In this paper, we suggest the aplication of the principal component analysis on the data of microfossil analysis and tests of rock mechanics. There are some problems about the past methods, for these data are multivariate. The principal component analysis is appropriate method for processing multivariate data statistically, finding out relationship between samples and then discovering classifiable standard which depends on qualitative consideration before. In this time, we have put in order the routine of this analysis on the geological and geotechnical data, then mention the sphere and some problems of common use.
The Japanese group of the IGCP Project 163 IGBA has been constructed to participate in the generation of an international data base of igneous rocks in the world. Geochemical major element data of the Japanese Quaternary volcanic rocks, which were compiled and managed by ARAMAKI and UI with a small DBMS of a mini-computer system, were analyzed as an example to show how to use the data base in petrology. The analytical system is composed of SPSS and TSAM; the former was used not only for statistical analysis but also for data modification, and the latter was for a trend surface analysis by using two-dimensional moving average method. Several descriptive statistics were calculated to evaluate the data, and cluster and factor analyses were performed to clarify the data structure. The Japanese Quaternary volcanoes were dividen into five areal groups, and the grouping was tested by computing canonical discriminant functions. It is assumed that these groups don't perfectly refrect geochemical composition, but distribution of several groups show interesting patterns for volcanic petrology.
A project has started to compile the informations on fossil vertebrate specimens deposited in Japanese institutions and museums. The data base will be constructed by using the data base management system FAIRS in the Data Processing Center of Kyoto University, and will be easily used by any Japanese researcher through the computer network. It contains keys on sample identification, type of specimen, taxonomic name including synonyms, locality, geologic age, radiometric age, biostratigraphic zone, preservation condition, depository, and bibliography. It has dictionary files on the taxonomic classification, locality names, geographical coordinates, map codes, and so on, which are used to transform original data offered by researcher into input data of the specific format. Inverted files will be created on several keys for quick and accurate retrieval process. In the future fossil invertebrate specimens will be included in this data base, numerical informations such as morphological measurements and indices of fossil specimens will be added, and image data such as photograph or its digitized data will also be managed by the same system. Then it will become the paleontological data base, and it will help paleontologists to make their researches faster and more detailed with less effort.
Recently, there are many cases to construct buildings and roads in the area avoided before on account of instability, so that the prevention works of landslide have been sharply important. Analysis of landslides and design of prevention works use to take a lot of time and labor, so the engineers can not be given enough time to do the routine works of landslides satisfactorily. On the other hand, microcomputers and their peripheral equipments have been cheaper and simpler as a result of the latest remarkable progress of microelectronics. So we have developed the interactive design system for prevention works of landslides by means of a microcomputer. Using this system, the analysis and design have been carried out rationally and almost automatically. In this paper, outline of the system is shown with some examples.
Geological observation and experiments have provided us very useful and important information with a variety of image data such as remote sensing and microscopic images. To utilize these data for researchers, it is highly desirable to develope a effective personal system which handle a variety of images with aids by computer. Depending on an understanding of the geological image processing, analysis, design and applications of the system are discussed with the context of organization, hardware and software planning.
In this paper, examples of digital image processing for the petrographic studies are presented. The functions of system for this image processing is characterized by measurements such as color index, grain size (area, perimeter, exetent), and grain shape, using the image data of microscopic pictures or hand specimens of rocks. Methods of the processing are as follows; 1.smoothing, 2.edge detection, 3.threshold, 4.boundary detection, 5.labeling, 6.topological analysis. This application described here is still in an experimental stage.
At the usual methods, geological data of borehole are generally collected and recorded from core samples and operational tests by geologist in situ. At this time, charecters of lithology and stracture are coded by observers following by the code book, so the quite lots of information are likely to be out of the data sets. Then, we have developed the ststem by which the image data collected from boring hole are recorded, degitized, reconstructed, analized and displayed.
As a subroutine subprogram PELTO was developed in order to interpolate randomly spaced two-dimensional data on a rectangular grid using the method given in Appendix of Pelto et al. (1968), the simple explanation is presented with the program list in Appendix.
As a subroutine subprogram SFS3L was developed in order to make a new set of grid data from a given set of grid data using de Boor (1962) 's method of spline interpolation, the simple explanation is presented with the program list in Appendix.