A deep learning approach for coastal downscaling: The northern Adriatic Sea case-study

Current regional-scale oceanographic operational systems may lack the resolution needed for coastal applications, where fine-scale dynamics such as river outflow and local processes are poorly represented. Artificial intelligence based techniques for interpolation and fine-scale reconstruction can be applied for studying coastal dynamics. In this paper, we propose a deep learning method based on a UNet-like architecture for coastal downscaling of marine ecosystem modeling products. Our method is applied to the northern Adriatic Sea, a marginal region of the Mediterranean characterized by strong spatial and temporal variability, where river inputs significantly influence the physical and biogeochemical state and dynamics, especially near the coast. To address these challenges, we trained a neural network on a reanalysis dataset, covering the period from 2006 to 2017, with a horizontal resolution of about 750 m, using as input the regional-scale products of the Marine Copernicus Service for the Mediterranean Sea. We demonstrate that our architecture is capable of recovering fine-scale features that are not captured by low-resolution modeling systems. Although this paper focuses on the northern Adriatic Sea, the robustness of the method, as demonstrated by the validation metrics, suggests that it can be effectively applied to other study areas.