Many models for the prediction of unsaturated hydraulic conductivity, using data from the soil water characteristic curve, have been proposed over the past thirty-five years. These models are called statistical models and were classified into three formulas by Y. Mualem and G. Dagan (1978). One of these is the generalized Burdine equation, another is the generalized Childs and Collis-George equation (generalized CCG equation), and the other is the generalized Mualem equation. However, the use of these models is restricted to the prediction of relative conductivity, which is the ratio of unsaturated to saturated conductivity. Therefore, if it is necessary to predict unsaturated conductivity, a value for saturated conductivity is also needed. The reason for this is that the models often have no physical basis for their assumptions. As a result, it is suggested that deviations between predicted and experimental values occur in many cases and could be diminished by the intentional use of the unknown parameters in these models.
In this paper, we propose a new model which can by itself be used to estimate unsaturated conductivity with a more exact reconstruction of the above-mentioned models. This is done by considering the effects of connection and correlation of the capillary pores in two cross-sections of soil slab, and also, by considering the effects of expansion and tortuosity in each pore. This model, assuming the tortuosity to be inversely proportional to the power function of the capillary radius, can converge to the generalized Burdine equation, the generalized CCG equation, or the Mualem equation (1976), corresponding to the different values of expansion and tortuosity parameters. Accordingly, the new model can be understood as a unification of the past statistical models.
Then, in order to realize the properties of the parameters which describe the pore construction, we apply a simple relation between soil water and suction to this model. The results show that the effect of expansion upon unsaturated conductivity, especially on the relative conductivity, is so small that it produces little difference between the generalized CCG model and the Mualem model. The results also show that the tortuosity has such a large effect that it is necessary to be precisely evaluated.

View full abstract

Thematic map information, which is made by Landsat MSS data processing, will be used as explanatory factors for the regional analysis or land evaluation. However, new problems will occur in the case of their use, derived from discussions on their production process.
(1) Information loss through the summarization process
(2) Applicability of the model used on both its precision and objective
(3) Spatial non-uniformity of predicted results
The actual data for the evaluation of agricultural land suitability, some of which are thought to include the above problems, are analyzed in order to examine the relation between the suitability score from the farsighted point and its evaluation factors including raw MSS data. The results are as follows.
(1) MSS data devaluate almost all of the other factors and contribute largely to explaining the suitability score.
(2) The suitability score in this case study can be explained mainly by the geomorphological data, estimated Biomass weight originating in MSS data, and MSS data.
(3) Among thematic information obtained by MSS data processing, there are such things similar to the existing map information but different, or as interpreted to be newly defined. Such information must start from checking whether it can be used or not. Moreover, when its contents are qualitative, the appreciation of categorization or ranking must be checked.
All of the considerations mentioned above can show one method for evaluating agricultural land suitability by using MSS data. This method has a merit of self-selecting effective factors at the analytical stage, and therefore, will be considered as desirable especially in developing countries with less information concerning land conditions.

View full abstract

Up to now, Landsat MSS data has been used for information extraction regarding the earth's surface, which corresponds from a scale 1 to a maximum of 50, 000 because of its lower resolution. It is already pointed out in the supervised classification techniques, which is one form of the information extraction method, that the wrong classification will occasionally happen owing to the mixture of land cover materials within the pixel data. From the considerations on such a mixed situation in a pixel, its positive applicability to grasp the mixing ratio can occur in the inverse.
In this paper, some methods for such an objective, which are related respectively to both (a) an algebraic approach by the generalized inverse matrix, and three statistical approaches by (b) a multiple regression analysis, (c) a canonical regression analysis as locally named and (d) correlation analysis, are discussed and checked concerning their advantages or disadvantages from the theoretidal viewpoint. The last two methods, (c) and (d), are a newly devised adaptation of the multi-variate analysis. The last method, (d), is thought to be the best according to the following results.
1) Methed (a) has constraints on the number and relation of the mixed elements. Moreover, it is difficult to make the expression including the existing data such as the geomorphological data.
2) All methods by the latter approach can build up a general expession.
i. Method (b) is not good in that the relation among mixed elements must be neglected and that the estimating precisions will be different according to the element.
ii. Method (c) or (d) can treat all elements at the same time and will give similar results. However, (d) is superior to (c) because it can show quantitatively the possibility of estimating by the prepared explanatory data.

View full abstract

For estimating the wave force for a rigid body such as concrete pole, concrete block etc, the method of O'brien-Morison is applicable. However, for long-flexible and floating body, we can hardly fi nd any study on such case.
The wave force caused by the motion of wave particles deforms a seaweed breeding rope and the rope is tensioned not only by friction force of wave particles which moves along the rope line but also by the periodical drag force of the wave particles across the rope line. Furthermore, the shape of grown seaweed tuft on a rope is deformed and its compactness is changed due to the alternating motion of the wave particles.
For a irregular and undefined body shape such as the above mentioned seaweed tuft, it is difficult to fi nd a suitable wave force estimation method. Thus, some experimental studies were conducted to ensure the wave force estimation method in a comparatively long rope for seaweed breeding.
As a result of the study, the author found the relationship among the wave force, wave characteristics and the specification of seaweed breeding system. Following are the major outcomes:
a. The relationship among Ad/L², FT²/(ρL⁴), hd/h and H/L has been clarified as shown in Fig.7. Using Fig., the wave force F can be estimated.
b. Found that F appears as a large value depending on the position of the float attached, and came up with a suitable way to arrange the float.
c. Found the suitable depth to set the breeding facility by rope to avoid the large shock under the transitional motion of the rope.

