A PROTOTYPE FOR THE DETECTION AND CLASSIFICATION OF SEISMIC EVENTS USING STA/LTA AND MACHINE LEARNING

In this study, the detection and classification of seismic events is a significant concern of this research. A volcano eruption is one of the natural disasters on Earth. Monitoring volcano activities is essential to analyzing and monitoring volcanoes before their eruption. This activity is beneficial in interpreting signals from a volcano before an eruption from the volcano can cause damage. Based on that, a tool has been developed to detect and classify volcanic seismic events. The combination of algorithm time series, which is STA/LTA and machine learning (LSTM), is being used to analyze data of seismic events. A dataset was collected from one of Indonesia's mountains during 2019 – 2021. The dataset will be classified into different classes based on the type of seismic events. Noise detection is implemented to classify true or false seismic events before continuing to detect and classify them. STA/LTA is used to remove noise signals from data seismic events. The next step is to use machine learning to classify labelling signals based on the type of seismic events. The experiments use a learning rate of 0.001 and 0.01. They show that tools can detect and classify signals of seismic events with an accuracy of around 0,70 – 0,80.