-演講時間: 115年6月3日（三） 14:00~16:00

-演講地點: E6-A203教室

-演講者: 王科植 副教授(國立臺灣師範大學資訊工程學系)

-演講主題: A Cross-Domain Collaboration Between Computer Science and Atmospheric Science: From Scientific Visualization to AI Climate Downscaling

-演講摘要: This talk shares my experience collaborating with atmospheric scientists over the past several years, from the perspective of a computer scientist working on scientific visualization and machine learning. I had the opportunity to start working with atmospheric scientists, and the collaboration gradually developed from small joint projects into shared research on AI-based climate downscaling. I will describe how this collaboration evolved in stages, and some of the practical difficulties of working across fields, which are often less about technical difficulty than about understanding what domain scientists really need and learning to speak each other's language. I will introduce a few representative works from our lab, including deep-learning emulators for scientific data and a visual analytics tool for examining what a climate-downscaling model has learned. I will also talk about our collaboration with Taiwan's Central Weather Administration on the CorrDiff downscaling model: for computer scientists, some of the work may seem like small engineering tasks, but the collaboration offers a valuable channel for our work to be actually deployed and used by domain scientists. Finally, I will mention our participation in CORDEX, an international project in which regional downscaling models are run on a common dataset and compared. Rather than presenting finished results, I hope to share some of the experience and reflections from doing cross-disciplinary research.

-講者簡歷:Ko-Chih Wang is an Associate Professor in the Department of Computer Science and Information Engineering at National Taiwan Normal University (NTNU), where he directs the Visual Data Analysis (ViDA) Lab. He received his Ph.D. in Computer Science and Engineering from The Ohio State University in 2019, advised by Prof. Han-Wei Shen. His research spans scientific visualization, visual analytics, and machine learning, with a focus on trustworthy deep-learning emulators and human-centered AI for large-scale scientific data. He served as General Chair of IEEE PacificVis 2025 and Local Organizing Co-Chair of Pacific Graphics 2025. His recent work spans ML-assisted scientific data summarization, explainable AI for climate downscaling, and large language models for visualization.