Journal of Hydrologic Engineering

Forum papers are thought-provoking opinion pieces or essays founded in fact, sometimes containing speculation, on a civil engineering topic of general interest and relevance to the readership of the journal. The views expressed in this Forum article do not necessarily reflect the views of ASCE or the Editorial Board of the journal.

In the quest for a best approach to estimate the spatial distribution of monthly precipitation for the state of Pernambuco, Brazil, this work tests different versions (choices of parameters and functions) of seven interpolation methods, totaling 26 distinct interpolation schemes. The performance of each method in estimating monthly precipitation was evaluated through a set of seven measures of performance evaluation through a cross-validation procedure. It was found that the trend surface analysis yielded the best overall interpolation results, followed by natural neighbor method, inverse distance weighting, and kriging. Performance ranking of all 26 methods is provided. Moreover, in order to assess the regional performance quality of the best identified method (trend surface analysis) in estimating monthly precipitation, a spatially explicit error analysis was also performed for each of the evaluation measures. Finally, temporal evolution of the error evaluation measures of the best performing methods is presented in order to demonstrate the seasonal impact of interpolation.

Infiltration trenches can help to accelerate infiltration of water into the deep vadose zone and groundwater table. A two-dimensional model was developed to characterize the performance of arbitrarily shaped infiltration trenches. This model employs a mixed form of Richards’ equation, which was solved using the coordinate transformation technique, and the reservoir equation to describe the effects of water ponding and evaporation through the soil–air interface. The model was validated using numerical and experimental test cases in the literature, and in all cases, acceptable results were obtained. Subsequently, the shape of an infiltration trench was optimized with respect to the amount of inflow water using the Taguchi technique. Running 25 combinations which were representative of 125 different trench shapes and volume of inflow water showed that a rectangular trench with a capacity of about 36% of the volume of inflow water is the best cross section of a trench. The predicted normalized recharge obtained by the Taguchi method and that of numerical modeling showed perfect agreement (about 0.01% error) for the combination that resulted in the most normalized recharge value.

A karst collapse column (KCC), a type of vertical structure that is common in the Carboniferous-Permian coal fields of North China, contains many graded broken rocks and often functions as a channel for groundwater inrush. This study used a custom-built apparatus to investigate the effects of the particle-size distribution on the seepage behavior of a sand particle mixture subjected to a high hydraulic gradient. Three different flow regimes were identified: (1) the Darcy regime, (2) the Forchheimer regime, and (3) the turbulent regime. When flow transition begins, the critical flow velocity, critical Reynolds number, and Forchheimer number increase as the nonuniformity coefficient $C_u$ (or the coefficient of curvature $C_c$) increases when the porosity remains constant, and the critical hydraulic gradient decreases. The permeability increases linearly, and the inertia factor decreases exponentially as $C_u\times C_c$ increases. Using new experimental data, the validity of five widely used empirical formulas for permeability and the inertia factor were evaluated. The results indicated that the empirically corrected formulas for permeability and the inertia factor yielded high prediction accuracy and can be used to predict the permeability and inertia factor of KCCs in practical engineering.

Quantifying the dynamics of snow depth is essential for understanding freshwater availability, mitigating flood and drought hazards, and monitoring the effects of climate change in cold regions. Here, a statistical approach for describing the dynamics of monthly snow depth loss (SDL) is developed and tested in 67 climate stations throughout southern Canada. The framework fuses an input selection scheme with multiple linear regression to approximate the SDL using a set of climate proxies, either explicitly or implicitly through modeling snow depth. Our findings suggest that statistical models—if properly developed and used—have the potential to form effective tools for describing the dynamics of SDL. In particular, the implicit statistical model, in which climate proxies are selected globally among all stations, provides an accurate model (expected $R^2=0.75$), which can outperform a frequently-used temperature-index model in a majority of stations. In addition, parameters of the statistical model can be regionalized efficiently (expected $R^2=0.71$ for the generalized model) using latitude, longitude, and altitude. This ability can provide a basis to extend the model application into ungauged sites.

A three-dimensional compliance monitoring network design methodology is presented to develop an adaptive and sequentially modified management policy that intends to improve optimal and justifiable use of groundwater resources in coastal aquifers. In the first step, an ensemble metamodel-based multiobjective prescriptive model is developed using a coupled simulation-optimization approach to derive a set of Pareto optimal groundwater extraction strategies. Prediction uncertainty of metamodels is addressed by utilizing a weighted average ensemble using set pair analysis. In the second step, a monitoring network is designed for evaluating the compliance of the implemented strategies with the prescribed management goals due to possible uncertainties associated with field-scale application of the proposed management policy. Optimal monitoring locations are obtained by maximizing Shannon’s entropy between the saltwater concentrations at the selected potential locations. Performance of the proposed three-dimensional sequential compliance monitoring network design is assessed for an illustrative multilayered coastal aquifer study area. The performance evaluations show that sequential improvements of optimal management strategy is possible by using saltwater concentrations measured at the proposed optimal compliance monitoring locations. Therefore, the salinity concentration data collected at the designed compliance monitoring wells can be used to collect feedback information in terms of salinity concentrations. This feedback information can be applied to improve the initially prescribed optimal groundwater extraction patterns while keeping the original management goal intact.

