Model selection for ocean transport

Lead Research Organisation: University of Edinburgh
Department Name: Sch of Mathematics

Abstract

The dynamics of the world's oceans are responsible for the transport of heat and important materials, such as carbon, around the globe. Understanding this transport is critical for many scientific endeavours, such as modelling climate change. Many transport models of varying levels of complexity have been previously proposed and yet little effort has been directed toward objectively assessing the relative explanatory and predictive strength of these models. The aim of this project is to explore and develop appropriate methods for comparing the performance of commonly used transport models in a systematic and principled way, using data and theory from Bayesian statistics. In particular, since the cost of implementing models grows rapidly with model complexity, it is necessary to determine when more complex models are worth the additional computational cost; this consideration of complexity versus cost introduces a sort of Occam's razor to model selection which needs to be made rigorous to inform practical decisions.

Initial methodological development is tested using particle simulations from an idealised model of a turbulent flow, as a caricature of ocean turbulence, and the common transport models compared are the so-called Brownian and Langevin models.

Publications

10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
EP/S023291/1 01/10/2019 31/03/2028
2284952 Studentship EP/S023291/1 01/09/2019 31/08/2023 Martin Thomas Brolly