FOREWORD

The XXIX Biennial Congress of the International Association of Hydraulic Engineering and Research (IAHR) was held at the Beijing International Convention Center (BICC) in Beijing, September 16-21, 2001. The central theme of the Congress is:

21^st Century: The New Era for Hydraulic Research and

Its Applications

The central theme was divided into five sub-themes, covering broad aspects of hydraulics, such as water resources, ecology and environmental hydraulics, forecasting and mitigation of water-related disasters, hydraulics of river, water works and machinery, and hydraulics for maritime engineering. More than 600 papers were presented and posted during the five days of the Congress. All papers that have been selected for presentations and posters on the XXIX IAHR Congress are included in 7 volumes (the John F. Kennedy student competition papers also included).

During the IAHR Congress some prominent and outstanding speakers invited by the Local Organizing Committee delivered 2 General Reports and 4 Keynote Lectures and other reports. As these excellent lectures dealing with many actual problems of hydraulic engineering and research could not be included in the Congress Proceedings. The LOC of XXIX IAHR Congress decided to publish a Post Congress Volume, which included mainly the contributions mentioned above. The authors of these contributions are: Prof. Forrest Holly, USA; Prof. Suo Lisheng, China; Dr. Torkil Tonch-clusen, Denmark; Mr. Zhu Erming, China; Dr. Wolfgang Kinzelbach, Switzerland; Mr. Oda Hideak, Japan; Dr. Zhaoyin Wang and Prof. Bingnan Lin, China; Dr. S. T. Su, USA; Prof. E. Plate and Prof. Rodi, Germany and so on. Besides, a few late papers are also included in this volume.

On behalf of the Local Organizing Committee the editor would like to deeply thank all delegates of the XXIX IAHR Congress for their attendance. We would also like to thank all the IAHR Council members for their confidence and help to the host of the XXIX IAHR Congress, 2001. The LOC would like to dedicate the Post Congress Volume to all the scientists and engineers and hope it may be reminiscence for those who have to take responsibility on the water related themes for mankind in future.

Guifen LI

Editor

Beijing, September 2001

CHIna_Times

Forest Holly

Professor, President of IAHR

Welcome

w Welcome to 29^th Congress of International Association of Hydraulic Engineering and Research

w Hosted by Chinese Hydraulic Engineering Society

w Organized by China Inst of Water Resources and Hydropower Research, and Tsinghua University

Recent events in US make this a difficult period for the world

w Many of you here today wondered if it was right to travel and engage in “normal” activity in this period of grieving for civilized discourse

w Some have asked if the Congress would be cancelled, as many international events have been

w Many of our colleagues who had expected to join us are not even able to do so

w Please join me in a few moments of silence for civilian victims of terrorism and strife not only last week in the US, but throughout the world

What can we do to help heal the situation?

w As water engineers and scientists, we are builders of civilization, not destroyers

w In our own relations with colleagues and peers, and especially with those we might not agree with, we must practice the qualities of civilized discourse and respect that we wish to see expressed in global relations among citizens of the world

Closing of welcome

w In this 29th IAHR Congress, even as we grieve for victims and our society, let us strive to combat a belief in the power of terrorism by affirming what is good and positive and constructive among civilized peoples throughout the world

w I wish you a technically productive and socially enjoyable week here in Beijing, and I declare the 29^th IAHR Congress to be open.

29th IAHR Congress

w Demographics: China 337; Japan 98; USA 40; Italy 37; Germany 26; UK 24; Netherlands 23; Iran 20; Korea 17; Canada 14;Switzerland 13; Austria 12; Russia, Portugal, Spain 11;

w Two invited general addresses

w Five technical themes, each with keynote speaker

w JFK Student paper competition

w Poster sessions

w Technical workshops

w Memorial Symposium: Pioneers of Modern Hydraulic Engineering in Asia (Thursday)

w LOC organized UNESCO support to enable attendance of 40 participants

Third World Water Forum

w World Water Council hosted First and Second World Water Forums, Marrakech and the Hague

w Gathering of global water interests

w Third forum to be held in Kyoto, Japan, March 2003, 8500 expected

w IAHR a strong supporter of this initiative

w Hideaki Oda, Sec General of WWF3, a Congress keynote speaker, Thursday

w Please visit WWF3 stand for more information

Busy two years for IAHR

IAHR Secretariat Relocation

w CM Delft Hydraulics generously hosted IAHR since its founding in 1935

w In 1996 CM CEDEX offered to host IAHR as we move into the new century

w Relocation achieved last month

w Appreciate patience of all during period of relocation and training of new staff

w Thanks to WL Delft and CEDEX

w Thanks to staff in Delft, especially Marjorie Keuning and Cheryl van der Zee with whom many of you have had direct contact

w Please sign commemorative books for each at IAHR stand

JHR Editor

w Prof. Marcelo Garcia, UIUC, begins term as new JHR Editor in October

w Thanks to Rodney White, who is retiring from HR Wallingford for his dedicated five-year term of service; JHR a top quality journal under his leadership

w Prof. Garcia unable to fly to Beijing; also Ippen Awardee

w Ippen lecture slot on Friday to be used for IAHR Member Forum

w Includes presentation of Prof. Garcia’s plans for changes in Journal editorial management

New IAHR Council Election Procedure

w No need to fear you’ll be named to nominating committee during Congress!

w After Graz, Council developed new election procedure for trial

w More democratic, less burdensome to Congress participants, still respects our constitutional boundary conditions

w My predecessor President Helmut Kobus appointed to chair NC in Iowa City July 2000

w NC worked very hard during past year

w Nominees published in Newsletter

w Large number of ballots already received

w Final ballot deadline Wednesday end of day

w If you’ve not voted, pick up ballot at IAHR stand or Secretariat

w Submit ballot in SIGNED envelope to Secretariat

w In this first trial, we learned of a few things needing adjustment

w Overall a positive change that is less complicated than it looks

w Friday’s GMA will include announcement of election results, and formal adoption of the consequent constitutional changes to Council composition.

w New Council takes office in October

Corporate Member Task Force

w Council working to actualize addition of “engineering” to our name

w IAHR historically and presently primarily an Association of researchers – this identity will not change – but we must offer value to members whose activity is primarily in applications

w Council member Gaele Rodenhuis organized Corporate Member Task Force, and Symposium on Managing Change in Research Institutes

w Invited Symposium, held in parallel with 29th Congress

w Council looking forward to facilitating further activities bringing together practitioners and researchers

Student Chapters

w Rapidly developing new area of IAHR affiliate membership, idea began in Graz

w U. Stuttgart; U. Illinois; U. Naples; U. Idaho; U. Iowa; Cal State U.; U. North Carolina; U. Minnesota

w 100 members at Stuttgart!

w Friday’s IAHR Member Forum will include presentation on activities of U. Stuttgart student chapter. Please attend

Affiliations with Other Water Organizations

w Council seeking ways to enlarge the connections, influence, impact of member activities

w Closer links with other organizations offers promising opportunities to do this

w EWRI agreement in effect, joint instrumentation sections conference in Colorado, ‘02, Journal info exchange

w COPRI agreement imminent, joint conferences on waves, tsunamis

w IWRA Investigative Task Force launched

w IAHS may offer opportunities for combined efforts

Is Hydraulic Engineering a Dinosaur?

w It’s been said that the 21st Century is the Century of Water

w Developed countries: water quality, allocation of water for irrigation, flood control, river restoration, wetlands restoration, fish passage, etc

w Developing countries: water supply, flood control, land reclamation, wetlands restoration, hydropower development, navigation, etc.

w Council met Friday with Mr. Wang Sucheng, Minister of Water Resources, China

w His excellency gave us a marvelous overview of water resource development in China

w The classic view is that it is the uneven distribution, not quantity, of water that is the challenge

w In China, it is the distribution AND the quantity that pose enormous challenges

w China is tackling this head on, and hydraulic engineers are leading the charge.

Why IAHR?

w IAHR is NOT a research institution

w IAHR is NOT a source of research funding

w IAHR has ONE professional employee

w But IAHR has 2500 volunteer members who benefit from exchanges and contacts with colleagues from around the world, facing a broad array of water challenges and opportunities

w “Them” is “Us”

w We can all learn from each other

w Task of IAHR Secretariat and Council is to facilitate and empower the efforts of volunteers sharing a common love of water science and engineering

w If you are not already a member of the family, please consider joining - applications at the IAHR stand

Closing of Address

w Thanks to LOC for what looks to be a highly rewarding and well-organized Congress

w Thanks to IAHR members for their patience during relocation of Secretariat and support of this fine association in a period of rapid change and opportunities

welcome speech at the Opening Ceremony of the 29^th IAHR Congress

Suo Lisheng

Professor, Vice-Minister of Ministry of Water Resources

Distinguished President Holly,

Distinguished Guests, Ladies and Gentlemen,

The opening of the 29^th IAHR Congress in Beijing, China at the beginning of the new century is a very significant event, for this congress has attracted so many world famous experts and scholars and will discuss and exchange ideas about subjects of rational utilization of water resources, water environment, water–related disasters, hydraulics of water works, and river mouths and coasts, summarize successful experiences and deficiencies and look forward to the challenging future, and will play an important role for the development of hydraulic research both in China and the world. On behalf of the Ministry of Water Resources, P. R. China, I wish to extend a warm welcome to all the more than 800 participants from more than 50 countries and regions all over the world! And also I wish to extend admiration and thanks to the efforts made by all of you in the field of hydraulic engineering and research and its applications.

Water is the source of life and lifeblood for the survival and development of human beings and without water there would be no human beings. Water is limited and is a natural resource that has no substitute. With the continuous development and progress of the human beings’ society, water has become an increasingly precious resource of strategic significance, and water resources issues and problems are concerned with the survival and development of human beings.

As everyone knows, China is situated in the Asian monsoon climate zone with uneven distribution of precipitation in time and space, frequent occurrence of flood, waterlogging and drought disasters and scarcity of water resources. The per capita volume of water resources in China is only 1/4 of the world average. For thousands of years, the Chinese nation has made indomitable struggles against water related disasters. Since the founding of the P. R. China, particularly, the Chinese Government has led the whole nation to construct a large number of water projects for flood control, farmland irrigation and drainage, water environment, hydropower generation and river transportation, and made brilliant achievements attracting worldwide attention. We not only have built more than 80000 reservoirs, but also are constructing the well-known Three Gorges and Xiaolangdi projects, have preliminarily established a flood control and disaster reduction system covering the whole country; the area of irrigated farmlands has increased to 54.7 million ha, thus providing a guarantee for the stable economic and social development. But, with the sustained and rapid economic development in recent years, the problems of water shortage, frequently flooding and deteriorating water environment have become more and more serious, and hence a stern challenge to us in the new century. Those problems have caused widespread concerns in all the social circles.

After decade’s years of practice of water harnessing, Chinese government has recognized that, in order to solve water issues and problems, we must change the traditional thinking for water harnessing, shift from the project oriented towards the resource management oriented and from the traditional towards the modern water resources undertakings, ensure sustainable economic and social development with sustainable utilization of water resources, and realize the safeties of water use, flood control, grain supply and eco-environment.

In the China Agenda 21, the Chinese Government puts the sustainable economic and social development among important strategic goals. The sustainable utilization of water resources put forward by the Chinese Government is a strategic issue for the economic and social development in China. In the Outline of the 10^th Five-year Plan for the Economic and Social Development in the P. R. China, it is pointed out that water conservation technologies and measures of various kinds should be extended comprehensively, with the enhancement of water use efficiency at the core, to develop a water saving agriculture and establish a water saving society; the bearing capacity of water resources should be fully considered in the construction of cities and layout of industry and agriculture, and agricultural and urban water saving should be intensified; planning and management should be strengthened, rational allocation of water resources for the whole river basin should be well conducted and water uses for domestic, production and ecological purposes should be coordinated; artificial precipitation enhancement, wastewater treatment and reuse and sea water desalinization should be actively conducted; groundwater resources should be rationally used and overdraft should be put under strict control; the reform of water management system should be strengthened to establish a rational managerial system and rational mechanism of water pricing.

The change of thinking for water harnessing and the huge challenges facing water resources undertakings provide a vast arena for the hydraulic professionals from China and even from all over the world to give play to their wisdom, and also open a new horizon for international hydraulic professionals to make cooperative studies in China. With China’s entry to WTO as a developing country, more international experts will surely be invited to take part in the water resources undertakings in China, and more Chinese hydraulic professionals will also take part in the undertakings of water resources and hydropower in the world. We welcome and encourage hydraulic professionals of other countries to come to China to carry out cooperative hydraulic studies, introduce more advanced technologies and experiences to China and make joint efforts with the Chinese people to solve water issues and problems.

The International Association of Hydraulic Engineering and Research is a very influential organization in the research and application of hydraulics. Since its founding in 1935, 28 congresses with seminars have been held, fruitful research achievements and examples of application were presented at each congress and the participants also included many scientists with initiative, important achievements, who made outstanding contribution to hydraulic research and set good examples for us. The Chinese hydraulic professionals have taken part in many activities of IAHR, not only making many friends but also learning many advanced international experiences through exchange.

I wish to take this opportunity to thank President Holly and other members of the IAHR Council for their help and support to the Chinese Hydraulic Engineering Society, and hope that close cooperation will continue between China and IAHR in the fields of sustainable utilization of water resources and so on.

I hope that this congress will provide a good chance for the participants to make extensive technical exchange, carry out academic discussions and look into the future, so as to further strengthen the friendship and cooperation among all the hydraulic professionals from different countries, promote hydraulic research in the world and the application of research achievements in different countries, thus making greater contribution to solve water issues and problems!

I wish the congress a great success and all guests’ good health and a pleasant stay in China.

Thank you!

Research Challenges arising from the World Water Vision

Dr. Torkil Jonch-Clausen

Chairman, Technical Committee, Global Water Partnership

Director, DHI Water & Environment

Danish Hydraulic Institute

Summary

Eighteen months after the presentation of the World Water Vision in The Hague a number of activities have been launched that will lead to achievement of the Vision for 2025. The third World Water Forum in Kyoto will focus on reporting on these activities. To achieve the Vision, a major contribution was expected through research and development efforts of the science and technology communities. The author reviews the research program of IAHR with comments on elements that could be given a different or greater emphasis based on his experience in the Vision exercise. Among these are:

Hydroinformatics

Indicators

Probabilistic methods

Data acquisition

Urban water management

Water resources management

Ecohydraulics and

Maintenance and development of research capabilities

The author concludes by asking whether it may not be time to resurrect the proposal of the World Water Commission to create a Research and Innovation Fund.

The World Water Vision: Where Do We Stand?

About eighteen months ago I had the pleasure of being with thousands of participants at the Second World Water Forum in The Hague. To-day I would like to report on the impact of that happening, of the World Water Vision exercise that led to it and on some of the many activities underway around the world to follow up on commitments made in The Hague. I will then turn my attention to reflecting on some of the implications that I see for the research agenda of IAHR that arise from the rapidly evolving approach to managing water.

Within weeks of the Forum, Kofi Annan issued his Report to the Millennial Summit of the United Nations. In it he referred to “a set of realistically achievable targets on water and sanitation” recommended by the Ministerial Conference of the World Water Forum. He urged the Summit to adopt the target of reducing by half, between now and 2015, the proportion of people who lack sustainable access to adequate sources of affordable and safe water. And they adopted this target!

Obviously also drawing on the Vision and Forum discussions, he stated that to arrest the unsustainable exploitation of water resources, we require water management strategies at national level and local levels. He said these should include pricing structures that promote both equity and efficiency. “We need a “Blue Revolution” in agriculture”, he said. “that focuses on increasing productivity per unit of water — “more crop per drop” — together with far better watershed and flood plain management”.

Just a month after the Forum, then USA Secretary of State Madeleine Albright referred to the Forum in her Earth day address. She said she intended to mobilise; to heighten public awareness; and issue a call for action; because the world has the capacity, and increasingly the will, to create water security for all. These are two examples of rather immediate impact of the Vision exercise and the Forum.

The World Water Vision

Those who participated in the Vision exercise wanted a world in 2025 in which almost every woman and man, girl and boy in the world’s cities, towns, and villages will enjoy safe and adequate water and sanitation and have enough food to meet their nutritional requirements. Their Vision could be achieved in a sustainable manner with a 10% increase in water withdrawals and consumption. Nevertheless, food production must increase 40%. This will only be possible within a sustainable water budget if people recognise that water is not only the blue water in rivers and aquifers, but also the green water—in soil. Recognition of its crucial role in the hydrological cycle will help make rain-fed agriculture more productive while conserving aquatic and terrestrial ecosystems. The percentage of water delivered to the domestic and industrial uses consumed by evaporation can be reduced, with most being returned after proper treatment to the ecosystems from which it is drawn. Domestic and industrial water reuse should become common, and new methods of ecosanitation not dependent on water as a carrier could be applied in many areas to reduce pollution and make full use of human waste as agricultural fertiliser. Natural and artificial wetlands can be used to improve polluted waters and treat domestic effluents. Countries that face water scarcities early in the century may invest in desalination plants—or reduce the amount of water used in agriculture, transferring it to the other uses, and importing more food.

Commitments

Many commitments were made at the Second World Water Forum to take actions that will help to achieve our Vision. The Netherlands Government has prepared a database of all such commitments that was released in Kyoto in June. They will be writing shortly to those who made commitments, thanking them for their participation in the Forum and asking them to share their progress with us, as inspiration to others to take similar action. This information will be used by the Water Action Unit of the World Water Council to track progress on these commitments along with other actions begun since. For we have passed from the stage of talking about what must be done, to one of action!

The Secretariat of the Second World Water Forum has worked closely with the World Water Council and the Secretariat for the Kyoto Forum. It has transferred all of the information in its data bases and the management of its web site so that these may serve the next event. The third Forum will be focussing on the actions that are being taken around the world as a means of raising awareness that solutions are available and actions are being taken by many. This will lead to greater commitment to action coming from the third Forum.

Focus Themes

There are some issues that will benefit from co-ordinated action. Three of these are:

Agriculture and environment security

Climate variability, climate change and water, and

Valuing water and financing infrastructure

It is useful to say a word about each of these.

One of the key challenges we face is to ensure food security for the increasing global population. Best estimates by many are that with more efficient irrigation, we can produce the 40% more food that will be needed with 15-20% more water consumed. Some feel that these estimates are too low. Others point out that in some regions, water withdrawals for agriculture and poor agricultural practices have already caused serious harm to the environment. Most recognise that this subject was inadequately addressed at The Hague. Follow-up began almost immediately in a meeting of stakeholders in Stockholm last summer. This led to a workshop on Food and Environmental Security sponsored by eight international organisations and hosted by IWMI in Colombo in December. Once again, all agreed that we have had enough talk, and that solutions will be found only by cross-disciplinary, multi-stakeholder actions taken at the country and basin level. A possible dispute between agriculturists and environmentalists is being replaced by a common effort to respond to the justifiable concerns of each community. It was agreed in Colombo that we have failed to recognise that resolution of this issue (and many other water-related issues) is not just a technical matter, but will be done by politicians. There therefore will be a special exercise to begin to address this factor. The stakeholders present in Colombo have agreed a “loosely co-ordinated” plan of action. Representatives of farmers and the private sector have agreed to join them. A step in their work plan will be to report on success achieved to date at the third Forum. This program of action was launched during the Stockholm Water Symposium in August.

The time horizon selected for the Vision exercise was 2025. While this helped us to focus on near-term problems and their solution, we ended up neglecting one of the biggest long-term drivers with respect to water resources management - climate change. The latest reports from the Intergovernmental Panel on Climate Change confirm that the process of climate change caused by human activities is well underway. They further tell that us that changes may be more severe than previously thought and that there are known serious consequences for the management of all natural resources, including water. A preliminary discussion of interested parties at an ad hoc workshop in Colombo noted that present water management approaches needed to better deal with climate variability. New water management strategies may offer solutions to some of the longer-term problems, and if widely adopted, create more resilience in our water management systems.

The United Nations University in Tokyo, together with the Secretariat of the third World Water Forum and the World Water Council hosted a more comprehensive workshop early last month. Its objective was to determine whether we might manage water resources better given the increasing predictability of climate variability and climate change and to develop commitment for action by all. There was unanimous agreement that we can manage water better using climate science, and that a program should be developed to this end. A workshop during Water Week in Stockholm this year has provided the framework and support for developing a Dialogue on Water and Climate.

A third challenge we face is that we simply are not seeing enough investment in water – neither in infrastructure, nor for that matter in urgently needed research into innovative solutions. The World Commission on Water rightfully was very concerned with this issue. It remains unresolved. It is also very complex, as investment requires returns. This links investment to the questions of valuing and pricing water, to the roles of the public and private sectors and of political stability and transparent governance. Many conferences have been held on the subject. Many books have been written on elements of the subject. Yet total investment levels at best remain stagnant while estimates are that they need to be at least doubled. The World Water Council and the Global Water Partnership have agreed to launch an initiative to create a partnership that will determine the scope of the problem on a national and regional basis and suggest optional approaches to specific local situations. Progress is being made on this front thanks to the co-operation of concerned international financiers. I am confident that we will be able to report concrete suggestions during the Forum in Kyoto.

World-Wide Movement

Thus there is a world-wide movement underway to act to resolve the water management issues displayed in The Hague. It involves cross-disciplinary partnerships and closer collaboration among international organisations. But most importantly, it is based on learning by doing at the local level, so that those so badly hurt by problems of lack of access to water, food and healthy environments may benefit from ACTION.

The Research Agenda

IAHR

When I was invited to speak at this conference, I did a little research about IAHR and its research agenda. I learned that scientific exchange is the main activity of IAHR. You develop technology transfer in several ways. Education is, of course, the classical way. You also encourage lifelong learning through exchange between individuals, national and regional groups using workshops, publication of books and manuals. You encourage management of the research program through exchanges among directors of hydraulic research laboratories. Corporate and national members are essential to your organisation. The transfer of new R & D results into engineering practice is accomplished through involvement of consulting and engineering organisations. Finally, you assure the participation of colleagues from developing countries through a Third World membership fund paid by IAHR members. This impressed me.

I was even more impressed when I read through the IAHR 1999 Research agenda by its comprehensiveness and the innovative approaches being taken in several sectors. Every area of concern being researched by you will increase the chances of achieving the World Water Vision. Nevertheless, I was motivated by your program to make some comments on areas I think of particular significance.

Hydroinformatics

I was delighted to find the emphasis placed in your research agenda on the concept of Informatics as opposed to Information and Communications Technologies (ICT) only.

This reflects the growing understanding of water professionals that our role is to inform and support water stakeholders before and after the decision-making process that is their responsibility.

In fact the evolution has been rapid. Without what might be regarded as simple ICT, it would have been impossible to develop the World Water Vision. The use of electronic mail and posting of documents on the Vision website to facilitate sharing of information and opinions that led to the creation of the Vision would not have been possible even two years earlier, especially in Africa and many parts of Asia.

Your agenda states that the rationale and purpose of hydroinformatics is to develop a new relationship between the stakeholders and users and suppliers of the systems. It offers the systems that supply useable results, the validity of which cannot be put in any reasonable doubt by any of the stakeholders involved. If the hydroinformation system is objective, the stakeholders may criticise a hypothesis of cultural practice (policies) leading to undesirable results, but not the system or tool. Thus the tool creates a possibility of negotiation and trade-offs based on merit and not on irrational sentiments. It was this type of hydroinformation system or tool that led to the agreement on water resources management that was part of the original Middle East peace agreement.

Hydroinformation correspondingly always works in a team, and may indeed create the sociotechnical means through which the team functions. The users become part of the system.

An example of this is the system through which the third World Water Forum is being created. The World Water Council’s second World Water Forum in The Hague was organised to be a participatory exercise. The Japanese hosts of the third World Water Forum proposed to the Council that the Forum should be created through participation. The main tool or system being used in the process is the Virtual Water Forum. Each group that proposes to hold a session in Kyoto in March 2003 is requested to open a page in the Virtual Water Forum. Through participation in discussions in a virtual session, interested individuals can shape the discussion that will eventually take place in the “real” Forum in 2003, even though they may not be able to be physically present themselves at that time. In addition, the degree of interest expressed in the virtual forum becomes a measure by which the Forum organisers may decide the importance (time and space) that should be given to the real session.

Usefulness of Indicators in Hydroinformatics

One of the principal concerns that is noted in your research agenda is how to make it possible for non-specialist stakeholders to have access to and to interpret the knowledge generated by specialists. A question that is often asked, for example by members of the boards of catchment management associations, is “how do I know whether we are making any progress in the management of our waters?” Or, “are we doing as well as or better than others, and why?”

Some “knowledge bases” have already been established. Others are being established, for example in the context of the Dialogue on Water for Food and Environment referred to earlier and in the Dialogue on Water and Climate. However it is unlikely that these will provide the kind of comprehensive and readily understandable information required by the non-specialist decision-maker. Similarly, those who cannot see “common-sense” relationships between various elements of water management and the environment will view sophisticated models with suspicion. Some indicator or indicators such as those provided in UNDP’s World Development Report might be developed to inform clearly and avoid suspicion.

Efforts are underway to develop such indicators for the environment in general and for water resources in particular. Most are familiar with the excellent work that was done by the World Resources Institute together with the World Bank, UNEP and UNDP in developing indicators for global ecosystems. They classified ecosystems under five main categories for agricultural, coastal, forest, freshwater and grasslands. At the same time they noted three other possible categories: mountains, polar and urban. Case studies illustrated the relationship between people and ecosystems. Their work led to and will be followed by the Millennium Ecosystem Assessment.

Another approach to developing a Pilot Environmental Sustainability Index was produced at the Davos Conference in 2000 by the Global Leaders for To-morrow Environment Task Force. The components in the system of indicators they developed were:

Environmental Systems

Environmental Stresses and Risks

Human Vulnerability to Environmental Impacts

Social and Institutional Capacity, and

Global Stewardship

This system was developed with the idea that, when refined, it could be used to measure environmental sustainability and even compare environmental sustainability practices of nations – an indicator similar to the UNDP’s development indicators.

The World Water Assessment Program is in the process of developing indicators for the state of the world’s water. This is not an easy task, as it must deal with the issues of scale and of boundaries that do not match the world’s administrative and national boundaries. A methodology is being developed that will attempt to develop indicators that will link water availability with water uses needs and demands while taking account of the ability to cope with water-related stress. Preliminary results of this work will be published at the third World Water Forum.

As has been stated by the National Research Council (USA): “Indicators used to report on a transition toward sustainability are likely to be biased, incorrect, inadequate, and indispensable. Getting the indicators right is likely to be impossible in the short term. But not trying to get the indicators right will surely compound the difficulty of enabling people to navigate through the transition to sustainability.” The task of developing these indicators will not be an easy one. Beyond a doubt it should clearly be on the research agenda of IAHR.

