Journal of Hydrologic Engineering

Channel geometry can affect the performance of the autoregressive (AR) model on river stage correction. To evaluate the effect, three ideal models were established by using the one-dimensional (1D) hydrodynamic model with imagined sets of data in assumed channels (i.e., rectangle, V-type, and complex type). To evaluate the performance of the AR model on river stage and discharge correction in a real system with unavailable observed discharges, an equivalent conversion technique by using the Manning’s formula was proposed to convert stages into equivalent conversion discharges. The results from the ideal models and real systems in the Fuchun River show that the AR model performs better on river discharge correction than on river stage correction, which is also verified at Shaowu station in the Minjiang River with available observed discharges. It indicates that the correlation of discharges is better than that of river stages. To transform corrected discharges to corrected stages, a transformation technique based on an assumed rating curve calculated by the Manning’s formula was employed. Furthermore, by combining the equivalent conversion technique, the AR model and the transformation technique, an equivalent conversion correction (ECC) method is proposed to improve the AR model performance on river stage correction. The results show that the ECC method can obtain better accuracy on river stage correction than the AR model.

Hydrologic Engineering Center-Hydrologic Modeling System model is used to estimate the water storage in the lakes of three dams due to rainfall events in three wadis located in the northern area of the United Arab Emirates (UAE). Like other arid and semiarid regions, rainfall events in the three wadis are limited, scattered, and random. For model calibration, the simulated results were compared with the observed water storage data for several storm events. A family of rainfall-runoff/storage curves was developed based on the duration and intensity of rainfall events. These curves can be used for prediction of surface water runoff in the three wadis and water storage in the dams in response to different rainfall events. The same rainfall event in wadi Bih generates almost twice as much of the surface water runoff generated in each of wadi Ham and wadi Tawiyean. This is mainly attributed to the large catchment area of wadi Bih as compared to the other two wadis. The sensitivity analysis revealed that the amounts of water storage in dams are highly correlated to the assigned curve number especially for the cases of wadi Bih and wadi Tawiyean.

Watersheds are featured by a variety of hydrological, meteorological, and ecological characteristics. Complexity and uncertainty are usually two major challenges during watershed classification which is one of the key processes in hydrological modeling. This study aims to develop an integrated rule-based fuzzy adaptive resonance theory mapping (IRFAM) system, by incorporating fuzzification and rule-based operation to more efficiently handle the complexity and uncertainty. The developed system has been tested with a case study conducted in the Deer River watershed in Manitoba, Canada. The results are further compared with the ones generated by the conventional adaptive resonance theory mapping (ARTMap) method. All subbasins are classified by IRFAM while ARTMap leaves five subbasins unclassified. Furthermore, another nine subbasins in the juncture between classified groups from the ARTMap classification results are relocated by IRFAM. The IRFAM system can take advantage of fuzzy set theory to generate full criteria combinations to match the input patterns and use the rule-based operation to screen the matched patterns into the target groups. Therefore, the developed system can effectively process the classification for the input patterns with a high degree of uncertainty and wide range in variations. Furthermore, the IRFAM can effectively help resolve the traditional difficulty in criteria generation, which is always affected by uncertainty due to insufficient references and historical records, and complexity occurs due to multifeatures. The improvement of classification efficiency and robustness will be directly beneficial to hydrological modeling and related watershed management which rely on classification results.

Past research shows that physically based distributed hydrological model has the advantage of better representing the basin characteristics and the hydrologic processes to potentially simulate/predict river basin flood. But how to physically derive model parameters directly from terrain data and to acquire channel cross-sectional size are still difficult jobs in physically based distributed hydrological modeling that prevented its operational application in river basin flood forecasting. To deal with these challenges, this paper first presents a physically based, distributed hydrological model for river basin flood forecasting/simulation, called the Liuxihe model. Then a method for estimating channel cross-sectional size was proposed that utilizes a readily accessible public data set acquired by remote sensing techniques, which could be employed by other physically based, distributed hydrological models also. Finally, a method for deriving model parameters was proposed that adjusts model parameters with initial model parameters derived directly from terrain data that is completely different from parameter calibration in lumped model, which also can be used in other physically based, distributed hydrological models. A medium-sized river basin in southern China was tested with the above model and methods, and 13 collected flood events were simulated with reasonable model performances. It can be concluded that parameter adjustment is still necessary and vital to improve physically based, distributed hydrological performance as there is no systematic and global referencing in deriving model parameters from terrain data. For the Liuxihe model, the highly sensitive model parameter is water content at saturation condition; the sensitive parameters are water content at field condition, river channel Manning’s coefficient, soil thickness, soil porosity characteristics, soil hydraulic conductivity, and hillslope Manning’s coefficient; and the less sensitive parameters are potential evaporation, evaporation coefficient, underground water recession coefficient, and water content at wilting condition. The results also showed that the method proposed for estimating river channel cross-sectional size is reasonable and could be applied widely, and the proposed Liuxihe model works well in simulating river basin flood events, thus presenting further evidence of the potential use of the physically based distributed hydrologic model for the operational use of simulating/predicting river basin floods.

