Journal of Hydrologic Engineering

Heavy rainfall-induced inundation is becoming more serious in urban areas making it necessary to urgently appraise and redesign the infrastructure system to drain the storm water more efficiently. For the complex urban drainage systems that include street, sewer, and ditch/river networks, a sophisticated urban drainage model is required to facilitate optimal planning and management. Although most existing models simulate the runoff generation process in a simpler manner and treat the drainage connections between runoff generation cells and the corresponding drainage links in a rigid/static manner, this study proposes a new hydrologically enhanced distributed urban drainage model. In the new model, the urban area is discretized into a four-layer network, i.e., two-dimensional (2D) grid network, 1D street network, 1D sewer network, and 1D ditch/river network. Physically based equations are utilized to describe water movement along the four networks, i.e., the 1D Richards equation is used to simulate the infiltration process along the vertical direction in the grid network, the 2D Saint-Venant equation is used to simulate the overland flow process along the planar direction in the grid network, and the 1D Saint-Venant equation is used to simulate the street, sewer, and river flows in the remaining networks. The new model incorporates the state-of-the-art physical descriptions about hydraulic and hydrological processes during the urban storm inundation period, which allows a more realistic depiction of the runoff generation processes, automatic alteration of overland flow routing path, direct and easier usage of gridded radar rainfall data readily available recently, and real-time hydrodynamic flux exchanges between surface and sewer pipes. The model is validated in a hypothetical application by comparing with the published literature results. Also, a real urban watershed application shows the capacity of the model to provide reasonable predictions of the outlet hydrograph, which indicates its potential for planning and real-time management of urban drainage systems.

As it is a critical process of watershed management, classification always faces challenges of inefficiency in handling complexity and uncertainty. This study attempts to fill this gap by developing an integrated adaptive resonance theory mapping system consisting of a two-stage adaptive resonance theory mapping (TSAM) approach and an integrated rule-based fuzzy adaptive resonance theory mapping (IRFAM) approach. To demonstrate their feasibility and efficiency, TSAM and IRFAM were compared with conventional adaptive resonance theory mapping (ARTMap) in two case studies in the Deer River watershed in Manitoba, Canada, which were classifications of watershed subbasins and types of land-cover to support hydrological modeling. Among the three approaches, IRFAM performed best in effectively processing the classification for input patterns with a high level of uncertainty and a wide range of variations, although it required predefined criteria. TSAM performed reasonably well by generating criteria for supervised classification on the basis of the internal relationship of the original data, indicating its advantage in handling an insufficient data situation with a low demand for subjective judgment. Consequently, the two developed approaches can be complementary and improve classification efficiency and robustness in dealing with systematic complexity and uncertainty and supporting watershed hydrological modeling.

A dynamic subtime-stepping method is described for solving contaminant transport problems that utilize higher-order implicit time-marching procedures with higher-accuracy nonoscillating spatial-discretization methods to resolve sharp-plume fronts in advection-dominated systems. Nonoscillating spatial-discretization methods for the advective term prevent unphysical oscillations and minimize numerical diffusion. Second-order temporal accuracy is achieved by using the Crank-Nicholson implicit scheme, however nonoscillating properties may be violated for Courant Numbers larger than one, leading to spurious oscillations. The proposed subtiming method allows use of small time-step sizes in critical portions of a domain, with larger time-step sizes in other locations. This locally limits the Courant number where required and still keeps the general solution free of time-step size restrictions. This technique makes it possible to apply higher-order nonoscillating schemes with higher-order temporal weighting for advection-dominated flows, even when the Courant number is much greater than one. Feasibility and applicability of the dynamic implicit subtiming method are demonstrated through three proof-of-concept example problems.

Soil moisture is an important component in hydrologic and meteorologic processes. Remote sensing devices, such as satellite radiometers, are useful tools to obtain soil moisture information over a large region. However, effective “ground truth” calibration data for satellite sensors are lacking. This paper presents a new approach on the basis of rainfall and river runoff hydrologic data to estimate satellite-based soil moisture. The catchment water storage change from the rainfall and river runoff has a strong link with the soil moisture information and the parameterization of the surface roughness parameters $h$ and $Q$. The proposed methodology is tested at the Brue catchment in southwest England. Two years of satellite data are used for calibrating and retrieving surface soil moisture, and 1 year of data is used for validation. This study indicates that the estimated daily soil moisture from satellite correlates well with the flow observations after applying the new calibration method, and good agreement ($R^2=0.74$) is shown between the water storage change and the soil moisture change. The results indicate that this new scheme could be a useful approach in improving the satellite soil moisture retrieval for hydrologic applications.

