In the Korean peninsula, which is located at the southern edge of north-east Asia, the 24th olympic game was held on Sept., 1988 in Seoul, the capital city of Republic of Korea. Since the Neolithic Age, the Han River of the 481 kilometer-long flowing through the heart of the Korean peninsula and bisecting Seoul into northern part and southern part, not only has been the nation's political, economic, industrial and cultural center, but also has played major roles in the national defence and security. The Han River also has seen brilliant cultures arise and flourish in its basin, which has left the area with abundant legacy of relics and scores of legends and myths.
To the Korean people, the river has been a fear complex of frequent flood damage. Perhaps this must have been a common conception for the most Asian. It seems for hydrologic characteristics of the rivers in Korea that rapid flow and meandering were intensive. For a long period of time, the Korean people lived with carelessness about the river management, thus there were many sufferings from floods and droughts damage which occurred in turn.
In the beginning of 20th century, a great flood attacked with unprecedented quantity in the Han River Basin. This flood occurred in 1925 and was recorded as the largest flood up to now. The purpose of this study, because the managements of the Han River Basin have been changed in several ways, is to analyze and to summarize its contents and to go far for river management in the future. The results of the study are summarized by the stages with definition of their major characteristics and activities as follows; Natural river condition with waterway transportation, (until 1910), Channel improvement works with water control (1910-1945), riverside development project with increasing of land use (1946-1960), water resources and hydropower generation project with water use (since 1960), comprehensive the Han River development projects with protection of environment, flood control, and water use (since 1980).

View full abstract

Flood routing techniques employing numerical models, especially those based on implicit finite difference schemes, have now been reasonably well accepted in practice. However, when they are applied to steep natural streams with slopes greater than 0.0005, the problem of numerical instability often arises, resulting in break-downs of numerical models. Major factors affecting numerical stability include streem slope, cross-sectional variability, tributary confluence, boundary conditions and initial conditions. The present article reviews and summarizes some of the research results obtained by the author in recent years. Stability criteria under various influencing parameters are given, and applications to field cases in Taiwan are also illustrated in this article.

View full abstract

A new method is developed in order to track and predict in short-term (15min-2 hr lead times) the spatial configurations, the velocity vectors and the rain intensity textures of the potentially flood producing rain fields that are detected on a radar scope. First, the rain field is decomposed into simple, tractable elements. Then a statistical adaptive forecasting scheme is developed in order to predict the evolution of each of these elements in time and space. The composition of these elements forms the complete rain field with respect to its spatial configuration, location and rain intensity texture at each prediction lead time. The method is currently being applied to the radar data of rain fields. Some preliminary application results are given.

View full abstract

It is important to clarify the mechanism of runoff process. The authors placed a small experimental basin in the Muroran district and measured rainfall, discharge, ground-water level, capacity of infiltration and water quality. In this paper, we reported the runoff characteristics of small watershed, which were obtained from the hydrological terms.

View full abstract

This paper studies primarily runoff simulation based on the concept of variable source area and the optimization technique for a mountain watershed located in Taiwan. The different hillslope lengths derived by virtue of kinematic wave theory, and the occurrence conditions related to the expanding and shrinking process between the overland and subsurface flow, are also introduced into the modeling procedure of runoff simulation. The proposed analytical method is finally applied to the study watershed in order to check its reliabilities.

View full abstract

The authors developed a new hydrological model for a mountainous basin. The applicability of the model is examined by comparing it with observed data from Shiozawa experimental basin in Japan.

View full abstract

To evaluate the role of non-homogeneous geologic condition on ground water runoff, spring discharge and electric conductivity of water (EC) were observed from April to December, 1988.
Although N and S spring are located very closely, each spring has different runoff pattern and EC value.
Considering the geologic condition, discharge and EC at N, S spring are estimated by multi-layer model.
It is indicated that ground water runoff system depeno on the depth of clay-layer between aquifiers.

View full abstract

It is usually considered that the runoff discharge consist of two components i. e. direct runoff and groundwater runoff. In this paper, these runoff are separated by Kinematic Wave Method and tracer method using δ¹⁸O data observed in Kanedaira test basin.
The results show that δ¹⁸O of precipitation has found only in channel precipitatin runoff and the rest of direct runoff and ground water runoff have δ¹⁸O value of groundwater. This implies that the meanings of runoff models especially physically based ones should be checked, because they usually consider that the direct runoff are originated from precipitation at that time.

