RIVER ENGINEERING USING INTEGRATED METHODOLOGY
The
Mankind has improved their living conditions by their ability to utilize the natural resources around them. It has been obvious that the natural resources from the rivers have been utilized by mankind ever since the beginning of the civilization. The extent of their benefits has been expanded continuously. The successes of river engineers have been due to their capability to sharpen their tools which have enabled them to better understand the rivers and thus, to harness more and more natural resources from the rivers.
In the very early stage of our history, the great philosophers could find the rules and trends of natural phenomena by their systematic observations of nature without the benefits of sophisticated instrumentation. Even though they couldn’t figure out quantitatively “why” a natural phenomenon would vary in a specific way; but, the rules, laws and principles they have established by qualitative observations, have enhanced our basic knowledge of natural phenomena significantly. We could approximately “predict” how a nature phenomenon would vary under certain conditions (or forcing). As the civilization advances, mankind’s demand on natural resources (including rivers’) has increased exponentially so has the concern about the conservation of environmental/ecological quality. In order to do their job efficiently, research engineers and applied scientists have devoted a great deal of effort in the advancement of the research tools.
The continuing pursuit of better understanding of natural phenomena and more effective utilization of natural resources (while upgrading the ecological and environmental quality), has driven mankind to advance the forefront of sciences as well as to sharpen the tools for conducting research and engineering throughout the history of human civilization. By careful measuring of natural systems quantitatively with instrumentations and describing them by mathematics, physical laws and principles guiding the variation trends of natural phenomena have been developed quantitatively by laboratory experimentations and validated by field observations. These laws have significantly enhanced our knowledge of the behavior of the nature, with varying degree of idealization.
With the advancement of mathematical modeling, analytic solutions of mathematical models of highly-simplified natural systems have provided us with the predictive capability of systems’ responses to various external forcings, under some idealized assumptions.
In the meantime, physical modeling methodology has been conducted in the laboratories by welldesigned experiments utilizing the latest and sophisticated instrumentations. It has produced more forcing-response relations, commonly called empirical laws, which have also provided us with additional predictive capabilities. Because not all idealization, linearization, and simplifying assumptions needed by mathematical models are required to develop the empirical laws by physical modeling, the physical modeling approach has been considered as more realistic. Consequently, they have been adopted by engineers to solve real-life problems more often than the highly idealized mathematical models. For many decades, engineering designs have relied on the scaled physical modeling to confirm the effectiveness of conceptual designs of experienced engineers. In fact, even at the present, a few less involved problems are still being studied or designed by this physical or scaled modeling methodology. This has been especially true in the investigations of river sedimentation problems, which are much more complicated than the river flow problems; thus more difficult to model mathematically.
The need of investigating realistic physical systems more accurately has accelerated the development of numerical modeling methodologies. With the aid of modern digital computers, more and more realistic physical systems can be simulated quickly. Details of the numerical modeling are to be presented in the next section.
In modern engineering design, one often needs to conduct a parametric trade-off analysis in order to identify a set of optimal design parameters, which is the “best” compromised choice to achieve the pre-selected multiple objectives while satisfying all the required design and other constraints. Some of these objectives and constraints are cost-effectiveness, safety of the community, protection of the ecological and environmental quality, and being acceptable by the laws, government, and people of the society, etc. If physical modeling is used alone to conduct such an investigation, numerous combinations of a large set of design parameters must be studied sequentially to determine their short- and long-term effects. Such a study would often be too costly in terms of funds and time. A more cost-effective approach is a well-planned integration of physical models, numerical models, and field measurements.
With advanced numerical solution methodologies and powerful computer technology, the mathematical models of nonlinear, realistic hydrosystems have been solved. Additional computational and information science and technologies such as GIS (Geographic Information System), GUI (Graphic User Interface), Scientific Visualization, Virtual Reality Display and Imaging, etc. have been applied to the development of pre- and post-processors of computational simulation models. The resulting software packages have been not only powerful and userfriendly, but also efficient and cost-effective.
