Journal of Hydrologic Engineering

In this study, the complete governing equations for unsteady open channel flow in fractional time-space are developed from the fractional continuity equation combined with the fractional motion equation, which is based on Newton’s second law of motion, by accounting for the acceleration terms in order to render physically interpretable hydraulic terms. Then the kinematic wave and diffusion wave approximations to unsteady open channel flow with physically interpretable terms in fractional time-space are developed from the fractional continuity and motion equations. Because the powers of all of the derivatives in the conventional governing equations of unsteady open channel flow in integer time-space are one, these conventional equations are derived as special cases of the fractional governing equations when the fractional powers in the fractional equations approach unity.

A new technique is proposed for calculating time-specific accumulative flux of surface runoff from a digital elevation model (DEM). The method incorporates ancillary datasets required to determine the depth of effective rainfall and a priori information on surface and underground storm sewer collection systems, permitting grid-based hydrologic analysis using a DEM. The diffusion wave equation is used to analyze temporal changes in surface runoff variables such as flow depth and velocity. Therefore, time-specific accumulative flux of surface runoff for each DEM cell, and for inlets and outfalls of the storm sewer collection systems, can be calculated and used to simulate the propagation of flood waves around storm sewers. The technique was applied to the Sangdong tailings dam in Korea that has both surface drainage and storm sewers. The proposed technique provides better time-specific hydrologic information on watersheds that contain both surface and underground storm sewer collection systems, which is omitted by the EPA’s Storm Water Management Model (SWMM) and the ArcGIS Hydrology toolset.

Sediment transport capacity of overland flows is crucial to understanding and modeling the erosion and deposition processes. The transport capacity for sediment mixtures is particularly complex because of the need for an algorithm to allocate the total available shear stress or stream power to individual size classes. The transport capacity for sediment mixtures may not even be unique, depending on whether the flow is eroding or depositing sediments unless all sediments are of a single size class. Using data from a set of flume experiments involving steep slopes and high sediment concentrations, a formula was proposed for the sediment transport capacity limited by available stream power and available accommodation space for sediments. Sediments used for the experiment were further sieved into five distinct size classes. The sediment transport capacity was measured for 25 flow-slope combinations to validate a new formula for individual size classes, and to propose and test a method to compute the sediment transport capacity for sediment mixtures from that for individual size classes. An independent set of experimental data was used to validate the new formula for individual size classes. The results showed that the formulation for sediment transport capacity was consistent for all size classes, and the sediment transport capacity was not as sensitive to the settling velocity as expected. With the effect of settling velocity moderated, the predicted and observed sediment concentrations at the transport limit showed excellent agreement with the Nash-Sutcliffe coefficient of efficiency (Ec) of 0.96 for individual size classes. The paper also shows that the transport capacity for sediment mixtures can be computed as an average of the sediment transport capacity for individual size classes weighted by the size distribution and moderated settling velocity. The sediment transport capacity can also be predicted using the average settling velocity for sediment mixtures for this set of experiments. An independent validation test showed that the predicted and observed sediment concentration agreed well for the experimental data with the coefficient of determination $R^2=0.85$, $Ec=0.71$, and a bias of approximately 25%.

The analysis of the dynamic response of aquifers to rainfall is a key issue for groundwater resource management. A data-driven evolutionary modeling approach, evolutionary polynomial regression, based on multiobjective optimization is used here in order to identify explicit equations that forecast groundwater piezometric levels as a function of past rainfall values and past measured values of groundwater table levels. This methodology is applied here to two aquifers located in the same climatic area of southeast Italy, representative of two completely different hydrogeological scenarios: a deep coastal karst aquifer and a shallow porous aquifer. An evolutionary polynomial regression approach using commercially available software returns highly reliable model that allow for describing the different hydrogeological behaviors of the two aquifers. These models can be used both for planning the management of groundwater resources and for obtaining new scientific insight about the aquifers, looking at the equations and at the variables identified by the software model.

This study presents the development of novel artificial neural networks (ANN) models for assessment of best management practices (BMPs) for controlling runoff and sediment yield from small agricultural watersheds. The ANN models integrate complex nonlinear effects of key climatic, topographic, drainage, and management characteristics and can evaluate BMP effectiveness without presumptions about their physical mechanisms or performance. Thirty-two ANN models were developed and tested. Penalty-related criteria and statistical model performance evaluation parameters were used to select the two models (one for runoff, one for sediment yield) with an optimum number of input parameters and hidden nodes. Event based monitoring data ($n=248$) at the outlet of seven watersheds (1.4–30.2 ha) in southwestern Wisconsin were used to train, validate, and test the models. Statistical parameters (e.g., ${\mathit{R}}^2=0.76–0.93$) suggested that the ANN models performed well. Sensitivity analysis for the BMP parameters showed that the runoff model was heavily influenced by length of grassed waterway and channel density; the sediment-yield model was mainly affected by upland crop type and tillage.