As your research agenda has noted, the presentation of comprehensive and understandable “encapsulated” information placed a great responsibility on those who prepare and provide it. Guidance on the ethics of this process will require the participation of philosophers as well as scientists.

Given the extent of work required in the field of hydroinformatics, I am inclined to take exception to the statement in your research agenda that “Hydroinformatics is a technology – not a science”. Like other sciences, it will evolve, but the basics will remain the same. We must develop means to inform all stakeholders of the issues concerning water resources management so that they may take informed decisions; not only decisions directly concerning water management but also those social and economic decisions that impact water resources and environmental sustainability.

Probabilistic Methods

Probabilistic approaches provide basic analytic tools to systematically integrate involved uncertainties, to quantify performance reliability of a hydraulic system, and to incorporate uncertainty and reliability information in decision-making for a more comprehensive project design and evaluation. What has been missing until now in conceptualising water management projects (as well as other infrastructure projects) is recognition of the number of associated risks. A recent research project that examined 60 infrastructure projects found that the complex relationship described in Figure 1 below. Systematic unbundling of these factors makes it possible to understand their importance and variation during the life cycle of a project.

As this research covered infrastructure projects in general, researchers associated with IAHR might usefully extend the analysis to examine risks related specifically to water infrastructure projects. This might provide information on risk reduction that we lead to increasing investments in the sector.

Data acquisition

The current research agenda refers mostly to technology issues related to the installation of data acquisition equipment in the field. Yet there is a world-wide decline in water monitoring. The number of monitoring stations for water flow and quantity in Africa, for example, declined 90% between 1990 and 2000 (Johnson et al, 2000). Fully understanding the complex interrelationships between man’s actions and global and local systems depends on better knowledge about such issues as minimum in-stream flows for maintaining biodiversity and groundwater recharge, maximum thresholds for common pollutants, and the relation of land use to hydrologic functions.

Rain and stream gauges around the world are disappearing, victims of loss of funding for monitoring programs. Better basic hydrological information about river discharge, flood frequency, dry season flows, condition of wetlands, and location of dams would help planners meet the growing human demand for water.

Increased funding for data gathering is essential. However given modern satellite and other technological systems, more research is needed to determine the appropriate density and type of monitoring stations necessary to-day to ensure the least-cost approach to obtaining the data required on weather, runoff, water quality and other parameters for integrated water management at different geographic scales. This too should be a challenge for IAHR member institutions.

Urban Water Management

I was pleased to read under your research agenda that urban drainage systems should be designed to convey storm water runoff and sewage flows of magnitudes varying from dry-weather flows to floods, control fluxes of pollutants resulting from human activities, and contribute to the general well-being of the human population. This is to be achieved within the framework of integrated management of urban waters, with minimal impacts on receiving waters in a cost-effective way, and under conditions of increasing populations of large cities. I presume that a further condition would be to have minimal impact on the upstream sources of water required for urban agglomerations. The objective of IAHR/IHA is to promote an ecosystem approach to the planning, design and operation of urban storm drainage. I could not agree more with this enlightened concept. However, I wonder if it goes far enough?

One of the concerns that I had while working on the World Water Vision exercise was the tendency to automatically link issues of sanitation to water. Thus we tend to say that there are over 1.2 billion people without access to safe drinking water, and in the next breath, that nearly half the world population lacks access to sanitation. In our minds when we say this, sanitation is water-borne sewerage. I think this is a serious mistake.

The developing world is falling further and further behind in the provision of means to dispose of its wastes. As communities develop water supplies, generally without provision for sewerage, the volume of waste grows. To the nearly three billion people that do not have access to sanitation must be added the billions of equivalent population from industrial and agricultural wastes. Where waste is collected, much of it is discharged without treatment. This situation impacts on health and on the environment. One of the effects is the pollution of surface and groundwater, sometimes rendering them unsuitable for domestic purposes even through treatment. The cost of correcting this situation through conventional collection and treatment processes will be thousands of billions of dollars. The less developed economies cannot afford these solutions or they have higher priorities for economic development. Inhabitants of rural and marginalized urban areas in particular are not able to pay, even if they were willing.

In the industrialised world, much of the urban infrastructure, particularly the sewage collection systems, will need to be replaced over the coming decades. In the constrained budget environment faced by all countries, it is not at all clear how these replacements will be financed, as the burden on taxpayers has often reached its limit. To improve the efficiency of sewage treatment and reduce the volume of residual pollutants reaching the environment, responsible authorities are having recourse to increasingly expensive treatment processes (e.g. chemical treatment) with increased operating costs. Many of the older sewer systems carry both sewage and storm water. Most of what is said above may appear to apply only to sanitary sewerage (the collection and disposal of human excrement). However there are clear linkages with the disposal of industrial and toxic liquid and even solid wastes.

The basic processes for the collection and treatment of wastes have been in use for two thousand years. Modern engineering and science have focused on methods to make them more effective and efficient, rather than on finding alternatives. Processes to treat or eliminate wastes at the source, rather than at the outlet from a transportation system, could drastically reduce the costs of sanitation. Indirectly this would reduce the costs of surface drainage. Such processes would likely ensure better protection of the environment.

While there is a wide range of sanitation issues that would benefit from shared research, the basic concept which needs challenging is that of water-borne sewerage. An alternative to this would open the field for innovative thinking with regard to the disposal of other wastes. Benefits that would result from finding such an alternative would include:

Elimination or reduction of the costs of construction, maintenance and operation of collection (sewer) systems, treatment works and final disposal;

Reduced water consumption;

Reduced costs of storm water drainage systems;

Elimination of residual pollutants currently discharged by treatment works; and as a consequence,

Protection of ground and surface waters, and their environment.

I invite researchers in this field to consider whether the approach to management of human waste that we began to use thousands of years ago is still appropriate and economical in this age of modern technology and natural resource constraints.

Water Resources Management

The IAHR Section for water resources Management has adopted, as a primary long-range goal, to promote the use of advanced technologies to address problems of environmentally sound water resources management, and has committed itself to encourage interdisciplinary approaches in hydraulic engineering with special regard to ecological concerns. Moreover, the Section wants to promote the adoption of appropriate methodologies for developing countries by education and training.

Most of those involved in the management of water now accept the principle of integrated water resource management (IWRM). I will not expound on those here. However, no matter how well-founded this principle is – and it is well-founded – what we often fail to recognise is that we have arrived at this principle and this approach in the countries of the North through a gradual process over the past 150 years. The economic development of the industrialised world took place in the first hundred years of this period driven by the “enlightened” belief that using science and under capitalism system nature could be controlled. It was in response to the Green movement in the North that this paradigm has shifted in the North.

As Tony Allan has pointed out, the underdeveloped countries of the South are now faced with improving their quality of life while at the same time being expected to shift from the water management paradigm under which developed countries evolved to that which the industrialised countries now espouse. This poses a moral dilemma for the North. But it also poses a research challenge for us – to develop cost-effective ways to provide for economic development under the principle of IWRM that calls for a sustainable environment.

One aspect of IWRM that proponents nearly always fail to recognise is that IWRM requires not only a demanding holistic professional and scientific approach but also an unprecedented level of political co-operation. We must recognise that water users and policy makers operate in political systems that determine whether or not the new paradigms can be assimilated that useful debate can take place. Political systems make sense to most of the players who live in them. Such systems have a political rationale. Decision-makers and water users will assimilate IWRM only if the innovation of integration is appreciated as a political process and not just as a technical, investment or information sharing process. Water policy will be transformed if it is politically feasible. Such innovation will be achieved by taking an inclusive approach and emphasising the institutional dimension of the inescapably political the integrated water resource management process. In brief, we must learn how to assess and influence political feasibility. Research into these processes will be sensitive, but is essential.

Ecohydraulics

The environment, especially in terms of water quality, pollution and protection of ecosystems is one of the major concerns of modern civilisation. Research in this field has significantly increased in recent years. However a lot of work remains to be done to improve our knowledge and capacity to understand phenomena, predict the effects of human works on natural ecosystems, and find solutions for maintaining acceptable water quality and biodiversity in our marine and continental environment. During the Vision exercise I often represented this dilemma by asking the question “How much water does a river need?”

All these reasons have led IAHR to create a new section on Eco-hydraulics to encourage collaboration between hydraulicians, biologists and chemists and others. I can only applaud this effort, and note a strong linkage to the issue of research in adapting to shifting paradigms of which I have just spoken.

Maintenance and Development of Research Capabilities

With a better understanding of the economic value of water and more public-private-partnerships, there will undoubtedly be more privately funded research. However many of the issues described in your research program will not be attractive to the private sector. One of the conclusions of the Vision exercise was that there should be an increase in public funding for research and development in the public interest.

The World Water commission saw that given the potential of the new technologies and the innate abilities of people, enormous gains could be made as new innovations occur in either institutional arrangements or technology application. The latter may be by the rediscovery and deployment of traditional technologies or the emergence of new technologies. A key to get the maximum benefit globally of these new developments would be how quickly they will be adequately evaluated, disseminated and adopted throughout the world.

Innovation also requires some assistance in incubation. The Commission recommended the establishment of an Innovation Fund that would help promote environmentally and socially desirable technical and institutional innovations. Some of the institutional innovations that could be considered for support included:

national “water stamps for the poor” programs;

time-bound subsidies for transition arrangements;

providing medium-term “bridging loans” in countries where long-term capital markets are not developed;

political risk guarantees for private operators entering risky markets;

new forms of mobilising NGOs and communities to improve services and protect the environment, and

exploring new approaches to negotiation of international water treaties.

In terms of technology the Commission also saw innumerable opportunities, which include innovations in particular “orphan” areas such as biotechnology for the food crops of the poor in water deprived areas. Finally, they saw geographic areas with problems that cry out for new approaches. For example, the Indo-Gangetic Plains have very large numbers of poor people and hunger, yet so much water badly distributed in space and time. It is a very big challenge to work out a water management paradigm for this area that is environmentally sound, socially responsible and economically productive.

Perhaps the time has come to lobby again for the creation of a Research and Innovation Fund with support from both the private and public sectors?

References

ALAN, ANTHONY. 2000. Millennial water management paradigms: making IWRM work. (SOAS- unpublished)

COSGROVE, WILLIAM J. AND FRANK R. RIJSBERMAN for the World Water Council. 2000. World Water Vision: Making Water Everybody’s Business. Earthscan.

MILLER, ROGER AND DONALD R. LESSARD. 2000. The Strategic Management of Large Engineering Projects – Shaping Institutions, Risks and Governance. MIT Press.

NATIONAL RESEARCH COUNCIL (NRC). 1999. Our Common Journey - A Transition toward Sustainability. Washington. National Academy Press.

UNDP, UNEP, WORLD BANK, and WORLD RESOURCES INSTITUTE (2000): World Resources 2000-2001- People and Ecosystems – the Fraying web of Life. World Resources I

THE MAIN PROGRESS IN THE HYDRAULIC RESEARCH

IN CHINA

Zhu Erming

Chairperson of the Executive Board, Chinese Hydraulic Engineering Society

Abstract

This paper summarizes the natural conditions, achievements of water resources undertakings and major existing water problems in China, emphatically expatiates upon the major progress in hydraulic research and, in consideration of the requirement of economic and social development for water resources, presents prospects for the development of water resources undertakings in the early 21^st century (2030).

NATURAL CONDITIONS AND ACHIEVEMENTS IN WATER RESOURCES UNDERTAKINGS

Natural Conditions

China has a vast territory ranging over 62° of longitude or 5200 km from east to west, and 52° of latitude or 5500 km from north to south, covering the tropical zone, sub-tropical zone, warm humid zone, temperate zone and cold temperate zone. The topography is complicated and presents three obvious steps from the “World Roof” at the western end to the eastern coastal plains. China has a large number of rivers, in which more than 1500 each have a drainage area of more than 1000 km² and have a total length of 430000 km. Most of the large rivers flow from the west towards the east with divides in between in the same direction, thus natural hydraulic links are lacking in the south-north direction. The climate is affected by monsoon and the precipitation varies greatly in time and space. The southwestern regions are rich in hydropower potential, and the northwestern regions are poor in rainfall and suffer frequently from droughts. The total volume of water resources in the country is 2800 billion m³ and the per capita volume is 2170 m³. As affected by the natural and climatic conditions, China suffers frequently from flood, waterlogging and drought disasters, the soil erosion is serious, and the eco-environment is fragile. The natural environment determines the position and role of water resources undertakings in the economic and social development and the eco-environmental protection in China.

Achievements of Water Resources Undertakings

The Chinese nation has a long history of water harnessing. The Dujiang Weir Complex in Sichuan Province, the Zhengguo Canal in Shaanxi Province, the Ling Canal in Guangxi Autonomous Region and the Beijing-Hangzhou Grand Canal were all built Anno Domini and water resources projects of various kinds such the Yellow River embankments were built in later periods. Since entering the 20^th century, particularly after the founding of the P. R. China, the water resources undertakings have seen a period of great development. In the last 50 years, a lot of water resources infrastructure has been built. By the end of year 2000, the total length of consolidated and newly-built embankments had amounted to 270000 km; more than 85000 reservoirs had been built with a total storage capacity of 518.3 billion m³, and 98 flood detentions zones had been established with a total detention and storage capacity of 97 billion m³, thus preliminarily creating a flood control engineering system for large rivers and lakes which can control normal floods, protect 40 million ha of farmland, 460 million population, more than 600 cities and major communication lines, industries and mines; the total annual capacity of water supply had amounted to 580 billion m³; the area of irrigated farmland to 55 million ha; the installed capacity of hydropower generation to 76.8 GW; the mileage of river transportation to 110000 km; and the soil erosion area that has been put under preliminary control to 800000 km², thus reducing annual sediment discharged into rivers by about 1.5 billion t. The water resources undertakings in China have played an important role in flood control, urban and rural water supply, hydropower generation, river transportation, water and soil conservation, ecological improvement, harbor, aqua-culture, recreation, etc, thus providing great supports for the sustainable, rapid and healthy economic development, social stability, promotion of the people’s living standard, smooth progress of reform and opening-up and the national economic safety and, at the same time, also providing extensive horizons for hydraulic research.

The Main Water Problems

Low capacity against flood and long term treats of flood and waterlogging disasters

At present there are 70% of the cities and 50% of the major embankments are not up to the national standards for flood control, the situation of flood control is still stern and the tasks of flood control are very arduous.

Uneven distribution of water resources in time and space and a big gap between water supply and demand

As affected by the monsoon climate, China’s precipitation varies greatly in time and space. The southern China is rich in water resources but poor in farmland, and the northern China is poor in water resources but rich in farmland. It is very pressing to implement rational allocation of water resources in order to solve the water shortage in the Huang-Huai-Hai region

Severe soil erosion and fragile eco-environment

According to the second national soil erosion remote-sensing survey, the water erosion area in the country is 1.65 million km² and the wind erosion area 1.90 million km². Therefore, the soil erosion is still very severe and the tasks to harness soil erosion are very arduous.

Serious water pollution and deteriorating water environment

By now about 38% of the river reaches of the seven major rivers in the country have been polluted to a varying degree, about 30% of industrial wastewater and 80% of domestic sewage are directly discharged into rivers, lakes and reservoirs without any treatment. To facilitate the harnessing of water pollution sources will be put on the important agenda of infrastructure construction in China.

MAiN PROGRESS IN HYDRAULIC RESEARCH

In the last tens of years, China has stick to the thoughts that sciences and technologies are the first productivity, economic development must rely on sciences and technologies and scientific and technologic work must be oriented to economic development, and the large-scale construction of water resources engineering has promoted hydraulic research.

Water Resources Research

The natural precipitation in China is distributed unevenly in time and space with great inter- and intra-year variations, and both consecutive wet and dry years often occur, which is the basic reason for the imbalance between land and water resources and the frequent occurrence of drought, flood and waterlogging disasters. In the last tens of years, China has achieved great progress in water resources assessment, allocation of water resources, water environmental protection, water legislation and water management. Structural and non-structural measures are applied to coordinate the relationship between water and the society, economy, ecology and environment and implement unified planning and management of development, utilization and protection of water resources. The multiple levels, multiple objectives and group decision large system analysis method was applied to the formulation of national mid to long term water plan, national groundwater plan, national water conservation plan, etc so as to promote the foresight for sustainable utilization of water resources. At the same time, the research projects of Macroeconomic Based Water Resources Planning and Management in North China and the Rational Allocation and Ecological Protection in Northwest China were implemented for the areas of serious water scarcity. In order to alleviate the water shortage in some areas, a number of inter-basin and inter-regional water transfer projects have been implemented, such as Datong River-Qinwangchuan and Yellow-Shanxi transfer projects. At present the South to North Transfer Project is being studied, which would cover the four major river basins of Yangtze, Huaihe, Yellow and Haihe.

At present coordinated planning is being made for the three large systems of water cycle, socio-economy and eco-environment of river basin with the dissipation theory, fuzzy mathematics and artificial intelligence in order to provide comprehensive guarantee of water resources for the regional economic development and eco-environmental protection.

Environmental Hydraulic Research

Water is an important element of the environment and water resources undertakings exert multiple impacts on the environment, in which the positive ones include: control of flood and waterlogging disasters, development of irrigation, improvement of navigation, hydropower development, water and soil conservation, water supply for human beings’ life and production, ecology and environment, and the negative ones include: inundation of large areas of land by reservoir construction, resettlement of large numbers of residents, layered distribution of water temperature in large-sized reservoir affecting water ecology, river barriers stopping fish migration, release of low temperature water from reservoir affecting water ecology; excessive upstream water use causing water shortage and river’s drying-up downstream, discharge of untreated wastewater upstream causing pollution of river channel downstream, and excessive overdraft of groundwater causing ground subsidence. Therefore, to strengthen water demand and use management, promote water use efficiency, protect and improve the eco-environment is an important task of water resources undertakings.

Water eco-environmental protection

w Conduct scientific zoning of water functions and formulate protection plans accordingly.

w Carry out targeted water quantity and quality evaluation and make forecast regularly.

w Strengthen the monitoring and information collection on water eco-environment and establish a data base system for water quantity and quality.

w Protect water eco-system and biological resources.

w Strengthen research and water pollution control, and accelerate the progress of water and soil conservation.

w Strengthen legislation for water ecological protection and protect water eco-environment by law.

Hydraulic works environmental protection

The impacts of hydraulic works on the environment are a complicated system of multiple variables, multiple structures and multiple levels, and are generally evaluated in a comprehensive way according to the varieties, components, elements and measures of environment, covering reservoir inundation, land occupation, induced changes in water and sediment regimes, river regime, storage and release, water table, etc, and impacts of construction. The Law on Environmental Protection of P. R. China was issued in 1979, the Regulations on the Management of Environmental Protection for Capital Construction Project in 1981, and the Regulations on the Environmental Impact Evaluation of Hydraulic Works in 1982, thus establishing a system of EIA for hydraulic works.

w The planning and design of hydraulic works should follow the policy of putting prevention first, combining prevention with remedy and implementing comprehensive harnessing so as to reduce as far as possible adverse impacts on the eco-environment. Reliable protection measures must be adopted for unavoidable impacts.

w In the planning and design, impacts on the eco-environment should be fully studied as an important element for the rational selection of design schemes, resident relocation plans and operation rules.

w Construction must not start before the EIA report for the project is approved.

w The system should be fully implemented that the environmental protection measures are designed, constructed and put into operation at the same time as the main works. The acceptance of environmental protection measures should be taken as a component of project completion acceptance.

w The monitoring and management of eco-environment during operation should be taken as a component of project management.

w In the formulation of the EIA report for hydraulic works, representatives from related areas and sectors, experts and the public must be invited to participate and an effective supervision mechanism should be established.

Major progress in water environmental research

w Observation has been made of pollutant diffusion and transportation for the main reaches of large rivers and studies have been carried out on the range and concentration distribution of offshore pollution belt and mutual impacts among multiple waste outlets.

w Studies have been made on the impact of regular waves on re-oxygenation coefficient for each system of the river dissolved oxygen model and the model for oxygen production of algae photosynthesis and methods for the determination of parameters have been determined.

w Studies of basic theories have been made on the pattern and rate coefficient of adsorption and resolution of suspended substances to metals in water of heavy metal quality.

w Long term, systematic studies have been made on thermal pollution mainly with physical models for cooling water of thermal power plants to put forward the measures to overlap intakes and outlets of cooling water. Studies have also been made with multiple hydraulic and thermal models for cooling water of nuclear power plants.

w Studies have been made on groundwater pollution and purification measures, offshore pollution, self-purification of pollutants, and analysis and simulation techniques for diffusion.

Studies of River Channel Evolution and Sediment Movement

Soil erosion is severe in many river basins in China. The rivers carry huge quantities of sediment and river sediment movement directly affects river configuration. Therefore, long term observation, studies and harnessing have been made for river sediment movement. The Chinese scholars have achieved a number of theoretical, semi-theoretical, semi-empirical and empirical results, and have made breakthrough progress in the studies of cohesive sediment running-up, high sediment content flow movement, density flow and mud-rock flow by combining the mechanical theories with statistics and physical chemistry.

In the aspect of river channel evolution, the theories of sediment movement and geomorphology have been applied to study the laws of scouring and sedimentation of natural rivers. The alluvial rivers are divided into five types: straight, curved, serpentine, braided and wandering, and harnessing measures have been adopted according to the evolution laws to achieve good results. The sediment movement at river mouth is affected by river flow, tidal flow and waves. The studies of river mouth are focused on the development of river mouth delta and the formation and variation of sand barriers and bars. Long term observation has been made to provide a scientific basis for the harnessing of river mouth.

In the aspect of engineering sediment, the focus has been put on the studies of reservoir sediment, water complex sediment, river channel and lake sediment, sedimentation at tidal gates and sediment induced abrasion of water turbine. A series of theories and experiences have been produced for sediment treatment, such as the theories of “storing the clear water and discharging the muddy to regulate water and sediment” for reservoir operation, “forward diversion and lateral sediment discharge” to separate water from sediment for the layout of water complex, “diverting flood for warping irrigation to improve soils, warping on the front and back to consolidate embankment” to utilize sediment resources, “carrying out transportation in static water and scouring sand regularly” to solve sedimentation of navigation routes.

While solving the sediment problems for river channel and hydraulic works, theoretical studies and technical development of mathematical simulation, physical model and prototype model observation have been promoted in the fields of river basin sediment generation, river channel modeling, laws of sediment movement at river mouth and coast, experiment on high sediment content water flow, and the results have been extended and applied effectively.

Studies of High Water Head Ship Lock Hydraulics

China has many large rivers, a large number of lakes, reservoirs and coastal lines connected to river mouths. Those waters provide favorable conditions for the development of river shipping and river-sea joint shipping. The total length of the rivers open to navigation is about 110000 km. The water resources in China vary greatly in time and space and the river sediment problem is serious, therefore, in order to develop river shipping, studies must be made on many scientific and technological issues. During the construction of ship locks of the Gezhouba Project, a lot of studies and experiments were carried out to solve the issues and problems, such as sedimentation affecting navigation, the harnessing of the Nanjinguan navigation channel and hydraulics of high water head ship locks. The Nanjing Hydraulic Research Institute established an experimental tank for non-steady flow and made 1:10 model experiments for extra-large valve in combination with mathematical models to optimize gate shape; at the same time, studies were made on the shape of valve intake, water inflow and ventilation of internal shaft in combination with prototype observation of the Gezhouba lock, etc to solve the problems of current vibration and cavitation erosion.

Studies of High Speed Flow Energy Dissipation

There are many high water head flood relief structures in China. Table 2-1 shows the existing high water head and large discharge relief structures and Table 2-2 shows those under construction and to be constructed. Because of the high dam, large discharge, high flow velocity and narrow channel, energy dissipation and scouring prevention are the most outstanding issues for high water head flood relief structures. Many types of energy dissipation structures have been developed through model experiments, theoretical analysis, prototype observation, etc. with consideration of practical conditions of engineering structures.

