Deep learning has emerged as a powerful tool for water quality monitoring and forecasting through the development of inversion algorithms capable of translating spectral reflectance into precise water quality parameters. However training these networks using remote sensing imagery is challenging due to insufficient data density and the paucity of cloud-free images, making it difficult to effectively train regional deep learning models.
To address this, this project proposes a novel approach by leveraging the eReefs model – encompassing hydrodynamic, biogeochemical, and optical simulations of the Great Barrier Reef – to generate comprehensive synthetic datasets of satellite imagery and create a virtual environment that mirrors the spectral response curves of satellite sensors. This project would simulate a decade's worth of satellite bands over the Great Barrier Reef region – offering a solution to the data scarcity issue and providing an abundant resource for pre-training advanced deep learning networks.