Data-Driven Modelling of Stratified Environmental Flows
Date
2019-08-31
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The research project studied is an integral part of the supervisor's Discovery Grant (DG) entitled "Process-based modelling of turbulent mixing in stratified natural waters.'' The purpose of the DG is to develop a new generation of models for turbulent mixing in natural waters, e.g. lakes, estuaries, and oceans, where the density of fluid typically increases with depth (i.e. stratified). Understanding stratified mixing is crucial for a range of important environmental problems, from modelling the ecological health of the Great Lakes to predicting oceanic heat uptake in a changing climate. Within the research project, a recent turbulent flow dataset is used to explore data-driven modelling strategies for stratified environmental flows. Although the governing equations for fluid flows have long been well known, solving them directly is prohibitively expensive for a majority of practical applications. As such, instead of solving the equations, the main purpose of the project is to obtain an optimal dynamical system model of complex flows based on a fully nonlinear data set, which will develop physical insights contributing towards the following questions [1]: 1. How does Re modify the dimensions and structures of the pancake vortices as they evolve in time? 2. How does Re affect the duration within which internal waves are emitted by the wake turbulence and the characteristics of wake-emitted waves? 3. How does the susceptibility to secondary shear instability within the pancake vortices vary with Re? Quantitatively, how much of the flow volume remains ‘turbulent’ at a given point in time? How much of the ‘turbulent’ region is susceptible to shear instabilities?
Description
Keywords
Citation
Tang, C. (2019). Data-Driven Modelling of Stratified Environmental Flows. Program for Undergraduate Research Experience (PURE), University of Calgary, Calgary, Alberta, Canada. 1-7.