Table 2-1 Existing high water head, large discharge flood relief structures in China

Basic data					Relief structures
No	Dam	Dam type	Dam height (m)	Discharge (m3/s)	Openings on dam			Spillway No.–width´ height	Tunnel No.–width´ height
					Surface Opening No.–width´ height	Middle Opening No.–width´ height	Deep Opening No.–width´ height
1	Shuikou	PG	101	51800	12-22´15		2-5´8
2	Panjiakou	PG	107.5	42600	18-15´15		4-4´6
3	Shuifeng	PG	106	40000	26-12´6.5			10-9´10	D8.6
4	Ankang	PG	128	37000	5-15´17	5-11´12	4-5´8
5	Yantan	PG	111	33400	7-15´21		1-5´8
6	Yunfeng	PG	113.75	24230	21-11.0´7.5		4-4.25´4.25
7	Ertan	VA	240	23900	7-11´11.5	6-6´5	4-3´5		2-13´13.5
8	Geheyan	GV	151	23458	7-12 ´18.2	6-6´5	4-4.5´6.5
9	Fengtan	GV	112.5	23300	13-14´12		1-6´7
10	Tianshengqiao I	ER	178	21750				5-13´20	1-6.4´7.5
11	Wujiangdu	PG	165	21350	4-13´18.5		2-4´4.4	2-13´18.5	2-9´10.44
12	Manwan	PG	126	20910	5-13´20		2-5´8		1-12´12
13	Baozhusi	PG	132	16060	2-15´17.3	2-13´15	4-4´8
14	Sanmenxia	PG	106	15100		12-3´8	8-3´8		2-8´8
15	Guxian	PG	125	13894	5-13´16.5	1-6´9	2-3.5´4.23
16	Huanglongtan	PG	107.5	13300	6-12´10		1-5´6	Emergency 10´12
17	Xin’anjiang	PG	105	13200	9-13´10.5
18	Dongfeng	VA	173	12580	3-11´7	2-5´6 1-3.5´4.5		Left 1-15´20	1-12´17.5
19	Baishan	GV	149.5	11000	4-12´12		3-6´7
20	Bikou	TE	101.8	9550				1-15´16	Left 1-9´8 Right 1-8´10
21	Liujiaxia	PG	147	9220			2-3´8	3-10´8.5	1-8´9.5

Table 2-2 High water head, large discharge flood relief structures under construction or to be constructed in China

Basic data					Relief structures
No	Dam	Dam type	Dam height (m)	Discharge (m^3/s)	Openings on dam			Spillway No.–width´ height	Tunnel No.–width´ height
					Surface opening No.–width´ height	Middle opening No.–width´ height	Deep opening No.–width´ height
1	Three Gorges	PG	183	102500	22-8´17	2-18´11 raft floating	23-7´9
2	Dachaoshan	PG	115	23800	5-14´17.8		3-7.5´10
3	Xiaolangdi	TE	154	17063				3-11.5´17	3-D14.5 3-D6.5 1-10´11.5 1-10.5´13
4	Xiluodu	VA	273	50311	8-12.5´18		7-5´6
5	Xiangjiaba	PG	161	48680	5-19´26	7-7´11
6	Longtan	PG	216	35500	7-15´20		2-5´8
7	Nuozadu	ER	258	35300				10-15´20	2-5´8.5
8	Goupitan	VA	225	26950	6-16´15	7-6´7	2-6´7
9	Xiaowan	VA	292	20683	5-11´15	6-6´5			2-10´12
10	Shuibuya	ER	232	15243			2-4´5	5-14´21.5
11	Pubugou	ER	186	9780				3-12´16	1-9´9 1-12´7.5 (transformed from diversion tunnel)

w Bottom flow (jump) energy dissipater. It dissipates energy through hydraulic jump to make upstream flow connect downstream flow steadily. This type of energy dissipaters is used in the Yanguoxia and Puxi projects constructed in the 1960s and the flood and sediment relief gates of the Gezhouba Project. Flaring pier-cushion pool joint energy dissipater is used for the high water head flood relief structures of the Ankang and Wuqiangxi projects constructed after the 1980s. Aerated pier-cushion pool and T-shaped pier-cushion pool joint energy dissipaters are also applied, which can reduce the length of cushion pool and promote the efficiency of energy dissipation.

w Submerged bucket dissipater. With this kind of dissipaters energy is dissipated through the rapid diffusion as induced by the spiral mixing function of bottom spiral flow and surface flow of overflow dam. Practices of years show that surface flow dissipates a small amount of energy and has a large speed, thus causing bank scouring over a long distance. Therefore, better dissipation effects can only be achieved with supporting dissipaters, in which flaring pier is one of the most effective ones.

w Ski-jump energy dissipater. With this kind of dissipaters, a trajectory bucket at the end of overflow dam is used to make the jets discharge in air to collide and diffuse and, after dissipating a part of dynamic energy, fall into a downstream cushion pool to dissipate energy through diffusion, turbulence and rolling. There are many types of trajectory bucket, and the common ones are continuous type, differential type, diffusion type, oblique deflecting type, torsional type, high-low sill type, etc. Multiple types of ski-jump dissipaters have been used for about 100 high dams, such as Xinfengjiang, Zhexi, Liuxihe, Wujiangdu, Baishan and Longyangxia projects, and have greatly reduced downstream scouring.

w Contracting energy dissipater. It is formed by sudden contracting side walls or blocks to create diffusion in both the longitudinal and verticdal directions, thus achieving significant energy dissipation. A flaring pier contracting dissipater is used in the Ankang Project to make the length of the cushion pool reduced by more than one third. For the overflowing dam of the Wuqiangxi Project, flaring piers are used to create longitudinal diffusion and traverse contraction of jet and bottom release openings are provided in the water free area of dam surface to increase release capacity and promote energy dissipation. In this way the length of cushion pool is reduced by 60%. Slit-type buckets of different contracting ratios and deflecting angles are used in the Dongjiang, Dongfeng, Longyangxia, Nanyi and Xiongdu projects to achieve satisfactory energy dissipation effects.

w Stepped joint energy dissipater. With this kind of dissipaters, water flow becomes aerated, slowed down, mixed and turbulent along the steps on dam slope or spillway surface to dissipate energy, and good effects are achieved in combination with energy dissipation from longitudinal diffusion created by flaring piers on dam crest. This kind of dissipaters is used in the Shuidong and Dachaoshan projects, etc and is being considered for some dams under design. The steps themselves cannot dissipate energy significantly, but can promote aeration to reduce cavitation and can achieve good results jointly with flaring piers.

w Dissipater in flood relief tunnel. It includes many types such as orifice plate type, vertical shaft type and vortex flow type. In the Xiaolangdi Project three orifice plates are used in the flood relief tunnels and energy is dissipated step by step by using the orifice plates to create contraction, then diffusion, and further contraction and diffusion repeatedly. In the Shapai Project, vertical shaft vortex flow energy dissipation in flood relief tunnels is applied, that is, reservoir water flows into a vortex chamber at the inlet of vertical shaft to create vortex flow entering the vertical shaft. Within the shaft frictional force of free vortex flow against the side wall and internal shearing and resistance forces of the flow dissipate a large amount of energy. After entering the horizontal free flow section, the vortex flow continues to dissipate energy significantly.

w Dam crest water fall-cushion pool energy dissipater. Cushion pool is applied to both the Liuxihe and Ertan arch dams, with which a spillway on dam crest throws the jet downstream far from the dam site. The jet dissipates a part of dynamic energy in air and most of dynamic energy is dissipated in the downstream cushion pool.

Studies of Technologies of High-velocity Flow Aeration for Cavitation Erosion Reduction

Because there are irregularities on the surface of spillway left from construction, cavitation erosion often occurs to high-velocity relief structures. Since the 1970s studies have been carried out on technologies of aeration for cavitation erosion reduction, applying the shapes of spillway surface and aeration measures recommended by model experiments to real projects and making prototype observation. Observation results further show that the designed hydraulic indexes are rational and the effects of aeration to reduce cavitation erosion are significant. In the last tens of years, research, design, construction and management agencies have made joint efforts to carry out experiments, prototype observation and theoretical analysis, thus achieving development and improvement in the studies, design and application of technologies of aeration for cavitation erosion reduction. Since the 1980s aeration cavitation reduction measures have been applied to almost all high water head relief structures in China. The Codes for the Design of Spillway issued in 1990 stipulates that aeration cavitation reduction measures shall be applied to water relief structures with a flow velocity of more than 35m/s. During the same period, in-depth studies have also been carried out on pulse pressure and current vibration of high-velocity flow with measures to reduce erosion adopted accordingly.

PROSPECTS FOR THE DEVELOPMENT OF WATER RESOURCES UNDERTAKINGS IN CHINA

Water resources engineering is important infrastructure for the national economic and social development and is thus of overall and strategic significance. With the economic and social development new requirements are being put forward and the water resources undertakings will enter a new era.

Requirements

w Provide safety against flood for the economic and social development.

w Provide water supply for food safety, urban and rural domestic use and production.

w Provide a good water environment and water ecology.

w Provide clean hydropower for the adjustment of energy structure and implement “West to East Power Transfer”.

w Harness the rivers for water transportation.

w Realize development of water resources undertaking coordinated with economy, society, ecology and environment.

Basic Thinking

The water resources undertakings in the new era should seriously implement the strategies of sustainable development and prospering the country by relying on sciences and education, correctly deal with the relationships between water resources undertakings with economic and social development, ecological improvement and environmental protection. The principles of comprehensive planning, taking all factors into consideration, and looking into the root causes while solving a problem should be implemented, and comprehensive measures of river and lake harnessing, hydropower generation, water and soil conservation, rational development, optimized allocation, high efficiency utilization, effective protection and strengthened management of water resources should be adopted to provide powerful support and an important guarantee for the coordinated economic, social and eco-environmental development and ensure a sustainable, stable and healthy development of the socialist modernization of China.

Goals and Objectives

The goals and objectives for water resources undertaking in the early 21^st century (2030) are:

w Establish a sound comprehensive flood control system. Through the establishment of flood control engineering system and non-structural measures, the standards of flood control and capacity against flood and waterlogging will be promoted step by step, and the standards of main protection areas against flood will reach a level consistent with the level of economic and social development.

w Establish a safe, reliable water supply system. The national capacity of water supply will reach 750 billion m³, water saving will be intensified, the area of irrigated farmland will be increased by 8 million ha, the grain yield per m³ of applied irrigation water will reach about 1.5 kg, and the water consumption per 10000 yuan of industrial output value will be reduced to about 10 m³.

w Establish an effective system for water and soil conservation and water resources protection. 50-60% of the water erosion area in the country will be put under primary control and soil erosion will be effectively reduced. Effective protection of water resources will be carried out based on the water functional zoning, the control of water pollution will be strengthened, and the water eco-environment of rivers, lakes and reservoirs will be improved step by step.

w Rationally develop hydropower resources in the western regions. A sound mechanism for the development of hydropower resources will be established, that is, conducting rolling development on the basis of river basin by implementing cascade and multi-purpose development. By 2030 the total installed generating capacity of hydropower in the country will reach 130-140 GW and the rate of hydropower potential exploitation will reach 34-37%.

w Establish a sound system for water resources management, a comprehensive legal system for water resources and the development, management and utilization of water resources by law will be achieved, unified management of water resources will be realized so as to ensure sustainable economic and social development with sustainable utilization of water resources.

In the 21^st century, the water resources undertakings in China will further develop vigorously and large-scale construction of water projects will open a great horizon for hydraulic research. We hope that the hydraulic professionals in the world will join us in the construction of hydraulic works and hydraulic research and make contribution to the solution of global water issues and problems.

Flood disaster management

Erich Plate1¹ and Zhao-Yin Wang²

¹ em.Prof. Universität Karlsruhe

² Tsinghua University, Beijing, China

Introduction

The concept of „sustainable development“ has been introduced as guiding principle for future development during the Rio Conference on Environment and Development in 1992. The concept requires an economic and social development of humanity which satisfies the need of the present generation without compromising the need of future generations to satisfy their own needs. It is generally understood to mean that the resources of the earth should be used sparingly, and that Some of the principles implied by this new paradigm of sustainable development have been summarized in the recommendations of IAHR that were initiated by Prof.E.Naudascher (IAHR,1997). Their application requires considerable effort and some rethinking from planners of hydraulic engineering works and hydraulic engineers.

When one looks at the interaction of humanity with natural rivers, the concept of sustainable development has been translated to mean that we should adhere to the ancient principle of „Living with rivers“ and to use this principle for our modern approach to river management. During the Second World Water Forum, in den Haag in March 2000, the members of the Japanese delegation organized a section on “Living with rivers”, which was based on a paper prepared by Dr. Rodney White for an IAHR task force (see also GWP, 2000).

A further prerequisite for sustainable development is that societies and nations are allowed to develop in peace and without destabilizing influences that divert the economic resources of a country into emergency handling and away from improving social structures and development of infrastructures. Among the destabilizing influences are natural disasters, and in particular floods. Therefore the paradigm of sustainable development implies that we learn to manage floods in a way which saves human lives and property while minimizing the impact of our measures for flood protection on the natural environment. As a subset of “Living with rivers” this implies a new way of looking at floods from the perspective of “living with floods”, which means that we should look at the problem of flood protection in an integrated way, involving experts from all fields of science and administrations touched by floods. A possible way of introducing this concept into the planning of flood protection measures was practiced after the den Haag forum, in a workshop in Maputo on the spring floods of the year 2000 in Mozambique. The meeting was organized by the Japanese Organizing Committee for the Third World Water Forum (to be held in Japan in 2003), in cooperation with IAHR. International experts of flood management interacted with regional experts and planners in identifying the most promising actions for improved flood management and presented their conclusions to the political leaders of the country in a high level ceremony. Prince William, the crown prince of Holland, also present at the meeting, called this approach the “Maputo spirit”. The Maputo Spirit requires to develop concepts on how to handle flood protection problems with due regard to satisfying all needs in a sustainable manner, i.e. so that environmental impacts of human actions are minimized.

Application of the principle of sustainable development as expressed in the concept of “living with floods” requires integrated planning of river measures, so that the benefits that can be derived from the river are optimized, with a minimization of the impact on the natural fluvial environment, and with due consideration of flood protection and flood management. This is the planning task of designing and constructing a flood protection system. Application of the Maputo spirit also requires to learn to live with the residual risk, which means that a disaster which occurs because the available protection fails – meaning that uncontrolled flooding occurs – is handled with minimum possible damage and losses of human lives. This is the task of operation of an existing system. Planning and operation are different sides of the same coin, but because they involve different actors and different concepts they need to be treated separately. This is an important statement, because in many discussions of flood protection systems the two tasks are treated as if they were the same.

Flood risk management for an existing system

Flood risk management in a narrow sense is the process of managing an existing flood risk situation. In a wider sense, it includes the planning of a system which will reduce the flood risk. These two aspects of flood risk management will be considered separately, starting with the management of an existing system, which consists of the processes indicated in Fig.1.

Fig.1: Stages of operational risk management (Plate, 2000)

Risk management for the operation of an existing flood protection system is the sum of actions for a rational approach to flood disaster mitigation. Its purpose is the control of flood disasters in the sense of being prepared for a flood and to minimize its impact. It includes the process of risk analysis, which provides the basis for long term management decisions for the existing flood protection system. Continuous improvement of the system requires a reassessment of the existing risks and an evaluation of the hazards depending on the newest information available: on new data, or on new theoretical developments, or on new boundary conditions, for example due to change of land use. The hazards are to be combined with the vulnerability into the risk. The vulnerability of the persons or objects (the “elements at risk”) in the area which is inundated if a flood of a certain magnitude occurs, is weighted with the frequency of occurrence of that flood. A good risk analysis process yields hazard or risk maps, which today are drawn by means of Geographical Information Systems (GIS) based on extensive surveys of vulnerability combined with topographic maps. Such maps serve to identify weak points of the flood defense system, or indicate a need for action, which may lead to a new project. Other weaknesses of the system become evident during extreme floods. For example, the Oder flood of 1997 has indicated (see for example Kowalczak, 1999) that weak points contributing to flooding of a city in a flood plain not only are failures of dikes, but also seepage through the dikes and penetration of flood waters through the drainage system, i. e. through backwater into the sewerage system or water courses inside the city.

Risk analysis forms the basis for decisions on maintaining and improving the system, which is the second part of the operation of an existing system. It is a truism that a system requires continuous maintenance to be always functioning as planned, and new concepts of protection may require local improvements of the existing system. A third part of the management process is the preparedness stage, whose purpose is to provide the necessary decision support system for the case that the existing flood protection system has failed. It is evident that no technical solution to flooding is absolutely safe. Even if the system always does what it is supposed to do, it is hardly ever possible to offer protection against any conceivable flood. There is always a residual risk, due to failure of technical systems, or due to the rare flood which exceeds the design flood.

It is the purpose of preparedness to reduce the residual risk, through early warning systems and measures which can be taken to mitigate the effect of a flood disaster. An important step in improving an existing flood protection system is the provision of better warning systems. Obviously, the basis for a warning system has to be an effective forecasting system, which permits the early identification and quantification of an imminent flood to which a population is exposed. If this is not accurately forecasted or at least estimated early enough, a warning system for effective mitigating activities cannot be constructed. Therefore, it is an important aspect that systems managers remain continuously alerted to new developments in flood forecasting technology, and to be prepared to use this technology to the fullest extent.

The final part of operational risk management is disaster relief : i.e. the set of actions to be taken when disaster has struck. It is the process of organizing humanitarian aid to the victims, and later reconstruction of damaged buildings and lifelines.

Flood protection as a dynamic process

Historically, flood protection underwent a number of development steps, depending on the type of flood: a flash flood obviously required different responses than a flood which inundates the lower part of an alluvial river. Flash floods have high velocities and tremendous erosive forces, and only extremely solid structures can withstand their destructive force. The only way for escaping a flash flood used to be to get out of harms way by moving houses and other immobile belongings to grounds which are so high that no floods can reach them. Later on, banks were strengthened with riprap or concrete linings against erosion. The damage potential of flash floods is confined to the direct neighborhood of the river, the total damaged area usually is not very extensive – although due to the high velocities the individual damage to structures or persons caught in such floods are very high. In recent times, flash flood caused large losses of life only of people unfamiliar with the potential hazard, such as tourists, which camp in the mountain canyons. Flash floods can be avoided by flood control reservoirs. However, this is usually an option only if can be combined with other purposes, such as hydropower generation, because the cost of the reservoir have to be compared with the potential benefits.

Floods in alluvial plains of large rivers are controlled by dikes and polders. Velocities are comparatively low, and the main danger to life is from the wide lateral extent of inundated areas, as has been experienced for example during the floods in Mozambique in February, 2000, in which a large part of Central Mozambique south of the Zambezi river was flooded. In the earliest days, people responded to such floods by moving the location of their cities and villages out of reach of the highest flood which they experienced, or of which they had clear indications, such as deposits on old river banks along the flood plain. Typical is the situation in the upper Rhine valley between Basel and Mannheim, where one finds the old villages and cities always on high ground or on the high bank of the old river flood plain. And if an extremely rare flood was experienced, which reached even higher, then people had no choice but to live with the flood damage. In other areas, people learned to live with frequent floods: for example, in Cologne the low lying parts of the city near the Rhine used to experience regular floods and they were prepared for it. Their method of protection is called today object protection: protection through local measures, such as building houses on high ground, perhaps on artificially generated hills, such as the farmers on the North Sea, by temporarily closing openings with sandbags or brick walls, or just by moving one´s belongings to a higher level of the house.

Population pressure and lack of other farmland made people to move into the flood plain, and to protect themselves by means of dikes: already the ancient Chinese started to build dikes along their large rivers to protect farmland and villages. The Herculean tasks of diking the Yangtse and the Yellow river, against floods of unimaginable magnitude, united the Chinese people into a nation in which no longer the individual was responsible for his own safety, but where flood protection became a national task. However, the protection by means of dikes cannot be perfect, as dikes can fail, and floods can occur which are larger than design floods. In recent times, the failure of the dikes caused some of the largest flood disasters in the world. The Oder river flood of 1998 (Bronstert et al. 1999, Grünewald, 1998) comes to mind, but even more striking are the floods of China, with the floods on the Yangtse a very illustrative example. Table 1 (from Wang & Plate, 2001) gives a summary of historical floods on the Yangtse river, which in 1998 experienced one of the largest floods of the twentieth century. Through a superhuman effort, the Chinese people were able to protect the vast area of the lower Yangtse flood plain from being flooded, and managed to reduce the number of casualties to the smallest number of any comparable floods in the twentieth century – in spite of a dramatic increase in population in the affected area.

Year	Discharge at Yichang Station (m³/sec)	Return period (years)	Inundated area (km²)	Grand levee breaches (No)	Death toll (persons)
1788	86,000	140	70 counties		10,000
1870	105,000	>200			30,000
1931	64,600	10	40,000	300	145,000
1935	56,900				142,000
1954	66,800	10	31,700	60	33,000
1998	63,300	8*	3,210	1	2292

Table 1: Major floods on the Yangtse river, with the highest ever observed flood in 1870. Modern hydrologic measurements started in 1877. The recurrence interval of the 1998 event in terms of maximum discharge is about 8. The Yangtse river experienced about 7 floods of approximately the same magnitude between 1896 and 1998, of which the last four in Table 1 are examples (adapted from Wang, 2000)

But the data of Table 1 also reveal one of the most fundamental features of rivers: in flood plains they are not stationary, but tend to shift their beds continuously. When the large rivers of the world leave their mountain confinement, they carry large amounts of sediment into the flood plain, and due to their lower velocity deposit the sediment on the plain. Without interference by man, the rivers build up alluvial fans: moving across a fan shaped area over which they spread their sediments – a rather complex process which only recently has found some theoretical discussion (Parker, 1997). This is in conflict with the demands of settlers, who want to have the state of nature to remain unchanged, so that property boundaries are maintained forever. In fact, a study by the University of Bern (Hofer & Messerli et al., 1998) of the effects of river floods in the delta of the Brahmaputra and Ganges rivers in Bangladesh showed that people were less concerned with the appearances of river floods, which they had learned to live with, but with the shifting of the river banks during floods, which destroyed land on one side of the river and built up land without owner on the other.

The effort of keeping the large rivers of China within the boundaries set by the dikes is an extreme case of man fighting the rivers, rather than to live with them. For by confining the river between dikes, one also confined the area on which sediment could be deposited, and a gradual increase of the river bed between the dikes is unavoidable. This is illustrated by the fact that the Yangtse flood of 1998 was a flood with a recurrence interval of only 8 years. Yet in terms of stages in the middle reach between the cities of Yichang and Wuhan it was higher than the stage observed in 1954, and in many places the highest stage ever recorded. Tulla in his works on the upper Rhine knew the sedimentation problem of the alluvial Rhine, and he found an at least temporary solution by straightening the river: this increased the erosive capacity, and in essence moved the sediment problem downriver: since the sediment was not deposited in the upper Rhine, it had to be deposited further downstream. Fortunately, the Rhine is a small stream by comparison with the large rivers of Asia, and the sediment problem proved to be manageable. The situation in China is different: against the floods of the large rivers, in particular the Yellow river, the Chinese won many battles, but they had to suffer many setbacks when the rivers breached their dikes and in extreme cases found a new bed by destroying all settlements in its new course.

The case of the Yangtse river is not only the story of a fight against nature of epic proportions, it also is an illustration of the development of the technology of defenses against floods. Protection of the vast fertile lands of East Central China against earliest floods was sought through dikes, and when these proved ineffective, the dike system was supplemented by polders, into which water was to be stored when the flood stage exceeded critical levels. But the relentless growth of the population forced people to move into the polders: today, the polders are inhabited by many thousands of people, and during the 1998 flood, the largest flood diversion basin – the Jingjiang Polder with a surface area of 920 km² and storage capacity of 6 billion m³, which had been the main reason for the reduced number of losses in 1954 as compared to earlier floods of similar magnitude - was not flooded because of the opposition of the people living in the polder.

The different examples of adjustment to floods serve very well to illustrate that modern options for flood management are not absolute, but depend on three variable factors: the available technology, the availability of financial resources, and the perception of the urgency of the protection, which is embedded into the value system of a society. As these factors change with time, the options which one has to consider also change, and new paradigms of thinking may require new solutions to old problems. When one looks at the time development of a protection system – not only against floods, but also against all kinds of other hazards – it is evident that this is a circular process, as indicated schematically in Fig.2. A state of a system may be considered satisfactory at a certain time, meeting both the demands on the river as a resource and for protection against floods. But new developments take place, leading to new demands on the river. Side effects occur, which impair the function of the system and which have not been anticipated. After some time, the system is considered inadequate, and people demand action to change the existing conditions.

In this circular process, the determining factor of technology is self evident. When J.G.Tulla planned his momentous correction of the Rhine river between Basel and Mannheim, he was planning a task for at least two generations, with people who would be ordered to work on the river with shovels and wheelbarrows to create the long lines of dikes along the river. In modern times, such a task would be finished in a few years, with only a few professionals, such as drivers of caterpillars and other large earth moving equipment, with modern geotechnical engineering skills guaranteeing long lasting earth dikes. Furthermore, the scientific basis for planning changes with the advance in scientific knowledge and the translation from science into engineering. Remedial measures have to be planned according to the new state of the art. Hydrologic inputs have changed, or better methods of calculation require a new evaluation of the flood potential (or the hazards).

Fig.2: The cycle of responses to changing value systems and changing environmental conditions for water management.

When we look for further technological development in flood control, many new possibilities have become available through modern communication technology. Of great significance is the development of modern forecasting and early warning systems. The possibilities of remote sensing are just being recognized, and the technology for converting forecasts from mathematical models of meteorological weather situations into warning systems is being explored at many locations. Indeed, great strides have been made in forecasting and warning for large rivers, with fairly long lead times between forecast and actual occurrence (see i.e. Wilke, 1997), and hydrodynamic models are available which can rapidly convert meteorological precipitation forecasts into flood forecasts (Moore & Jones, 1997, Goeppert et al. 1998), whereas for forecasting flash floods, which requires to localize usually randomly occurring convective storms, the success has not yet been high (see i.e. Quiby and Schubiger ,1998, for an example of forecasting in the Alps). However, forecasting and warning is only one aspect of the possibilities of communication technology – it also permit the dynsamic operation of flood control systems. A reservoir for flood control can be controlled on the basis of forecasting results to provide maximum protection by chopping off the peak of the flood wave, or a series of barrages, such as on the Rhine river, can be operated through remote control to provide maximum storage in the system of barrages.

There also is the human influence on the system. The catchment may have changed: a rural area which was heavily wooded some years back is now cleared for agriculture, a patch of land used for agricultural purposes is converted into urban parking lots, agricultural heavy machinery compacts the soil and changes the runoff characteristic of a rural area. Today as always, a further important criterion is the availability of funds, i.e. the financial resources which can be allocated to flood protection; resources which usually have to come from public funds and are in competition with other needs of society.

But finances are not the only issue. Decisions for flood protection also depend on the changing value system of the society, starting with the solidarity of the non – flood endangered citizens of a country with those endangered by floods. For example, in the not so distant past the infringement on the natural environment by engineered river works usually has been accepted as the price to pay for the safety from floods. However, in recent times flood protection by technical means faces serious opposition, not so much because of concern about the long range geomorphic adjustment of the river (which is bound to occur sooner or later), but generated more directly from the fact that dikes and land development cut off the natural interaction of river and riparian border. The reduction of wetlands and the impairment of riparian border fauna and flora in many – particular in the developed – countries causes great concern of environmentalists and has led to a backlash against flood protection by dikes and reservoirs. For example in some parts of Germany people are actually talking about removing some of the existing flood protection works, and moving the dikes further away from the river is a technical approach favored by many. In other countries, complete removal of existing dams has been talked about as a means of giving back to nature what used to be hers, (but also because some people find the failure risk of a dam unacceptable). Pristine nature is assumed to have a right of its own that needs enforcement, in order to reduce the steady decline of rare species, and recreate habitats for wild life, which in the past were given up in favor of human development.

The recognition that the adjustment process is open ended - is a transient only in the stream of development - is part of the principle of sustainable development: while revising or constructing a flood protection system to meet our needs, this principle requires us to remember, that future generations may have other needs and other knowledge, and that we should not cast our solutions into immutable solidity, such as producing irremovable gigantic concrete structures, or permanently degraded soils. For a discussion of issues involving sustainable water resources management on the basis of the original Brundtland report (WCED 1987) see Jordaan et al.(1993) and Loucks et al. (1998)

Planning a flood protection system

Development of a flood protection system traditionally is a sectoral task, and in many countries of the world flood management is still seen as a task which has to be solved in a regional or a sectoral context. Regional: Agricultural areas are protected or not protected according to agricultural needs, flood control of a city is done without regard to the surrounding country side. Sectoral: highway planning takes little cognizance of the needs for flood protection. Roads can have a significant effect on floods. In rural areas of Germany, the tendency is to pave the roads used by farm vehicles, without due regard to the fact that the roads act as channels for rainwater. Or the roads can act in the opposite way by inhibiting runoff. For example, in the Mekong delta highways have been built on dams to keep them operational when one of the frequent floods occurs. However, the design has not given sufficient regard to the need of draining the water off after the flood, and the dams act like retention structures.

