was made. Results showed that with some historical information included, the DGNM is more adaptive and accurate than the Muskingum method. Compared to the complicated dynamic wave model, the DGNM is relatively simple to implement and reasonably accurate. Then, the DGNM and $m$-DGNM were further applied to the Qingjiang River for flood routing. Reasonable results obtained suggest that they are highly suitable for practical flood routing applications.

Reliable water shortage evaluation in irrigation districts is based on more accurate water supply and demand data series. The objective of this paper is to propose a stochastic simulation model to more accurately simulate a water supply-demand time series (WSDTS). This stochastic simulation model is a two-dimensional contemporaneous-temporal dependence model (C-TDM) of natural WSDTS based on a time series of annual rainfall and reference crop evapotranspiration (${\mathrm{ET}}_0$) from 1970 to 2013 in the Luhun irrigation district in Henan Province, China. The results reveal that the statistical properties as well as the contemporaneous dependence relationship (CDR) and temporal dependence relationship (TDR) of the WSDTS are well-maintained in the simulated data. All correlation coefficients exceed 0.95. The root-mean square error (RMSE) of the C-TDM indicates that the model performs better than either a contemporaneous dependence model (CDM) or a temporal dependence model (TDM). Furthermore, the copula-based stochastic simulation model is used to evaluate the water shortage risk in the Luhun irrigation district. The joint probability distribution (JPD) of a rainfall-poor, rich-${\mathrm{ET}}_0$ scenario is 0.24, suggesting that a serious water resources shortage is unlikely to occur. In addition, the results indicate that considering both the TDR and CDR in stochastic simulation models can provide more conservative and reliable forecasts for water shortage risk evaluations in irrigation districts.

Rainfall-derived infiltration and inflow (RDII) is a major issue causing surcharge flows in many municipal sanitary sewer systems. This paper demonstrates a statistical method for characterizing the rainfall-derived inflow (RDI) originating from residential foundation drains (also referred to as weeping tiles), as well as the importance of having site-specific data. The results differentiating the contribution of RDI from residential weeping tiles (WT) to total RDII are demonstrated for a case study site in a residential subdivision in London, Ontario, Canada. This research used statistical linear regression analyses with bootstrapping methods to quantify the RDI and its flow duration. It was found that the RDI from WT contributed up to 85% of the total RDII in the sanitary sewer. By disconnecting WT at this site, the RDI generated as a result of rainfall events was reduced by a minimum of 78% in volume and 32% in flow duration. Thus, this paper presents a novel method to quantify RDI and its duration from statistical perspectives, which provides better supporting evidence and guidance for RDI projects.

In water resources and environmental management, hydrologic indexes are often valued as decision support tools because of their practical interpretability. This is true with the streamflow drought index (SDI), which is considered to be a relevant tool for assessing the availability of water resources at the watershed level. Hence, the future of freshwater resources at the watershed scale could be better understood by achieving a realistic projection of SDI. This study used a process-based watershed modeling approach to describe a framework for SDI projection. Specifically, the Soil and Water Assessment Tool (SWAT) model was used to simulate distinctly two watersheds located in the state of Texas, the Trinity and the Neches River Basins. The SWAT model was calibrated with monthly streamflow data for the period 1990–1995. The model was subsequently validated with two decades of discharge data (1996–2015). The evaluation of the SWAT performance during the calibration and validation stages showed acceptable values of efficiency criteria for both watersheds (i.e., Nash-Sutcliffe efficiency ranging from 0.56 to 0.65; index of agreement from 0.79 to 0.92). The calibrated model was used to simulate runoff for the future period 2041–2070 using inputs retrieved from a future climate scenario. However, the SDI calculation requires knowledge of the probability distribution of cumulative discharge data. A Kolmogorov-Smirnov’s goodness-of-fit analysis was conducted for both observed and simulated cumulative discharges. A lognormal distribution was considered for estimating time series of SDI. For the period 1996–2015, the SDI values recovered from the SWAT simulations matched closely with those derived directly from the observed discharge data ($0.52\le R^2\le 0.91$ for the Neches River, and $0.79\le R^2\le 0.89$ for the Trinity River). This result demonstrated the capacity of the analytical procedure to capture and project realistically SDI signals. However, analysis of the ${\chi }^2$ statistic of the SDI patterns for the past and the future periods did not reveal any significant difference.

A key challenge in enhancing flood forecast relies in the difficulty of reducing predictive uncertainty. The precipitation-dependent hydrologic uncertainty processor (HUP) is a flexible model independent Bayesian processor that can be used with any hydrologic model to provide probabilistic forecast. This study investigates the use of HUP with different hydrologic models for hydrologic uncertainty quantification in a flood forecasting scheme for a semiurban watershed of southern Ontario, Canada. The purpose is to better understand predictive uncertainty and enhance flood forecasting system reliability in semiurban conditions. HUP is based on Bayes’ theorem, and it updates the prior distribution given available information at the forecast time to obtain the posterior distribution that is close to future unknown actual value. In this study, the hydrological model (HYMOD) and the modèle du Génie Rural à 4 paramètres Horaire (GR4H) were selected to work with HUP, and the Bayesian processor was calibrated using a number of selected flood events from 2005 to 2014. The performance of the processor was assessed by graphical tools and performance metrics, like reliability plots, Nash-Sutcliffe efficiency (NSE), and continuous ranked probability score (CRPS). Results show that HUP provides a robust framework and a reliable analytic-numerical method for the quantification of hydrologic uncertainty, the actual values are well captured by the uncertainty bounds, the CRPS values are relatively small, and reliability curves lie close to the bisector. The comparison between the NSE calculated from the output of the sole deterministic hydrologic model (HYMOD/GR4H) and from the median of the predictive distribution produced by HUP-HYMOD/HUP-GR4H, demonstrates that HUP has the ability to improve the deterministic forecast. For low peak flow events, HUP combining with different hydrologic models presents similar predictive performance, while for high peak flow events, a well performed deterministic model is required in HUP to produce an accurate probabilistic forecast.