View full abstract

Laser-Doppler Velocimeter (LDV) is applied to the simultaneous measurements of the particle and the fluid velocities as well as the particle concentration in the solid-liquid two-phase flow. The solid particles used in this study are polystyrene beads with a mean diameter of 1.25mm. Backscattered-light mode for the particle and the forward-scattered-light mode for fluid are adopted to detect simultaneously and independently the laser-Doppler signals from the particle and the fluid. Three interpolation schemes by Tsuji and Morikawa (1982) are re-examined for the signal processing of the defective parts in the fluid time-sequence signals. It is shown that the linear interpolation scheme is adequate for spectral estimation by the maximum entropy method.
The instantaneous relative velocity is calculated as the difference between the particle and the fluid velocities when they are uncorrelated. It is found that the added energy dissipation due to the relative velocity is negligibly small compared to the viscous dissipation by the eddy-eddy interaction.
Comparisons are made between the fluid turbulence statistics for single-phase and two-phase flows. Results indicate that the probability density of fluid velocity in the two-phase flow shows an asymmetric distribution, and that the fluid turbulence intensity and the viscous dissipation increase in the presence of particles with diameters larger than the Kolmogoroff microscale.

View full abstract

This paper deals with some theoretical problems still left unsolved for pipe network flow analysis. There are the two well-known direct solution procedures suitable for electronic computation, i. e., the Newton-Raphson Method (NRM) and the Linear Theory Method (LRM). The author demonstrates the following facts: 1. The numbers of iterations required for convergence are all the same for different sets of trees and cotrees, in a directed-graph if initial values are estimated through a linear approximation of friction terms. 2. Each element of the Jacobian matrix with LRM is consistent with the value of the corresponding element of the Jacobian matrix with NRM, multiplied by a given constant. This produces much difference in convergence of solution. The solution with NRM converges rapidly with second order while the solution with LRM oscillates near the true solution. The averaging method to smooth out the oscillation guarantees the convergence of second order as well as NRM. The abovedescribed points are also examined through computations. Finally, the author proposes that the two methods should be unified as the NRM with an initial estimate obtained by a linear approximation to frictional terms.

View full abstract

In this paper, the methods to separate reasonably the baseflow component of a small mountain river flow using the water quality are investigated. The relationships between the chemical characteristics of the river water and the runoff phenomena are first discussed. As a result, a concentration-discharge equation for the river flow is obtained, and the point that there exists a close relationship between the time variable aspects of water quality and that of the ratio of direct flow to baseflow component is also elucidated. The baseflow component is separated using a steady and an unsteady water quality model derived on the basis of the above discussion, and the results of separation are checked with the field observation data. It seems that the unsteady water quality model provides a better baseflow separation than that of the steady one.

View full abstract

A plate loading test is sometimes performed on pavement, and the uppermost layer of it is destroyed when vertical load is increased. One of the design criteria of the pavement by the rational method is to estimate the tensile stress or the tensile strain at lower face of the uppermost layer on the axis of the plate circle. An extensive presentation has been made for these estimations with tables and graphs. However, diagonal crack from just outside the perimeter of the plate was observed in the test, which is not in concordance with the tables and the figures. In order to accout this fact, detailed calculations were made on the basis of elastic multi-layer theory, and contours of the principal stress and the principal strain were drawn. One of concluding results is that the crack of the uppermost layer can be considered to have originated from a point where the principal tensile straints maximum, and this point is found not always at lower face of the uppermost layer but, in some cases, at the proximity of the plate perimeter. The point of the maximum principal strain moves depending on the ratio of Young's moduli and on Poisson's ratios.

View full abstract

Nowadays, it is obligatory, for a number of embankment dams to have their inspection gallery made of mass concrete for operational safety. For these inspection galleries, water tightness is most essential, and therefore, the possibility of the crack initiation in the concrete during construction due to hydration heat of cement must be removed. Accordingly, it is significant to know the temperature -variation in the concrete due to hydration heat during the placement. In this paper, the analytical method for the distribution and variation of the temperature in the gallery concrete using the boundary integral element method are shown. The results of the proposed method are discussed by comparing with measured temperature behavior of the same gallery concrete as used in the analysis. These investigations showed that the proposed analytical method is very powerful for simulation of temperature variation in the gallery concrete and proper evaluation of the effects of the pipe cooling.

View full abstract

Influent to the facility for treatment of wastewater from rural settlement has big changes in load both hourly and seasonally. For example, according to the research at K-facility which treats the wastewater from 45 households in Hyi5go Prefecture, the quantity of wastewater in summer was more than twice as much as that in winter and also most of the daily load concentrated within the four hours from 6 to 10 a.m.
This paper describes the stability of treatment with Intermittent Aeration for the high rate removal of nitrogen, considering the actual fluctuation of the quality of wastewater and the influent pattern of load. The results of the basic experiment in the laboratory were as follows:
1) According to the experiment under the condition of 1 day-retention time using the artificial wastewater, in which concentration of BOD and total nitrogen was 170 ppm and 35 ppm, respectively, high rate removal of nitrogen higher than 85% was accomplished stably.
2) Removal of nitrogen from wastewater by microbial system consists of both denitrification as N₂ gas and physical separation by absorbing of activated sludge. In case of treatment with Intermittent Aeration, it can be said that the former is much more than the latter because of the fact that concentration of total nitrogen in effluent water is much less than that of inorganic nitrogen in influent water.
3) When at least 1 day-retention time was kept, it was possible to maintain good quality of effluent water by the adequate control of operation (mainly by changing the aeration time) against the seasonal change of influent.
4) In case of operation with 2 cycles per day, the maximum fluctuation of load between cycles was estimated to be 1.5 times from the influent pattern of load at K-facility. The experimental results showed that even if such change in load was given between cycle the removal rate of nitrogen was not much influenced.

View full abstract