When we look at flood protection from the point of view of a modern decision maker, developing a new project starts with a set of guide lines which are based on the value system of the present society. In this setting, and in countries like Switzerland or Germany, environmental protection and flood management are tasks of similar importance, and the optimum flood control system is a compromise between these two conflicting objectives. To illustrate this process, the case of integrated planning for flood safety and a healthy environment as part of a sustainable project is shown schematically in Fig.3 (adapted from A.Götz, Swiss Institute for Water Resources. Personal communication).

Fig.3: Integrated project planning for considering flood safety and ecology as complementary objectives (adapted from A.Götz, oral communication, 1999)

The societal goal of sustainable development is converted into a set of objectives: objectives for the safety, and objectives for the preservation of natural functions. If an analysis of the existing situation is showing that the existing conditions meet the objectives, then the only action required is to keep it that way, i.e. to maintain the conditions and to prevent intrusion of external demands that could alter the situation to the negative. For example, to prevent settlement of a flood plain, it might be necessary to set up legal barriers. If the existing situation does not meet the objective, a process has to be initiated for improving the situation.

The next stage is the decision process for finding the best alternate which meets the objectives of the design. There are many cases when it is impossible to meet all requirements, in particular when constraints are set, which might be financial, social, or political. Then it is necessary to change the objectives, to make them to conform more to reality. In this manner, many well meaning nature preservation objectives had to be overruled, or protection objectives had to be set aside. Finally, the alternative selected is implemented, and the project is completed.

The response to the reassessment of the flood danger is the phase of project planning for an improved flood disaster mitigation system. Experts involved in risk management have to ensure that the best existing methods are used to mitigate the damages from floods: starting with a clear understanding of the causes of a potential disaster, which includes both the natural hazard of a flood, and the vulnerability of the elements at risk, which are people and their properties. The project planning aspect of risk management is summarized in Fig.4., which basically consists of the two parts: risk assessment, which yields the basis for decisions on which solution to use, and the implementation phase, which involves a great deal of activity ranging from the fundamental decision to go ahead to the complex of detailed design and construction. When this is accomplished, the flood management process reverts to the operation mode described in the first part of the paper.

Fig.4: Project planning as part of risk management

Hazard maps, as used for operational risk management are also the foundation on which decisions for disaster mitigation are to be made. Risk assessment does, however, not stop at evaluating the existing risk, i.e. with the analysis of the risk. The risk analysis process has to be repeated for each of the structural or non-structural alternatives for mitigating flood damage. Good technical solutions integrate protection of rural and urban areas, through coordinated urban storm drainage projects, stream regulation in rural and municipal areas with bridges and culverts designed to pass more than the design flood. Structures including reservoirs and dikes are usual technical options, but other possibilities adapted to the local situation also exist, such as bypass canals and polders on rivers. Risk assessment, for example, also includes to investigate the option to do nothing technical but to be prepared for the flood if it strikes: i.e. to live with the situation as is and be prepared for the floods.

It is obvious that the process of evaluating the risk depends on the technical or non-technical solution contemplated, and therefor, the risk mitigation step is not an independent third step in series with the second, but it interacts and the two are interdependent: the technical or non-technical solution is evaluated, the new hazards determined and the decision basis is enlarged by this analysis.

The decision for which of the possible alternatives to use depends on a number of factors, among which the optimum solution in the sense of operations research is one important factor. The classical approach for optimizing a cost function can easily be extended, at least formally (Plate, 2000) to the case of flood protection systems. But there might be other compelling reasons for deciding on a particular alternative, even if it is not cost effective for flood protection. One of these reasons might be the expected loss of human lives. Without question, the foremost purpose for flood protection must be the saving of lives, and cost effectiveness can only be of secondary concern.

Conclusions

A framework was given for classifying the different processes of flood management. It was found useful to distinguish three levels within flood risk management: the project operation level, the project design level, and the level of engineering decision making involving estimating the risk in the setting of a cost benefit analysis. The risk management process for the operation has been described extensively in previous papers – for example in Plate, (1997). Details therefore have been omitted – as have details of the structures design level, on which the writer has published a number of comprehensive papers. It was an interesting exercise to identify the different processes which contribute to the three different levels, and it was particular important to identify the changing conditions under which flood protection has been approached during different times. It was concluded that the natural environment is always changing, due to natural processes such as geomorphological modifications of a flood plain, or due to human interference, such as using the flood plain for agricultural purposes and cutting the flood plain into different regions by building dikes. Under such conditions, sustainable development is difficult to achieve, and the efforts which the Chinese population is making for preventing the large rivers of China to behave like natural rivers are cited as examples of non-sustainable development, which implies that the fight against the huge floods of the Yangtse and Yellow River will never be completely won, and also the less dramatic changes of smaller rivers like the Rhine need to be constantly observed and solutions for flood control adjusted to the changing conditions.

Acknowledgement

The paper is a modified version of a paper presented by the first author for theEuropean Conference on Advances in Flood Research, Potsdam, November 1 – 3,2000.

References

Bronstert, A. , A.Ghazi, J.Hljadny, Z.W.Kundzevicz and L. Menzel, (1999): Proceedings of the European Expert Meeting on the Oder Flood, May 18, Potsdam, Germany. published by the European Commission

Göppert, H., Ihringer, J., Plate, E.J., Morgenschweis, G., (1998): Flood forecast model for improved reservoir management in the Lenne River catchment, Germany. In: Hydrological Sciences Journal vol. 43, 1998, No. 2, pp. 215 - 242

Grünewald, U.et.al. (1998): The causes, progression, and consequences of the river Oder floods in Summer 1997, including remarks on the existence of risk potential. German IDNDR Committee for Natural Disaster Reduction, German IDNDR Series No. 10e, Bonn

GWP, (2000): Global Water Partnership: Towards water security: a framework for action. Presented at the Second World Water Forum, The Hague, Netherlands, March 2000.

Hofer,T.und B.Messerli, (1997): Floods in Bangladesh.- Institut für Geographie, Universität Bern, Bericht f. Schweizerische Behörde für Entwicklung und Kooperation

IAHR (1997): “Criteria for equitable allocation of benefits from water storage projects” News Bulletin , International Association for Hydraulic Engineering and Research, 1997/3

Jordaan, J., Plate, E.J., Prins, E., Veltrop, J., 1993: Water in Our Common Future: A research agenda for sustainable development of water resources. Paris: Unesco 1993

Kowalczak, P. (1999): Flood 1997 – Infrastructure and Urban Context in A.Bronstert et al.(eds.) Proceedings of the European Expert Meeting on the Oder Flood, May 18, Potsdam, Germany. published by the European Commission, pp.99-104

Loucks, D.P. et.al 1998: Task Committee on Sustainability criteria, American Society if Civil Engineers, and Working Group UNESCO/IHPIV Project M-4.3 Sustainability criteria for water resources systems. ASCE, Reston, VA. USA

Moore,R.J., and D.A.Jones (1997) Linking hydrological and hydrodynamic forecast models and their data, in R.Casale et al. (eds) Proceedings of the First European Expert Meeting on River Basin Modelling (RIBAMOD) published by the European Commission, pp.37-54

Parker,G. 1999: Progress in the modelling of alluvial fans, Journal of Hydraulic Research Vol.37, pp.805-826

Plate, E.J. 1997: Dams and safety management at downstream valleys. In Betamio de Almeida, A. and Viseu, T. 1997: Dams and safety management at downstream valleys. Balkema, Rotterdam, pp.27 – 43

Plate, E.J., (2000): Flood management as part of sustainable development to be published in the Proceedings of the International Symposium on Flood Defence Universität – Gesamthochschule Kassel , 20. –23. Sept. 2000

Quiby, J.C., and F.Schubiger (1998): Quality assessment of the meteorological forecasts for localized flash floods. In R.Casale et al. (eds) Proceedings of the First Workshop on River Basin Modelling (RIBAMOD) published by the European Commission, pp.73-80

Wang, Z.Y. and E. J. Plate (2001) recent flood disasters in china, (to be published in J. Water and Maritime Engineering)

WCED, (1987): World Commission on Environment and Development. Our Common future. Oxford University Press, Oxford, UK

Wilke, K. (1997) Forecast systems for large rivers – the Rhine River Catchment, in R.Casale et al. (eds) Proceedings of the First European Expert Meeting on River Basin Modelling (RIBAMOD) published by the European Commission, pp.105-126

SUSTAINABLE GROUNDWATER MANAGEMENT

Wolfgang Kinzelbach

Institute for Hydromechanics and Water Resources Management, ETH,

Zürich, Switzerland

Contents

w Global ground water situation

w What does sustainable mean in terms of groundwater?

w Tools for recharge determination

w Models and uncertainty

w Interfacing to economics

w Conclusions

Global water availability

w Accessible runoff 13000 km³/a

w Human withdrawals 4000 km³/a

w Human in stream use 3000 km³/a

w Groundwater available 2500 km³/a

w Groundwater used 800 km³/a

w Depletion of groundwater reservoirs 100 km³/a

Averaging is misleading

Stored Volume vs. Renewal Rate of fresh water resources

Surface water (lakes and rivers):

w Volume 104,000 km³

w Renewal rate 30,000 km³/a

Groundwater:

w Volume 10,000,000 km³

w Renewal rate 3,000 km³/a

Structure of Water Use

w Agriculture 69% (90% consumptive)

w Industry 23% (20% consumptive)

w Domestic 8% (20% consumptive)

Groundwater is special…

w Groundwater use is much smaller than surface water use, but

w Groundwater is a strategic resource for drinking water in the arid and semi-arid world

w Groundwater is practically the only resource available year round

w Sustainability problems are most severe in groundwater both in the context of quantity and quality

w The feasibility of increasing the resource is low

SUSTAINABILITY CONSTRAINTS

w Abstraction < Recharge

w Limitation of drawdowns (vegetation, subsidence, collapse of fractures)

w Prevention of saltwater intrusion/upconing

w Prevention of soil salinization, salt backflow

w Guarantee of minimum downstream flow (wetlands, vegetation, users)

w Prevention of groundwater pollution

GENERAL PRINCIPLE

w Withdrawal (Consumptive use) < Recharge (from precipitation and surface water infiltration)

w Considering the downstream: Withdrawal < Recharge–Minimum downstream requirements

Main Cause for Water Table Decline:

Large Scale Irrigation with Groundwater, Examples:

w Ogallalla Aquifer, USA

w North China Plain

w Karoo Aquifers, South Africa

w Aquifers of the Arab Penninsula

w Chad Basin aquifer

w Northern Sahara Aquifer System (SASS)

Typical decline rates 1 to 3 m/a

SALTWATER UP CONING

SALINIZATION DUE TO HIGH GROUNDWATER TABLE

SCIENTIFIC TOOLS FOR SUSTAINABLE AQUIFER MANAGEMENT

Methods for the determination of groundwater recharge

w Environmental tracers

w Remote sensing

Models and coping with uncertainty

w How to use models

w Quantification of uncertainty

Interface to socio-economic analysis

w Common pool

w Discounting

ENVIRONMENTAL TRACERS FOR RECHARGE DETERMINATION

w Tritium

w Tritium-Helium 3

w Chlorinated-Fluorinated Hydrocarbons

w SF6

w Chloride

Combination of methods for determination of recharge

w Water balance method (hopelessly inaccurate)

R = P – ET

w Chloride method (hopelessly local)

R = (D + c_p*P)/c_R

SCIENTIFIC TOOLS FOR SUSTAINABLE AQUIFER MANAGEMENT

Methods for the determination of groundwater recharge

w Environmental tracers

w Remote sensing

Models and coping with uncertainty

w How to use models

w Quantification of uncertainty

Interface to socio-economic analysis

w Common pool

w Discounting

Why Models

w Interpretation of data

w Interesting quantities only indirectly known

w Predictive capability

w Ease of scenario analysis

w Integration of all knowledge in one framework

w Creating coherence in projects

TYPICAL WEAKNESSES OF MODELS

w Uncertainty of parameters

w Non-uniqueness of calibration

w Unknown hydrological future

w Uncertainty of conceptual model

w Way out: Stochastic approach

Quantification of Uncertainty in Recharge Rate

w Given uncertainty in transmissivities and observed heads

SCIENTIFIC TOOLS FOR SUSTAINABLE AQUIFER MANAGEMENT

Methods for the determination of groundwater recharge

w Environmental tracers

w Remote sensing

Models and coping with uncertainty

w How to use models

w Quantification of uncertainty

Interface to socio-economic analysis

w Common pool

w Discounting

Tragedy of “the commons”

Alternative Approaches to Sustainability Needed

w Traditional neoclassical approach to optimal resource use (economic efficiency):

Maximize the Present Discounted Sum of Net Benefits

–Used Extensively in Cost-Benefit Analysis of Projects and Economic and Environmental Policies

w Main deficiency: costs and benefits in the distant future make no difference

w Alternatives to the traditional approach give more value to the future

Options (Potential in km³/a) (compare to 4000 km³/a)

w Water saving (agriculture1000, industry 160, households ...)

w Water conservation methods including rain harvesting (1000)

w Change of diet (?)

w Change of economic activity and import of „virtual water“ (presently ?)

w Desalination (presently 20)

w Inter basin transfer (100)

w Reallocation of people, population policies (presently 6)

w Gaining time by non-sustainable exploitation (presently 100)

Conclusions

w Sustainable management of aquifers is a burning problem

w New scientific tools are available to support the definition of sustainable water use

w Modeling will play a major role

w The stochastic approach allows us to stay humble

w Natural science has to interface to economics and implementation in order to be really useful

Rational Approach to Turbulence : THE CONTROL EQUATIONS OF INCOMPRESSIBLE TURBULENT FLOWS

Ge Gao¹ and Yan Yong²

¹ Beijing University of Aeronautics and Astronautics, CHINA

² Florida Atlantic University, USA

Abstract

This paper presents a preliminary study of incompressible turbulent flow using a unilateral statistical average scheme. As the ensemble average is taken on two groups of turbulence fluctuations separately, this average scheme is able to capture the first-order statistical information of the fluctuation field. Continuity, momentum, and mechanical energy equations are derived for the fluctuation field based on this valuable information. Concepts of orthotropic turbulence and momentum transfer chain are used to model correlation terms, and eventually leads to a complete set of equations of incompressible turbulence. These equations preserve the nonlinearity of typical turbulence and contain no empirical coefficients and wall functions. The mechanical energy equation, derived in the form of a series to reflect the typical multi-scale nonlinear phenomena, is able to describe statistical mean flow and coherent flow. Four benchmark turbulent flows, namely, plan jet, round jet, flat plate flow with laminar-turbulent transition, and backward facing step flow, are simulated for both statistical mean flows and coherent structures to verify the adaptability of the newly derived equations.

Introduction

Turbulence is one of the most challenging topics in the area of fluid mechanics. Such a problem has been the center of endless debate since its complexity has brought many different views, all of which, in common, have their complexity as well as inability to solve the problem. Currently, there are two major movements in the study of turbulence. The first one, orientated toward statistics, tries to model the evolution of averaged quantities of a flow. This community, following the glorious trail of Taylor and Kolmogorov, believes in the phenomenology of cascades, and strongly disputes the possibility of any coherence of order associated with turbulence. The second movement, orientated toward coherent structures, considers turbulence as coherence among chaos from a purely deterministic point of view.

Both aforementioned movements have their own perplexing questions. The coherent structures need primary theoretical explanation. The statistical models, on the other hand, have to face the closure problem, which is introduced by the conventional Reynolds average, and must rely on empiricism to introduce extra equations and empirical coefficients. Since there are no universal coefficients in any existing model to encounter the variety of flow cases, the closure problem has never been flawlessly solved. And herein raises the question whether the closure problem is a constitutive nature of turbulence or only a specific problem introduced by the Reynolds average. The follow-up questions are whether

Turbulence can be solved within the framework of Navier-Stokes equations, or new physical principles need to be introduced, and whether the present existing mathematical tools are good enough to obtain purely analytical equations of turbulence. The authors' answer to the above questions is as follows: the closure problem is really a specific problem introduced by the Reynolds average, and rational equations of turbulence can be obtained within the framework of Navier-Stokes equations.

It is the authors’ belief that the main reason that turbulence remains unsolved for more than a century is lack of direct first-order statistical information of turbulence fluctuations. The conventional Reynolds average scheme does not recognize this information, leaving researchers with no other choices than relying on higher order statistics. The present approach introduces a unilateral average scheme for the purpose of effectively extracting this information. In this approach, turbulent fluctuations are divided into two groups based upon a set criterion. The nonzero ensemble averages of individual group of fluctuations are used to define the momentum transfer chain and orthotropic turbulence. The former describes the momentum transfer from the mean flow to fluctuation flow, and then from fluctuation flow to molecular motions. The latter reflects the physics that turbulent fluctuations constitute non-isotropic viscosity for the mean flow. Both concepts direct the modeling of the correlation terms arising from the unilateral average and give rise to additional equations for the closure problem. As the number of unknowns is equal to the number of derived equations, no any empirical coefficients or wall functions are needed in this approach. Four benchmark flows, i.e., boundary layer flow (flat plate) and transition, free shear flow (plan/round jet), and separated flow (backward facing step) were simulated to verify the adaptability of the newly developed equations of turbulence. The same set of equations used in all flow conditions has produced promising results regarding laminar-turbulent transition, plan/round jet anomaly, and separation flow reattachment. Meanwhile, the coherent flow structures for the four benchmark flows cases are also simulated by using the same set of equations.

Unilateral Average Scheme

In spite of its complexity, turbulent flow, in the minimum scale in which the concept of continuum mechanics is still valid, satisfies the Navier-Stokes equations. In the case of a large Reynold's number, the nonlinearity of N-S equations usually leads to an infinite number of solutions for a set of given initial conditions. Let u_i be one of the random samples in the solution space. The corresponding fluctuating velocity is defined as

(1)

As a note, u, with or without sub-indexes, always denotes a velocity vector in the present analysis. The mean velocity in the above equation is obtained from the following ensemble average

(2)

in which N is the number of possible solutions of Navier-Stokes equations. Now, we divide the solutions into two groups according to a certain criterion. The fluctuation components in the two groups are denoted by and . The ensemble average of the fluctuations of the first group

(3)

is called the first drift velocity, where N_I is the number of solutions in the first group. Likewise, if N_II is the number of solutions in the second group, the second drift velocity is defined as

(4)

As the ensemble average of all fluctuations

(5)

is zero,

(6)

in which M_I =N_I /N and M_II =N_II /N, and

(7)

M_I and M_II are slowly-varying weighting functions of x and t. To have a better view of the drift velocities derived in (3) and (4), we use an assumed fluctuating flow shown in Figure 1 to observe their physical meanings under the time average. For simplicity, this fluctuation flow has only a finite time span. As the time average over the entire span is zero, it does not give any information of the fluctuating flow. However, the information can be extracted from the time average of those positive fluctuations. The question is whether the mean value of positive fluctuations provides sufficient information. If not, can additional information be found from the negative fluctuation field? As shown in the figure, . itself apparently does not cover all the information of the fluctuation field. The weighted average and , however, satisfy:

(8)

As the weighted drift velocities are symmetric, the information of the fluctuation field can be effectively represented by alone. For convenience, the weighted drift flows are called drift flows in short.

Fig. 1. Schematic fluctuation field with time average.

As the mean value of pressure fluctuations _i' vanishes throughout the turbulent flow, the weighted pressures and are also symmetric, i.e.,

(9)

Momentum Equations of Drift Flow

We now write the N-S equation for the i-th solution u_i as

(10)

The ensemble average of (10) for all the solutions is

(11)

where

(12)

Subtracting (11) from (10), we obtain the momentum equation for the i-th fluctuation

(13)

The weighted ensemble average of (13) for the fluctuations of the first group is

(14)

where

(15)

The weighted ensemble average of (13) for the fluctuations of the second group is likewise

(16)

where

(17)

The condition for (14) and (16) to be valid is that the M₁ is a slowly varied function of x and t. Such a condition should be met in the division of the fluctuation field.

It can be proved that the three correlation terms given in (12), (15), and (17) are related as

(18)

Substituting (18) into (14) and (16), respectively, we obtain

(19)

(20)

It is noted that the only difference between (19) and (20) is of the opposite signs of the last terms on the right of the equations. It reconfirms that weighted drift flows and are symmetric. Only one drift flow, say , is, therefore, needed in the present analysis.

Modeling

As usual, we model as the divergence of the turbulence stress tensor, i.e.

(21)

The momentum equation of the mean flow is, therefore, written as

(22)

We now model

(23)

in which

(24)

It is important to note that any type of divisions of the fluctuation field will lead to the same set of equations (19) and (20). The division, however, becomes unique once the modeling equation (23) is introduced. The rationale behind this modeling equation can be explained by substituting (23) into (19) and (20)

(25)

(26)

In the above equations, all terms, except for , are fluctuation related. is a necessary term in the momentum equation for the drift flows. Without it, and would be constantly zero if they were initially zero. This diffusion term represents the essential mechanism of laminar-turbulent transition in flat-plate flow. The modeling equation (23) seems to defy the common belief that the correlation terms of the fluctuation field is of a flow property rather than a fluid property. We notice that the correlation terms on the left hand side of (23) are for two groups of turbulent fluctuations. The division criterion, however, may well depend on the fluid property. As can be seen in the following development, the correlation terms of the two groups of fluctuations provide unique channel to transport momentum from the mean flow to the fluctuation flow.

Naming , , and , we now rewrite (22), (25), and (26) as

(27)

(28)

(29)

The above equations clearly demonstrate a momentum transfer chain that begins from the mean flow. The momentum of the mean flow is first transferred to the drift flows through divergences of laminar and turbulence stresses of the mean flow. These divergences have the same effect on the two drift flows. They not only drive the two drift flows into opposite motions, but also assure a definite connection between the two drift flows and their symmetric properties. The next level of the momentum chain is the transfer of the momentum of the drift flows, again through divergences of stresses of the drift flow, to molecular motion in a form of heat.

Due to the symmetric property of two drift flows, we drop the superscript I and II in the following discussion for convenience.

Orthotropic Constitutive Relationship

To determine the turbulence stress tensor, we first observe the following constitutive relationship between stresses and strains:

(30)

For an anisotropic turbulent flow, 36 constitutive coefficients C_ij are independent. These coefficients vary with the change of coordinate systems. As the drift flow is largely influenced by the turbulence mean flow, it is reasonable to assess that the drift flow field constitutes an orthotropic environment for the mean flow. In the orthotropic turbulent flow, there are also 36 nonzero constitutive coefficients. If the coordinate system is set in the three principal material axes n₁, n₂, and n_3, where n₁ is in the mean streamline direction, and n₂ and n₃ are on the normal plane to the streamline [3], the constitutive relationships becomes

(31)

To model the orthotropic turbulence viscosity, we introduce mean displacement vector of the fluctuation field, and define eddy viscosity tensor as

(32)

The constitutive equation with symmetric property is written as

(33)

where I, J, and K, taking the same values as i, j, and k, have no implication of summation, and

(34)

Equation (31) shall be written as

(35)

It is seen that there are only 6 independent constitutive coefficients for orthotropic turbulence flow. The relationship between principal orthotropic coordinates 1-2-3 and arbitrary xyz coordinates is shown in Figure 2.

Fig. 2. The relationship between principal orthotropic coordinates 1-2-3 and arbitrary xyz coordinates.

The strain-stress relationship in the xyz coordinates can be written as

(36)

where [T] is a coordinate transformation matrix between the xyz and principal material coordinate systems. The above equation is the same as (30), which has 36 nonzero viscosity coefficients.

Mechanical Energy Equation

Vector is defined as the averaged displacements of particles of the first group of fluctuations. Its direction is assumed to coincide with the direction of the maximum tensile stress of the mean flow. Its magnitude can be determined from an independent mechanical energy equation that describes the relationship between the kinetic energy change of the mean flow due to the dragging effect of the drift flow and the work done by the drift flow. As the absolute velocity of drift flow in a fixed coordinate system is , the difference of kinetic energy of the drift flow at x and x+can be expanded into the following Taylor’s series

(37)

The first term on the right hand side represents the mean flow kinetic energy change over displacement due to its own variation. Likewise, the third term represents the drift flow kinetic energy change over displacement due to its own variation. The second term represents the kinetic energy change due to the interaction of mean flow and drift flow, in which we are only interested in the following term for Newtonian fluids:

Since the constitutive relationship for Newtonian fluids is only related to the mean flow strain, this term stands for the mean flow kinetic energy change caused by the drag of drift flow. It is worthy to notice that the left hand side of the momentum equation of drift flow (28) is the acceleration of the drift flow conveyed by the mean flow and the right hand side is the force applied to unit mass of the drift flow. The work done by this force over mean displacement should be equal to the change of kinetic energy of the mean flow due to the drag of the drift flow, i.e.,

(38)

The independent equation (38) signifies dynamic equilibrium of the mean flow kinetic energy change and work done by the drift flow. The series form on the left hand side of the above equation corresponds to numerous scales of turbulence coherent structures up to the N-th order. For N=1, we have energy equation of the first-order coherent structure

(39)

This is the mechanical energy equation concerning the first order intermittent coherent flow. As is normally a first-order small variable, the statistical mean flow of weak turbulence can be obtained by removing intermittency through a smoothing scheme. In this smoothing scheme, eddy viscosity at one grid point is the average of those of adjacent grid points. For strong turbulence, the statistical mean flow cannot be obtained by the smoothing scheme because is no longer small. Instead, higher order terms needs to be included in (39) in order to obtain the mean behavior of turbulence. If N is infinity, we can obtain statistical mean flow in an absolute sense. It is seen that statistical mean flow and coherent flow are controlled by the same equation (38) but characterized by different order N in the equation. For moderate turbulence, N=2 can lead to statistical mean solutions.

For strong turbulence, the mean flow is in a state of being heavily cut and stirred. The entire kinetic energy of the mean flow becomes turbulence energy. This state occurs, for example, in the region near vortex tail of bluff body or the reaction zone of strong combustion. In this case, N=3 is needed for statistical mean solutions.

Care must be exercised when applying the mechanical energy equation. When calculating statistical mean flow, one may start with the first order energy equation. If coherent behaviors appear, higher order N should be used.

In the numerical computation, we need to use the coefficient of substance C_s such that

(40)

We assume that the displacement vector has effective length . For simplicity, we use a cubic element with dimension of 2 to describe the coefficient of substance. Such a cubit element has volume V=(2)³ and surface area S=6´(2)² . Now, we use V_e to indicate the volume occupied by turbulence flow, and S_T to indicate the area of those surfaces subjected to turbulence surface stresses. The coefficient of substance is then defined as

(41) In a general three-dimensional flow, the cubic element is full of turbulent flow, and all its 6 faces subjected to turbulence surface stresses. As V_e=V and S_T=S, C_s=1. In a two-dimensional flow defined on the x-y domain, the two surfaces perpendicular to the z-axis experience turbulence strength but not turbulence stresses. In this case, V_e=V, S_T=4´(2)² , and thus C_s=2/3. If the center of the element is located at the solid boundary, the affective volume V_e and S_T are both cut in halves. Therefore, C_s=1´1/4=1/4 for a 3D boundary layer flow, and C_s=2/3´1/4=1/6 for a 2D boundary layer flow. It is noticed that the length of is about 0.8d, where d is the boundary thickness. The adoption of the same coefficient of substance throughout the boundary layer can still lead to accurate results.

