This paper presents a methodology and a case study for estimating the vertical profile of ground water permeability from a time series of water level fluctuation in a well, which is used for daily life. A numerical simulation model for seepage flow in unsaturated and/or saturated condition proposed by Irie et.al. (1999) is used for the analysis. The model applicability is examined by both of a laboratory experiment and the model sensitivity analysis which shows the effect of soil parameters other than the permeability on the daily fluction of water level to be small. A field study was carried out for a domestic well on the Kanto-loam tableland at the west side of Yokohama City. The vertical averaged transmissivity is evaluated for a variety of water stages by simulating the daily fluctuation of water level at different seasons. The layer structure of the ground and permeability of each layer are estimated from the vertical difference of the average permeability. The estimation results are fairly consistent with the existing geological data obtained in the area.

View full abstract

Tank model coefficients were optimized at 12 basins. The “Basin Characteristics”, which was concerned with gradient, soil, geology and land use, were derived using GIS data for each basin. The relationship was investigated between the model coefficients and the “Basin Characteristics” using Multiple Regression Analysis. As a result, it was realized that each coefficient could be optimized with a one-dimensional equation that consists of several basin characteristics. According to the results of trial application at 2 basins, it became obvious that this method is useful in constructing a tank model for long term runoff.

View full abstract

Few studies have specifically shown the effects of soil heat flux on ground surface heat balance. In this study, four practical calculation methods for soil temperature and soil heat flux were examined by using data collected at the Cold Region Meteorological Observatory of the Kitami Institute of Technology. The results indicated that the calculated soil temperature values were influenced by initial conditions more than by boundary conditions at the ground surface in cases in which observed values were used for determining the boundary and initial conditions, and a cubic equation for subsurface field was found to be better than a constant for describing initial conditions. On the other hand, the calculated values for soil heat flux were more influenced by the surface boundary conditions than by the initial conditions, and a linear equation was found to be better than a constant for describing surface boundary conditions during constant intervals. Next, a conduction-convection ratio was proposed as an index to show the influence of the soil heat flux. It became clear that the influence of the hourly averaged soil heat flux in the observatory was sometimes negligible around sunrise and sunset. Finally, referring to published references, the distribution ratios of sensible, latent and soil heat flux corresponding to different types of land cover were plotted on triangular coordinates. It was found that the influence of the soil heat flux was not always negligible when the calculation interval was one day or longer.

View full abstract

Probability Density Function (P.D.F.) for maximum amount of rainfall in a year was investigated for several duration time of rainfall such as annual, daily, hourly and ten minutes. The rainfall date was supplied from the climate observation site of the National Weather Service at Sapporo, Sendai, Maebashi, Hikone, Matsuyama and Fukuoka. At first, 5 types of P.D.F. Shown in table 2 were chosen and the accuracy was evaluated by mean square error defined by eq. (3). Secondly, dividing by the rainfall data into two groups such as for determining of parameters and evaluating the estimating accuracy of rainfall amount, the type of P.D.F. was evaluated from practical point of view. The result shows that the accuracy increased from type I to type V in every duration of rainfall time and the sites. Secondly, using by logarithmic normal distribution characteristics, the synthesization of P.D.F. was attempted. The standardization of rainfall was carried out by eq. (6). The result shows the same straight line in plane of probability index and standardized rainfall in each rainfall time. In addition, the straight line of the above 6 sites coincides with the same line. This method is very useful if the return period is relatively short comparing with number of rainfall data.

View full abstract

Annual and monthly water loss data, namely precipitation minus runoff, from experimental watersheds of all over the world, are collected and stored in the database. The data are released to public through WWW. Annual water loss is regarded as annual actual evapotranspiration, Et. By combining the database to Geographic Information System, GIS, the world distribution of Et is obtained. The important point is that the database is based on the observational results. The map of Et is consistent with empirical evidences. It can be used as verification data of the models that calculates the fluxes or as ground truth for the products of satellite remote sensing. Annual Et is plotted against annual precipitation, Pa. Et increases with Pa almost along 1:1 line, however, the relationship departs at certain Pa. This value is different among the region, and is an important parameter which governs the regional hydrological environment. The maximum annual Et in wet tropics is around 1800mm. It is estimated that Et slightly decrease with Pa after Et reaches maximum because energy available for evapotranspiration decreases with precipitation.

View full abstract