As a result, more and more computational simulation models have been developed and applied to the river engineering research and design. Lately, some of these models have also been applied to support the decision making in planning and management. It is not surprised that computational modeling has been the methodology of choice by river engineers recently. It has been even called the research and engineering tool of the 21st century.
The lack of quality control of some of the ever increasing number of computational models on the market has caused serious concerns about their accuracy and reliability, however.
All numerical models should be mathematically verified first by analytic solutions linear or nonlinear. Once the model is proven mathematically correct, it should be evaluated through a series of physical process validation, which is based on laboratory experimental results for the purpose of determining whether the model is capable of reproducing the basic physical processes relevant to the real-life problems to be studied by the selected model.
Before a mathematically correct and physically capable numerical model is to be applied to an investigation of a real-world problem, one more step, the Application Case & Site-Specific Validation, is required. This step is to begin by using an appropriate portion of field data collected at the study site to calibrate the site-specific values of the model parameters. Then, the calibrated model is used to predict field characteristics and processes under prescribed forcing. The predicted results are then compared with those measured data at the same site under the same forcing conditions. If a reasonable agreement between the model simulations and the field measurements is achieved, the numerical model is validated for the application to the study case at the specific site. During the valuation test(s), it is very important that the calibrated model parameters cannot be changed or tuned. One is not recommended to apply this model to the study of a similar problem at a different site, nor a different case at the same site, especially at a significantly lapse of time, e.g., a few years later or for a different event.
To conduct an application case and site-specific validation successfully, one must have a sufficient amount of high quality field data collected at well-designed locations with proper spatial distribution. Due to the fact that the numerical model represents an idealized and simplified system of the real-life problem, it is not expected that the numerically simulated results and the field measurements are in perfect agreement. The more important or useful results are the trend of spatial and temporal variations of the natural systems under specific forcings being studied. Therefore, one should make certain that the trends of field property variations predicted by the computational model are reasonably accurate and reliable.
Today, to carry out river engineering projects, one needs to take many factors into account before finalizing the comprehensive plan. These factors include hydrology, climate change, hydrodynamics, structural mechanics, geo-technology, sediment transport, morphology, pollutant transport, water quality, ecology and environmental quality, cost-effectiveness, effects on economy, society, and stakeholders, among others. In addition, the project’s effects on connected areas within a watershed or river basin, and its impact during both short- and longer-period of time must be considered as well. It is no longer acceptable to solve a river engineering problem in a small area temporarily, because from the lessons learned in the past this localized and temporary solutions by the lowest-cost bidder, have brought more problems and adverse impact to the affected larger region for a long time. As a result, it has cost much more to correct the mistakes.
Therefore, a comprehensive, multidisciplinary study has to be conducted, before the “best” compromised plan is achieved, in order to satisfy a large number of objective functions as much as possible and in the meantime, meet the required constraints. Due to the fact that some of the objectives and/or constraints are mutually contradictory, sophisticated optimization methodologies have to be applied to this complicated decision-making process. Take a simplified river engineering project planning for example, before making the final decision, the project managing engineer needs to make sure that at lest the engineering design shall:
• meet the required technical specifications based on structural mechanic and river hydrodynamic considerations,
• satisfy the constraints of ecologic and environmental quality,
• meet the safety standards required during extreme events climatically, hydrologically and geologically,
• bring no unacceptable effects to the connected regions within the same watershed or river basin in the foreseeable future,
• meet the constraints demanded by the stakeholders affected, or at least acceptable by them,
• be cost-effective including not only the initial construction, but also the routine operation and long-term repairment and maintenance,
• be acceptable by the society including governments, social and cultural groups, etc, and
• meet other specific requirements important to the site, but too numerous to list comprehensively.
To carry out this decision-making task of river engineering project planning and design one needs to apply a large number of computational models. Examples are:
• models for free surface flows, river bend aggradation degradion, bank erosion and failure, river course migration and meandering, morphological changes, etc.