A diffusion wave equation, which is derived from the Saint-Venant equations for one-dimensional, gradually varied, unsteady open-channel flow, describes the wave propagation in open channels. Therefore, it is important to solve the diffusion wave equation accurately and efficiently. In this paper, a numerical solution for a diffusion wave equation is developed by using the Differential Quadrature Method (DQM). The performance of DQM is tested against two other numerical solution methods, the finite difference method (FDM) and finite volume method (FVM). In order to demonstrate the applicability of DQM, first a hypothetical example is solved with both DQM and the two other numerical methods. Then, the DQM is applied to a real flooding event that occurred in Aksu River, Sutculer Basin, located in Mediterranean Region, Turkey. The measured flow rates are routed through the Aksu River by the diffusion wave equation and the outflow is obtained by DQM. Finally, this flood event is also solved by explicit and implicit approximations of FDM and FVM, and the results are compared to the solution obtained by using DQM. Based on the comparison of the results, it is concluded that DQM provides results close to those obtained using FDM and FVM but with higher computational speed fewer nodes and less memory usage.

In this paper, the wavelet analysis was linked to the genetic programming (GP) concept for constructing a hybrid model to detect the seasonality patterns in the rainfall–runoff time process. This approach was used to determine the dominant input variables of an artificial neural network (ANN) rainfall–runoff model via a sensitivity analysis. In this way, the main time series of two variables, rainfall and runoff, were decomposed into some multi frequency time series by the wavelet transform. Then, these decomposed time series were imposed as input data to the GP to optimize the input structure of ANN model. This methodology was utilized in daily and monthly timescale modeling for two watersheds with distinct climatologic regimes. The obtained results were compared favorably to ANN and GP models. The obtained results showed that the proposed model can monitor both short and long term patterns due to the use of multiscale time series of rainfall and runoff data as the GP inputs. Moreover, using the proposed sensitivity analysis, the number of input variables in the ANN modeling was decreased.

The coincidence of flood flows of the mainstream and its tributaries may determine flood peaks. This study analyzed the risk of flooding as a result of such flood coincidences by considering flood magnitudes and time (dates) of occurrence. The Pearson Type III (P3) and log-Pearson Type III (LP3) distributions were selected as the marginal distribution of flood magnitude for annual maximum flood series; the mixed von Mises distribution was selected as the marginal distribution of flood occurrence dates. Two four-dimensional (4D) copula functions were developed for the joint distribution of flood magnitudes and occurrence dates. The upper Yangtze River in China and the Colorado River in the United States were selected to evaluate the method of computing risk. The coincidence probabilities of flood magnitudes and dates were calculated, and the conditional probabilities for the Three Gorges Reservoir (TGR) were analyzed. Results show that the von Mises distribution can fit the observed flood dates data well. The X-Gumbel copula was selected for risk analysis. On the basis of the proposed model, the coincidence and conditional probabilities for any return period were obtained.

Previous studies have reported limitations of the efficiency criteria commonly used in hydrology to describe goodness of model simulations. This study examined sensitivity of model performance to the objective function used during automated calibrations. Nine widely used efficiency criteria were evaluated for their effectiveness as objective function, and goodness of the model predictions were examined using 13 criteria. Two cases (Case I: Using observed streamflow data and Case II: Using simulated streamflow) were considered to accomplish objectives of the study using a widely used watershed model (SWAT) and good-quality field data from a well-monitored experimental watershed. Major findings of the study include (1) automated calibration results are sensitive to the objective function group—group that work based on minimization of the absolute deviations (Group I), group that work based on minimization of square of the residuals (Group II), and groups that use log of the observed and simulated streamflow values (Group III)—but not to objective functions within the group; (2) efficiency criteria that belong to Group I were the most effective when used as objective function for accurate simulation of both low flows and high flows; (3) Group I and Group II objective functions complement each other’s performance; (4) with regard to the capability to describe goodness of model simulations, efficiency criteria that belong to Group I showed superior robustness; (5) for the study watershed, use of the long-term interannual calendar day mean as baseline model did not improve capability of an efficiency criterion to describe model performance; and (6) even for ideal conditions where uncertainty in input data and model structure are fully accounted for, identifying the so-called global parameters values through calibration could be daunting as parameter values that were significantly divergent from predetermined values produced model simulations that can be considered near perfect even when judged using multiple efficiency criteria.