There is a strong analogy between fractal geometries and scale invariant processes. Fractal geometries are self-similar at different scales. Similar to fractal geometries, solutions of scale invariant processes at different space-time scales are self-similar. This unique property of scale-invariant processes can be employed to find the solution of the processes at a much larger or smaller space-time scale based on the solution calculated on the original scale. Here, we investigate scale invariance properties of hydrologic processes as initial-boundary value problems in one-parameter Lie group of point transformations framework. Scaling (stretching) transformation has unique importance among other Lie group of point transformations, as it leads to the scale invariance or scale dependence of a process. Scale invariance of a process allows using the same mathematical model for the process at different scales and facilitates finding the solution at any scale using the solution at the original scale. In this study, the process parameters and source/sink terms are regarded as state variables of some (secondary) processes that underlie or couple with the original process. Then under the scaling transformations, the invariance conditions for the resulting system of processes at time-space scales that are different from the original time-space scales are investigated. The conditions to be satisfied by the form of a governing equation and its parameters, as well as the initial and boundary conditions of the process, are established in order for the process to be scale invariant. Also, the self-similarity of the solution of an invariant process is demonstrated by various numerical example problems.

Due to an increase in deforestation and construction activities in the Himalayan region, there is an increase in sediment laden flow in many rivers, such as the Ganges, which originate from the Himalayan region. Many cities, located along these rivers, withdraw river water through abstraction wells located close to river banks and it is of growing interest to know how the sediment load increase in the source water may affect the turbidity, also a measure of sediment load, in the abstracted water. With this in mind, a study has been carried out to assess the variation of turbidity in the source water and the abstracted water at a river bank filtration site in Haridwar, located on the bank of the River Ganges and at the foothills of the Himalayas, for the period 2005–2006. It is found that the turbidity removal coefficient varies with the natural logarithm of influent concentration of the source water. A theoretical development, based on the use of the Sakthivadivel and Einstein model, is also provided to justify the removal coefficient’s logarithmic variation with the influent concentration.

With an increasing tendency of the flood disaster frequency and intensity in China, risk analysis of design flood volume for the Three Gorges Dam Project (TGP) should be of paramount importance for about one-half of the population and gross domestic product in China. How to predict the floods of Yangtze River with high precision is always a difficult problem, for the vast drainage basin and numerous tributaries. After a brief introduction of extreme value distribution and compound extreme value distribution, the TGP design flood volume predicted by the Chinese Hydraulic Design Code (CHDC) recommended annual maximum data sampling based Pearson type III extrapolation method is compared with the stochastic simulation technique based on the important sampling procedure (ISP) and the multivariate compound extreme value distribution (MCEVD), respectively. Comparison results show that the 100-year joint return period of the 3-day flood volume predicted by ISP and MCEVD is greater than 1,000-year return volume predicted by CHDC recommended method.

A study was conducted to enhance the historic regression relations that predict the peak discharge from breached embankment dams. Forty-four dam breach case studies were collected and added to an existing database resulting in a composite database of 87 case studies. The composite database was evaluated and a statistical analysis performed using regression techniques. Peak outflow $\left(Q_p\right)$ prediction expressions from breached embankment dams were developed as a function of the height of the dam $\left(H\right)$ , the volume of water behind the dam $\left(V\right)$ , the embankment length $\left(L\right)$ , the average embankment width $\left(W_{\text{ave}}\right)$ , and a combination of these variables. The multivariate regression analysis indicated that a series of expressions may be formulated relating peak outflow as a function of $H\cdot V\cdot L$ and $H\cdot V\cdot W_{\text{ave}}$ . The newly developed expressions derived from the expanded database appear to reduce the conservatism in predicting the peak discharge from a breached embankment, reduce the prediction error, and reduce the uncertainty bandwidth while improving the prediction correlation. The available data are limited and the quality of the composite database is quite variable. The study results strongly suggest that the art and science of dam breach forensics (i.e., data acquisition, quality, and availability) has not changed since inception and must be improved to enhance regression prediction credibility.