(27), (28), and (38) plus two continuity equations

(42)

and

(43)

constitute equations of incompressible turbulent flow. It is worthy to point out that these equations do not contain any empirical coefficients. Weighting coefficients M₁ and M₂ for the division of turbulence fluctuations into two groups are not explicitly appear, and thus need not to be determined.

Numerical Examples

In order to verify the adaptability of the newly derived equations of incompressible turbulence, we compare computational results and experimental data for the following four benchmark flows. Both mean flow and intermittent coherent flow are calculated for each flow field. A standard SIMPLE scheme and staggered grid scheme are used for the finite difference formulations. A second order central difference form is used throughout the computation to avoid possible numerical error contamination. As a result, computational stability relies on physical viscosity rather than numerical viscosity. All computations were performed on personal computers.

Plan Jet

Two-dimensional plane jet flow is a primary example to verify the spreading characteristic of typical free shearing flows. Many turbulence models, including the k-e model, use plan jets to adjust their empirical coefficients. There is no any empirical coefficient in our equations. Therefore, the computational accuracy of the spreading rate of jets is completely relies on the adaptability of the newly developed equations of turbulence.

In the computation, the width of the jet nozzle is 2-cm wide and Re=3´10⁴. The computation zone is a symmetric half domain, with length 1.0 m and width 0.3 m, divided by the jet central axis. The mesh size is 40´40 and the coefficient of substance C_s=2/3. Figure 3 shows the patterns of jet flow and its induced flow obtained through the first-order energy equation (39) and smoothing scheme. The straight streamlines within the jet boarders demonstrate that the low-order coherent structures and weak disturbance have been suppressed by the smoothing scheme. Figure 4 shows the velocity vectors between 0.4 to 0.96 m along the x axial direction. It is seen that the velocity beyond the jet is quite small. But there are still some coherent vortices outside the jet since the coherent velocity is much larger than the mean velocity there. The spreading rate, defined as the ratio of coordinates x/y at those locations where the velocity is one half of the maximum velocity along the y direction, is 0.105, which fell into the range of 0.10-0.11 measured in experiments [3].

Fig 3. Flow pattern of plane jet.

Fig. 4. Velocity profiles of plane jet.

Figure 5(a) through (k) are the computational results of equation (39) without using the smooth scheme. The mesh size is 200´150. This set of results vividly displays the step by step formation process of entrainment vortices along the jet boundary. The intermittent coherent patterns are very similar to experimental data [5, 6] and the photo shown in Figure 6.

Fig. 5 (a)

Fig. 5 (b)

Fig 5. Streamlines of coherent structures in plane jet.

Fig. 6. Photo of coherent structure of a jet.

Round Jet

Motivated by the plan/round jet anomaly, we simulated a round jet flow using the theory developed in this paper. We first transform all equations from Cartesian coordinates to cylindrical coordinates. The computational zone is one half of the 1-m long and 0.2-m wide physical domain. A 41´61 mesh is used in the computational zone with uniform grid lengths in the direction of the central axial. Much denser grids are used near the region smaller than 0.02 from the central axis in the y direction to accommodate strong shear force and large velocity gradient. The coefficient of substance Cs=2(r-h)/3(r+h), where r is the radius of the center of the control volume, and h is the half width of control volume. The simulated results show that the spreading rate for the round jet is 0.086, which again falls in the range 0.086 ~0.093 measured from experiments [4]. The plane jet/round jet anomaly has been successfully eliminated. Figure 7 clearly shows the jet flow and a surrounding vortex. Shown in Figure 8 (a) is the velocity field in which the flow with injection is very weak. Figure 8(b) is a detailed version of the velocity field. As the momentum is transferred in the horizontal direction, the surrounding flow is sucked into the jet. The speed of the original jet flow is gradually reduced because of the momentum loss. Constant sucking and mixing lead to continuous expansion of the jet flow, speed reduction, and flux increase along the downstream direction. Figure 9 gives the comparison of velocity profiles in the y-direction at various locations. A self-preservation is clearly evident.

Fig. 7. Streamlines of a round jet flow.

Fig. 8(a). Velocity field of a round jet flow.

Fig 8(b). Detailed Velocity field of a round jet flow

Fig. 9. Self preservation of velocity profile in round jet.

Boundary Layer and Transition

In the simulation of flat-plate boundary layer and transition, one normally encounters the difficulty to deal with laminar-turbulence transition in two directions. One is the change from laminar flow to fully developed turbulence along the wall direction. Another one is the change from the laminar sublayer to the region of logarithmic law in the direction normal to the wall. This is a real challenge to our newly developed equations of turbulence because they contain no empirical coefficients and no wall functions for possible adjustment.

Now, we consider a two-dimensional flow over a 6-meter long flat plate. The computational zone between the solid bottom and border of the top free flow is 3-meter wide. An 81´61 mesh is applied with an exponential distribution along the y-direction and a dense uniform distribution around the transition zone in the x-direction. The nearest 3-4 grids to the solid wall are within the laminar sublayer. Reynolds number is varying based on the length factor along the wall direction. Within 10mm thick near-wall region, the velocity at entrance has a laminar profile of 1/2 power, above which is a uniform velocity distribution with Re=1.0´10⁵ at the entrance. The Reynolds number at the exit is Re=2.2´10⁷. The coefficient of Substance C_s=1/6. Initial mean flow velocity is set to be 1 throughout the computation zone, while the initial drift flow is set to be zero. In the computation, some orderly phenomena, including repeatedly ejecting and rolling up, and intermittence, appear in the fully developed turbulence zone as unsteady quasi-periodic swings around a steady turbulence mean velocity profile of 1/7 power. It seems that this type of quasi-periodic swings symbolizes some coherent patterns, which is simulated by the closed equations developed in this paper. To eliminate the unsteady swings, a smoothing scheme is implemented to average the eddy viscosity coefficients of the nearest neighboring grids along the normal direction of mean flow streamline. Such a practice effectively leads to a smooth mean flow profile, in which the transition is shown around Re=8´10⁵. Shown in Figure 10 is the evolving process of the calculated mean velocity profiles and the thickness of the boundary layer around the transition zone. The comparison of the calculated mean velocity profile in the fully developed turbulence zone and Klebanoff's experimental data [7] is shown in Figure 11. To the authors' knowledge, all simulated details, such as the thickness of boundary layer, friction coefficients, logarithmic velocity profile, form factor, and turbulence stresses, are in good agreement with experiments.

Fig. 10. Velocity profiles in the transition zone of boundary layer.

Fig. 11. Comparison of simulated and experimental velocity profiles

When employing the second order energy equation, more steady and smooth mean flow profiles are obtained without using the smoothing scheme. All quasi-periodic variations of coherent flow disappear. This indicates that the coherent structures in the boundary layer of zero pressure gradient are dominated by the first and second order terms in the energy equation.

Figure 12(a) through (d) are several selected intermittent flow patterns of the boundary layer. In the computation, the same mesh, Reynolds number, and first order energy equation are adopted without using the smoothing scheme. The intermittent patterns are drawn by subtracting 0.8 U_e from the mean velocity at each grid point, where U_e is the freestream velocity. As the horseshoe vortex and bursting are typical three-dimensional coherent phenomena in boundary layer, they cannot be obtained from a two-dimensional calculation. But the two-dimensional results have clearly shown that the first-order energy equation (38) is the control equation for primary coherent structure.

Fig. 12 (a)

Fig. 12 (b)

Fig. 12. Intermittent patterns in boundary layer.

The above results demonstrate that the present set of equations is able to calculate broad range of boundary layer flows from laminar to turbulent through transition. Of cause, the boundary layer flows with pressure gradient, flow on rough surfaces, and three-dimensional coherent structures should be studied in details in the future.

Separation Flow over Backward Facing Step

The fourth example is a separation flow. According to case 0421 published in the 1980 Stanford Turbulence Conference [8], we calculated a flow over a backward-facing step. The separation flow is quite complex since there are curving accelerating flow, decelerating flow, adverse pressure gradient and favorable pressure gradient. There is extremely strong turbulence within the recirculational vortex tail zone due to the quasi- periodic vortex shedding and breaking, which leads to violent turbulence and drift velocity being much larger than mean velocity. There also exists a narrow strip of negative viscosity upon the upper border of the vortex, making the far down-stream velocity profiles much full.

The computational zone is 20-m long and 4-m wide. The step itself is 4-long and 2.5-m wide. A uniform mesh 100x60 covers the entire computational zone. Such mesh does not consider the effect of the boundary layer because the first grid near the wall is much thicker than the boundary layer. Such treatment may lead to larger friction on the non-slip boundary. However, it will not severely affect the shape of the recirculational vortex. Figures 13(a-p) demonstrate a complete process of vortex formation, growth, braking-up, and shedding. As shown in Figures 13(a-c), a single vortex gradually grows until it reaches the length of 7. Once the reattachment length reaches 7, an additional vortex core appears. In this oscillating double vortex structure, the second one continues drifting toward the downstream direction, gradually narrowing the width of the recirculation zone between the two cores. It eventually breaks away from the vortex next to the step and becomes multiple small-scale vortexes shedding downstream as shown in Figures 13(d-p). The remaining single vortex starts growing again, marking the beginning of the next quasi-period of vortex formation, breaking-up, and shedding process. For the present Reynolds number the shedding vortex length is about 35% of the total vortex length. If Reynolds number increases, the shedding tail will be longer. The calculation reproduces the vortex shedding process and explains that the shedding vortex tail zone is mainly controlled by higher order terms of the energy equation.

Fig. 13 (a)

Fig. 13 (b)

Fig. 13. Vortex evolutions behind a backward facing step.

When the third-order energy equation (38) is adopted, the vortex shedding and oscillation completely disappear and a steady vortex of length 6.8 is obtained (Figure 14), which falls in the experimental length data 6.5-7.5. If using the second-order equation, oscillation still exists in the vortex tail zone because the drift flow is stronger than the mean flow there. The source of turbulence in the tail zone is the vortex breaking rather than shearing stresses of the mean flow. The above results clearly show that the first-order energy equation (39) dominates primary coherent structures and, therefore, constitutes a large eddy model, and the higher-order equation (38) constitutes a mean flow model

Fig. 14. Statistical mean flow pattern of backward-facing step flow

obtained by higher-order energy equation.

Discussions

The objective of the present study is to present new equations of incompressible turbulence derived based on the physics of turbulence. These equations should be able to describe statistical mean flow and coherent flow simultaneously. This objective is difficult to achieve by the traditional modeling based on the Reynolds average. As the Reynolds average completely loses the first-order statistical information of turbulence fluctuations, the second order information becomes the only tool to study the full influence of fluctuations upon the mean flow. Such a practice inevitably encounters many difficulties. Empirical coefficients and the concept of gradient diffusion introduced in traditional turbulence have never been able to declare generality because they are lack of sound mathematical and physical bases. The research presented in this paper takes a very different path to study incompressible turbulence. The major theme is the introduction of the unilateral average scheme that divides turbulent fluctuations into two groups. The nonzero mean of each group represents the first-order statistical information of the complex fluctuation field. In the derivation of governing equations, we introduce dimensionless weighting coefficients M₁ and M₂ such that weighted drift flows and are symmetric. More importantly, M₁ and M₂ do not explicitly appear in the governing equations, and, therefore, need not to be determined. The symmetric property of the two weighted drift flows clearly demonstrates that the universal symmetric law found in nature also exists in such a highly random event as turbulence.

Perplexing turbulence research for a long time, the closure problem of endless higher order correlation terms is indeed associated with infinite number of turbulence scales created by nonlinear instability. As the Reynolds equations cannot be closed by any mathematical means, new closure method should be developed based on physical understanding of turbulence. The endless cascade down process of eddies and the decaying process of coherent structures from lower-order to higher-order are merely apparent disorder phenomena. Behind these phenomena, the authors believe that there is a deterministic momentum transfer chain starting from the mean flow to the drift flow, and eventually to random molecular motion. The physical meaning of modeling equation (23) may be clearly seen from the drift flow momentum equation (25). Molecular and turbulence viscosity terms in (22) are attributed to the momentum loss in the mean flow. The negative signs of these two terms indicate that they are source terms in the drift flow momentum equation (25). All the diffusive momentum of the mean flow has been transferred to the drift flow. It should be noted that the laminar viscosity term in the mean flow momentum equation does not directly lead to molecular dissipation because the moving direction of laminate in turbulent flow is not tangent to mean streamlines. The laminar diffusion process is thus carried out on the smaller scales of the drift flow rather than large scales of the mean streamlines. The laminar viscosity term in the mean flow momentum equation is a source term in the drift flow momentum equation. The numerical computation has shown that if the laminar viscosity term is excluded from the drift flow momentum equation, the transition process is no longer possible. The interaction of the laminar viscosity term and the nonlinear convection term in the mean flow momentum equation is the key factor for transition. in (25) transfers momentum of the drift flow to molecular heat motion through laminar viscosity. The term in (25) transfers momentum of the drift flow to molecular motion through cascade down process. These two terms complete the final stage of the momentum transfer chain. The authors believe that this modeling consists with physical reality of turbulence.

The conventional concept of Boussinesq’s eddy viscosity cannot properly treat many complex flows, such as flow with large curvature, solid-like rotation, and flow with strong body force. The shortcoming is due to the anisotropy of typical turbulence. The conventional models not only have to use empirical coefficients to correct the modules of turbulence length scales, but also cannot determine directions of the length scales. The isotropic eddy viscosity model, therefore, cannot treat anisotropic turbulence properly. Although the secondary moment models introduce anisotropic stress tensors, many cardinal problems, such as accuracy and generality of the scale equation and generality of empirical coefficients, still exist as the focus of difficulties. The unilateral average precisely preserves all the necessary information of turbulence in the momentum equation of the drift flow and the mechanical energy equation. In order to obtain the mean displacement vector, the following three assumptions are made:

w The direction of the mean displacement vector consists with the direction of the maximum tensile stress of the mean flow.

w The drift flow is orthotropic. The principal axis of the orthotropic coordinates coincides with the direction of the mean flow streamline.

w The anisotropic eddy viscosity tensor plays its role in the mean flow and drift flow simultaneously.

The merit of these three assumptions needs to be verified by numerical computations.

The concept of orthotropy is popular in mechanics of composite materials. One of the simple examples is the wood laminate. There is obvious similarity between wood grains and water grains. Introduction of the orthotropic concept into the modeling of turbulence viscosity greatly simplifies the matrix of viscosity coefficients through reducing 36 coefficients of three-dimensional anisotropic matrix to 6. Multiplication of the drift velocity and the mean displacement length vector provides 9 tensor components, in which six shear stress coefficients compose three engineering shear stress coefficients plus three normal stress coefficients. When the coordinates deflect from the orthotropic coordinates, 36 non-zero coefficients appear again like a usual anisotropic case. The eddy viscosity tensor and turbulence stresses obtained in this study are very different from the ones obtained based on the conventional isotropic eddy concept. The more series terms of the energy equation are used, the larger the displacement vector is. A large displacement vector will lead to accurate statistical mean solutions.

The independent mechanical energy equation (37) gives rise the relationship between the mean flow energy and the drift flow energy. It is known that the effect of the drift flow on the mean flow is to resist the mean flow motion. The work done by the drift flow over a displacement should be equal to the loss of the kinetic energy of the mean flow over the same displacement. This equation is unlike the conventional mechanical energy equation that is dependent on the momentum equation. The present mechanical energy equation is a bridge connecting the mean flow and the drift flow

The series form of the energy equation provides multiple discrete length-scales to describe infinite layers of turbulence structures and eddies. It is noticed that (37) is an algebraic equation for displacement vector. Its solution is discrete. So are the eddy viscosity tensor and turbulence stresses. In order to smooth turbulence mean stresses, we need to average the displacement vector at the end of computation. If the average is performed at each iteration step, the computation would give rise incorrect results. One of the examples is the simulated round jet flow, which failed to show the correct spreading rate. Apparently, the averaged solutions of (37) is no longer the solution of the original differential equation. This phenomenon deserves further investigation in view of philosophy and methodology in the study of nonlinear science.

The ability of simultaneously simulating mean flow and coherent flow of turbulence shows that newly developed turbulence equation has provided a hope to unite statistical and structure movements. Since the calculation of coherent flows can be performed on meshes of 10³ to 10⁵ grids, the original aim of large eddy simulation model is also realized. Of cause, for coherent structures, detailed study of three-dimensional calculations is needed.

Conclusions

The essential theme of the present approach is the application of the unilateral statistical average scheme to turbulent fluctuations. As the first step toward solving the closure problem, the first order statistical information is extracted from the fluctuations to characterize orthotropic turbulence and the momentum transfer train. It is assumed that the turbulence fluctuations constitute an orthotropic environment for the mean flow. This assumption not only reflects anisotropy of typical turbulence, but also considers the mean flow effect that dictates the preference in the turbulence environment. The momentum transfer train is introduced based on the physics of the momentum transfer from mean flow to drift flow through viscosity and finally to molecular heat. This physical description enables one to model correlation terms arising from the unilateral average and remove the need of empirical coefficients and wall functions in the equations of turbulence. The same set of equations is able to produce promising results in the numerical computation for quite different turbulent flow conditions, ranging from various mean flows to coherent flows.

The calculations of four kinds of benchmark turbulent flows, i.e. free shearing flow, boundary transition flow and separation flow, have proved that the set of equations may provide precise statistical mean results as well as vivid coherent structure flows on sparse meshes. Since several difficult problems have been solved by using the present equations, such as transition, plane jet/round jet anomaly and vortex shedding phenomenon, the wide adaptability of the equations has obtained initial proof. It is promising that the set of equations may be used for further theoretical and engineering study of turbulence.

As turbulence is one of representative problems of nonlinear science, the methodology used in the present study of turbulence, such as unilateral average, the symmetry of the drift flows, orthotropic eddy viscosity, momentum transfer chain, the series form of independent energy equation and its discrete solution, may provide reference of mathematical-physical method for general nonlinear science.

Acknowledgements

The authors wish to express their sincere appreciation and gratitude to Xiaoyan Pu and Weibing Li for their valuable assistance in numerical computation. The authors also wish to thank Professor W.L. Chow of the Department of Mechanical Engineering at Florida Atlantic University for his initiative and promotional efforts in this research.

References

Schlichting, H., Boundary Layer Theory, McGraw-Hill, New York, 1968.

Klebanoff, P.S. Characteristics of turbulence in a boundary layer with zero pressure gradient. NASA, Report 1247. 1956.

Wilcox, D. Turbulence Modeling for CFD, DCW Industries, Inc., 1993.

1980-1981 AFOSR-HTTM- Stanford Conference on Complex Turbulent Flows.

Sedimentation studies in China

-Present and Future

Zhao-Yin Wang¹ and Bingnan Lin²

¹ Dept. of Hydraulic Engineering, Tsinghua University and International Research and Training Center on Erosion and Sedimentation, Beijing 100084, China.

² International Research and Training Center on Erosion and Sedimentation, P.O.Box 366, Beijing 100044, China

abstract

Sediment management is always most challenging in the hydraulic engineering on sediment-laden rivers. Chinese people have accumulated abundant experiences in solving the problems like watershed erosion, heavily sediment-laden rivers, reservoir sedimentation, estuarine and coastal sedimentation, and debris flows and control strategies. This paper summarizes the sediment issues, research approaches and management strategies. The paper outlines a vision for the future sediment research in the new century, especially reporting on an ambitious plan of validation of sediment studies performed for the Three Gorges Project. The need for multidisciplinary approaches to conducting research is also discussed.

INTRODUTION

For milleniums, China has been confronted with serious sedimentation problems. An outstanding example is the strong aggradation that takes place as the Yellow River carrying annual sediment load of about 1.6 billion tons enters into its lower course on the flat Huan-Huai-Hai Plain. In time, a perched river was formed that frequently breached its levees. In the 2000 years preceding the 1950’s, there were on the average two breaches every three years, causing at each time very heavy losses in lives and properties. Another major sediment-laden river is the Yangtze River in Central China. The average annual sediment load is 462 million ton at Cuntan near the inlet of the Three Gorges Project Reservoir. Historically, the Yangtze River floods also subject the riparian people to heavy losses in lives and properties.

In recent decades, in an attempt to harness the major rivers, China has been carrying out a large program of hydraulic constructions on such sediment-laden rivers as the Yangtze, the Yellow, and the Pearl. As these rivers either have alluvial beds in their upper reaches or are fully alluvial in their lower reaches, many sedimentation problems are encountered in the various stages of planning, design and operation of the projects. Research leading to the solutions of these problems constitutes the bulk of the present-day sedimentation studies in China. Outstanding examples of the said hydraulic constructions include Gezhouba Project and Three Gorges Project (TGP) on the Yangtze River and Xiaolangdi Project on the Yellow River. In addition to the applied research, some fundamental studies have also been carried out.

More often than not, natural rivers are irregular in cross section, profile and alignment. Furthermore, the sediment they carry may vary in physical properties from place to place and its quantity of occurrence may also vary both spatially and temporally. These along with highly variable hydrological conditions make a large alluvial river a very complex system affected by many factors. Solutions of sedimentation problems arising from planning and design of large projects to be built on these rivers are bound to depend on a multitude of variables. Simulation or modeling methods are thus often employed to provide solutions to planning and design engineers, because more variables can thus be taken into account with less assumptions made. The sate-of –the-art of the science of sedimentation consists mostly of the results of laboratory and theoretical studies carried out under more or less simplified conditions. Verification with prototype data is limited. A pressing question to be answered is then “how realistic is our present-day knowledge in sedimentation?” Plans have been drawn up in china to collect field data for the study of impacts of the Three Gorges Projects on the Yangtze River with respect to sedimentation after closure of the dam in 2003.

Besides the traditional research interests of sediment transportation in rivers, the science of sedimentation is growing up into a more interdisciplinary science related to environment, economy and ecology. This paper reports the main achievement of the sedimentation management in the past and indicates the development of the science in the future.

Soil Erosion and Watershed management

Environmental health encompasses the maintenance and quality of the natural sources: soil, water, air and biota. According to the preliminary statistics conducted for the world, the annual erosion of surface soil from river basins amounts to 60 billion tons, of which 17 billion tons are discharged into the oceans. In the process, as much as 5 to 7 million ha of farms are annually ruined. Moreover, aeolian erosion of bare lands has intensified desertification after depriving the ground of good top soils. Eroded soil contains nitrogen, phosphorous and other nutrients and deposit in lakes and reservoirs, that contaminate the waters through eutrophication and other biological as well as chemical processes.

China is a country suffering severe erosion, with 1.82 million km² suject to water erosion, 1.88 million km² to aeolian erosion and 1.25 million km² to glacial erosion. In other words, more than 46% of the territory of the country are being eroded. The soil eroded from the Yangtze River basin is about 2.2 billion tons per year and that from the Yellow River basin 2.3 billion tons per year. The loess plateau in the central China is known for high rates of erosion and sediment yield, which result in numerous heavily sediment-laden rivers, such as the Yellow, Weihe and Huangfuchuan Rivers. From 1950 to 1996, Chinese people have improved 0.7 million km² of the water-erosion land, but in the mean time, the total water-erosion land has increased from 1.16 million km² to 1.82 million km². As a result the untreated erosion land remains at 1.13 million km² (Wang, 2000). The main causes for the increase of erosion land are deforestation, overgrazing, harvest of medical herbs, slope tillage, development of mining and urbanization. For instances, people in the Xiaojiang Watershed on the Yun-Gui Plateau of China cut trees and burn the wood for iron and copper production in 1958. The forest cover was reduced from 23% to 18% due to logging and the erosion rate was almost doubled. In the Yangtze River basin alone, the area of slope land people are still plowing is more than 11 million ha. The Shenfu-Dongsheng coal mining removed 162 million tons of soil in the period 1998-2000 and thus intensified the erosion rate greatly.

Erosion causes many problems. More than 5 billion tons of fertile soil are lost every year due to erosion, in which the amount of nitrogen, phosphorus and potassium is about 50 million tons, much higher than the annual national production of fertilizer (30 million tons). The farmland deteriorates and grain production is reduced by erosion, too. Moreover, the nutrients are transported with the sediment into the environmental waters and causes eutrophication and red tide. More than 240 harmful algal blooms occurred in the 1990s, causing huge economic losses. Erosion exerts the highest ecological stress on the vegetation cover. In the northern part of the loess plateau, the vegetation can hardly develop because the extremely intensive erosion carries off the top soil, on which the vegetation relies. In the areas with vegetation, such as the upper reaches of the Yangtze River, erosion damages and destroys the vegetation and scars the land surface.

To study the laws of soil erosion Chinese scientists have conducted laboratory and field experiments with artificial rainfall systems. They revealed the relationships of the erosion rate with the splash power of rain-drops, soil structure and composition, surface sealing and crusting, vegetation, slope and topography, as well as human activities (Cai et al., 1998). Based on the widely-applied numerical models like USLE, RUSLE, CREAMS, ANSWERS etc, Chinese scientists developed many soil erosion models for the typical erosion land, such as the loess plateau and the upper Yangtze River basin, and combined the models with GIS (Jin, 1995, Hu et al., 2000).

The main erosion control strategies applied in China are building sediment barriers in gullies and terracing the slope farmland in the arid and semi-arid areas, building sediment-check dams and reforesting the hills in the wet areas, and comprehensive reclamation of small river basins in both arid and wet areas. Comprehensive reclamation of small basins is the main strategy employed in the loess plateau. The area is arid and semi-arid and, therefore, reforestation alone is difficult to achieve the objective of erosion control. People built many sediment barriers and created productive warped farmland. Terraced fields enclosed with borders 20-cm high may trap almost all rainfall water and greatly reduce erosion. Impounding water with dam on rivers provides water for drinking, irrigation and reforestation. People plant grass on the slope and trees around the fields and roads. As a result, the sediment fed into the rivers by erosion from the loess plateau has been greatly reduced since 1984 (Gu,1994). Fig.1 shows the annual water and load of the Yellow River in the period from the 1950s to 1990s measured at the Lijin hydrological station (Wang and Liang, 2000). The reduction in sediment load in the 1980s and 1990s is due mainly to sediment trapping by reservoirs and sediment barriers, terraced fields and reforestation.

