Surface circulation properties in the eastern Mediterranean emphasized using machine learning methods

Abstract. The eastern Mediterranean surface circulation is highly energetic and composed of structures interacting stochastically. However, some main features are still debated, the behavior fine-scale dynamics their role in shaping general yet unknown. In following paper, we use an unsupervised neural network clustering method to analyze long-term variability different mesoscale structures. We decompose 26 years altimetric data into clusters reflecting patterns weak strong flows with either strain or vortex-dominated velocities. cluster more persistent western part basin, which active than due flow along coast, extended bathymetry engendering continuous instabilities. that reflects a dominated middle including Mid-Mediterranean Jet (MMJ) pathway. temporal analysis shows frequent intermittent occurrence could explain previous contradictory assessment MMJ existence using in-situ observations. Moreover, prove Levantine Sea becoming as activity showing positive trend.