Recent geoelectrical resistivity interpretation procedures require more careful and reliable inforinations. New electrical prospecting techniques using high density data, high resolution inversion and visualization of simulation would become increasingly important tools for geoexploration. This paper begins with a brief review of the model optimization in electrical prospecting method and of the application of a stcitistical information criterion. In details, in the interpretation procedure using the digital linear filter algorithm we adopt AIC (Akaike's Information Criterion) for the evaluation of model optimization in the ground water exploration by vertical electrical sounding method. Next, four kinds of field interpretation have been chosen to discuss the application of AIC evaluation to practical problems including the so-called equivalent resistivity layer model.

View full abstract

If the noises in data are Gaussian and independent, LSQ (least square) estimates of regression parameters are unbiased and of minimum variance. However, if the data are contaminated by outliers, LSQ will result inaccurate estimates of parameters. Robust technic is widely used to avoid the difficulty. The technic requires penalty factors for each data and estimation of parameters is done by LSQ. Specification or selection of penalty function form is left for data analyst. When we adopt distributions which have heavier tails than that of Gaussian, ML (maximum likelihood method) will reduce influences from outliers and will result better estimates of parameters. Once ML estiamtes obtained, AIC (Akaike's Information Criterion) can help us in selecting a proper model of regression from candidates. In this article, Type 7 Pearson system distribution is used as a heavy-tail distribution. The distribution naturaly links Cauchy, approximate Laplace and Gaussian distributions. Consequently the regression procedure with this distribution is robust and has high adaptability to data contaminated by outliers.

View full abstract

There are many operating and closed quarries of tuff in Ooya region Utsunomiya City, Japan. In March 1990, a rock-ceiling of the underground quarry suddenly sank and collapsed. It was confirmed that the field acoustic emission increased in number before the sinking, just like the pre-seismic activity of natural earthquakes.
Our definitions of the field acoustic emission include the acoustic emission by progressive cracks in rock mass, the concussion by rock splinter that is separated from ceiling or wall, and the vibration that is caused by buckling of the rock pillar.
We predicted the area and the time of the sinking by the results of observation and data analysis of the field acoustic emission.
The basic features of data analysis are as follows;
(1) To use the activity factor (amplitude multiplied by vibration duration time) which is effective to represent the microseismic activity more than the frequency.
(2) To construct database system including the each observed field acoustic emission data, the location data of observation point, and the topographical map of this region.
(3) To draw the microzonation map of sinking potential to estimate the sinking area by the calculated center and the scale of each observed field acoustic emission.
(4) To introduce the hyperbolic function to estimate the probability of sinking.

View full abstract

Methods of pattern recognition by digital processing are applied to analyze edges, lines, density of lines and standard deviations of the LANDSAT Thematic Mapher image in Mt. Takakuma area of South Kyushu, Japan. Mt. Takakuma area is composed of the sediments of Cretaceous to Quaternary in about 20 kin times 24 km square area and centrally intruded by the Takakumayama Granite of Miocene age. Peripheral area of the Granite includes the auriferous tungsten deposits which are mined out. Edges and lines extracted in image mostly coincide to the geological structures of faults, joints, cracks, sedimentary bedding strikes, and so on firstly because the frequency of the direction of lines are concordant to the general tendency of the studied geological structures, secondly because the lines and edges are more or less similar in result, and thirdly because some of the long continuing edges and lines are concordant to the major faults. Areal features represented by the textural standard deviations and densities of lines coincide to the geological zones of formations. The areal features by the computer pattern recognition is useful for the geological analysis. Combined image of lines, density of lines, and standard deviations are newly proposed for the advanced structural analysis.

View full abstract

Expression of geological maps seems to be variable even within an defined area, and this is considered to be due to the variety of the purposes and view points for use. Based on the present mapping process, such view points are divided into two cases. One is to express the geological configuration as accurately as possible, and the other is to express any geological concepts or relations considerably. Any points of view for mappingg is conclusingly expressed on the coordinates of these two points. Based on the numericalized geological model, various models on the coordinates may be convertible each other, or from fundamental one to specialized one.
To establish the fundamental geological model, not only excellent interpolation methods but also some rough models which may be based on the previous geological maps must exist. Therefore, it is required to establish the algorithms of this step for computer mapping in future.

View full abstract

This paper presents the algorithm for the automated inference of the stratigraphic sequence and its“Prolog”program, based on Shiono and Wadatsumi's (1988) principle, which formalized the logical process to arrange the order of geological events using the mathematical notation related to the binary relation.
The algorithm is as follows:
(1) Geological bodies observed at outcrops are listed and each body is described as body (a) .
(2) Two types of geological relation are described in terms of the binary relations: on (a, b) when we observe an outcrop showing that a is on b, and upper (a, b) when we obtain a chronological datum suggesting that a is upper (or older) than b.
(3) It is checked whether the relation“on”is antisymmetric.
(4) The relation“on”is extended to the relation“upper”by renexive and transitive closure.
(5) It is checked whether the relation“upper”is antisymmetric. If“upper”is antisymmetric, it is a partia ordering.
(6) It is checked whether every pair of geological bodies is comparable under the relation“upper”. If every pair is comparable, “upper” is a total ordering.
(7) If“upper”is a total ordering, we can arrange geological bodies in a linear sequence: that is, the stratigraphic sequence, using a proper sorting method.
The present study will be useful to construct the fundamental part controlling the computerized-geological mapping system “GEOMAP-CAD” (Wadatsumi et al., 1987) and to enrich its knowledge-base.

View full abstract

A well data management system is developed on a personal computer to maintain well data for the analysis of sedimentary basin which is the basis of resource assessment. Microsoft (MS) Excel, a spread sheet software, is used for well data input and management, and PC-ARC/INFO, a geographic information system (GIS), is used for graphical presentation. Well information, such as well location and vertical subsurface data is filed on the shrewd sheet, whereas two dimensional geographic data such as topography are maintained in PC-ARC. With the help of some in-house software, the data in MS-Excel are converted to a format acceptable for PC-ARC/INFO. The locational information and descriptive information are combined on PC-INFO, and PC-ARCPLOT (a part of PC-ARC/INFO) can produce the graphical output of well data both on maps and cross-sections. Furthermore, isodepth map and stratigraphic section indespensable for analysis of sedimentary basin can be edited on graphical outout by PC-ARCEDIT (a part of PC-ARC/INFO) .

View full abstract

GEOLIS FD is a floppy disk version of GEOLIS ( Bibliographic data base for geology of Japan) and is distributed without cost for scientific use by Geological Survey of Japan. This article describes a retrieval program for searching data from the GEOLIS FD. The functions of the program are 1) to search data from GEOLIS FD using logical combillation of keywords, 2) to format outputs, and that 3) output device file can be selected by user. The program in written by ANSI C language and does not use any nonstandard system call of MS-DOS. Consequently the program runs on almost all Japanese MS-DOS based personal computers.

View full abstract