In order to investigate the magma accumulation process of Izu-Oshima Volcano in the eruption preparation period, we conducted 14 relative microgravity surveys from 2004 to 2009. The gravity changes tend to decrease near the northern margin of the summit caldera, and rates of negative gravity changes reached as much as 0.015 mgal/year. Assuming the Mogi model, we estimated the pressure increase at a depth of 3.65 km with increasing volume of 7.6 × 106 m3/year, which could account for the observed data. However, the uplift rate calculated with the estimated increasing volume would be five times as large as the observed uplift rate by the GPS network, so these estimated parameters cannot explain both gravity changes and uplift changes simultaneously. Izu-Oshima Island could be a volcano having an extraordinary magma accumulation system that cannot be expressed as a simple physical model. More accurate microgravity survey with the uplift observation should be conducted in the future to clarify the magma accumulation system of this complicated volcano.
As one of the forecasts on volcanic phenomena, the tephra fall forecast has been operationally disseminated from the Japan Meteorological Agency (JMA), since 31 March 2008. JMA and the Meteorological Research Institute (MRI) have been developing a system of tephra fall prediction to improve the tephra fall forecast. In this paper, we focus mainly on the technical method of quantitative prediction of tephra fall in this system. The method of tephra fall prediction consists of three steps: The first step is to set the model of eruption column, which leads to the initial condition. The eruption column is composed of tracer volcanic-ash particles with virtual masses. The next step is to calculate the time evolution of the tracer particles with the JMA mesoscale tracer transport model. This is the Lagrangian description model using the meteorological fields predicted by the JMA Mesoscale Model (MSM) and also by considering the processes of three-dimensional advection, horizontal and vertical diffusion, fallout, dry and wet deposition. The final step is the output for the amount of ash-fall where the surface density of the ash-fall mass is transformed from each virtual mass of the deposited tracer particles. The above-mentioned method was applied to the event of the eruption at Asama volcano on 1 February 2009 with the eruption cloud echo height observed by weather radar in the first step and finer horizontal grid size than that of MSM in the final step. From the qualitative point of view, the predictions of ash-fall areas and principal distribution axis are almost correct in comparison with observations. Also, from the quantitative point of view, the amount of ash-fall can be predicted in the same order along the principal distribution axis.