Journal of Hydrologic Engineering

Hydrologists widely use regional flood frequency analysis to predict flood quantiles for ungauged and sparsely gauged target locations. The analysis involves: (1) use of a regionalization approach to identify a group of watersheds (region) resembling the watershed of target location by searching in watershed-related attribute space, and (2) use of information pooled from the region to perform regional frequency analysis (RFA) for flood quantile estimation. Conventional regionalization approaches prove ineffective for identification of regions in situations where there are outliers in the attribute space. This paper proposes an entropy-based clustering approach (EBCA) to identify regions by accounting for outliers and recommends applying a recently proposed RFA approach on the regions for quantile estimation. The EBCA yielded seven new homogeneous regions in four major river basins (Mahanadi, Godavari, Krishna, and Cauvery) of India. The regions are shown to be effective compared to six existing regions (used by an Indian government organization) and those delineated using global K-means clustering and region-of-influence approaches, in terms of regional homogeneity and utility in arriving at reliable flood quantile estimates for ungauged sites.

Using recent advancements in the statistical extreme value theory, this study proposes a methodology for detection of change in flood return levels under climate change. Nonstationary scaling of regional projected peak flows with global warming is first tested by a likelihood ratio test. For nonstationary possible future realizations, the authors then investigate how long the stationary historical design magnitudes or return levels of floods will remain valid, taking into account the uncertainties in the estimation of observed and projected return levels. Although some flood projections are found to be nonstationary, many are stationary in nature. No coherent change in flood return level across the projections is detected in the case study of floods in the Columbia River using available streamflow projections. Most projections yield flood quantiles that are not likely to be critical in the coming century. However, for some simulations detection is achieved, with earlier detection in design magnitudes of lower return periods. A possible worst-case scenario considering the maximum of all the projections shows detection of change in floods of higher return periods in the 21st century.

The aim of this study is to compare the performance of a symbolic regression combination method based on gene expression programming (GEP) with different neural network combination methods when used in the development of multimodel systems. The two different neural network combination methods used in this study are the multilayer perceptron neural network (MLPNN) and the radial basis function neural network (RBFNN). The methods were used to combine the results from different types of rainfall-runoff models to test the multimodel combination system in catchments located in Thailand and New Zealand. Comparison of the results revealed that the GEP performed better than neural network methods in the case of the catchment located in New Zealand. Nevertheless, the RBFNN method outperformed the GEP and the MLPNN combination method in the case of the catchment located in Thailand. However, which combination method produces better results in the multimodel combination is not conclusive. The results suggest that the selection of the best combination method to be used in conjunction with the multimodel approach may depend on the catchment type.

Urmia Lake, the largest lake in Iran, is an important water body and habitat for a variety of different species. Data of ten climate models among the atmosphere ocean general circulation models (AOGCMs) were assembled for the 2021–2050 period under A2 and B1 emission scenarios. First, both the 30-year historic (1961–1990) and 30-year future monthly temperature and precipitation in the region were generated and weighted by the beta function ($\beta$) that assigns weights to AOGCMs based on the model evaluations. Then, the cumulative density function (CDF) in each month was determined. Precipitation and temperature values at 25, 50, and 75% probability were extracted from the CDFs. Subsequently these values were introduced to the Long Ashton Research Station Weather Generator model (LARS-WG) to downscale and produce time series of temperature and precipitation in the future, subject to the uncertainty of climate models. Then, these monthly time series at different risk levels were introduced into the artificial neural network (ANN) model and future monthly runoff time series were generated. The findings indicated a decrease in the flow into the lake of the order of 21, 13, and 0.3% under the A2 scenario and an increase of 4.7, 13.8, and 18.9% under the B2 scenario with respect to the observation period at 25, 50, and 75% risk levels, respectively. The results of this study highlight the necessity to immediately and effectively cut back on the water use in the basin not only to remedy the projected climate change effects in the future, but also to ensure gradual reclamation of the ongoing environmental crisis.

This study investigates the impact of event characteristics on runoff dynamics during extreme flood events observed in an $8.5\text{-}{\mathrm{km}}^2$ experimental watershed located in South Korea. A high-quality data set containing the 31 most extreme flood events with event rainfall in excess of 50 mm were analyzed using an event-based rainfall-runoff model; the revitalized flood hydrograph (ReFH) routinely used for design flood estimation in the United Kingdom. The ReFH model was fitted to each event in turn, and links were investigated between each of the two model parameters controlling runoff volume and response time, respectively, and event characteristics such as rainfall depth, duration, and intensity, and also antecedent soil moisture. The results show no link between the parameter controlling runoff volume and any of the event characteristics, but identified a dependence between response time and rainfall depth. These results show that the linear unit hydrograph fails to adequately represent a reduction in watershed response time observed for the more extreme events. A new and dynamic link between the unit hydrograph shape and rainfall depth is introduced. The consequence of the observed nonlinearity in response time is to increase design peak flow by between 50% for a 10-year return period, and up to 80% when considering the probable maximum flood.

Stormwater wet detention ponds hold a permanent pool of water and offer many beneficial uses including flood mitigation, pollution prevention, downstream erosion control, increased aesthetics, and recreational uses. The reduction efficiency of nutrients is generally low in stormwater wet detention ponds in many urban areas. To enhance nutrient reduction, floating treatment wetlands (FTWs), which are a new best management practice (BMP) in stormwater management, can be installed in wet detention ponds to offer an innovative solution toward naturally removing excess nutrients and aiding in stormwater management. This study assesses nutrient reduction in three Floridian stormwater wet detention ponds where FTWs have recently been installed. Both episodic (storm event) and routine (nonstorm event) sampling campaigns were carried out at the three ponds located in Ruskin, Gainesville, and Orlando. Most notably, nutrient reduction rates after installation of the FTWs reached levels of 33% for total nitrogen at the Ruskin pond during storm events, 71% for total phosphorus at the Gainesville pond during storm events, and 17% for total nitrogen at the Orlando pond during nonstorm events.