One-dimensional (1D) models based on the Saint-Venant set of equations (1D–SVE) are frequently used for large- and regional-scale hydrodynamic modeling due to their low implementation costs and small computation time compared with more advanced two-dimensional (2D) or three-dimensional (3D) models. A simplified description of the flow transformation, which is a weakness of the most commonly used methods to solve 1D-SVE, and the lack of accurate river topography can make it difficult to achieve reliable flow transformation results for a long section of a river. Mapping of floodplain retention and determining the active flow area are issues that significantly affect the results obtained from 1D numerical models. The methods described in this study attempt to reduce the retention mapping error and facilitate the determination of the active flow area in models based on 1D-SVE. On the basis of high-resolution digital elevation models (DEMs) a volume–water level relationship was introduced into the continuity equation. As a result, the error of water volume retention mapping in the applied hydrodynamic model based on 1D-SVE was decreased by 33% for the case study section of the Warta River in Poland. It was also shown that the proposed methodology for determining the active flow area in 1D models can be used for both wide natural floodplains and narrower embanked floodplains. It allows mapping the active flow area by introducing an additional model-specific parameter $C_T$ (slip velocity in Pasche’s method).

Bulletin 17B and its proposed update, Bulletin 17C, continue to recognize difficulties in determining flood frequency estimates among streamflow records that contain flood peaks coming from different flood-generating mechanisms, as is the case in the western United States [Interagency Advisory Committee on Water Data (1982). “Guidelines for determining flood flow frequency: Hydrology Subcommittee Bulletin 17B.” Reston, VA: USGS]. In the “Future Studies” section of Bulletin 17C, the Work Group identified the need for “the identification and treatment of mixed distributions, including those based on hydrometeorological…conditions” [England, J. F., Jr., T. A. Cohn, B. A. Faber, J. R. Stedinger, W. O. Thomas, Jr., A. G. Veilleux, J. E. Kiang, and R. R. Mason, Jr. (2018). “Guidelines for determining flood flow frequency—Bulletin 17C.” Chap. B5 in USGS Techniques and Methods, Book 4. Reston, VA: USGS]. This study provides a general statistical framework to perform a process-driven flood frequency analysis using a weighted mixed population approach. Furthermore, it allows for accounting for both sampling and mixing uncertainties. Analyses are based on 43 long-term USGS stream gauges in the western US with at least 50 years of annual peak flow data, at least 25 of which are generated by atmospheric rivers (ARs). Visual and quantitative goodness-of-fit assessments are made to evaluate the performance of the weighted mixed population approach with respect to the observations. Thirty-four (80%) of the 43 sites have similar flood frequency curves from both the homogeneous (single) and heterogeneous (weighted mixed) population methodological approaches. Yet nine (20%) of the sites have notably different quantile estimates in the upper tail of the distribution. Two important factors contribute to the overall differences in the flood frequency estimates among these sites, regardless of their physiographic locations. The best goodness of fit in the upper tail of the distribution, the portion of most concern in designing flood flow structures, is found when (1) potentially influential low floods (PILFS) are identified, and/or (2) the composite distribution contains markedly different at-site log-unit skews (shape parameter) among the AR/non-AR subpopulations compared with the single homogeneous population. Furthermore, the weighted mixed population confidence intervals tend to be wider than the single population in both tails of the distribution, due primarily to the reduced sample size from separating the observed flow series into AR/non-AR subpopulations and the contributions from the mixing fraction of ARs. However, we found similar interval widths throughout the remaining distribution, implying that our simulation framework can capture the improved procedures for quantifying quantile estimate uncertainties described in Bulletin 17C in addition to the mixing ratio uncertainties.

In southwestern Georgia, increased groundwater withdrawal for irrigation in the Lower Flint River Basin (LFRB) is greatly affecting the supply of freshwater to downstream regions during drought years. This study was undertaken to understand the sensitivity of groundwater components to changes in irrigation pumping scenarios during droughts in the highly connected karst Upper Floridan Aquifer (UFA) and assess the effectiveness of irrigation shutdown on stream-aquifer fluxes. This study was done using the United States Geological Survey (USGS) MODular Finite-Element Groundwater model (MODFE). The model was run for the water year 2011 and changes in simulated groundwater-budget components for monthly irrigation pumping rates of 50%, 75%, 100%, 125%, 150%, and 200% of that indicated by metered irrigation systems were quantified. The effect of irrigation shutdown in the vulnerable Spring Creek subwatershed was also investigated. Results show that increasing irrigation pumping in the study area decreased aquifer storage, induced recharge to the UFA (vertical leakage), and induced recharge from streams into the UFA. During the months of May–September 2011, doubling of irrigation pumping induced nearly double the recharge rate into the UFA from streams and effected a 10% decrease in discharge rate from the UFA to streams. Analysis of irrigation shutdown indicates that eliminating irrigation pumping from the Spring Creek watershed alone facilitated an increase in baseflow of $46\times {10}^4{\mathrm{m}}^3/\mathrm{d}\mathrm{a}\mathrm{y}$, or about two-thirds the total increase in baseflow that would be realized from eliminating irrigation pumping within the entire LFRB.

