Objective: The long coastline and proximity to tsunamigenic zones mandate the requirement of an effective Tsunami forecasting system for India to minimize loss of life and property. Methods/Statistical Methods/ Statistical Analysis: In this paper, we discussed a tsunami forecasting model for Bay of Bengal using Artificial Neural Network (ANN) and network of eight Tsunami stations. The ANN algorithm at each station characterizes the tsunami detected, while velocity and direction are obtained from the specified arrangement of buoys. The effectiveness of the ANN algorithm in characterizing a tsunami is discussed using actual time-series data. Findings: It is observed that modification to the ANN algorithm in7 can characterize a tsunami effectively, in terms of its amplitude and period. Analysis of the methodology is carried out using simulated data for obtaining the direction and velocity of tsunami. Improvements/Applications: Presently, the Indian Tsunami Warning System (ITWS) issue warnings based on the possible scenario selection from its exhaustive event database on the basis of inputs collected from various sources. Applications: This methodology could augment capabilities of ITWS by providing additional inputs on peak amplitude, direction and velocity of a detected tsunami for the proper scenario selection. This in turn helps in disseminating more reliable warnings.
Keywords
Artificial Neural Network for Tsunami, Bottom Pressure Recorders, Characterization of Tsunami, Sumatra Earthquake on 12th September 2017, Tsunami Direction and Velocity, Tsunami Warning.
User
Information