INVESTIGATION ON REAL-TIME FORECASTING SYSTEM OF STORM SURGE AT THE NORTHERN TOKUSHIMA COAST

We investigated the real-time forecast system of a storm surge to reduce its disaster. Typhoon properties such as central atmospheric pressure were forecasted before several hours by two methods. The one is forecasting by using neural network technique based on the past typhoon data and the other is utilization of typhoon forecast information by the Meteorological Agency. As a result of the comparison of both methods, it was found that the latter gave more correct forecast of typhoon properties. Based on the forecasting values of typhoon, we carried out the real-time forecasting of a storm surge and examined the practical utility of this system.