View full abstract

Operational comparisons are presented using the optimization techniques to identify the parameters in Sugawara' s Tank model which has been widely employed for long-term runoff analysis. Themathematical optimization techniques tested herein are both the Powell and Newton methods. Computational burden is too stringent in the application of the Powell method to parameter optimization in the Tank model, because a great number of runoff computations are required. The efficiency of optimization performance resulting from use of the Newton method clearly depends on how effectively the sensitivity coefficients can be derived. An important feature of the proposed approach is the theoretical derivation of sensitivity coefficients which can directly be used in the optimization scheme of the Newton method.
A vector differential equation in terms of storages in the four tanks is numerically integrated using the transition matrix which is computed by expanding the matrix exponential. The elements of this transition matrix play a significant role in eliminating additional computations involved in the solution of sensitivity coefficients. On the basis of the simulation results, the Newton method coupled with sensitivity coefficients appears to have potential performance advantages for optimizing the parameters in the Tank model.

View full abstract

The estimation of effective rainfall is the most important part in the runoff calculation. It is very usefull to show the loss curve being used in the calculation, if there is no information on the effective rainfall of that basin. This paper gives the standard loss curves for 4 geological features, that are obtained by using the synthetic Tank Model whose parameters are idetified for geological features. These curves are compared to the results of runoff separation by filtering method, and used in the actual calculations with the object of verification. The results show good accordance.

View full abstract

An attempt of run-off prediction is carried out using an artficial intelligence system making use of the fuzzy theory in it's inference process. A new type of an artficial intelligence system which can treat “fuzzy sets” has been developed. The system consists of a fuzzy inference engine and a fuzzy knowledge base. Several rules for run-off prediction are placed in the fuzzy knowledge base. Run-off prediction simulation is tried out with real data from Nahari river area of Kohchi prefecture. Some promising results are obtained as an results of prediction simulation.

View full abstract

In this paper, the authors propose a system of nonlinear flood forecasting by the filter separation AR method.
Firstly, we compare the flood prediction of three methods (a regression model, a unit hydrigraph and a quasi-physically based model cited from the paper of Loague and Freeeze, 1985) with one of the filter separation AR method on data sets from the Washita River Experimental Watershed.
Secondly, we calculate and compare the modeling efficiencies of real time forecasting by the filter separation AR method with the generalized storage function method to which Karman filtering theory is applied and the tank model method.

View full abstract

Recently, weather radar has been fully utilized and detailed real-time information over a wide range of area has become obtainable with hight time-space resolution. But the accuracy of the radar hyetometer is not always high enough to use for flood forecasting.
In this paper, a forecasting method by using radar data is proposed, in which, a radar constant is included in the system parameters and identified together by Kalman filtering. The application of this method to the Onga river basin shows that the forecasting of water satges by using radar data is possible with the same accuracy of that by rainfall data on the ground.

View full abstract

Distrubuted rainfall-runoff model was developed using Radar AMeDAS Raingage Composite Map of Precipitation as input. It consists of two submodels: one for direct runoff and other for base flow. The direct runoff submodel is a distrubuted model which routes direct runoff in each grid through a channel network to outlet using the kinematic wave model. In this submodel, a watershed is considered as meshed grids connected by a channel network which is derived from altitude data at grid points. The base flow submodel is a lumped storage-drainage model. The model makes it possible to take the effects of areal distribution and movement of storm into account. Sample calculations with respect to Uono River basin resulted in a good corespondence between predicted and observed hydrographs.

View full abstract

Even identical rainfall intensity witl different distribution of rain drops diameters can result in different received power by rader, so it is difficult to determine precise rainfall intensity for all types of rainfall from received power by radar. In addition, because a radar beam passes 500-3, 000M Above ground level, rainfall intensity obsevred by radar does not necessarily coincide with the ground rainfall intensity. In thie paper the authers discussed the method of improving coherency between the current scene of a radar raingauge and ground rainfall by uaing past observed data of a radar raingauge and ground raingauges, and could find techniques yielding a certain degree of success.

View full abstract

This paper describes real-time calibration of B, β in the radar equation used to transform the radar reflectivity into the rainfall intensity. The calibration is intended to obtain a higher degree of accuracy in estimating areal rainfall and forecasting flood runoff. Various methods of calibration are compared in terms of the accuracy of hydrograph in the Yura River basin (1, 880km²) located in the north of the Miyama Radar Raingage system, Ministry of Construction.

