Journal of Hydrologic Engineering

Surface depressions significantly influence hydrologic processes. In traditional semidistributed models, however, the entire area is connected to the outlet of a watershed and surface depressions are often lumped as a single depth to control runoff water release. As a result, hydrologic processes related to topographic characteristics of depressions cannot be directly simulated. The objective of this study is to quantify the impact of depressions on rainfall-runoff processes and the mechanism of dynamic water release from depressions. To achieve this objective, a new semidistributed depression-oriented hydrologic model (HYDROL-D) is developed. Unlike the traditional methods, a unique modeling framework is proposed in HYDROL-D to facilitate separate modeling for the depressional area (DA) and nondepressional area (NDA) of each subbasin. A DA is further divided into time-varying contributing area (CA) and ponding area (PA). A depression-dominated delineation (D-cubed) algorithm is utilized to provide specific characteristics of surface depressions (e.g., maximum depression storage and maximum ponding area at various depression levels for all subbasins), which are processed to define the characteristic areas (i.e., NDA, DA, CA, and PA), varying functions of CA and PA, and the hierarchical control thresholds for water release in HYDROL-D. The model was applied to a site in North Dakota, compared with the widely-used hydrologic modeling system Hydrologic Engineering Center, Hydrologic Modeling System (HEC-HMS), and calibrated using the observed data. The results demonstrated that because of the discontinuity caused by depressions, only a small portion of the study area contributed water directly to the outlet. The new modeling framework was able to effectively account for the dynamic processes associated with surface depressions. In addition, compared with the single threshold, multiple hierarchical control thresholds showed a better ability to reveal the mechanism of dynamic water release from depressions.

Watershed scale modeling is commonly used to evaluate nutrient fluxes to surface waterbodies and the appropriateness of regulatory intervention, such as the implementation of a total maximum daily load (TMDL). The Soil and Water Assessment Tool (SWAT) is a widely used watershed-scale model employed to simulate streamflow and nutrient fluxes, including phosphorus (P), from nonpoint sources, which has become ubiquitous throughout nutrient management legislation in the US. A number of studies have shown that manure spread on fields has high potential for nutrient loss in surface runoff directly following application. The authors evaluated the influence of the seasonality of hydrologic fluxes on P losses from dairy manure spread fields during this especially labile period across three mechanistic soil P loss models: SWAT v2012, SWAT v2016, and JoFlo. Calibration was performed using a DDS algorithm for SWAT v2012 and JoFlo and results of the SWAT v2012 calibration were applied to SWAT v2016. Calibration results were non-ideal; however, we propose that our methodology appropriately isolates the processes of concern and furthers the debate on what remains a critical open question. SWAT v2016 and JoFlo show a strong influence of seasonality with respect to P losses from manure-spread field, in agreement with recent field-scale studies. SWAT v2012, commonly utilized for establishing nutrient management policy, demonstrated a limited influence of seasonal hydrology on P losses, suggesting issues in the underlying mechanisms driving labile P mobilization. Our study reinforces the idea that development of hydrologic model structure should strive to parallel the intent of environmental policy.

The emergent hydrologic and land-use controls of coastal-urban stream water quality were evaluated by using similitude and dimensional analysis, considering southeast Florida a prototype of growing coastal-urban environments. The goal was to test a fundamental hypothesis that the coastal-urban stream water quality processes represent emergent ecohydrological-biogeochemical similitudes (parametric reductions). The in-stream total nitrogen (TN), total phosphorus (TP), and biomass ($\mathrm{Chl}a$) were normalized by their immediate upstream reach concentrations to formulate the dimensionless numbers of $\mathrm{TN}/{\mathrm{TN}}_0$, $\mathrm{TP}/{\mathrm{TP}}_0$, and $\mathrm{Chl}a/\mathrm{Chl}{a}_0$. Stream dissolved oxygen (DO) was normalized by its saturated concentration (${\mathrm{DO}}_{\mathrm{sat}}$) to obtain the dimensionless $\mathrm{DO}/{\mathrm{DO}}_{\mathrm{sat}}$ number—avoiding a misleading scaling by upstream concentrations in the presence of a DO sag phenomenon. The emergent controls of stream water quality were represented by a small set of dominant driver dimensionless numbers. For each water quality indicator, nine original variables (including predictors and response) were reduced to three to four important and mechanistically meaningful dimensionless numbers. The hydrologic control number (role of watershed hydrology versus the external Everglades) and salinity number (ratio of downstream to upstream salinity) exhibited the key controls on stream $\mathrm{TN}/{\mathrm{TN}}_0$ across the wet and dry seasons. In contrast, the land-use number (ratio of agricultural plus vegetated lands to built lands), hydrologic control number, and salinity number dominated $\mathrm{TP}/{\mathrm{TP}}_0$ and $\mathrm{Chl}a/\mathrm{Chl}{a}_0$ incorporating the two seasons. However, $\mathrm{DO}/{\mathrm{DO}}_{\mathrm{sat}}$ was controlled by the hyporheic exchange number (role of watershed groundwater versus surface hydrology) and land-use number in the wet and dry seasons, respectively. The formulated similitude and dimensionless numbers provided important insights and understanding that may help achieve healthy coastal-urban streams.

