Improving long-term monitoring around the Shetland and Orkney archipelagos with high resolution modelling and data science. 

High-resolution coastal ocean modelling is essential for understanding and managing complex coastal systems under increasing environmental and socio-economic pressures. We present the development of an unstructured FVCOM numerical model for the Shetland and Orkney archipelagos in the north of Scotland, with a hindcast run covering a 30-year period at unprecedented high resolution (~70m around the Shetland coast and hourly output), nested in the Scottish Shelf Model and fully-forced with 5.5km CERRA atmospheric data at hourly frequency. The unstructured grid allows to resolve the complex coastline and bathymetry that characterizes these areas. This region is paramount for the aquaculture industry, with Shetland alone making up for more than 20% of the Scottish salmon and more than 80% of Scottish mussel production, yet its energetic circulation, complex bathymetry, and strong coastal–ocean interactions make monitoring and prediction of potential environmental impacts particularly challenging. Combining numerical modelling with data science tools, we explore the system variability and complexity. This allows the identification of emergent patterns, dominant modes and changes that may otherwise be overlooked. Our work helps supporting more effective long-term monitoring and sustainable use of marine resources in a region increasingly affected by climate change.