View full abstract

In this paper, the results of the field observation of rainfall which had been carried out by the authors, are shown and the effects of height or direction of the mountain on rainfall are discussed. With these investigation, we calculate the wind field in mountanous area by using the digital maps and spectral analysis. Combining the calculated wind field and the Kessler's model for rainfall, we simulate the precipitation field in mountanous area. Comparing the observed data and calculated rainfall pattern, we confirm that the above-mentioned calculation can simulate the actual orographic effect of rainfall precisely.

View full abstract

The runoff processes throughout a year including rainy and snowmelt season in two small neighbouring basins are studied. The obtained results are as follows:(1) It is shown that river runoff is composed of several runoff components from different vertical soil layers in a slope by using absorbance of ultraviolet 260nm as an index of water quality, also that they are sensitive to accumulated rainfall or rainfall intensity.(2) Geographic effects on runoff characteristics including runoff coefficient, peak discharge and the relation between accumulated rainfall and runoff are clarified.(3) A synthesized numerical model of runoff is proposed and applied to each basin in order to verify the serveral results of the field observation stated above.

View full abstract

In this paper, by applying the filter-separation AR method, the response characteristics and non-linear separation laws of rainfall in snow season are compared with those of non-snow season and discussed. The results of analysis are as follows:(1) There is no significant difference between the response function of inter flow system in snow season and those of non-snow season.(2) The response function of groundwater flow in snow season shows an about half peak discharge and continues a longer period than those of the non-snow season.(3) In snow season more effective rainfall are supplied to the groundwater system than non-snow season.

View full abstract

Snowmelt runoff analysis consisting of three submodels is developed. These submodels are the model for estimation of basin-wide snow water equivalent using snow covered area, the model for estimation of basin-wide snowmelt rate and the runoff model. In each model we use the information on snow covered area derived from LANDSAT MSS data.The meteorological data from AMeDAS is used also. We apply these models to the Uono River Basin in NIIGATA, during the snowmelt season in 1979, 1981, 1982. The results of model applications are in good agreement with observed data.

View full abstract

The present paper discusses the effect of the lumping scale on runoff. The procedure of this study is (1) to decide the first order stream in a small mountainous basin by using digital map statistically and theoretically.(2) to constract a basic equation expressing the contributing area concept.(3) to compare the runoff characteristics of small and large basins (4) to predict the runoff in large basins from runoff characteristics in small basins.

View full abstract

This paper presents an automatic basin model generation system. The data to be inputted are height data on a regular grid which covers the basin and position and height data of channel network. The system produces a basin divide and subdivides the catchment into many small slopes taking streamlines (the steepest ascent lines) as dividing lines. Though in most cell methods developed to date for hydrologic application, the direction of flow is assumed to be parallel with axis, such assumption seems too simplistic in considering water-flow processes on three-dimensional catchment. The system presented here attaches importance to direction of water-flow. A quasi-three-dimensional slope runoff model to be coupled with this basin model is also presented.

View full abstract

The kinematic wave method is applied to evaluating the concentration time of overland flow in the basin composed of inclined planes and river channel. Firstly, the relationship between concentration time and geographical features of the basin is made clear for rainfall of constant intensity. Secondly, we evaluate quantitatively the effect of unsteadiness of rainfall on concentration time and peak discharge. The result of computation shows that concentration time is greatly influenced by the equivalent roughness coefficient of planes, and the concentration time defined in this study for unsteady rainfall is almost equal to that for rainfall of constant intensity.

View full abstract

A physically-based conceptual model for continuous modelling of Hortonian overland flow and infiltration under variable rainfall has been proposed and tested at a point, with the values generated by a complex finite-element infiltration model. The model has been shown to predict both the volumes and positions in time of overland runoffs, from a real 146mm/day rainfall pattern quite well, for three simulation time intervals and three values of top-soil saturated hydraulic conductivity. An interesting feature of the model is its ability to give realistic determination of rainfall excess, compared to the methods currently available.

View full abstract

LATERAL MODEL, a simulation model of surcharged flow in urban sewer pipe systems, is presented. A number of lateral pipes are connected to the top of sewer pipe in order to drain storm and sewage water from inlet to the sewer pipe. A pressure-control effect that is caused by movement of stormwater into or out of the lateral pipes and has been ignored in overall traditinal simulation models, is taken into account. An equation for determining a pressure wave celerity of the surcharged flow is derived. Run-off experiments are made and the adaptabilities of the model is investigated comparing the experimental results with the simulated ones.

