Towards Advancing Streamflow and Peak Flow Prediction with Machine Learning: Identifying Infrastructure at Risk

Due to climate change and its impact, the need for adaptive strategies for natural disaster mitigation and resource management has never been more urgent. Central to this is water resource management, which is essential for sustainable human activities, ecological balance, and the mitigation of natural hazards like floods. Streamflow is a crucial element of water resource management and plays a vital role in planning and building water infrastructure, implementing emergency response plans, supporting flood mitigation initiatives, and regulating agricultural and industrial use. However, accurate prediction of streamflow still remains a challenge due to the complex non-linear and non-stationary interaction between hydrological and climate processes and human action. These challenges highlight the need for innovative and reliable prediction methods. Due to advancements in technology, data availability, and computer power, Machine Learning (ML) is showing promise in this field. However, there are still some gaps, such as including physics in ML models, utilizing the power of spatial and temporal signals, and data synergy. This dissertation fills these gaps by developing robust ML frameworks that enhance both scientific understanding and practical implementation for streamflow prediction.Advisor: Tirthankar Roy