Multi-Scale Coastal Temperature Prediction for the KAUST Reef Environment

KAUST Research Program

This project uses the KAUST reef environment as a testbed for advancing coastal temperature prediction capabilities that support reef monitoring and environmental decision-making. The work addresses coastal thermal processes across different spatial and temporal scales through two complementary tasks: ultra-high-resolution coastal ocean modeling and AI-based subseasonal forecasting. The first task will develop and validate a 50–100 m unstructured coastal ocean modeling system using in situ observations, satellite-derived sea surface temperature products, and other environmental observations to resolve reef-scale circulation and thermal processes. The second task will develop AI models for subseasonal prediction of coastal water temperatures using environmental observations, satellite products, and subseasonal weather forecasts, with a focus on identifying environmental drivers and assessing the predictability of marine heat stress several weeks in advance. Together, the tasks will establish a scientific and technical foundation for future reef forecasting, environmental monitoring, early-warning systems, and an integrated operational forecasting and decision-support platform.