MC 6460
Speaker
Miles Couchman,York University
Title
Data-driven explorations of turbulent mixing in stratified flows
Abstract
Understanding how turbulence enhances the irreversible mixing of scalars in density-stratified fluids is a central problem in geophysical fluid dynamics. Of critical importance is accurately parametrizing how turbulence transports heat within the ocean, a leading area of uncertainty in global climate models. We here present a series of data-driven approaches for extracting novel insight into turbulent mixing processes based on local analyses of the flow. First, we describe an unsupervised clustering technique for analyzing spatiotemporal distributions of mixing in oceanographic data, highlighting that traditional methods of analysis may significantly underestimate mixing generated by extreme, localized events. We then consider complementary direct numerical simulations of stratified turbulence, revealing that extreme mixing events are often correlated with the formation of stable, anisotropic density interfaces embedded within the turbulent flow. Finally, we introduce an automated method for classifying experimental videos of stratified flow instabilities, uncovering new dynamical regimes and turbulent transition pathways. Collectively, our findings suggest that rare, but extreme mixing events have the potential to dominate bulk mixing statistics. Current parametrizations of oceanic heat transport may thus be skewed by undersampled measurements, resulting in a focus on the most common, but not necessarily the most significant, mixing events.