View full abstract

TUBE MODEL, a surcharged flow simulation model in urban sewer pipe systems, is presented. A highly deformable sewer pipe is used in which the pipe wall has a nonlinear elastic relationship. One of the principal parameters of the model, β, that is a variable indicating the degree of nonlinear elasticity of the pipe wall, is determited comparing the flow equations of TUBE MODEL and LATERAL MODEL that is a fundamental and realistic model of surcharged flow. It is derived that a pressure-wave celerity of surcharged flow in TUBE MODEL nealy equals to that in LATERAL MODEL. Numerical experiments are carried out and the adaptabilities of TUBE MODEL are investigated comparing the simulated results by the two models.

View full abstract

In order to take measure to reduce flood damages resulting from urbanization, a large number of detention basins have been constructed in many urban catchments.
In the present paper, a simplified method to calculate the runoff from a urbanized river basin with many detention basins using a linear reservoir model is proposed, and the hydrodynamic effect of detention basins can be expressed by an equivalent roughness.

View full abstract

We have examined the stochastic analysis of the precipitation in Tokyo for the flood control plan, considering the two different heavy rains.The results are as fol lows.(1) Probability distribution: the stochastic analysis of precipitation in Tokyo supports ‘the square root exponen-tial-type distribution’ more than the Gumbel's and the logarithm normal one.(2) Separation of rainfall types;the Cleveland-type rainfall intensity formula agrees with the local heavy rains data better than the typhoon's and the total ones.(3) Two hyetograph model;‘the equi-risk line’ agrees with the typhoon-type rainfalls very well, also the traditional ‘central peak shaped’ rain-fall model does with the local heavy rains as the same.(4) Applications: the equi-risk line model has been applied to convert the amount of infiltration into the equivalent storage one.

View full abstract

Hydrologic maps are presented, which are used to estimate values of the parameters in the expressions of efficiency of urban storage facilities to control flood, water supply and quality. The expressions have been derived by the authors. AMeDAS, the automated meteorological data acquisition system of the Japanese meteorological agency, supplies hourly rainfall data at more than 1300 observatories deployed all over the Japanese islands, and enables to calculate the parameters. Examples to explain how to use the maps are also presented.

View full abstract

The quality of groundwater in unconfied aquifers is going to become from bad to worse with the increase of the amounts of pumping for agriculture or fish farming. It is especially one the urgent problems to prevent from the saltwater intrusion into the aquifers in many coastal lowland.
In this study, it is tried experimentally to analyze the mechanisms of saline intrusion and to estimate the effects of its check.
The main subjects of consideration are as follows: 1) the calculation of the time-dependent flux of salinity and of the flow velocity of saltwater, 2) the estimation of the defensible effcts due to the artificial structures, and 3) the valuation of coefficients on the basis of the one-dimensional dispersion analysis.

View full abstract

This study presents the experimental and numerical results on the dispersion properties at the salt water mixing zone in a confined aquifer. The method of shifting particles and the implicit finite difference method are applied to the numerical computations of the convective-dispersion and the seapage flow equations. The salt concentration distribution and the seaward flow rate of fresh water obtained by the present numerical method are in good agreement with those by the laboratory experiment. It is shown that the mixing zone spreads wider towards the fresh water effluence at the steady state. At the unsteady state of salt water intrusion, it is also shown that the stagnation point of flow appears in the fresh water region and that the large dispersion takes place at the tip of the salt water wedge when the salt water advances.

View full abstract

Seawater intrusion is examined on the two-layered unconfined aquifers. Numerical solutions obtained by Residual Flow Procedure show that Dupuit's approximation and Herzberg relation are hardly satisfied near the downstream end. With apprication of velocity shape factor, analytical solutions are derived on flow rate and on the shapes of free surface and fresh-saltwater interface. Analytical positions of fresh-saltwater interface agree with experimental position and numerical one. From the simuled results, it is clarified that seawater easily intrude to the two-layered aquifers in case that the layers in higher permeability overlie on the layers in lower permeability.

View full abstract

The behaviors of unconfied groundwater and saline intrusion in the estuary basin of the Yoshino River are investigated by computer simulation method based on observations. The objective region is nearly rectangular with an area of about 50km². The monthly field survey of the groundwater levels and the chlorine-ion concertrations in 31 wells was continued for about 3 years from October, 1983. The contents of this study are 1) the hydraulic and stochastic analysis of the groundwater levels, 2) the identification of aquifer constants and the simulation of water balace based on the Tyson-Wever's equation, 3) the flow simulation by FEM and 4) the time-dependent analysis of the saline intrusion by FEM. The behabiors of groundwater and its salinity were estimatedin the quasi-three dimensional analysis. The most suitable aquifer constant, T=986.2m²/day, was calculated and the present state of groundwater flow was reproduced with good agreement. The salinity movement was discussed by using the FEM.

