On Thursday, October 27, 2022, the Selected Topics Meteorology Guest Lecture (ME-4097) was successfully held. On that day, the Meteorology Program held a Meteorology Selected Topics lecture featuring a guest lecturer online, open to the public as a means to broaden knowledge regarding the application of deep learning in predicting typhoon phenomena.
The Guest Lecture began with an introduction and opening remarks from the Head of the Meteorology Program, who serves as the course instructor for the Meteorology Seminar. The event continued with a presentation on the topic “Applications of Deep Learning for Typhoon Prediction” by Dr. Iyan E. Mulia, PhD, lasting approximately 50 minutes.
The presentation began with a general overview of tropical cyclones (typhoons), including their history and intensity. This was followed by an introduction to deep learning and Generative Adversarial Neural Networks (GANs) as a type of deep learning model. The GAN deep learning model achieves accuracy comparable to Numerical Weather Prediction (NWP) and is far superior to standard parametric models. The computational time for GAN to predict a typhoon event is approximately 2 seconds. These wind field and pressure predictions by GAN are expected to facilitate efficient storm surge forecasting.
The event then continued with a 30-minute Q&A session. During this session, students and general attendees were invited to ask questions and discuss the material presented earlier. The next part of the event was the presentation of certificates to the speakers by the lecturer in charge of the Meteorology Selected Topics course, followed by a closing session featuring a group photo.
