Classification of Types A and B Volcanic Earthquakes using Neural Network for Monitoring Mount Merapi Activities

Abstract

Mount Merapi is one of the most active volcanoes in the world. Seismic activity at Mount Merapi was divided into tectonic, volcanic A and B, avalanches, and multiphase. At the volcanology and mitigation station in the area of ​​Mount Merapi, the vibrations received by the seismometer installed on the mountain were then transmitted to the station to be interpreted in the form of a seismograph. Determination of the type of earthquake at the Mount Merapi station was done manually by analyzing the shape of the wave formed. Utilizing the earthquake waveform, it could be used to determine the type of earthquake using the Nntool toolbox in Matlab. So that the determination of the type of earthquake no longer needed to be done manually. The classification process began with training the system to understand earthquake-type classes. After the system understood the earthquake data belonging to each category, the classification process could be carried out. The types of earthquakes analyzed in this research were volcanic earthquakes of types A and B. The results of these classifications were used as a determinant of the type of earthquake that occurred. The results were obtained in the form of an introduction to the types of volcanic earthquakes type A and type B which were carried out automatically with a total accuracy of 91.83%. Type A earthquake recognition accuracy was 100%, type B earthquake recognition accuracy was 84.21%, the earthquake recognition accuracy of non-A & B type was 94.44%.