• models for river bank erosion, local scours, channel migration, and instream structural designs, and their sustainability,
• models for water quality, pollutant/contaminant transport with fate or rate analyses of chemical and biological reactions,
• models for impact assessments to the health of ecologic systems and environmental quality,
• models for cost analysis for initial construction, routine operation and repairs and longterm maintenance,
• other models may be needed to meet additional site-specific requirements, and
• computational models for multi-objectives and multi-constraint optimization and compromised optimization are also needed to provide decision support in making river engineering planning and design decisions.
The
Before an integrated model or its module is applied to the study of a real-world problem in a natural river, one needs to calibrate the model parameters to insure that they are capable of representing the site-specific characteristics unique to the study site. Because there are no two sites of natural rivers that are the same identically, the calibration is key to insure the model use can indeed predict the response of the river consistent to the physical observation. After the model parameters have been calibrated, the model is ready to be validated by comparing the model simulated results with additional data measured at the study site. If the agreements are reasonable, especially between the trends of field property variations, the model is finally validated for conducting the field studies of the river.
This is the approach, which has been used
to complete several river engineering studies successfully. Some typical
examples are: the Mississippi River, Victoria Bendway
submerge weir system design for enhancing the navigatibility
(Jia/Wang, 2002; Xu/Wang/Jia
2003), the
The
With the confirmation of CCHE3D model’s acceptable accuracy and reliability as well as its capability in simulating real-life problems in natural environment, it is selected to evaluate the technical effectiveness of a variety of conceptual designs of the weir system. In order to obtain the design tool for applications to general river bendways, a simplified bendway was constructed in a laboratory (fig. 5). The model’s site-specific parameters were re-calibrated by the datameasured.
In this case, the experience from numerical simulations was used to design the data collection locations and the grid-resolution, to reduce the data collection costs without compromising the quality and quantity of the data collected for conducting the calibrations and validations of the model for this study. By varying the weir length, height, and its direction relative to the flow, the distance between weirs and the total number of weirs required to maintain the safety of navigation, and the curvature of the bendway, a large number of flowfields were obtained with significant savings of funds and time. From statistical regressions a large number of empirical relations have been established, from which one can identify the most favorable weir length, height, angle, etc. under each special case. Finally, by an optimization analysis, considering all relevant design parameters, a set of compromised design parameters can be determined to result a “best” possible design based on hydrodynamics only, which is similar to the one shown in fig. 6 The comprehensive design requires one or more step, i.e. to conduct a multi-objective and multi-constraint optimization analysis to take the economics, environmental quality, culture, politics, stakeholders’ conditions etc. into consideration. And finally a river engineering plan can be decided.
Figure 1. Data Collection for Application Site-specific Validation of Computational Model
Figure 2. 3D Flowfield Simulated by CCHE3D Model
Figure 3. Transversal Velocity
Fields Affected by a Submerged Weir
Shown at Selected Channel Cross-sections
Figure 4. Quantitative Site-specific Validation of CCHE3D
Figure 5. Physical Model Designed and Tested for Conducting Calibrations of a Computational Simulation Model, CCHE3D
Figure
It has been shown that to carry out today’s river engineering research and designs with everincreasing complexity, one should utilize the Integrated Methodology, which is a combination of Computational Simulation Models, Laboratory Experiments, and Field Measurements. The Laboratory Experiments of selected basic physical processes relevant to the project are needed to validate the Computational Models’ capability of simulating the basic physical processes important to the project. The Field Measurements have to be sufficient in both quality and quantity, so that a portion of them can be used to calibrate the model parameters for representing the unique and site-specific characteristics, and the remaining portion is used to validate whether the model is capable of predicting reasonably accurate and reliable responses of river system to prescribed forcings. Once the Computational Model has been verified and validated, it can be used to assess the performance or effectiveness of several conceptual designs for multiple or single objectives from both short and longer-term points of view. A huge number of cases can be run in a relatively short time and cost a lot less than those required by using the traditional research methodologies of laboratory testing and/or field monitoring of the performance of conceptual designs.