This study investigates four widely used satellite and gauged-based precipitation products for hydrological evaluation in the poorly gauged Tigris River basin (TRB), with an area of $\mathrm{445,656}{\mathrm{km}}^2$, using the Soil and Water Assessment Tool (SWAT) watershed model. The multiple precipitation data sources (PDSs) evaluated in this study include Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks-Climate Data Record (PERSIANN-CDR), Multisource Weighted-Ensemble Precipitation (MSWEP), Asian Precipitation Highly Resolved Observational Data Integration towards the Evaluation of water resources project (APHRODITE), and National Oceanic and Atmospheric Administration (NOAA) Climate Prediction Center (CPC) data. The SWAT model was calibrated using three approaches to identify more realistic parameters. Overall, APHRODITE could capture the spatiotemporal distributions of daily precipitation with a correlation coefficient of 0.65, root-mean square error of 0.62 mm, and percent bias of 19.1%. In addition, APHRODITE also captured the monthly streamflow with reasonably accuracy for selected streamflow stations in the TRB with mean Nash-Sutcliffe efficiency of $>0.65$. Conversely, MSWEP overestimated and CPC underestimated the observed mean climatology, respectively, and had similar effects on monthly streamflow simulations. Among all the selected precipitation products, the relative performance of CPC is poor in comparison to other data sets.

In recent years, there is an ongoing debate regarding the causes of observed streamflow change across the Midwest. In the Wabash River basin in Indiana, streamflow trends due to changes in land management and climate are potentially masked by the effect of flood control reservoirs constructed in the mid-1960s. To investigate the impact of environmental changes on streamflow, historical daily data were used to estimate daily naturalized streamflow for six sites in the Upper Wabash basin. Annual statistics calculated from daily average streamflow show that reservoir management has resulted in 36% and 38% decreases in annual maxima series (AMS) and Richards-Baker flashiness index (RBI), respectively, and a 21% increase in annual minima (MIN) for the Wabash River at Lafayette between 1971 and 2000. Analysis of the naturalized streamflow time series from 1968 to 2014 indicates that there is an increasing trend in MIN and mean annual flow (MAF) across the watershed, as well as an increasing trend in the RBI and AMS. These trends are consistent with streamflow trends detected in other studies throughout the region.

Erosion is one of the key factors in environmental management and river restoration. Different mechanisms influence riverbank stability and mass failure, such as seepage erosion. The present study investigates the effects of soil dispersion as one of the soil chemical properties on seepage erosion to evaluate the instability criterion for riverbanks. The middle Atrak River in Iran, where extensive riverbank erosion, seepage cavity, and lumped mass failure have been observed, was selected as a case study. Results indicated that there was significant correlation ($r=0.90$) between the soil dispersion and the volume of the seepage undercutting cavity. Further, the results of numerical simulations suggested that where the depth of the undercutting cavity was large, the safety factor of riverbank stability was about 25% less than that at a location with less undercutting. Attempts were also made to understand the effects of groundwater level on the stability of eroded riverbanks due to seepage. It was also observed that in locations with larger cavity dimensions, an increase in subsurface water table up to 3 m may cause the safety factor to be 5% higher than that at locations with smaller cavity dimensions, although this value is not significant in geotechnical engineering.

This study investigates challenges of and solutions for estimating regional flood discharges from a limited number of stream gauge records with varying lengths. Routine application of standard analytical flood frequency analysis methods using available hydrologic data can lead to biased flood-hazard assessments. This paper shows that more analysis is needed for common flood-frequency methods (e.g., USGS regional regressions from at-site fitting, or regional frequency analysis using L-Moments) applied to the Weser River Basin ($\sim \mathrm{46,000}{\mathrm{km}}^2$) in northern Germany. Daily average discharges and the 10 largest instantaneous peak discharges were obtained for 19 stream gauges located along rivers draining areas more than $\mathrm{2,000}{\mathrm{km}}^2$. These discharge records are analyzed to identify temporal sampling biases and redundant gauges. The discharge data are processed further to (1) estimate instantaneous peaks from daily average discharges (the adjustments from daily to instantaneous discharges reach as high as 1.39 for gauges with smaller drainage areas); (2) eliminate unrepresentative data for cases in which reservoirs may have changed flow conditions, because mean annual maximum discharges were 21% lower after reservoir construction for the longest gauge record; and (3) pool the data to increase sample size (to 840 years of pooled data compared with 52–94 years for each of the gauges) for extreme value distribution fitting. The processed data are used for regional frequency analysis with L-moments to estimate flood discharges (i.e., 100-year and 500-year) for all river reaches draining more than $\mathrm{2,000}{\mathrm{km}}^2$ throughout the basin. These processing steps result in discharge estimates with differences ranging from $-31\%$ to 71% for 100-year discharges and from $-41\%$ to 107% for 500-year discharges compared with at-site estimates.