A U.S. EPA (EPA) model was developed for the Cathedral Run stormwater wetland (Philadelphia, Pennsylvania). This research presents a formal sensitivity analysis of hydraulic and hydrologic model parameters contributing uncertainty with the multiobjective generalized sensitivity analysis (MOGSA) algorithm. The parameters identified as significant include: percent routed ($P_R$), subcatchment soils, subcatchment width, wetland soils, and the flood weir coefficient. These results suggest that this model is well parameterized for detailed simulations of stormwater control installations, and contests the existence of a globally sensitive set of parameters. This research demonstrates that detailed models of stormwater control installations are significantly affected by uncertainty related to parameters beyond traditional calibration (i.e., runoff generation) parameters. The authors present a monitoring design based on wetland water surface elevation. The simplified monitoring scheme obtained statistically significant calibration data as determined through MOGSA. The generalized likelihood uncertainty estimation (GLUE) algorithm was then applied to develop marginal posterior model parameter distributions and two-dimensional (2D) probability spaces using a formal Bayesian likelihood function. The GLUE results demonstrate the importance of uncertainty and equifinality within the context of stormwater wetland modeling.

Climate change and sea-level rise are expected to alter the likelihood of extreme events, such as floods, within the design lifetime of infrastructure components. Critical civil infrastructure facilities, including wastewater treatment, transportation, and energy, need site-specific flood contingency plans that reflect the effects of changing climate. This study developed a sensitivity analysis method to assess future flood risk by estimating flood frequency under conditions of higher sea level and streamflow response to increased precipitation intensity. The method was applied to an ungauged location on a tidal estuary in the Mid-Atlantic region as a case study. One-dimensional (1D) unsteady flow analysis using a hydraulic analysis software developed by the U.S. Army Corps of Engineers was used to predict discharge and water surface elevation along the estuary reach, subject to prescribed boundary conditions of upstream discharge and downstream water surface elevation. A current-climate flood-depth frequency curve was estimated for the study site based on simulations of high-flow events in the 18 years for which simultaneous upstream and downstream records were available. The simulations were repeated, applying additive water surface elevation (WSEL) perturbations at the downstream boundary (to represent anticipated sea-level rise) and multiplicative event discharge perturbations at the upstream boundary (to represent anticipated change in watershed hydrology). The perturbations were applied separately and together. Revised flood-depth frequency curves were calculated for each set of perturbations. For this location, the 1% annual exceedance (100-year) WSEL is 1.18 m (3.9 ft) higher, and the 0.2% annual exceedance (500-year) WSEL is 1 m (3.3 ft) higher, than current climate in the worst-case scenario, 1.22 m (4 ft) of sea-level rise and a 30% increase in event discharge. For that scenario, the current 1% exceedance (100-year) WSEL has a 67% probability of exceedance (1.5 year). The results indicate that the effects of the upstream and downstream changes are not additive. This research will help infrastructure stakeholders be aware of the flood risk and vulnerability while environmental changes are underway.

Many hydraulic projects as reservoirs, ponds, paddy fields, and soil and water conservation engineering projects have been constructed upstream of the Wudaogou station basin in northeast China. This study investigates the impact of climate variability and hydraulic project construction on the runoff decrease through the concept of climate elasticity. The elasticity of streamflow in relation to precipitation and potential evapotranspiration is calculated by four methods: two nonparametric estimators, two parametric estimators. Results show that for the period 1936–2010 the two nonparametric climate elasticity methods show a smaller value of the elasticity of streamflow in relation to precipitation (1.85 and 2.47) but a larger value of the elasticity of streamflow in relation to potential evapotranspiration ($-0.02$ and $-0.48$) than the parametric estimators. It is estimated that climate variability is responsible for 47.3% (40.3 mm) of the streamflow reduction between 1936–1956 and 1957–2010, while hydraulic project construction contribute to 52.7% (45.1 mm) of the reduction. From 1957 to 2010, the amount of construction for water conservancy has been growing. To further investigate the effect of hydraulic project construction, the study period (1957–2010) is split into three subperiods using two faint jump points (1975 and 1987), and the effect of hydraulic project construction on the runoff during each subperiod is evaluated. The results indicate that as the cumulative storage of hydraulic projects increases, hydraulic project construction plays an increasingly important role (from 12.8 to 90.8%) in streamflow change. Almost all the results show that streamflow changes due to hydraulic project construction are less than the cumulative storage of the hydraulic projects, which verifies the reliability of the theoretical calculation.

Rescaling of the geometrical sizes and the value of hydraulic conductivity in the classical problem of steady two-dimensional (2D) potential seepage through a rectangular earth dam with an empty tailwater is shown to result in a mathematically equivalent problem of seawater intrusion with a sharp interface into a confined horizontal aquifer, which discharges fresh groundwater to the sea through a vertical segment of the beach. The shape of the interface, the vertical and horizontal sizes of the static intrusion wedge, and its cross-sectional area are written in an explicit form, using the Polubarinova-Kochina formulas, rectified. The densities of the two liquids and the aquifers’ hydraulic conductivity and thickness, as well as the incident hydraulic gradient serve as input parameters. With reduction of the incident groundwater gradient far upstream from the intrusion zone (due to, e.g., freshwater abstraction by wells), the sizes of the wedge rapidly increase. The analytical solution has been validated with recent sand tank experiments.