The author wants to emphasize that both the laboratory and field measurement are important to river engineering investigations, especially the field measurements. In order to conduct river engineering studies successfully, more field data of high quality are needed for both calibrations of model parameters and validation of model’s capabilities in predicting accurate and reliable river characteristics. The most cost-effective and “best” compromised decision can be obtained only by using a comprehensively verified and validated Integrated Methodology. A the science and engineering advance, the society shall demand better utilization of river’s natural resources without affecting the ecological and environmental quality. In addition, the river engineering decisions have to be made so that they are cost-effective; safe, even during some extreme climatic, hydrological and geological events; acceptable legally, socially, and by stakeholders; etc. Therefore the level of complexity of the task is going to be further increased continuously, the river engineering researchers are challenged to continuously sharpen their research and design tools.
This research was supported in part by the USDA Agricultural Research Service under a Specific Cooperative Agreement No. 58-6408-2-062, which was monitored by Dr. Matt Romkens, Director of the USDA-ARS National Sedimentation Laboratory, and by the University of Mississippi. The valuable assistance from Dr. Yafei Jia, Research Professor, Dr. Tingting Zhu, Post Doctoral Research Associate and Miss Janice Crow, Administrative Coordinator, all of NCCHE, are acknowledged.
In order to provide
the interested readers with additional information, several relevant publications
are included in this Bibliography.
Holly
F.M., Jr., and J. L. Rahuel
(1990). “New
numerical/physical framework for mobile-bed modeling,
Part I: Numerical and physical principles.” J.Hydr.
Res. IAHR, 28(4), 401-416.
Jia Y., and
S.S.Y. Wang (1999). “Numerical model for channel flow and
morphological change studies,” J. Hydr.
Engrg.,
ASCE, 125(9), 924-933.
Jia Yafei and
S.S.Y. Wang (2002). “Simulation of Flows around a Submerged Weir in Channel
Bendways,” Proceedings of the 5th International Conference on Hydroscience and Engineering,
Nguyen,
Dan K., Yu-E Shi, and Sam S.Y. Wang. (2006) “2D Shallow-Water Model Using
Unstructured Finite- Volumes Methods,” Published in the Journal of Hydraulic
Engineering, Vol. 132 (3): 258-269.
Osman A.M. and C.R. Thorne (1988). “Riverbank stability analysis, I:
Theory.” J. Hydr. Engrg., ASCE, 114(2),
134-150.
Wang
S. S.Y. and W. Wu (2004). “River Sedimentation and Morphology Modeling
– The State of the Art and Future Development”, Invited Keynote Address,
Presented at and Published in the Proceedings of the Ninth International
Symposium on River Sedimentation, on CD-Rom,
Wang
Sam S.Y. and Yafei Jia
(2005). “On the
Methodology to Develop Reliable Models for Water System Research”, Acta Geophysica Polonica, 53(4): 487-499.
Wang
Sam S.Y. and Weiming Wu (2005). “Computational Simulation of River
Sedimentation and Morphology – A Review of the State of the Art,” Invited
Paper, Published in the International Journal of Sediment Research, Vol.
20, No. 1, pp. 7-29.
Wang Sam S.Y. and Yafei Jia (2005). “On the Methodology to Develop Reliable Models for Water System
Research,” Invited Paper, Published in the Acta
Geophysica Polonica,
53(4): 487-499.
Wang
S.S.Y. and Y. Jia (2005). “A Systematic Procedure for Flow
Model Verification and Validation,” Presented at and Published in the
Proceedings of the World Water & Environmental Resources Congress, ASCE,
Wang Sam S.Y.
(2006). “Challenges in Verification and Validation of Computational Hydroscience Models,”
Keynote, 2nd Sino-American Workshop on Advanced Computational Modelling in Hydroscience and Engineering, On CD Rom,