Eiji Ikoma

Development of a data mining application for huge scale earth environmental data archives

The amount of earth environmental data has increased explosively because of recent advances in observational techniques. However, such valuable data has not been used adequately because researchers can not handle the volume. In this study, we collaborated closely with some researchers in the field and developed a support system for analysing various kinds of natural phenomena that have been difficult to analyse in the past. This was enabled by developing an analysing toolkit for a huge amount of earth environmental data using the data mining technique, and developing an easy-to-use interface for non-computational researchers. Moreover, we discuss some new knowledge that was acquired during the development of this system.

Implementation of Disaster Prevention-Oriented Information Service Platform of XRAIN on DIAS

In recent years the phenomenon of localized burst has increased in Japan. The disaster caused by local heavy rain is often serious, and disaster prevention measures have become urgent. X band MP Radar (called XRAIN) was Installed throughout Japan because of high spatial resolution, which was higher than that of conventional radar, the shortened observation time was useful for localized burst. However, the quantity of data of XRAIN are on a macroscale approximately 1,000 times bigger per day than conventional radar, thus, it was not easy to develop an environment that accumulated data in data-producing, and information sharing was difficult. Therefore, on DIAS, which is a earth environment data platform with the largest volume of storage in Japan, an information platform that acquired and accumulated and shared XRAIN data was developed for the purpose of information sharing. In this study, we addressed three problems in handling large-capacity data: 1. Processing delay, 2. Overlapped data, 3. Low presentation speed. We implemented solutions to these problems and develop a stable information platform on DIAS. As a result, DIAS contributes to the study of flood disaster prevention measures and is the most important information platform for XRAIN. In this paper, we introduce problems and methods are suggested for the solution, in addition, the implementation result that occurred in developing this information platform is reported.

Proposed Method for Disaster Prevention-Oriented Synthetic Rainfall of C-band radar and XRAIN

In recent years, the phenomenon of localized bursts of rainfall has increased in Japan. For this reason, collecting and sharing rain information quickly is the most important issue in terms of disaster prevention measures related to water disasters. XRAIN, which is suitable for the observation of localized heavy rain, enables high resolution observations and drastic shortening of observation times. However, the observation area of XRAIN does not cover all of Japan. Conversely, C-band radar, which is suitable for wide-area observation, covers the entire country of Japan, but it has difficulty to observing localized heavy rain. Both radars have opposite characteristics. For this reason, we examined a method that complements observation information from XRAIN with that of C-band radar to determine the rain situation in all of Japan, including localized heavy rain. In this study, we also proposed a method that synthesizes observation data of XRAIN and C-band radar and uses our proposed disaster prevention-oriented method. As a result, our method made it possible to display observation information throughout Japan, including localized heavy rain, in real time. The proposed method and implementation results are described in this paper.

Development of an earth environmental digital library system for soil and land-atmospheric data

We propose and examine new methods for automatic data loading system and flexible user interface system with many features such as 3D visualization. We implement the earth environmental digital library and operate it on the Web. Though our system is focusing the limited users like earth environmental researchers, more than 8000 hits per month describe the practical usefulness of it.

Development of an Earth environmental digital library system for soil and land-atmospheric data

Recently, as interest in the Earth environment increases, research on a digital library system for environmental data such as remote sensing data is getting more popular. However, no existing system is sufficient for practical use in the field of Earth environmental research because of the wide variety of data and poor user interface systems. We propose and examine new methods for automatic data loading and a flexible user interface system with many features such as 3D visualization. We implement the Earth environmental digital library and operate it on the Web. This method can load 1100 kinds of Earth environmental data from 30000 files into the digital library almost automatically. Although the system is focused on limited users like Earth environmental researchers, more than 8000 hits per month describe the practical usefulness of it. The distribution of the contents of accessed data shows the demand for various kinds of those data.