View full abstract

Two methods, one uses the second-order filter and the other the extended Kalman filter are proposed, which are used to predict the concentration distribution and to identify the parameters of the one-dimensional stochastic convective-dispersion equation is discussed. It is shown that the proposed methods identify the parameters effectively and predict the concentration distribution accurately. It is clarified that the computational time for the second-order filter requires about two times longer than that for the extended Kalman filter, whereas the accuracy of prediction of the second-order filter is almost the same as that of the extended Kalman filter.

View full abstract

In conventional analysis of the regional groundwater flow, the model parameters T, S are identified by a direct comparison of the computed head values with the observed head data. However, we should keep in mind that the computed values are macroscopic solutions, while the observed data include local random-components caused by the heterogeneity of aquifer. This study aims to propose a new method for the parameter identification, which takes full advantage of the statistical property on uncertainty of actual head behavior.

View full abstract

Most of inverse problems of groundwater flow are focused on determining the distribution of transmissivities over a flow domain from head measurements in wells under given boundary conditions. However, for groundwater management, it is important to esimate boundary flux. In this study, the technique proposed by Carrera is employed to determine the distributions of boundary fluxes as well as transmissivities from head data. The effect of zonation on the transmissivity estimation is also examined in a model where a low-transmissivity belt exists. The results show that the technique used herein is useful to estimate the boundary flux. On the other hand, the transmissivity estimation is quite sensitive to the zonation.

View full abstract

This paper is motivated by the social needs with respect to groundwater protection and prevention of land subsidence when developing the underground space in urban area. The hydrological assessment of groundwater environment is enumerated in realizing of a large scale and deeper underground space because it permits more reasonable design. The construction of underground space may locally invite the groundwater damming and whithdrawal around underground works^and it may possibly create a new groundwater sink of aquifer in addition to the channeling flow along a tunnel extension. The water budget in underground space is evidently associated with groundwater inflow and air conditioning, but it may also be influenced by the relative humidity under a given temperature.

View full abstract

Unsaturated water percolation was investigated by measuring the soil temperatures at twelve different vertical points. Warm/Cold pure water of 65 1 was given on a lysimeter as an artificial rainfall once a week. From the time variations of soil temperatures due to the warm/cold water, we attempted to know how fast and how far the water went down in the lysimeter. By estimating the effect of the thermal diffusion in soil on the experimental data of soil temperatures, we have reached the conclusion that the rain water percolated to 8 cm deep within ten minutes; however, after that, it went down very little.

View full abstract

A numerical scheme for computation of the convective diffusion of solute in saturated-unsaturated soil layers with arbitrarily. shaped boundaries has been developed applying the boundary, fitted coordinate method. Examples of computation on convective diffusion of solute, with surface and point or localized sources in an S-shaped hillslope are given. For a plain source like acid rain, the solute concentration can heardly observed in the throughflow discharge caused by the first rainfall, but it increases largely as soil moisture increases by the convection of rain water in unsaturated soils. For point sources, even in the initial stage of infiltration near the ground surface, the solute is diffused laterally, although the flow velocity is nearly vertically and the diffusing cloud of solute is convected downslope.

View full abstract

Organochlorines have been widely detected in the groundwater in Japan. In order to resolve the mechanism of groundwater pollution, migration and leaching of the undiluted organochlorines were examined in water and porous media. It proves that undiluted trichloroethylene (TCE) readily infiltrated in unsaturated zone due to the physical properties, greater in density and less in surface tension and viscosity than water. In saturated zone TCE migrated and remained stagnant as isolated in pore space of the media. These phenomena reappeared in the Hele-Shaw model. Leachate concentration of TCE due to the rainfall infiltration and groundwater flow amounts to several thousands times as much as the tentative standard for drinking water. Besides, a new monitoring method by analysing the gaseous contents in the surface soil has been developed and was applied to the field observation in the waste landfill site.