Reforestation is effective in wet areas (South China, for example) if mass movement of soil is controlled by multiple check dams. The success of reforestation relies more on the agricultural policy than technology. The change of ownership from community to private households has incited incentives in farmers on reforestation since the 1980s. After the 1998-flood event of the Yangtze River the state-funded reforestation projects of the upper Yangtze River watershed was sped up. In many mountainous areas people are still burning wood for cooking and heating. Planting shrubs and fast-growing trees in selected zones providing the local people fuel wood is an effective measure to protect the forest.

Fig.1 Variation of annual water and sediment load of the Yellow River (Sediment load is almost halved since 1985)

Training of heavily sediment-laden rivers

A river is considered as heavily sediment-laden if the ratio of annual sediment load to water is over 1 kg/m³. The Yellow, Weihe, North Luohe, Yongding, Liaohe and Daling Rivers in China, the Colorado River in USA, the Ganges and Burahmaputra Rivers in India and Bangladesh, and the Nile River in Sudan and Egypt are several examples of heavily sediment-laden rivers. These rivers are dynamic and difficult to harness. With the load/water ratio about 37 kg/m³ the Yellow River is one of the most heavily sediment-laden rivers. The history of river training of China is essentially a history of the people's struggle against the Yellow River floods because this river is the most disastrous and challenging. The experience of the Yellow River training is perhaps one of the most painstaking and brilliant chapter of Chinese history.

The Yellow River and Flooding Disasters

The Yellow River is the second longest river in China (5,464-km) with total drainage area of 752,000-km². The river watershed is mostly arid and semi-arid with long-term average annual runoff depth only 77 mm and total annual runoff 58 billion m³. The average annual sediment load before the 1980s was 1.6 billion tons, ranking first in the world with the recorded highest of 3.9 billion tons. The sediment in the lower Yellow River is mainly composed of silt and is liable to be suspended. The rainfall occurs mainly in July, August and September with high rainfall intensity.

The river carries a huge amount of sediment produced by soil erosion from the loess plateau. The sediment deposits on the channel bed and the estuary. In time, a perched river was formed that frequently breached its levees. From BC 602 to 1949 the river witnessed 1,593 dyke bursts, flooding vast areas in 543 years and claiming millions of human lives. The river shifted its major course (600-700 km long) 26 times with the apex around Zhengzhou with devastating calamities, left numerous old channels, among them 8 major shifts (5 natural and 3 human-caused) with river mouths alternating between the Bohai Sea and the Yellow Sea. For its wild behavior, the Lower Yellow River was dubbed “the sorrow of China”. Fig.2 shows the migration of the river in the period from BC602 to 1855 and the abandoned channels (Li 1992).

The Yellow River floods are very disastrous. For instances, a rainstorm in the middle reaches during 6-8 Aug. 1843 generated the biggest flood in the history. The crest discharge at Sanmenxia was recorded at 36,000 m³/s. Twenty-seven counties and 15,000 villages were flooded and thousands people were killed. From 1796 to 1855 the grand levees were breached 22 times and the major task for the river training was to close the breaches in this period. A flood from the upstream magnified by a heavy rainfall in the lower reaches breached the grand levee at Tongwaxiang on 18 June 1855. The river poured out and inundated 8 counties, and finally pirated the Daqing River Channel. Consequently, the Yellow River shifted its major course from south to north and flowed into the Bohai Sea. Thousands people were killed by the flood and several million people lost their shelters and farmland. In 1933 the river swelled again by heavy rainfall during 5-10 Aug. with a crest discharge of 22,000 m³/s. The flood caused 54 levee breaches, inundating 67 counties of flooded area of about 8,600 km² and killing 18,293 people. In 1958, 198 mm rainfall within 5 days enhanced the crest discharge at Huayuankou to 22,300 m³/s. The grand levee was not broken but 1,700 villages on the floodplain were flooded.

Fig.2: Migration of the Yellow River and the abandoned channels

Training and Regulation Strategies

The Yellow River training has a history of more than 3,000 years. Levee construction was the major strategy of flood control. The Qin Emperor united the country and linked the flood defense dykes into an entire levee system about 2,200 years ago. The lower Yellow River was confined within the levees but sediment deposition raised the riverbed and made the river frequently shifting its courses. People developed many strategies to harness the river, among them the wide channel and narrow channel theories are the most influencing. The wide channel theory is to confine the river within a wide river valley with levees and divert the flood with diversion channels. Wang Jing- a minister of the Han Dynasty- was the major practitioner of the strategy. From BC168 to AD69, the river was active; it flooded and changed its courses several times. Wang Jing implemented a large-scale training project from AD69. He completed and enhanced the dykes, built many diversion channels and weirs. The river was confined by the enhanced levees of tens of kilometers apart. The riverbed silted up at a low speed of about 1-cm per year. In the following 800 years the river was calmed and no big flood disasters occurred (Li 1992).

The second strategy is to narrow the river and confine the flood within the stem channel in order to raise the velocity and keep high carrying capacity of the flow, preventing sediment from depositing and even scouring the bed. Pan Jixun- a minister of the Ming Dynasty- was the most outstanding advocator and performer of the strategy. From 850 to 1600 the river behaved active again. Pan regulated the levee system, blocked many branches of the river and made the river flowing in a single channel in the lower reaches in the period 1565-1592. The river became relatively stable in the following decades.

Since the 1950s the Yellow River Water Conservancy Commission (YRCC) has been the leading institute for the river training (Wang, 1989). The nation has spent $1 billion for flood control and saved $500 billion in flood loss (Chen 1999). The riverbed profile has been maintained stable for half century, with only parallel rising following the extension of the river mouth into the sea (Zhang and Xie, 1985). The main strategies are to reduce flood discharge with reservoirs, enhance the capacity of the river channel by enhancing and reinforcing the levees, and retending floodwater with detention basins. They are referred in short as: upper reaches storing, lower reaches discharging and two sides retending.

The strategies employed are listed as follows: (1) Trapping sediment and controling flood with reservoirs- Dam construction and flood regulation with reservoirs is the most effective strategy. Eleven reservoirs have been constructed on the river with total capacity of about 57 billion m³. More than 10 billion tons of sediment were trapped reducing the sedimentation rate in the lower reaches. The newly constructed Xiaolangdi Reservoir with sediment-trapping capacity of 7 billion m³ may control sedimentation of the lower Yellow River for 20 years; (2) Wide river valley with reinforced levee system-The river in the Henan reaches is 5-20 km wide to accommodate sediment and retend water if rare floods occur; (3) Narrow channel to prevent sediment from deposition- The river width in Shandong reaches (0-600 km from the river mouth) is 0.4-5 km to maintain high sediment carrying capacity of the flow; (4) Enhancing and reinforcing levees-A total length of 1320-km of levees has been raised by about 9-m in the past 50 years, with a total amount of 400 million m³ of earth work and 4 million m³ of rock masonry work; (5) Diverting flood with flood-detention basins- Five flood detention basins were constructed. The Dongpinghu basin detended floods in 1954,1957,1958 and 1982 and effectively reduced the discharges to the lower reaches.

New Strategies to Train the Yellow River

On the 5 August 1996, a flood discharge of 7,860 m³/s washed through the lower Yellow River, breaking dikes and destroying 2,898 villages and 212 towns on the floodplain. The recurrence period of the flood with peak discharge of 8,000 m³/s is only 2 years according to 1950-1996 data. It created the highest stage in historical records and caused the highest economic loss. The main reason for the highest flood stage is the quick siltation of the main channel in the recent decades. The water flow was much lower than before because the peak discharge of floods was clipped by reservoir regulation. The river channel was in much less chances scoured by turbulent flood, thus the main channel is silted up at high rate, namely 0.1-0.2 m/year. In many sections the main channel is filled up. The flood did not flow down the river in a well-defined channel but flowed randomly within a valley confined of up to 10 km in width by the main levees. The flood took 17 days to travel the 800-km from Huayuankou to Lijin although the average time for floods of the same discharge to travel over the same distance in 1950-1990 was only 7-8 days. Moreover, the population on the floodplains within the levees has quickly increased to 1.7 million with rapid economic growth on the land, which explains the high economic loss.

The amount of water diverted annually from the river has increased to more than 30 billion m³ (70% of the total runoff). Less and less water flows and carries sediment down the river to the sea. Sedimentation control and maintaining the water conveying capacity of the channel are the main aims of the river training. Besides the traditional and currently-employed strategies, scientists and engineers have suggested and tested many new methods. Among them dredging and scouring sediment with seawater are the most promising auxiliary measures.

Lin et al (1999) proposed to take advantage of the topography and divert seawater laterally from the Bohai Bay to pumping plants that would lift the seawater for injection into the Yellow River. In a scheme proposed, the diverting canal is about 50 km long and the point of injection is chosen at Lijin about 110 km upstream of the existing river mouth. As the seawater diverted is practically free of sediment, it would degradate the river channel downstream of the point of injection. A local drop in water surface at the point of injection would be produced and would give rise to retrogressive erosion extending far upstream. This would stop further rising of the riverbed. In a scheme in which a seawater discharge of 1,500 m/s is injected at Lijin, it is estimated that retrogressive erosion would extend upstream by a distance of 330 km. Further upstream, the riverbed would be under the control of the newly completed Xiaolangdi project. Also noticeable is the fact that discharge of turbid seawater from the river mouth would assume the form of density current that could climb over the mouth bar and travel to more distant points of the Bohai Sea before deposition takes place. This would effectively stop or slow down the seaward extension of the estuary (Lin et al., 2000). The foregoing idea is presently under further investigation.

Dredging can be used to: (a) remove the sand bar from the river mouth; (b) widen and deepen a short length of shrinking channel at special locations; (c) raise the elevation of surrounding ground and reinforce the dykes with the dredged sediment. The Yellow River has shifted its delta channel (70-100 km from the river mouth) 11 times due to sedimentation and extension of the channel since 1855. The recent shift of the delta channel from the Diaokouhe channel to the Qingshuigou channel occurred in 1976. The ending reach of the mouth (16-18 km long) shifted to the Chahe Channel in 1996 mainly for the purpose of creating land to facilitate oil well drilling. The 100-km long Qingshuigou Channel has been used for 24 years, much longer than the average life of the previous delta channels, partly due to the dredging of the mouth sand bar in the 1980s and 1990s.

RESERVOIR SEDIMENTATION MANAGEMENT

Sedimentation of Reservoirs and Major Strategies

The problem of reservoir sedimentation peeps in Table 1, in which only some major reservoirs of capacity larger than 100 million m³ are presented (Qian, 1991). For small reservoirs the percentage of capacity loss due to sedimentation is even higher. The major strategies to control sedimentation and restore the capacity of the reservoirs are: storing the clear and discharging the turbid; flushing by draw-down and flushing by emptying the reservoir; and making use of density currents as well.

Table 1 Capacity loss of large reservoirs due to sedimentation in China

Reservoir/River Total capacity Year surveyed Reservoir sedimentation

(mi. m³) volume (mi.m³) (%)

Sanmenxia/Yellow 9640 1960-1981 5518 57.20

Yanguoxia/Yellow 220 1961-1978 160 72.70

Qingtongxia/Yellow 620 1966-1977 485 78.20

Liujiaxia/Yellow 5720 1968-1986 1078 18.85

Gongzui/Daduhe 357 1967-1987 286 80.11

Fenhe/Fenhe 721 1959-1988 327 45.30

Hongshan/Laohahe 2560 1960-1977 475 18.60

Guanting/Yongding 2270 1953-1985 612 26.96

Danjiangkou/Hanjiang 16050 1968-1986 1129 7.06

Sediment transportation in the Yangtze and Yellow Rivers occurs, of 60-90% of annual sediment load with 40-60% of annual runoff water, in 2-4 months of the flood season. The Three Gorges Project (TGP) on the Yangtze River is planned for flood control, power generation and inland navigation. For all these purposes, it is important to maintain an adequate storage in the reservoir. The main strategy to control sedimentation is to draw down the pool level from 175 m to 145 m in the flood season from June to September when the sediment concentration is high and allow the turbid water wash through the reservoir to the downstream. The reservoir stores water from October when the income water becomes clear. Fig.3 shows the typical variation process of sediment concentration at Yichang -the dame site of TGP- and the operation scheme of pool level for sedimentation control. By storing the clear and releasing the turbid, less sediment deposits in the reservoir while the reservoir is still able to store enough water for power generation in the low flow seasons. Numerical models proved that after 150 years operation sedimentation in the reservoir will reach an equilibrium, 17 billion m³ of the capacity will be lost due to sedimentation but 22 billion m³ can be permanently reserved.

Fig.3 Typical variation process of sediment concentration at the dame site of TGP

and the operation scheme of pool level for sedimentation control

For reservoirs on rivers with high sediment concentration and low water runoff draw-down flushing and empty flushing are employed. Low level outlets are open in the flood season, so to draw-down or empty the reservoirs and create riverine flows along the impounded reaches, which scour and release the sediment deposited in the reservoirs. Retrogressive erosion is induced by draw-down and empty flushing, which may extend the flushing far upstream of the dam. A successful example is the Sanmenxia Reservoir on the Yellow River, in which draw-down flushing caused retrogressive erosion to a 100-km long reach upper from the dam and effectively preserved the storage capacity. The Hengshan Reservoir on the Changyuan River in the Shanxi Province is a gorge type small reservoir with a capacity of about 13 million m³. The area is arid and there is almost no flow in the non-flood seasons. The reservoir was used to store water during the flood season and provide water for irrigation in the non-flood seasons. It was silted up quickly in the first 8 years and 30% of the capacity was lost due to sedimentation. Then, the reservoir was emptied for flushing sediment in the flood season of 1974 and 1979. Consequently, about 2 million m³ of its capacity was regained.

The Bajiazui Reservoir is on the Puhe River in Gansu Province. Density currents consisting of fine sediment occur in the reservoir if the inflow concentration is over 200 kg/m³. The high density mixture flows under the overlying clear water to the dam through the narrow and deep reservoir channel, then it is vented out of the reservoir through the bottom outlets. The ratio of the outflow concentration to the inflow concentration is about 100%. In other words, the high concentration of sediment can be discharged out of the reservoir without loss of the stored water. The Heisonglin Reservoir on the Yeyu River is another example, which discharges density currents with the ratio of the outflow concentration to the inflow concentration up to 91%.

Physical Modeling

The strategies controlling reservoir sedimentation are usually studied with physical and numerical models before it is employed. The Three Gorges Project (TGP) on the Yangtze River is the largest reservoir in China with the highest power generation capacity in the world. The TGP project benefits shipping allowing 10,000-t tows as well as passenger boats of 3,000-t class to sail to the inland metropolis Chongqing 620 km upstream of the dam. Thus twin flight of five locks each and a ship lift are provided in the left abutment of the dam to serve as permanent structures of navigation. Altogether 21 physical models have been built for the study of sedimentation problems related to the project, among them six are for the reaches downstream of the dam; one is for the Hanjiang River for a preliminary verification of the technique of model testing presently employed. The rests are for the sedimentation control of the reservoir and navigation channels. As the river is meander, the model would often take up the whole width of a laboratory, so that a large laboratory is generally needed to house a single model. Lighter materials, such as lucite, bakelite, coal powder and nut shell (treated), are used as model sediment in order to achieve similarity of sediment transport. In the case of TGP, movable-bed models are required to predict sedimentation in the lock approaches, training of difficult reaches in the backwater region, deposition in the great river port of Chongqing and others. Prediction of sedimentation in 80 years or more is usually required. This means that the models would have to be run continuously for several months in a row.

River models are as a rule distorted, except for the case in which the presence of structures has a strong influence on sedimentation. An example in point is the sedimentation in the lock approaches of TGP. In this case, undistorted or slightly distorted models are required. Physical model for the study of sedimentation is designed to satisfy the criteria of similarity with respect to gravitation, resistance, turbulent diffusion, sediment entrainment and deposition, capacity of sediment transport and bed deformation. The procedure is by and large conventional, save for the way to approximate similarity in the entrainment of sediment. Observed bed load in the sand range was plotted against mean velocity in the cross section. Two such plots were made, covering a range of transport rate of about 1.3 to 2,000 kg/s. The lower ends of these plots are extended downward somewhat to yield a velocity of 0.6 m/s, corresponding to a sediment transport of about 1 kg/s. This velocity is taken as the threshold at which sediment begins to move. It is scaled down to the model value for the model sediment to match. This is a rather common practice adopted in China.

In 1979, the late Dou proposed another procedure for physical modeling of sediment-laden flow (Dou, 1979). This procedure features unification of time scales for suspended and bed loads. Still many more schemes exist for physical modeling, including the tidal model of sedimentation for the Bay of Hangzhou and the large models for the study of the Yellow River.

Numerical Modelling

The Three Gorges Project on the Yangtze River is planned for flood control, power generation and inland navigation. For all these purposes, it is important to maintain an adequate storage in the reservoir. Therefore, estimating reservoir deposition or loss of reservoir capacity is of great importance. In 1972, Han presented a one-dimensional mathematical model of sedimentation based on the equation of non-equilibrium transport of sediment, which was first presented by Dou (1963) from a heuristic approach without formal derivation. Later, Han (1979) and Lin et al (1983) presented different derivations and arrived at the same expression:

(1)

where h is the depth of flow, S the mean concentration of sediment in a cross section, q the water discharge per unit width, the fall velocity of sediment, the mean concentration of sediment corresponding to the capacity of sediment transport and a variable coefficient. Also subscripts x and t denote the independent variables in partial differentiation. On the basis of this equation of non-equilibrium transport of sediment as well as the equations of momentum and continuity of the mean flows of water and mixture, Han developed a 1-D model for the computation of steady flows over a movable bed. In this model, Han specifies the values of to be 0.25 for the case of deposition and 1.0 for the case of erosion on the basis of considerable amount of field data. This model has been widely applied to the long-term forecast of deposition in the reservoir of TGP and degradation of the alluvial channel downstream of the dam. Depositions in the reservoir for a period up to 120 years have been computed (Han et al, 1993).

The coefficient was analyzed for two-dimensional cases (Zhou, 1990). Curves were given for the evaluation of the coefficient. In 1998, Zhou and Lin (1998) presented an extended 1-D model that can also yield approximately lateral changes of cross-sectional area. This model has also been applied to the computation of deposition in the lower part of the TGP reservoir. Many versions of 2-D models with sediment have been developed and applied in China, using Cartesian and boundary fitted coordinates. The former model has been employed for the case of movable lateral boundary. A 3-D model has also been developed. Its use is still limited.

The reservoir of TGP has three characteristic levels, namely, the normal pool level (NPL), the flood control level (CFL) and the dry season control level (DCL). At the beginning of the flood season in June the reservoir is to be drawn down to FCL for the releasing of floods. At the beginning of the dry season in October, the reservoir is to be impounded to NPL for full generation of power and for 10,000 ton tows to sail about 600 km to Chongqing. As power is generated, the reservoir will be gradually drawn down to DCL. It may be further drawn down to FCL when flood provides enough depth of flow for the safe passage of the tows.

Lin (1992) suggested a new scheme of reservoir management to lower the reservoir pool from 145 m to 135 m in the flood season when an incoming flow is between 45,000-m/s and 56,700-m/s. On the basis of the long-term average, there would be 7 days a year during which the pool in front of the dam would be lower than 145 m. During the short period the operation of the locks would have to be suspended. This scheme, if adopted, should start applying not later than the 11^th year after the commission of the project. In this way, deposition may be shifted forward and more sediment would then be carried by the flow and be discharged out of the reservoir. Thus there would be less deposition in the reservoir. Mathematical modeling indicates that by adopting the proposed scheme, an increase of storage in the amount of 3.5 billion m between elevations of 145 and 175 m may be obtained for flood control. This is 15.9 % of the initial storage of 22 billion m available between the same elevations. Except for the short suspension of lock use, the scheme is particularly good for navigation. It would increase the depth of flow in the reach of fluctuating backwater, and even more important it would also greatly reduce the deposition in the Chongqing harbor. As this scheme adopts another FCL at 135 m, it is dubbed double FCL scheme. Additional variants of the scheme is studied and proposed by Zhou et al. (2000).

ESTUARY SEDIMENTATION

Mouth Bars and Sedimentation in Navigation Channels

The Yangtze River is the navigation artery of China. Sedimentation at the river mouth generates the so-called mouth bars, which jam the navigation channels. As shown in Fig.4 the Chongming Island appeared in the river mouth 800 years ago and bifurcates the river into the North Branch and South Branch. Until the 18^th century the North Branch was one of the main discharge channels. Due to the Coriolis effect the mainstream shifted to the South Branch and the North Branch shrank in the past century. Furthermore, the South Branch is bifurcated into North Channel and South Channel again by the Changxing Island, which was born by sedimentation at the wide South Branch after a 100 years flood in 1860 (Le et al, 1998). A sand bar, which is named Jiuduansha Shoal, appeared after the 1954 flood and bifurcates the South Channel into North Passage and South Passage. Nowadays, 50% of the river water flows through the North Channel. The rest 50% flowing into the South Channel is divided again by the Jiuduansha Shoal half-to-half into the North Passage and South Passage. Nowadays, the North Passage is the main navigation channel because it is relatively stable.

Fig.4 The Yangtze River mouth and the dredging project improving the navigation channel

At the river mouth the runoff velocity is low and sediment deposits forming a huge sand bar under water. The mouth sand bar makes the navigation channel shallow (-7 m only) and large container ships can not navigate through it to the Shanghai Harbor and Nanjing Harbor. In order to create a 12.5 m deep navigation channel the government launched a project with a budget of $2 billion to dredge the north passage in 1998, as shown in Fig.4. Parallel levees and numerous groins are to be built along the north passage. Sediment is dredged from the channel and filled onto the enclosed Jiuduansha Shoal and East Hengsha Shoal to create land for harbor construction. The project will be constructed in three phases: in the first phase the channel will be dredged to 8.5 m deep, allowing the 3^rd and 4^th generations of container ships to navigate into the river; in the second phase, the channel will be dredged to 10.5 m deep and in the third phase to 12.5 m deep and a new harbor in the river will be created. The project will be completed in 2008 and the shipping throughout of the river mouth will then increase from 196 million tons to 350 million tons, including 16 million TEU containers. Dredging will continue to maintain the deep channel at an intensity of 10-15 million tons per year after the completion of the project.

Shrinking River Mouths

The Bohai Bay is an inland-sea water of China with 12 major rivers, including the Yellow, the Haihe, the Yongding and the Ziya, pour into it. Since the 1970s the water runoff to the river mouths have been greatly reduced because the quickly increasing water demand causes over-diversion of the river water. Moreover, about 600 reservoirs and thousands of wells, numerous dams and tide locks constructed in the past decades have changed the river flow, sediment load and fluvial processes. Consequently, many of the estuarine channels have been shrinking quickly. The runoff water of the Haihe River flowing into the Bohai Bay reduced from 7.3 billion m³ in 1950s to 0.17 billion m³ in the 1980s. The Yongding River began to be dry in 1960s and has become an emergency floodway.

Although the sediment load from the rivers to the estuaries is limited (2 million tons per year), a tremendous volume of sediment is carried into the river mouths by sea currents, which is initiated by waves from the surrounding silty coast. Because of the reduction of runoff discharge, the sediment deposited in the river mouths is hardly scoured away and the river outlets are clogging up. Studies find that there is a high turbidity belt along the coast, which acts like a sediment transportation belt and brings continuously sediment into the river mouths. For instances, a volume of 18-million m³ sediment deposited in the 11-km long channel below the Haihe Tide Lock from 1958 to 1989 (Fang, 1996). The channel was narrowed from 250 m in 1958 to 100m in 1990. And the channel bed at the tide lock was silted up by 6 m. The bed elevation of the mouth bar raised from -3.2m to +0.5m. The outlet of the Yongding River is filled with sediment at a rate of 3.64 million m³ per year. The river mouth sedimentation poses a great flooding risk to the highly industrialized zone.

Laboratory experiment and numerical modeling have been conducted to study the laws of sediment transportation and deposition at the river mouths. The present strategies are mainly dredging and scouring sediment by using tidal water. Dredging is undertaking at many river mouths but the dredged mouths are quickly silted up again in the next year. In the period 1981-1994, people dredged 7.36 million m³ of silt from the Haihe River mouth before the flood seasons but in the meantime more than 9 million m³ of sediment from the sea deposited in the river mouth. People store sea-water by using the tide lock during flood tide and released the sea-water to scour sediment during ebb tide. The strategy does not work well because the stored sea-water always carries sediment thus causes sedimentation in the upstream channel of the lock. Double guiding dike is a new strategy to cutoff the high turbidity belt, which is about 5 km wide along the coast depending on the wind and waves. If the strategy is adopted, high turbidity water can not enter into the river mouths and the river mouths will be prevent from siltation. The strategy is still under investigation.

Debris Flow and Debris Flow Control

Debris Flow Disasters

Debris flow is a wide-distributed and frequently occurred sedimentation disasters in China. Chinese people call the phenomenon as "dragon" which stand for power and irresistible. More than 800 counties (about 40% of the total ) have witnessed debris flows and more than 100 cities and towns have been hit by debris flows. According to an investigation, there are more than 10,000 debris flow gullies throughout the country. Rainfall debris flow frequently occurs in Yunnan, Sichuan, and Gansu provinces. The Xiaojiang watershed in Yunnan province has a drainage area of 3,220 km². There are 107 debris flow gullies in the area. Annually 100-2,000 debris flows take place and 10-40 million tones of solid material are carried into the Xiaojiang River from the gullies. Glacial debris flow takes place mainly on the Qinghai-Tibet Plateau. The Guxiang Gully in the plateau watches more than 10 glacial debris flows every-year. A glacial debris flow of huge scale occurred in the gully in 1953, with depth 40-95m and discharge about 28,000 m³/s (Du and Zhang, 1985).

Debris flows played many tragedies. A debris flow occurred in the suburbs of the Xichang City of Sichuan Province in 1981, which damaged 5 streets and caused more than 1,000 casualties. On July 8, 1984, a debris flow from the Guanmiao Ravine carried 60 huge stones of diameter 5-10 m and 430 stones of diameter 2-5 m and rushed down to the Nanping County town at a velocity of 9.2 m/s. It cut half of a three story building and destroyed a 1 m thick concrete wall of a prison. There are 1,368 debris flow gullies along the railways in China. About 300 debris flow disasters happened in the past 5 decades, which buried 41 railway stations. The railway transportation was cut off for 7,500 hours (Shen et al., 1991). On July 9, 1981, a debris flow with a “dragon head” 8-m high flowed from a gully to the Chengdu-Kunming railway at a velocity of 13.2 m/s. It carried rocks of several meters in diameter and the density of the mixture was estimated at 2.32 t/m³. The debris flow damaged the 110 m long Liziyida Bridge on the Dadu River, destroyed a pier and an abutment, overturned the No.422 passenger train and killed 300 passengers. More than 5 million dollars were lost and the railway was blocked for 384 hours. Debris flows also destroy forests and results in temperature difference enlarged. The debris flow areas become drier in dry season and suffer from more intensified rainstorms in wet season.

