Development of Real-Time Tsunami Early Warning System Dashboard Based on Tunami-F1 and Machine Learning in Sunda Arc, Indonesia

The majority of tsunamis in Indonesia are nearfield tsunamis, therefore a quick tsunami early warning system is essential to minimize the risk. This study aimed to build a prototype of the dashboard for tsunami early warning systems that can provide tsunami prediction within the first five minutes. This prototype of the dashboard has the main function to predict the occurrence of tsunamis based on hydrodynamic modeling of wave propagation and tsunami artificial intelligence (AI) modeling which runs simultaneously when receiving earthquake data with tsunami potential. The AI model needs just a few seconds to predict tsunami heights and arrival times at affected locations, while the hydrodynamic wave propagation model needs a few minutes to complete the task. However, the results from the latter model are more accurate and complete, such as sea level elevation and the estimated time of arrival. The result of this study shows that the dashboard prototype can quickly and accurately display the results of tsunami predictions by combining the results of the two models. The results and other data are then compiled into a tsunami bulletin and distributed to competent authorities. This study is expected to be used as one of the considerations in developing a dashboard system for tsunami early warning by applying AI in addition to tsunami propagation modeling.