View full abstract

In order to clarify the mechanism of solute migration in an unsaturated zone of soil stratum, the basic experiments on the upward and downward movements of water and solutes (NaCl, Sr, Co, Cs) under the artificial environment (rainfall-evaporation cycles) were carried out. The experimental apparatus used in this study consists of a large soil box, nozzles of rain-fall, infrared lamps, thermometer sensors, recorders and sampling tubes. The vertical profiles of solute concentration, temparature and water content in the unsaturated zone above the groundwater are measured for different groundwater depth. The analytical solutions of governing equations at the steady-state are derived. The comparison between solutions and experimental results shows a good agreement. The results describe that the solute migration depends upon not only vertical movement of moisture but also a depth of capillary fringe.

View full abstract

Some geohydrologic investigations before disposing of radiowastes into underground facilities are necessary, because the safety assessment should be required assuming that hazardous radionuclides might be released into geoenvironments from the radioactive contaminated materials. The flow velocity of groundwater, the recharged origin and the basin have to be determined through in-si tu observations. In this study, the ³H-³He dating method has been researched and developed in some technical points to estimate the age of groundwater. This method was preliminarly applied to the confined aquifer in Mishima la-va flow to estimate the groundwater flow velocity and reasonable results compared with past some investigations were drawn.

View full abstract

Surges in open channels was computed by using Ameins' four-point difference scheme which is the most efficient method for flood routing problems. The method bas been applied to surges, which suggested the optimum value of θ=0.6, however, the effect of change of Δx/Δt was not discussed. Price has proposed, comparing the results with exact analytical solutions for the monoclinal wave, that the optimum value of Δx/Δt is equal to kinematic wave speed. In this paper, comparisons are made between the results of numerical solutions and laboratory experiments. The results of calculations demonstrated good agreement with results of laboratory experiments. The numerical instability is observed for the large value of Δx compared to Δt.

View full abstract

Moving hydraulic jump condition in a channel of constant water depth is examined by reviewing past theoretical, numerical and experimental studies and conducting experiments anew. It is concluded that the moving hydraulic jump condition is H_b/h_o>0.61-0.64, where H_b is the height of bore front and he is the constant water depth in the downstream region of the bore.

View full abstract

A three-dimensional computational model, based on finite volume method and k-ε turbulence model, for predicting steady state confluence flow is presented. The confluence region is covered by a curvilinear grid which is generated numerically using a quasiconformal mapping method. The computation is done in the transformed domain which is a parallelipiped. The computed results agree well with the experimental results. The performance of a levee which facilitates two streams to mix gradually and thereby to reduce the superelevation and the strong secondary flow in the main stream, is demonstrated.

View full abstract

The characteristics of river bed variation at the contraction during the flood are investigated exprimentally. Moreover its characteristics for natural river which have several types of contraction are investigated by the numerical simulation on the basis of one dimensional bed variation method. From the experimental results, it is found the tendency that the bed level approaches to the bed level on equilibrium condition of sediment discharge for flood discharge. As the results, the bed level lowers during rising stage and then phenomenon of bed aggradation during the falling stage occurs at the up-stream end of the contraction. It is found from the numerical simulations that the bed variation of the contraction is influenced by the approach length, length of contraction and duration of the flood.

View full abstract

An algebraic stress turblence model is developed to reproduce turbulent flows with turbu-lence-driven secondary currents in ducts of arbitrary cross-section. The model is obtained as an extension of the k-ε model to include non-linear effects of the mean rate of strain. The expression of the non-linear terms in the model is derived from Launder-Reece-Rodi model with neglecting advection and diffusion terms. Then, the model is applied to a compound channel flow to test its validity. Comparison between calculated results and experimental data reveals that the present model can reproduce mean flow field as well as Reynolds stress distributions.

View full abstract

Comprehensive field observations have been conducted to investigate characteristics of flood flow in the Hinuma River. The observations revealed that the water stage had irregular profiles due to longituidinal variation in the roughness consisting of luxuriant vegetations on the flood plains. The irregularity of flow profiles couldn't be reproduced well by the conventional method of non-uniform flow computation. In consideration of these facts, the method of flow computation was improved so that it would be possible to predict flow profiles in channels with luxuriant vegetations with more accuracy, compared with the existing methods.

View full abstract

Laboratory tests were performed to evaluate, the lateral momentum transfer between two flow regions having different velocities. In the tests, media with very high porosity were put on a part of a test channel in such a manner that lateral discontinuity in velocty was made. The experimental results showed that the mixing coefficient f expressing intensity of the momentum transfer changed remarkably depending on the position, width and permeability of the porous media. On the basis of the mechanism of fluid exchange between the two regions, prediction methods of the mixing coefficient f were developed.

View full abstract