Studies on Mechanism of Debris Flow

Quite a few scientists adopt one or several constitutive equations to study the mechanism by assuming the debris flow a homogeneous non-Newtonian continuum. It is successful in some cases. For instances, the visco-plastic conceptualization of debris flow with fine materials explains the high gravel-carrying capacity, laminar flow, velocity profile with a plug, and the roll waves and intermittence of debris flow (Yano and Dido, 1965, O’Brien and Julien, 1988, Wang et al., 1990). The dilatant fluid model interprets the mechanism of supporting force for the moving gravel and stones and the particles velocity profile (Bagnold, 1954, Takahashi, 1978, Savage and McKeown , 1983).

To study the mechanism and control strategies of debris flow, Chinese Academy of Sciences established the Dongchuan Debris Flow Observation and Research Station on the Jiangjia Ravin of the Xiaojiang Watershed in 1988. Kang (1985) reported that debris flows in the Jiangjia Ravine occur during or after rainstorm in summer. A typical debris flow begins with torrential flood. Following erosion of the gully bed the flow develops from low-viscous turbulent flow into high-viscous laminar flow as the specific weight of the flowing mixture increases from 1.1 g/cm³ to 1.9 g/cm³. Highlight of the process is intermittent flow with a series of waves rolling downstream when the density reaches 1.9-2.3 g/cm³.

Field studies find that there two types of debris flow with distinguished dynamic characteristics and mechanisms: 1) Viscous debris flow, which is composed of clay, sand and gravel and exhibits non-Newtonian features. Such type is characterized by the striking phenomena of intermittent flow, “paving way process”, low resistance and drag reduction, extremely high superelevation at bends and well-mixed deposit materials. 2) Two-phase debris flow, which is composed of stones and gravel as the solid phase and the fluid mixture of water and low concentration of clay and sand as the liquid phase. Typical two-phase debris flow exhibits high, steep head consisting of rolling, colliding and noising large gravel.

Recent studies indicate that the resistance of debris flow can not be approached by using the constitutive equations because the viscosity and other rheologic parameters represent much greater resistance than the real debris flows. Fig.5 shows the Manning’s roughness n of viscous debris flows (o) and water flows (△) in the Jiangjia Ravine as a function of the depth. Drag reduction must occur in debris flow because the debris flow velocity is higher than the clear water flow at the same flow depth. The rate of drag reduction of viscous debris flows R_D is shown as a function of the gas concentration in the debris mixture. The rate of drag reduction R_D is defined as

(2)

in which n_w and n_d are the Manning’s roughness n of water flow and debris flow. For the viscous debris flow in the Jiangjia Ravine, the rate of drag reduction is as high as 60%. In other words, debris flows are 2 times faster than water in the same gully although debris mixture has much higher viscosity than water. The drag reduction is perhaps due to the paving way process (30%) and air cushion (30%) (Wang et al., 2001).

Fig.5 (a) Bed roughness of viscous debris flows (o) and water flow (△) in the Jiangjia Ravine as a function of the depth of the flows; (b) The rate of drag reduction of viscous debris flows as a function of the gas concentration in the debris mixture

Prediction and Warning of Debris Flows and Control Strategies

Prediction of rainfall debris flow can be made by using the 10 minutes rainfall intensity I₁₀ and a parameter of accumulated precipitation P_a, which is defined by (Chen, 1985):

(3)

in which P₀ is the precipitation just before the most intensive 10 min. rainfall, P_i is the precipitation on the day i-days before the debris flow. Debris flow occurs if:

(4)

The critical value P_c is different for different areas For the debris flow gullies in the Xiaojiang Watershed on the Yunnan Plateau, for instance, P_c equals 60 mm. By combining the results and rainstorm forecasting, debris flow can be roughly predicted.

Many detecting and warning systems have been developed for preventing railways, highways, bridges, factories and mines from debris flow disasters. For instance, vibration detector receives vibration induced by debris flow and transmits warning signal to the protected objects. Debris flow level detector can send warning signal to the protected objects when a debris flow is over a given stage. UJ-2 type ground sound wave probe and warning system were developed by the Chengdu Institute of Mountain Hazards Studies in 1984. The system can automatically work for 3 months with a group of batteries. The system had successfully sent warning signals of 12 debris flows to the protected area 2.8 km downstream.

The most important strategy to control debris flow is building dams. The Daqiao Creek lying on the right side of the Xiaojiang River was an active debris flow gully. Five detention dams reduce solid material transported into the Xiaojiang River and basically control debris flow. Nowadays lattice dam, window dam, slit dam and comb-shape dam are widely employed because, on the one hand, these dams can trap large boulders and mitigate debris flow disasters, and on the other hand, they have much longer life span because the siltation rate is much lower than normal dams (Kang, 1996). Debris flow flume is an important strategy to protect railways and highways from debris flows and is widely employed in China. In Gansu Province alone there are 25 flumes which guided debris flows across over the highways and railways and consequently protect the transportation arteries from damages. Another measure is to establish debris silt basins. By guiding debris flow into a given area, the farmland, hydraulic works and dwelling area are protected from disasters of debris flow. A 2.5 km-long diversion channel has guided debris flows from the Jiangjia Ravine to a deposition area and prevented the Xiaojiang River from blocking for twenty-two years. In some debris flow gullies willow piles are planted in the gully bed, with 1 m underground and 0.5 m emerged. Sediment carried by debris flows is trapped by the willow brush. While more and more sediment deposits in front of the willow piles the willows grow up and become strong. Thus debris flow is controlled. Comprehensive debris flow control projects are conducted in many debris flow watersheds. The projects are composed of reforestation, building dams, constructing debris diversion channels and basins on the debris cone and constructing reservoirs on the main stem and tributaries. The projects are very successful (Wang, 1999).

SEDIMENT STUDIES IN THE NEW CENTURY

Validation of Sediment Studies

In addition to the conventional hydrological gauging, collection of field data for future validation of sedimentation studies carried out for the Three Gorges Project (TGP) was started in 1993. The timing is a year ahead of starting the construction of TGP. It is to collect the data of the Yangtze River before the river is disturbed by the construction of the project. Reaches both up- and down- stream of the dam were observed. For the period from 1993 to 2009, a fund of over 30 million USD has been appropriated. Additional funds would be appropriated after 2009 for field studies to monitor the scheme of reservoir filling. The observation will cover, among other things, deposition in the reservoir, depth of water in the fluctuating backwater reach for the passage of large tows, deposition in the lock approaches, degradation of river channel downstream, especially the part of the river down as far as to Hankou. All these field studies will help validate or modify the results of the present studies and place people in a better position to answer how realistic is the present knowledge in sedimentation. It would take at least 10 years, i.e., up to 2019, to complete the initial phase of verification. Our knowledge on sedimentation should then be advanced by a large margin.

Closure of the TGP dam will be effected in 2003, when the pool will reach the elevation of 135 m. The normal pool will be raised further to 156 m in 2007. Although structurally the reservoir will then be ready for impoundment to 175 m in the year 2009, considerations based on sedimentation studies do not recommend hasty impoundment. It is suggested in the feasibility study, a period of several years was allowed for sedimentation observation before full impoundment to pool 175 m. Sedimentation engineers have recommended schemes for impoundment in steps. Execution of these schemes is to be monitored by field studies as well as real-time mathematical modeling. The main concern is the deposition in the upstream metropolis Chongqing. Under natural conditions, there is deposition in the port of Chongqing during the flood season. During the low flow season from November to March in the following year, the deposition is scoured away, thus maintaining equilibrium between erosion and deposition. It is estimated that change in this process will begin when the reservoir pool reaches approximately the elevation 160 m. Should harmful deposition be found thereafter, mitigation measures would be applied and observation continued.

Theoretical Studies

Sediment theories have been developing for about a century. Traditional sediment research programs have fallen out of step with the needs of the profession. Sediment engineering is moving toward becoming more multidisciplinary. The sediment researchers have to think bigger and broader. Unsteady sediment transport, environmental sedimentation and eco-sedimentation, and economic sedimentation are the new directions of theoretical studies in the new century.

The main theories and formulas of sediment dynamics were established based on steady and uniform flows. Nevertheless, the theories and formulas often fail to apply in engineering projects because sediment in nature is transported by unsteady and non-uniform flows. It is more often so following development of the application scope and requirement of high accuracy estimation of the rate of sediment transport. The parameter in Eq.(1) represents the mean concentration of sediment corresponding to the capacity of sediment transport. But it is calculated with the formulas for steady and uniform flows. In the numerical models, a coefficient is introduced to offset the error. However, new formulas, which can be directly applied in unsteady sediment transportation, are needed. Sediment-removing capacity is defined as the capacity of the flow to remove sediment from per unit length of a river section to other places per time. Differing from the well-defined sediment-carrying capacity, which is the feature of the mean flows and represents the amount of sediment load the flow can transport through the channel, the sediment-removing capacity is the feature of unsteady, non-equilibrium flows and represents the capability of the flow to change the channel shape and location. The sediment-removing capacity is found proportional to the fluctuation intensity of discharge of the unsteady flow (Wang and Wu, 2001). The movement of a river channel within the fluvial plain is defined as the river motion. The speed of the river motion is a function of the sediment-removing capacity.

Environmental sedimentation is an important aspect of sediment studies. Sediment-oxygen demand, contents of ammonia, nitrate, phosphorus, silica in sediment, sulfide, methane, and heavy metals, and the chemical interaction between water and sediment are the major indicators of sediment quality. A primary interaction is the exchange of solutes between the sediment and the overlying water. The flux-the transport of mass-of dissolved and particulate chemical species to and from the sediment are important components of the chemical and biological cycle that take place. For example, the consumption of dissolved oxygen by organic matter that settles to the sediment in the spring is usually the primary cause of summertime oxygen depletion in the bottom water of lakes and estuaries (DiToro, 2001). Eutrophication and harmful algal bloom (red tide) are often related to the nutrients from sediment. The study and modeling of the bio-geo-chemical process is one of the new job of sediment researchers.

Vegetation-erosion dynamics is a new interdisciplinary science, studying the laws of evolution of watershed vegetation under the action of various ecological stresses, especially soil erosion. By introducing the mathematical expressions of various ecological stresses, the vegetation development and the erosion process is modeled. The vegetation of a watershed or an area may exists in three states: vegetation developing and erosion reducing, vegetation deteriorating and erosion increasing, and the transitional state between the two. Human may change a watershed from one state into the others, the effort required depends on the distance of the present position to the destination position on the vegetation-erosion chart. The theory may predict the vegetation development of an area under the action of various ecological stresses and answer the questions: whether it is possible to permanently change the landscape by human activities and how much effort is required to achieve it (Wang et al., 2001).

Sediment is also regarded as a kind of precious resources (Qian, 2000). Sediment depositing in reservoirs cost capacity, but sediment transported to the estuaries can be used for land creation. The so called artificial peninsula project at the Yangtze River mouth will create 300 km² land by trapping the sediment from the river, which is of economic value of $10 billion. The Yellow River mouth channel was artificially shifted to the Qingshuigou-Chahe Channel in 1996 to create a land oil field with the sediment. Sediment value-addition is defined as the sum of all economic values (positive or negative) due to sediment transportation and deposition. The sediment at the abandoned Yellow River mouth is scoured by waves. A part of the scoured sediment is transported into the mouths of the rivers flowing into the Bohai Bay (see Chapter 5), causing shrinking of the mouths and enhancing the dredging cost. Therefore, wave breakers for protection of the Yellow River delta may not only stop the beach erosion and the land loss, but also greatly enhance the sediment value-addition. The study of sediment value-addition will help people for better sediment resources management.

CONCLUSIONS

The main erosion control strategies applied in China are building sediment barriers and terracing the slopes in the arid and semi-arid areas, sediment-check dams and reforestation in the wet areas, and comprehensive reclamation of small river basins in both arid and wet areas. The Yellow River was unstable and disastrous. It has been harnessed by employing reservoirs, levees, wide river valley and narrow channels, and flood diversion basins. The sedimentation problems in the TGP project and operation schemes are studied with physical and numerical models. With the strategy of storing the clear and releasing the turbid a permanent capacity of 22 billion m³ can be preserved. Numerical calculations recommend a double FCL scheme to flush more sediment and preserve 3.5 billion m³ more permanent capacity. Dredging of the Yangtze River mouth bar is performing for improving the navigation and allowing the 4^th and 5^th generations of container ships to navigate into the river. The shrinking of the river mouths by the Bohai Bay is a result of the movement of the turbidity belt along the bay, which is generated by silty beach erosion and tidal currents. The mechanism of debris flow is studied by field investigations and laboratory experiments. Flume, dam train, debris diversion channel and basin, and comprehensive control project composed of reforestation and engineering measures are proved effective in reducing and controlling debris flow disasters.

In the new century the sediment approaches and theories will be validated through continuous observation of sedimentation of the TGP reservoir with appropriated fund of over 30 million USD. The studies will place people in a better position to answer how realistic is the present knowledge in sedimentation and promote the discipline of the science by a large margin. The science of sediment transportation is moving toward becoming more multidisciplinary. Unsteady sediment transport, environmental and ecological sedimentation, and economic sedimentation will perhaps become the new directions of research in the new century.

Acknowledgement

The study is supported by the National Natural Science Foundation (No. 59890200) and the Ministry of Science and Technology of China (G1999043604).

References

Bagnold R.A., 1954, Experiments on a gravity free dispersion of large solid spheres in a Newtonian fluid under shear, Proc. Roy. Soc. London, 225 A, 49-63.

Cai Qiangguo, Wang Guiping and Chen Yongzong, 1998, Process and Modeling of the Soil Erosion and Sediment Yield in the Loess Plateau of China, Science Press (in Chinese).

Chen Xiaoguo, 1999, Flood control situation and strategies of the lower Yellow River, The Yellow River Water Conservation Commission (in Chinese).

Chen Q., 1985, Formation and characteristics of the debris flow at the Jiangjia Gully of Dongchuan in Yunnan, Memoirs of Lanzhou Institute of Glaciology and Cryopedology, No.4 pp.70‑79.

DiToro, D.M., 2001, Sediment Flux Model, John Wiley & Sons, New York/Chichester/ Weinheim/ Brisbane/ Singapore/ Toronto.

Du Ronghuan and Zhang Shucheng, 1985, A large‑scale debris flow in the Guxiang Gully, Tibet in 1953, Memoirs of Lanzhou Institute of Glaciology and Cryopedology, No.34.

Dou, G.R. 1963, Suspended sediment transport of in tidal flows and bed deformation calculations, Chinese Journal of Hydraulic Engineering, No.4 (in Chinese).

Dou, G.R. 1979, Physical model study of total sediment transport in river, Chinese Science Bulletin, v.24, No.14 (in Chinese).

Fang Xiufang, 1996,Suggestion on rebuilding the Haihe River Gate, Water Resource in the Haihe River, No.5.(in Chinese).

Gu Wenshu, 1994, Basic situation and law of water and sediment variation in the Yellow River basin and discussion of the measures for controlling soil loss, Soil and Water Conservation in China, No.7, pp.8-14.

Han, Q.W. 1979, On non-equilibrium transport of non-uniform suspended sediment, Science Bulletin, v.24, No.17 (in Chinese).

Han, Q.W. et al., 1993, Report on computation of and research on sediment deposition in TGP reservoir. Compendium of reports on key technical problems in sedimentation and navigation for TGP, Ministries of Water Resources and Communications, pp 488-564, Wuhan Industrial University Publishers, ISBN7-5629-9795-1/TV*2 (in Chinese)..

Hu L., Li R., Yang Q. and Shao M., 2000, Regional soil loss assessment model based on GIS, Journal of Basic Science and Engineering, Vol.8, No.1, pp.1-8.

Jin Desheng (eds), 1995, Experiments and simulations in geomorphology, Earthquake Press (in Cbhinese).

Kang Z., 1985, Characteristics of the flow patterns of debris flow at Jiangjia Gully in Yunnan, Memoirs of Lanzhou Institute of Glaciology and Cryopedology, No.4 pp.97-100.

Kang, Zhicheng, 1996, Debris flow hazards and their control in china, Science Press of China.

Le Jiazan et al., 1998, Analysis of the fluvial process of the Yangtze River mouth, Report of Shanghai Institute of Navigation Channel Survey and Design (in Chinese).

Li Jianshen, ed., 1992, Chinese Flood Control Series-General Introduction, Chinese Water Conservation and Hydropower Press (in Chinese).

Lin, B.(P.) N., Huan, J., and Li, X. 1983, Unsteady transport of suspended load at small concentrations, ,J. of Hydraulic Engineering, ASCE Proceedings, v109, n.1.

Lin, B. 1992, Preliminary study on a proposal to reduce training work for Chongqing and to enhance flood control capability of TGP reservoir----Double FCL scheme for reservoir management”, IWHR and IRTCES (in Chinese). Also Selected Works of B. Lin, China Water Resources and Water Power Publishers, 2001, ISBN 7-5084-0448-3/Z.39.

Lin Bingnan, 2000, Sedimentation in lock approaches of the Three Gorges Project, Report of International Research and Training Center on Erosion and Sedimentation.

Lin Bingnan, Zhou Jianjun, and Zhag Ren, 1999, Diverting sea water to scour the estuary- a quest for new approach to regulate the lower Yellow River, International Journal of Sediment Research, Vol.14, No.4, pp.1-10.

Lin Bingnan, Zhou Jianjun and Zhang Ren, 2000, Diverting seawater to scour the estuary and to regulate the lower Yellow River, Engineering Science, Vol.2, No.4, pp.25-33 (in Chinese).

O’Brien J.S. and Julien P.Y., 1988, Laboratory analysis of mudflow properties, J. of Hydraulic Engineering, ASCE, Vol.114, No.8, pp.877-887.

Qian Zhenying, 1991, Water Conservancy in China, Water Resources and Hydro-Power Press (in Chinese).

Qian Zhengying, 2000, Sellection of papers on water resources by Qian Zhengying, Chinese Water Resources and Hydro-Power Press, pp.206-216 (in Chinese).

Savage S.B. and McKeown S., 1983, Shear stress developed during rapid shear of dense concentrations of large spherical particles between concentric cylinders, J. of Fluid Mechanics, vol.127, pp.453-472.

Shen Shouchang, Zhang Meng, Li Fuxiang, Zhu Chonghai, Chen Guangxi and Wang Jiying, 1991, Debris flow along railways in China, Proceedings of the International Symposium

Takahashi T., 1978, Mechanical characteristics of debris flow, J. Hydraulic Div., ASCE 104, pp.1153-1169.

Wang Huayun, 1989, My River Training Practice, Henan Science and Technology Press (in Chinese).

Wang Zhanli, 2000, General Situation of soil and water loss and integrated control in China, Journal of Catastrophology, Vol.15, No.3, pp.18-25 (in Chinese).

Wang ZhaoYin, Lin Bingnan and Zhang Xinyu, 1990, Instability of non-Newtonian open channel flow, Acta Mechnica Sinica, Vol.22, No.3, pp. 266-276.

Wang, Zhao-Yin, 1999, Mountainous Hazards in China, German IDNDR-Series 18, German Committee for International Decade for Natural Disaster Reduction, pp.1-48.

Wang Zhao-Yin & Liang Zhiyong, 2000, Dynamic Characteristics of the Yellow River Mouth, Earth Surface Process and Landforms, Vol. 25, pp. 765-782.

Wang Zhao-Yin, Cui Peng and Yu Bin, 2001, The mechanism of debris flow and drag reduction, Journal of Natural Hazards, No.3, pp.37-46 (in Chinese).

Wang Zhao-Yin and Wu Yongsheng，2001，Sediment-removing capacity and river motion dynamics， International Journal of Sediment Research，Vol.16，No.2，pp. 105-115.

Wang Zhao-Yin，Wang Guangqian，Hu Shixiong and Gao Jing，2001， Vegetation-erosion dynamics and its applications， Submitted to Catena for possible publication

Yano K. And Daido A., 1965, Fundamental study on mudflow, Bull. Disaster Prev. Res. Inst., Kyoto Univ. Japan 14, part 2 pp.69-83.

Zhang Ren and Xie Shunan, 1985, The depositional profile of the abandoned channel and the main cause of continuous aggradation of the Lower Yellow River, Sediment Research, No.3, pp.1-10.

Zhou, J., 1990, New approach for the solution of the equation of non-equilibrium transport of suspended sediment, Prize paper, Proceedings, 7th Congress of APD, IAHR, Beijing, China.

Zhou, J. and Lin, B. 1998, One-dimensional mathematical model for suspended sediment by lateral integration, J. of Hydraulic Engineering, ASCE Proceedings, v. 124, n. 7.

Zhou J., Lin B. and Zhang R., 2000, Double FCL scheme for the management of TGP reservoir, Chinese Journal of Hydraulic Engineering, n.9 (in Chinese).

Hydraulic Engineering for Poverty Reduction and Resource Sharing

S. T. Su

Past President, Chinese American Water Resources Association;

Roland Way Oakland, California 94621 USA.

Tel: 510 636-2147 Fax: 510568-2205 E-Mail: ssu@harza.com

Poverty

Today we have more than two billion people in the world living in extremely poor conditions. No clean water, no electricity, not enough food, lack of sanitary facilities and no access to health care when they get sick. They live on less than a dollar a day and have a life expectancy of less than 40 years. This is the real tragedy of mankind -poverty!

How do we solve the problem of poverty in the world?

Many experts and policy makers conclude that economic growth is the best solution to solve the poverty problem. Economic growth is necessary to achieve higher living standards. Furthermore, the economic growth and development will integrate the poor regions with the global economy and improve the distribution of income and wealth.

The main channels through which economic growth contributes to poverty reduction are:

w Economic growth will shift resources to more efficient uses, create wealth and increase social stability;

w Expansion of employment opportunities combined with an increase in capital investment in building infrastructures, water and energy facilities resulting increased labor productivity and income;

w Better social services in the fields of basic education, health and hygiene as a result of increased government revenues.

There have been dramatic results on poverty reduction by economic growth. In Asia, faster economic growth and relatively balanced income redistribution resulted in a remarkable reduction in the percentage of people living in poverty. For example, the percentage declined from 30% to 10% in China, 60% to 15% in Indonesia, and 18% to 2% in Malaysia, over the 20-year period from 1970 to 1990.

Hydraulic Engineering is essential to poverty reduction.

Hydraulic engineering is the backbone of a sound economic growth. A well-designed economic growth would utilize local and global resources more efficiently to first meet people's basic needs and then increase people's wealth and quality of life. The basic elements of this economic growth are water, food, sanitation and electric power. This is why hydraulic engineering is so critically needed to lay the foundation for economic growth.

Hydraulic engineering is a very effective weapon in the fight for reducing and eliminating poverty. Although applying hydraulic engineering to poverty reduction has been demonstrated in the past, we need to do more and do a better job. We need to have a better understanding of the interactions between the structures and their surroundings, the effects of climatic changes, the social impacts, and the ecological context for sustainable development.

Global Resource Conflicts

The global demand for basic resources such as water, energy, and minerals will continue to grow in the decades ahead. Such growth will be driven by a combination of population increase, higher living standard and economic expansion. The human community is expanding by roughly 80 million people per year; at this rate, total world population will reach 6.8 billion people by 2010 and close to 8 billion by 2020.

Why do we need to prevent from global resource conflicts?

Unfortunately, the world's supply of the basic resources is rather limited. According to a recent study by the World Wildlife Fund, the earth lost nearly one-third of its available natural wealth between 1970 and 1995 as a result of human activity, more than in any other period in history.

Among many of the resource shortages, water shortage would have the most dangerous consequence and could lead to a global war. In areas of North Africa, the Middle East, and South Asia, the supply of fresh water is already proving inadequate for many human needs and creates a dangerous situation politically.

Hydraulic Engineering is essential to peaceful resource sharing

Take the water shortage as an example. Hydraulic engineering can provide practical solutions to water shortage by creating more supplies, providing better distribution systems, and enabling more efficient use of water. New technologies in desalination, loss reduction and irrigation will be the key to the solutions.

Peaceful and equitable water sharing depends on optimal water management agreed by river basin stakeholders from different countries. Hydraulic engineering will be the common link and the building blocks of deriving such an agreement.

Our goal for the 21^st century!

Let us, as hydraulic engineers and scientists, show the world what a difference that we can make by coming up innovative and practical solutions to poverty reduction and peaceful resource sharing. I urge we all take up this challenge as our goal for the 21st century.

MEASUREMENT OF EVAPORATION FROM BARE SOIL AND A NEW APPROACH TO SURFACE RESISTANCE

Shriyangi Aluwihare¹ and Kunio Watanabe²

^{1 2} Hydroscience and Geotechnology Laboratory, Faculty of Engineering, Saitama University 255, Shimo-Okubo, Urawashi, Saitama 338-8570, Japan

ABSTRACT

A new open chamber, which provides reasonably accurate estimate of evaporation, is explained. The uniqueness of this chamber is that it is completely open at the inlet and thereby the effect of chamber on the natural profiles of temperature, humidity and turbulence is minimized. On the other hand, a detail numerical model incorporating moisture and heat fluxes, which simulates bare soil evaporation together with surface resistance is briefly described. Simulated evaporation rates from an experimental bare plot were compared with the measured rate by the open chamber system. Although the model estimated slightly higher evaporation rates, the diurnal variations show a similar trend. The model is calibrated with the measured and simulated soil moisture of top 0 - 5 cm layer and a reasonable agreement is found. The simulated surface resistance does not show a proper relation with the soil moisture, but it can be proposed that hysteresis is present in the surface resistance-soil moisture relations.