Constructing hydrological models for large urban areas is time consuming and laborious due to the requirements for high-resolution data and fine model detail. An open-source algorithm using adaptive subcatchments is proposed to automate Storm Water Management Model (SWMM) construction. The algorithm merges areas with homogeneous land cover and a common outlet into larger subcatchments, while retaining small-scale details where land cover or topography is more heterogeneous. The method was tested on an 85-ha urban catchment in Helsinki, Finland. A model with adaptive subcatchments reproduced the observed discharge at the catchment outlet with high model-performance indices emphasizing the strength of the proposed method. Computation times of the adaptive model were substantially lower than those of a corresponding model with uniformly sized high-resolution subcatchments. Given that high-resolution land cover and topography data are available, the proposed algorithm provides an advanced method for implementing SWMM models automatically even for large urban catchments without a substantial manual workload. Simultaneously, the high-resolution land cover details of the catchments can be maintained where they matter the most.

In this study, the hybrid wavelet-M5 model was introduced to model the rainfall-runoff process via three different data division strategies (75%–25%, 60%–40%, and 50%–50%) for two different catchments at both daily and monthly scales. The performance of the wavelet-M5 model was also examined in the case of multi-step-ahead forecasting. In this way, first, the rainfall and runoff time series were decomposed using the wavelet transform to several sub-time series to handle the multiresolution characteristic of rainfall and runoff time series. Then the obtained subseries were applied to the M5 model tree as inputs. The obtained results showed the better performance of the wavelet-M5 model in comparison with individual artificial neural network (ANN) and M5 models so that the obtained determination coefficient was 0.80 by the hybrid wavelet-M5 model, while it was calculated as 0.23 and 0.19 by the ANN and M5 tree models, respectively. It was also concluded that the wavelet-M5 model could lead to better performance in the multi-step-ahead forecasting issue since the catchment showed a semilinear behavior because the error would be constant in linear models.

Drought occurs across all climates worldwide, but has different impacts from one region to another. Because water is the most basic of human needs, drought and water shortage can result in severe socioeconomic difficulties that take years to overcome. While drought has traditionally been classified on the basis of deficiencies in precipitation, with the advent of climate change and the increase in evapotranspiration that is occurring, efforts have been made to focus instead on derived soil moisture anomalies. A less recognized factor modulating soil moisture anomalies is the use of engineered water supply, through canals or river diversions, and also through groundwater pumping. This paper attempts to formulate a multivariate metric of sustained water stress severity that attempts to incorporate these additional sources of water supply in engineered subbasins. The proposed index uses entropy theory as the basis of its formulation, and was compared against traditional indexes, and the differences were noted. Remotely sensed vegetation products were used as an independent means of assessing the validity of the proposed index. It was found that the proposed index collapses to the more traditional drought indexes in subbasins without significant unnatural water supply, while presenting a more accurate picture of water stress in subbasins where engineered water supply is present, as interpreted by the vegetation (or its lack thereof) the water supply is able to sustain.

The soil water content usually appears to be distributed in distinct layers during the process of homogeneous soil infiltration. However, existing infiltration models rarely consider the partitioning characteristics of the wet zone. The wet zone is typically assumed to be divided into a saturated zone and transient unsaturated zone during infiltration. Here, variations in the depths of the saturated zone and transient unsaturated zone during the infiltration process were studied using commercially available software. The wet zone partition function and its fitting parameters are presented. This function can characterize variations in the ratio of the saturated zone depth to the wetting front depth over time. The fitting parameters were predicted by a back-propagation (BP) neural network. The improved Green-Ampt model was established by introducing the established wet zone partition function. The infiltration results simulated by the improved Green-Ampt model, the software, and the traditional Green-Ampt model were compared with the measured values. The consistency between the measured and calculated values of the three models decreased from the improved Green-Ampt model to the software model to the traditional Green-Ampt model. The wetting front depth and the infiltration rate calculated by the traditional Green-Ampt model were larger than their measured values, which means that the effect of the transient unsaturated zone during the process of infiltration must be considered. The improved Green-Ampt model can provide analytical solutions and accurately simulate the infiltration process, which is beneficial for engineering calculations.

The Nonparametric Multivariate Standardized Drought Index (NMSDI) based on precipitation and soil moisture data in conjunction with copula functions is of primary concern in this study. We are the first to investigate bivariate return periods of the NMSDI using the two typical drought characteristics (duration and severity) at 10 stations in Konya Closed Basin (KCB) in Turkey. As a result, lognormal and log-logistic distributions were identified as the most suitable distributions for drought duration and severity series according to five commonly used goodness of fit tests. Various types of copulas were considered in modeling the joint dependence between duration and severity series at each station. Our results from the five goodness of fit tests and tail dependence assessments showed that BB6, BB7, and BB8 copulas outperformed the joint modeling of duration and severity series in the KCB. The bivariate return period analysis revealed a high risk for southeastern and southwestern regions in the KCB for the 3-month NMSDI series while north to northwestern regions could be exposed to high risk for the 6-month series.

This study aimed to estimate water quality–based minimum environmental flow (Eflow) of a river under the impact of different plausible scenarios. The water quality model QUAL2K was deployed to simulate levels of dissolved oxygen (DO) and biochemical oxygen demand (BOD) in a river during the dry season. Hypothetical scenarios of pollution were generated by varying BOD in contributing drains, and climate change scenarios were generated by altering air temperature over the study region. DO and BOD levels were simulated under the impact of various scenarios. Corresponding to each scenario, minimum Eflow throughout the river stretch—that is, the headwaters flow (quantity and quality) meeting desirable river water quality standards (as proposed by the Indian Central Pollution Control Board)—was explored. Water quality charts were developed for direct and user-friendly estimation of Eflows. The proposed approach was applied to the Yamuna and Bhadra rivers in India. The results reveal the inefficiency of existing river flow conditions in maintaining permissible water quality standards. Adverse effects of pollution load, upstream diversions, and climate change are highlighted.