To better understand and monitor the hydraulic effects caused by clogging near groundwater production wells, a study was conducted in the unconfined sand aquifer used by the City of North Battleford, Saskatchewan, Canada. Detailed investigations were carried out near one production well between the spring of 2007 and the fall of 2008. Results from four pumping tests conducted over this time period indicated no significant temporal changes in hydraulic conductivity within approximately 2 m of the pumping well. In contrast, long-term monitoring of hydraulic head differentials did reveal increases with time, and suggested that clogging began to accelerate within a radius of 1 m of the well after about one year of continuous pumping. Continuous monitoring of hydraulic head in pumping wells and nearby piezometers, combined with breakpoint analyses of head differentials and specific capacity tests, will be more effective than standard pumping tests for detecting temporal and spatial trends in aquifer and well screen clogging.

Water flow through engineered channels is important for decision making given its close ties to availability for allocation. However, planners often rely on estimates for natural streamflow, then use stream-by-stream assumptions and aggregation to estimate allocatable flows rather than directly assessing flows through engineered channels. Further, synthetic flows based on historical records can be unreliable when parameter nonstationarity due to effects like climate change is likely. This case study of the Wailoa Ditch, a major engineered surface water supply system on Maui, Hawaii, uses a natural experiment based on Maui’s declining rainfall to demonstrate and validate that both problems can be addressed. For Wailoa, synthetic and actual flow characteristics differ by less than 5% when historical records are adjusted to reflect changing rainfall. Direct simulation of Wailoa’s flows reproduces modern conditions more accurately than stream-by-stream approximations. Precipitation-based scenario analysis suggests that under the influence of both decadal oscillations and climate change, Maui is far more likely to experience water supply shortages on its main engineered system in the future than in the past.

In the context of changing climate, this study assessed the effects of global warming on water resources in Tanzania for the mid-21st century. Climate projections from six global circulation models under the most extreme emission scenario (RCP 8.5) were used as inputs to the soil and water assessment tool (SWAT) to examine the effects. The results were analyzed both at spatial (country-level, watershed-level, and subbasin-level) and temporal (annual and seasonal) scales concerning potential and actual evapotranspiration, surface runoff, water yield, and soil moisture. At the country level, the results showed a 0.8–27.4% increase in annual precipitation, which led to a general increase in evapotranspiration ($-2.2–7.3\%$), surface runoff (12.6–94.1%), water yield (7.5–73.4%), and soil moisture (2.9–20.7%). Watershed-level analysis showed 2.4–31.5%, $-2.6–6.8\%$, 18.4–159.7%, and 3.2–22.8% changes in average precipitation, evapotranspiration, surface runoff, and soil moisture, respectively. While no distinct spatial trend was identified for evapotranspiration and surface runoff, soil moisture was projected to increase in the majority of the areas during both the wet and dry seasons. Probability analysis at the subbasin level showed that less than 15% of the area, during the wet season, experienced decline in precipitation, soil moisture, potential and actual evapotranspiration, surface runoff and water yield, while less than 86% of the country experienced decline during the dry season. Overall, the climate change was found to create a more favorable condition for agricultural production during the wet season by increasing the supply of water, but in the dry season, less favorable conditions were projected.

Sediment dynamics of lowland rivers are of importance in building resilient strategies to manage environmental change. Yet the effects of natural and anthropogenic disturbances on sediment dynamics are poorly understood. Here a low-frequency suspended sediment sampling data set is used to assess the spatial and temporal variations of suspended sediment fluxes in the River Thames (United Kingdom). Sediment rating curves (SRCs) were used to analyze both the spatial and the temporal variation of catchment-suspended sediment transport. SRC exponents for the River Thames were found to be between 0.21 and 1.13. The 95% confidence interval was also determined through a bootstrapping technique. The seasonal and interannual variability of SRC parameters were analyzed to reveal seasonal and secular changes. The results are used to quantify the seasonal flushing effect, in which suspended sediment concentrations are typically substantially higher during the first floods after the summer dry period. The suspended sediment concentrations of the River Thames during the first floods after summer are estimated to be around 1.5–2 times those of other floods, for a given water discharge. A decrease in the flushing effect which began in the 1990s is observed (around 50% of its original magnitude), which may be attributable to changes in catchment and river channel management.

Digital topographic data derived from airborne light detection and ranging (LiDAR) provide high-resolution information for characterizing the morphology of channels and riparian zones, and can be used to map the spatial distribution of channel incision resulting from flashy runoff response in small urban streams. Here the researchers present a methodology for extraction and analysis of transects throughout the drainage network in order to quantify longitudinal patterns of bankfull channel geometry and to identify the spatial distribution of channel incision. Owing to local gaps in LiDAR coverage and anomalies in channel and riparian zone morphology, the set of extracted transects are censored to remove those that would otherwise yield spurious results. Two procedures are then utilized to identify the bankfull stage along individual transects: (1) an iterative spline procedure based on minimizing slope angle and distance from the channel for points that exceed a minimum elevation threshold above the thalweg, and (2) a slope-difference algorithm that identifies local maximum values of the change in slope by evaluating a fixed distance backward and forward from each point on the transect. Each of these approaches generates an independent estimate of bankfull stage, depth, and width for individual transects. These are supplemented by a third estimate by a human operator. All three approaches are compared along a study domain with a deeply incised channel. Longitudinal trends in bankfull depth and width derived for all three approaches reveal that the major trends are detectable regardless of which method is used. The method can be used to identify differences in channel geometry between incised urban streams and less-incised streams.