Keywords: Open chamber, Numerical model, Evaporation, Surface resistance, Hysteresis

INTRODUCTION

Surface evaporation is one of the main processes in the air-land energy exchange, but due to its complex nature many unknowns remain to be solved. Evaporation process through bare soil surfaces is mainly controlled by the atmospheric conditions (Mohamed et al. 1997), surface soil wetness and moisture transport in the soil layer due to temperature gradients and soil moisture. Because of the involvement of these complex processes, for a deeper understanding and accurate quantification of evaporation from a bare soil need detail models of high vertical resolution and frequent measurements. Evaporation can be estimated by numerical models and by several experimental approaches as well. Most of the research efforts in evaporation are based on searching for new measurement techniques. Numerical models of heat and moisture budgets have been developed and applied (Camillo et al. 1986) to various problems. Recently most of the models incorporate the surface resistance, which quantifies the restriction to evaporation. Although functional relationships were produced between surface resistance and near surface soil water content (Camillo et al. 1986), actual formulation of surface resistance parameter needs proper modeling of atmospheric conditions and moisture transport due to temperature gradient and soil moisture. In this paper, a detail numerical model of heat and moisture budgets, incorporating near-surface air-land energy exchange and resistance parameters to evaporation, which simulates evaporation from a bare soil surface, has been described. On the other hand, chamber technique is one of the common methods in measuring evaporation. The errors related to evaporation estimates with chambers are mainly due to the alteration of natural profiles of radiation, turbulence, temperature and humidity (Mohamed et al.1997). The open chamber system (Aluwihare et al. 2000) described in this paper minimizes its influence on natural environment and thus gives better measurements. Intensive diurnal measurements were performed during a drydown period of 3 days after artificial wetting of an experimental bare plot. A comparison between the evaporation obtained by the numerical model and measured by the open chamber has been carried out. The diurnal variation of surface resistance has also been simulated and presented briefly.

THE OPEN CHAMBER SYSTEM

Figure 1 schematically illustrates the open chamber system used for measuring evaporation. The evaporation measuring equipment is based on the idea that when an air stream is injected to the chamber, the vapor flux from the surface into the chamber increases the absolute humidity of the extracted air. The evaporation rate is calculated in the following manner;

where, E is the evaporation rate, Q is the volumetric flow rate of air in the chamber system, and are the absolute humidity of the air after and before passing through the chamber respectively. A suction arrangement is used in passing the air through the system to avoid pump effects. The system mainly consists of two sections; an open chamber and a set of equipment for measuring evaporation. The interior dimensions of the Perspex chamber are; length 120 cm, width and height 50 cm each and the chamber is made of two 60cm long sections which can be connected in the field. The bottom of the chamber is open. The uniqueness in this chamber is that it is completely open at its inlet. To sample the inflow air for the estimation of its average relative humidity and temperature, a small amount of air is sucked by a tubel arrangement at the open end as shown in Figure1. Forty four (44) tube inlets are installed at the cross points of the wire net as shown. A small pump is used in sucking air through the tubes, provided at the entrance of the chamber. All tubes are connected to a small `box` type container, which is used to mix air for average measurements of average relative humidity and temperature of inflow air. As shown in Figure.1, the sampling arrangement is spread throughout the cross section of the chamber, facilitating sampling of the entire air profile at the inlet. The inlet has the same cross sectional area as the entire box, which reduces the resistance to flow. A guide box has been used at the entrance of the chamber to facilitate the incoming air to be properly guided. The ceiling of the chamber has eight measuring holes to insert sensors to measure pressure and air velocity etc. within the chamber at any level above the surface. All external pipes are insulated to avoid any air temperature changes while passing through the system. During experiments with the new equipment, the measurements were taken in 20 seconds intervals and finally a 5 minutes average value was stored in the computer. The wind speed in the system can be regulated with the pump situated at the extreme end of the equipment. The flow rate was measured by a flow meter (F) mounted on the pathway of the flow. A radiation meter is also mounted in the chamber, which is directly connected to the data logger. Inlet and outlet temperature, humidity, flow rate and radiation values were directly recorded with the computer.

THE NUMERICAL MODEL

A numerical model incorporating moisture and heat fluxes between a certain depth in subsurface and a reference height above soil surface is used in simulating the evaporation from a bare soil surface. The model is primarily based on the Milly (1984) equations. The basic equations for mass and heat fluxes are given respectively by;

where is the liquid water density, K is the hydraulic conductivity, is the matric head,, are the matric head diffusivity and temperature diffusivity of water vapor in the system defined according to Milly (1984), is the transport coefficient for absorbed liquid flow due to thermal gradients, k is the unit vector opposite gravity. accounts for the combined effect of simple Fourier heat diffusion and latent heat transport by temperature-induced vapor diffusion. L is the latent heat of vaporization of water, is the specific heat of liquid water, g is the gravitational acceleration and T and are the soil and reference temperatures. Following Milly(1984), the equation for subsurface mass transfer;

where is the volumetric water content, is the volumetric air content, is the density of water vapor and t is the time. The corresponding equation for heat transfer is,

where, C is the volumetric heat capacity of soil and W is the differential heat of wetting. When a local equilibrium exists between the liquid and vapor phases, the vapor density is given by,

where is the saturation vapor density, ψ is the relative humidity, g is the acceleration of gravity, R is the gas constant for water vapor and T is the absolute temperature. The soil thermal properties are derived following the approach of Campbell (1985) and the functional dependence of K, on is defined with Van Genuchten (1980) equations.

Boundary conditions

The upper boundary conditions applied in solving equation (4) and (5) with actual atmospheric fluxes under no rainfall conditions. By conservation of mass at the surface,

where, E is the evaporation rate at the surface which can be defined by the following relation,

where, and are the absolute humidities at the soil surface and the screen height from the surface and aerodynamic resistance , which is assumed to be equal in both vapor and sensible heat transport, can be expressed in the following form, (Camillo et al.1986);

where u is the wind speed at screen height, k is the Karman constant and is the roughness length. P₁ and P₂ are integral stability functions of for bare soil surfaces, where L is the Obukhov’s length defined in Brusaert (1982), in terms of u_* - the frictional velocity. Evaluation of stability functions P₁ and P₂ depends on whether the atmosphere is neutral, stable or unstable defined by the conditions, or respectively. The reader is referred to Brusaert (1982) and Camillo et al., (1986) for the approach in deriving P₁ and P₂ values. In this model P₁ and P₂ were solved by an iterative technique incorporating successive substitution. Considering energy conservation at the bare soil-atmosphere interface;

where, is the net solar radiation, LE is the latent heat and H is the turbulent sensible heat, defined in the following manner;

where, is the volumetric specific heat of air, and are the temperature at the surface and screen height respectively.

Surface resistance

Surface resistance can be defined as the resistance for the water vapor to diffuse from the source to a point just outside the surface. Assuming the topsoil layer to be wet during each simulation period, the surface resistance can be formulated as follows;

where, is the saturated absolute humidity at the soil surface temperature.

Method of solution

One-dimensional forms of partial differential water and heat equations were solved by the Galerkin Finite Element method. The soil domain consisted of a series of homogeneous layers of variable thickness, starting from 1 mm near the surface with thickness increasing with the depth. A fully implicit backward difference scheme was adopted in time discretization. The water and heat equations are solved alternatively to maintain the tridiagonal nature of the matrix. At the end of each time step the surface resistance was calculated using equation (12) The air temperature, relative humidity, net radiation and wind speed measured at the screen height should be available as inputs to drive the model. Evaporation and surface resistance, together with temperature and moisture profiles are the outputs of the model.

FIELD EXPERIMENT

To collect data to run the numerical model and to carryout the evaporation measurements using the open chamber, an experimental plot of size 1.2 x 0.5 x 1.2 m was selected in the premises of Saitama University - Japan. The plot was artificially filled with fine homogeneous sand (particle diameter 0.15 mm), and was isolated from the surroundings by plastic sheets. The plot was then saturated and drainage was allowed for a period of 24 hours with top surface covered with a plastic sheet for equal soil moisture distribution. Eight thermistors were used in measuring the soil surface temperatures, while another 8 were used at depths 2, 5, 10 and 15 cm in measuring the relevant temperatures below the soil surface in the plot. Meteorological data such as temperature of the air, relative humidity, net radiation were continuously measured at a point 25 cm from the surface. The open chamber was placed on the plot and measurements were started from 16 ^th October 2000 for a period of three days.

RESULTS

The accuracy of the open chamber

The accuracy of the chamber system was described elsewhere, (Aluwihare et al. 2000) a full description is not included in this text. But for a better understanding of the degree of accuracy provided by the chamber system, a comparison between the evaporative water losses from a pan kept under the chamber, recorded by a balance and the open chamber is produced here. The cumulative evaporation rate measured by the equipment and calculated by the balance recordings in the field is illustrated here. Figure 2 shows an average difference of 5% between the two methods before rain. During the period following the rainfall to the end of the experiment showed an average difference of 10.5%. The occurrence of a relatively high error after rainfall might be due to the instantaneous and spatial variations of humidity in the atmosphere, which is usual after a rainfall. The chamber affected net radiation was approximately 6% differed from the unaffected. The pressure inside the chamber relative to the ambient atmospheric pressure was found to be negligible.

Calibrating the numerical model

A good check on the model can be obtained if it produces physically reasonable results everywhere, any time. The only available soil moisture data was the 0-5 cm layer value. As the area of the experimental plot was restricted, it was not feasible to gravimetrically measure subsurface soil moisture. The soil moisture data of top 0-5 cm layer, is compared with the simulated and the results are shown in figure 3 for day 2 of the experiment. The soil moisture profiles show a physically realistic behavior and relatively good agreement.

The simulated and measured evaporation

The simulated and measured diurnal course of evaporation from the soil shows reasonable trend, although the simulated values show a higher evaporation rate (Figure 4). The subsequent heating and cooling of the open chamber itself might have contributed to the difference involved in the two profiles. As the numerical model is also subjected to several assumptions the difference shown between the evaporation profiles can be expected.

Surface resistance to bare soil evaporation

For a deeper understanding and quantification of the restriction to evaporation during drying, the surface resistance parameter should be well estimated. Figure 5 gives the relations between the surface resistance and soil moisture for day 2 of the experiment. It is visible that the surface resistance cannot be directly represented as a function of soil moisture. It is questionable that the surface resistance has a hysteresis effect on soil moisture on a given day as a hysteresis loop can be clearly identified.

CONCLUSIONS

A detailed numerical model incorporating heat and moisture fluxes has been developed to simulate bare soil evaporation. The simulated evaporation showed a reasonable agreement with the evaporation measured by a new open chamber system although the model showed higher evaporation. The subsequent heating and cooling in the chamber itself and several assumptions made in the model might have resulted in the difference. It is questionable that the surface resistance parameter can be represented as a function of soil moisture. We can suggest that the surface resistance parameter has a hysteresis effect on soil moisture in a course of a day.

References

Aluwihare, S., Mohamed, A. A. and Watanabe, K. (2000). Open chamber system for measuring evaporation: Pro. 55 ^th Annu. Nat. Conf., Jap. Soc. of Civil Engineers, III-A223.

Brutsaert, W. (1982). Evaporation into the atmosphere, Kluwer Academic Publishers.

Camillo, P. J. and Gurney, R. J. (1986). A resistance parameter for bare soil evaporation models. Soil Sci. 141(2), 95-105.

Campbell, G. S. (1985). Soil Physics with basic, Elsevier Science B. V.

Milly, P. C. D. (1984). A simulation analysis of Thermal effects on Evaporation from soil. Water Resources Res. 20(8), 1087-1098.

Mohamed, A. A, Watanabe, K. and Kurokawa, K. (1997). Simple method for determining the bare soil resistance to evaporation. J. Groundwater Hyd. 39(2), 97-113.

Van Genuchten, M. TH. (1980). A closed form of equation for predicting the hydraulic conductivity of unsaturated soils. Soil Sci. Soc. of Ame. J. 44, 892-898.

A Model for Simulation of Ice Jams

Kailin Yang¹ , Zhiping Liu² , Chujun Chen³ , Cuijie Liu⁴ and Guifen Li⁵

¹ Senior Engineer, China Institute of Water Resources and Hydropower Research, Beijing, 100038.

² Professor, China Institute of Water Resources and Hydropower Research, Beijing, 100038.

³ Professor, Northeast China Investigation Design and Research Institute, Changchun.

⁴ Senior, Northeast China Investigation Design and Research Institute, Changchun.

⁵ Professor, China Institute of Water Resources and Hydropower Research, Beijing, 100038.

Abstract

A model, based upon the assumption that the surface ice pans or floes entrained at the leading edge of an ice cover are well-distributed under the cover, is presented to simulate of the evolution of ice jams. This model can simulates the water temperature variation along the river; frazil ice concentration distribution; surface ice transport; ice cover progression and stability; deposition and erosion on the bottom of ice cover; and the thermal growth and decay of the ice cover thickness. For verification a case of the ice jams that occurred in the Baisan reach is studied using this model.

Keywords: Model, ice jam, river and Baisan reach.

Introduction

Ice jam is a local river ice process. Ice jams can occur not only during the freeze up period but during the breakup periods. When the progression of an ice cover stops at a cross section duo to the high flow velocity where there is an open water reach, and if a large amount of the upstream incoming ice, including surface ice and frazil in water sweeps under the leading edge of the ice cover and is deposited under the ice cover, an ice jam will be initiated. If the formation of an ice jam is caused by frazil granules, it is called as frazil jam or a hanging dam. Ice jams are the most dramatic events on northern rivers. They can cause water levels far higher than open water flood levels, with obvious consequences on riverside communities and nearby structures. Significant progression on the prediction of ice jams has been made in the last couple decades but much work still needs to be done.

Several models have been developed to simulate ice jams , which includes the two typical models: Beltaos and Wong’s model (1990), Lal and Shen’s (1991). Beltaos and Wong’s model is a one-dimensional, steady-state model that solves two simultaneous differential equations to compute the water surface elevation and the thickness of a jam as functions of river distance. Using the model, Beltaos and Burrell (1992) studied the ice jams in the Restigouche River caused by the surface ice pans or floes coming from upstream during the breakup period. Lal and Shen’s model is a dynamic model, which is composed of two major parts: a one-dimensional unsteady flow model and a one-dimensional thermal and ice condition simulation model. The model can simulates the water temperature variation along the river including the supercooling condition; frazil ice concentration distribution; surface ice transport; ice cover progression and stability; deposition and erosion of the undercover ice cover; and the thermal growth and decay of the ice cover thickness. The model is modified by Shen and Wang (1993), consider frazil and anchor evolution, and the transport and accumulation of the surface ice discharge entrained at the leading edge of the ice cover under the cover. Shen and Wang’s model may be in theory used to predict the evolution of frazil jams, but it is very difficult to use the model in practice because the sizes of the surface ice pans or floes are hard to be obtained.

This paper, based upon Lal and Shen’s model (1991), studies the transport, deposition and erosion of the ice particles under the ice cover, including the surface ice pans or floes entrained at the leading edge of the ice cover. A method is presented to simulate the frazil jam processes and verified by field data, including the water level, flow discharge and the sizes of the ice jams.

Mathematical Models

The mathematical models for the unsteady-state flow, the water temperature variation, border ice growth and surface ice transport as well as ice cover progression and stability used in this paper are similar to Lal and Shen’s model (1991). The models for the transport, deposition and erosion of the frazil particles, including the ice pans or floes entrained at the leading edge of the ice cover, are given as below.

A frazil jam will be initiated when a large amount of the surface ice is entrained at the leading edge of an ice cover of which the progression stops. Since the sizes of the surface ice pans or floes are different significantly and vary with the change of the weather, up to now one knows little for the distribution and motion of the pans or floes under the ice cover. To facilitate the analysis, an assumption is made that these pans or floes are well-distributed under the cover as the frazil particles suspended in water and the flow velocity of them is the same as that of the water. Thus the transport, deposition and erosion of the pans or floes can be described as the frazil particles.

For a one-dimensional flow, the distribution of the frazil particles under the ice cover or jam can be determined by the following transport equation

(1)

where = frazil concentration; = cross section-averaged water velocity; B = width of water surface; = net flow cross-sectional area; =distance;= time;= ice density; =probability of deposition of frazil particles reaching underside of ice cover; =buoyant velocity of ice crystals; and= rate of erosion of frazil accumulation layer per unit area. The first term on the right of Eq.(1) represents the rate of deposition of the frazil ice accumulation on the underside of the ice cover or jam, and the second term represents the rate of erosion of the frazil ice accumulation. The initiation of the movement of frazil deposition or erosion on the underside of the ice cover is governed by the critical velocity criterion, . If , or the flow velocity is greater than , the deposition event will occurs; if , the erosion event will occur; and if , neither the deposition event nor erosion one will occur.

An ice cover or jam may be divided into three layers: the snow layer or snow cover, the frazil accumulation layer or the lower layer, and the solid ice layer between the snow layer and frazil accumulation one.

The rate of change of solid ice thickness may be expressed as

(2)

(3)

(4)

where =latent heat of fusion of ice;=thickness of solid ice layer; = heat exchange coefficient between air and snow cover surface; = temperature of snow cover surface;= air temperature; = turbulent heat exchange coefficient between ice and water; = freezing point of water, ;=cross section-averaged water temperature; =thickness of snow cover; =thermal conductivity of snow; =thermal conductivity of ice; =porosity of frazil accumulation; and =thickness of frazil layer. If there is no snow on the solid ice layer, should be replaced by the heat exchange coefficient between air and the solid ice surface,, and thus represents the temperature of the solid ice surface.

The turbulent heat exchange coefficient,, may be determined by

(5)

where =depth of flow.

The rate of change of the thickness of frazil accumulation layer,, may be written as

(6)

The first term on the right of Eq.(6) is the rate of change of the thickness of the frazil layer caused by the deposition and erosion, and the second term is that caused by the heat growth and decay of the solid ice layer. When the water temperature is great than , the rate of change of may be obtained by

(7)

The surface ice flow discharge entrained at the leading edge of the ice cover,, is

(8)

where ,,, and = thickness of solid ice layer; thickness of frazil layer, cross section-averaged flow velocity, width of the open water surface in the river reach, and concentration of surface ice at leading edge of ice cover, respectively.

Provided that the surface ice pans or floes entrained at the leading edge of the ice cover are well-distributed as the frazil particles in the flow, the equivalent concentration, , of the ice particles in the flow, consisting of the surface ice pans or floes and the frazil crystals, at the leading edge may be approximated as

(9)

where =frazil concentration in water at leading edge of ice cover and = flow discharge at leading edge of ice cover.

With the models above the evolution of a frazil jam may be simulated. In the following the practical case of the frazil jam in the Baisan river reach is studied by this model.

Case study

BASIC SITUATION

The Baisan river reach is a part of the Songhua River in the northeast of China, where the weather is very cold in winter and snow is frequent and intense. Before the Baisan hydropower plant is built, the Baisan reach is a place where ice jams usually occur during the freeze up and during breakup periods in history. To assure the design, construction and operation of the Baisan hydropower plant, the engineers of the Northeast China Investigation Design and Research Institute carried out the field observation and measures on the ice regime in the winter between 1963 and 1964 (Pan 1964). As a result, the regional ice regime is fairly well understood.

The longitudinal profile of the Baisan river reach, measured in 1963 and in 1964, is illustrated in Fig.1. The reach is about 12 kilometers long from Song 7 to Song 35 and the widths of the water surfaces in the reach vary between 100 and 220 meters. There are the deep pits over 5 meter both between Song 15 and Song 17 and between Song 19 and Song 24, respectively. The profiles of four typical cross-sections are shown in Fig.2 .

Figure 3 illustrates the day-averaged temperatures recorded in the Baisan region in the winter that lasted from October in 1963 to April in 1964. The first snow was on October 3 in 1963 and the day-averaged temperature was still beyond zero degree. On November 8 the day-averaged temperature dropped down below zero and kept. Two days later, the surface ice pans appeared. At beginning, the areas of the ice pans were about 1 square meter, the thickness’ were about 0.1 to 0.4 meters. The surface ice concentrations were between 10 and 40 percent during the freeze up period. The ice bridge was initiated at Song 32 in the morning on November 26 . There a bend existed, the water depth was big, the flow velocity was slow and the border ice grew rapidly. Before the ice cover appeared, the width of the open water surface in the Song 32 section had reduced to 30 meter from about 170 meter and the surface ice concentration was about 30 to 40%. Those are the reasons why the ice cover at first formed there.

The Song 6 section, located on the upstream of the Song 7 section, started to be covered by ice on December 9, 1963. From then, little ice was transported to the cross-section Song 7.

Fig.1. Longitudinal profile of Baisan reach

Fig.2 Profiles of cross-sections

Fig.3. Day-averaged temperatures from October 1 in 1963 to April 30 in 1964

Next day after it appeared at Song 32, the ice cover progressed upstream to Song 19. Because the bed slope was big and the flow velocity was higher than 1 m/s, the leading edge stayed there for about two days. A large amount of the surface ice swept under the leading edge and were deposited on the bottom of the ice cover. As a result, the heavily thick ice jam, about 4 to 6 meter, was initiated between Song 19 and Song 24. At about 10:00 in the morning on November 29, the leading edge of the ice cover moved to Song 17. The flow velocity at Song 12 was always over 1 m/s so that the ice cover couldn’t progressed upstream for a long time. Even if it went upstream, the leading edge of the ice cover would turn back as the incoming ice was not adequate enough. The river surface between the Song 7 and Song 12 sections was only partially covered by border ice before the breakup period.

On forming during the freeze up period, the ice cover and jams in the Baisan reach could have hold in place by sections of intact ice cover before the breakup period in the end of winter. During the break up period, in April of 1964, the premature event occurred and then the thicker ice jams were initiated soon, which resulted in severe ice flooding.

BOUNDARY CONDITIONS

The boundary conditions include both the variation of the water level at Song 33 or at the downstream boundary and the flow discharge, water temperature and the magnitude and concentration of surface ice as well as the frazil ice concentration at Song 7 or at the upstream. At present the field data available are: a) the complete data for the water temperature, flow discharge and water level as well as thickness and concentration of the surface ice at Song 28 in the all winter; b) a partial data for the thickness and concentration of the surface ice at Song 7 from November 29 to December 9, 1963; and c) the complete data for water level at Song 33.

The relationship between the concentration of the surface ice and the air temperature at Song 7 may be linearly approximated as

(10)

The averaged thickness’ of the surface ice pans at Song 7 are usually about 0.1 meter.

The simulation in this paper is carried out from November 9, 1963, to the end of March, 1964. It is obvious that the boundary conditions at Song 7 are not complete and the lack of the data have to be added by those measured at Song 28. The concrete method is as given below:

w that the data for the flow discharge at Song 7 are approximated by those measured at Song 28;

w that the data for the water temperature at Song 7 from November 1 to 29, 1963, are replaced by those at Song 28, and from November 29 to before the breakup period the water temperatures at Song 7 are taken as zero since little ice is transported downstream after the river reach on the upstream of Song 7 had been fully covered by ice;

w that the data for the concentration of the frazil ice in water at Song 7 are approximated as zero because there is no data available.

Effect of snow cover

In the Baisan region snow felt frequently in the winter between 1963 and 1964. The ice cover was fully covered by a thick layer of snow during the all winter. As is well known, a light snow cover has a thermal conductivity at least one order of magnitude lower than that of ice, so it has a correspondingly higher insulating effect. In computation the thickness of the snow cover is determined by using the field data.

MODEL CALIBRATION

The composite Manning’s coefficients for the river bed and ice cover are calibrated using the data for the flow discharges and water levels measured. The initial values of the composite Manning’s coefficients vary between 0.02 and 0.0435 along the Baisan reach. In computation the Manning’s coefficient for the bed in a river section is assumed to equal the initial value of the Manning’s coefficient for the ice cover.

For the deposition of the frazil ice on the underside of the ice pans or floes that are transported downstream on the river surface, is assumed to be 0.001 m/s; for that of the ice pans or floes entrained at the leading edge of the ice cover, is assumed to be 0.0015 m/s .

Whether an ice cover progresses upstream or not is governed by the critical Frude’s number, . If , or the Frude’s number is greater than , the ice cover will progress upstream; otherwise, it won’t. In the computation is taken as 0.09.

In the simulation, the observed time of the ice cover initiation is used as the time of ice cover. The location of the ice cover initiation is set at Song 32.

The physical constants that are not subjected to much calibration are as given below(Shen and Wang 1993):

=0.4; ; =; =

; =; =

Results of simulation

Based upon the above conditions, the event of the ice jams that took place during the freeze up period in 1963 is simulated. The results are as given below.

Figure 4 demonstrates the curves of the water temperatures versus time at Song 7 and at Song 28, respectively. The origin of the time, , starts from on November 9, 1963. The curves of the flow discharges versus time at Song 7 and at Song 28 are shown in Fig.5.

The longitudinal profiles of the water levers are shown in Fig.6 and in Fig.7, in which the solid line represents the value of the simulation and dotted line represents the field data. The origin of the distance,, starts from Song 7.

The longitudinal profiles of the width-averaged thickness’ of the ice jam on December 6, 1963, and on March 25, 1964, are demonstrated in Fig.8 and in Fig.9, respectively. The results demonstrates that the shape of the computed longitudinal profile of the ice jam is similar to the measured one but their magnitudes have the significant difference at some position. Moreover, after it reached Song 14, the actual leading edge of the ice cover didn’t move upstream, but the theoretical leading edge of the ice cover will move to the upstream of Song 8.

Fig.4 Curves of water temperatures Fig.5 Curves of flow discharges

Figures 10 and 11 illustrate the curves of the water levels versus time at Song 17 and at Song 28, respectively. Obviously, the theoretical curve is well coincided with the curve plotted by the field data.

Fig.10. Water level at Song 17 Fig.11. Water level at Song 28

The simulation shows that on the given conditions the times when the theoretical leading edge of the ice cover moves from Song 32 to Song 19 and then to Song 14 are close to those ones observed on the spot.

CONCLUSIONS

This paper studies the evolution of ice jams. Based upon the assumption that is that the surface ice pans or floes entrained at the leading edge of the ice cover are well-distributed under the cover as the frazil ice in water is, a model is presented to simulate the dynamic processes of ice jams. The model has been verified by the field data for the ice jam measured in the Baisan reach. The results are satisfactory.

Acknowledgements

under contracts noThis research was supported by the National Natural Science Foundation of China . 59739170

References

Pan Shu Fang et al.(1964). “Field Measure on ice regime in the Baisan reach in the Songhua River during 1963 and 1964 “. Northeast China Investigation Design and Research Institute report.

Beltaos, S., and Burrell, B.C.(1992). “Ice breakup and Jamming in the Restigouche River, New

Brunswick”. Proc. IAHR Ice Symp., Banff, 437-449.

Beltaos, S., Wong, J.(1990). “Ice Jam Configuration: Second Generation Model”. National Water Research Institute Contribution 90-65. Burlington, Ont., Canada.

Lal, A.M.W., and Shen, H.T.(1991). “A mathematical Model for River Ice Process”. J. Hydr. Engrg., ASCE, 117(7), 851-867.

Shen, H.T., and Wang, D.S. (1993a). “A River Ice Simulation Model RICEN-Model Formulation and Program Guide,” Report No.93-7, Civil and Environmental Engrg. Dept., Clarkson University.

Shen, H.T., and Wang, D.S. (1993b). “Simulation and Analysis of Upper Niagara River Ice Conditions,” Report No.93-9, Civil and Environmental Engrg. Dept., Clarkson University.