As one of the applications using GEO Grid, for the purpose of reducing volcanic disaster, we made a pyroclastic flow computer model which uses the highly accurate data of the ASTER sensor on an earth observatory satellite (with a spatial resolution of 15 m) above the sea-level. In this way GEO Grid is also tackling the technical development of such simulations for reduction of natural disasters.
Pyroclastic flow occurred 9500 times or more in 5 years, from 1991 to 1995 in the Unzen volcano of Nagasaki Prefecture. There were 43 victims in the pyroclastic flow of June 3rd, 1991. Pyroclastic flow of the Unzen volcano is considered to be most dangerous, as unstable parts of growing lava domes collapsed and flowed down at high temperature (above the 600 ℃) and high speed (above 100 km/h). For reducing damages caused by volcanic disasters, a "map (paper)" was drawn in the form of a disaster map (hazard map) based on the past activities of the principal active volcanoes in the entire country. In the future, the information of "the geographical information system (GIS)" will be used in superposing function of various data and drawing a "real time hazard map" responding to a demand of a local area.
It is possible to simulate an energy cone model on the Web browser in GEO Grid (Fig.1). This simulation requires the point of column collapse height (Hc) and the equivalent coefficient of the pyroclastic flow friction (H/L) as input parameters, and estimates the range of the energetic arrival of pyroclastic flow. Presently, simulation can be performed with nine volcanoes namely, the Merapi volcano (Indonesia), the Fuji volcano, the Unzen volcano, the Kirishima volcano, the Sakurajima volcano, the Yotei volcano, the Usu volcano, the Tarumae volcano and the Bandai volcano (Fig. 2). By adding satellite data such as three dimensional elevation model, even in midst of eruption, it can always use an updated topographical data around volcanic activities. It is possible to operate high speed processing with the grid technology. Processing of data in a brief time of 10 seconds - 3 minutes is also possible. Application of this simulation is not just restricted to the pyroclastic flow, but also to various volcanic and geological disasters such as avalanches and landslides. Opening the model to public from 2007 has been planned. We are aiming to enable the use of this simulation for any volcano in the world for researchers and to prevent disasters anytime and anywhere. Executing numerical simulation by granular flow model of lava flow is planned in the future. By this granular flow model of lava flow, quick correspondence in case of volcanic eruption becomes possible and expected to be useful as a deciding material for evacuation of inhabitants.




