Sunmin Kim

I am an Associate Professor at Kyoto University's Department of Civil and Earth Resources Engineering, where I lead research on intelligent water systems and climate-resilient hydrology. My work bridges data-driven machine learning approaches with physically-based hydrological modelling to address pressing challenges in flood prediction and water resource management.

My research focuses on applying artificial neural networks — including convolutional and recurrent architectures — to real-time river stage prediction and rainfall forecasting. I am particularly interested in investigating how reinforcement learning algorithms can be applied to the operation of multiple dam reservoirs to enhance short-term flood control performance.

Another major thread of my work concerns climate change adaptation: using General Circulation Model (GCM) outputs to project future changes in extreme precipitation and water availability across major river basins in Japan and Asia. Recent work also investigates Probable Maximum Precipitation (PMP) estimation under future climate scenarios using large ensemble simulation data.