The University of Southampton
University of Southampton Institutional Repository

Marine ecosystem model analysis using data assimilation

Marine ecosystem model analysis using data assimilation
Marine ecosystem model analysis using data assimilation
Numerical modelling of the marine ecosystem requires the aggregation of diverse chemical and biological species into broad categories. To avoid large bias errors it is preferable to resolve as many explicit state variables and processes as possible. The cost of this increased complexity is greater uncertainty in model parameters and output. When comparing models, the importance of quantifying both bias error and the variability of unconstrained solutions was revealed as two marine ecosystem models were calibrated to data. Results demonstrated that all prior parameter information must include realistic error estimates if model uncertainty is to be quantified. Five simple ecosystem models were calibrated to observations from two North Atlantic sites; the Bermuda Atlantic Time-series Study (BATS) and the North Atlantic Bloom Experiment (NABE). Model-data mists were reduced by between 45 and 50%. The addition of model complexity (a parameterised microbial loop, a variable chlorophyll a to nitrogen ratio and dissolved organic nitrogen) led to larger improvements in model performance at BATS relative to NABE. Calibrated parameter values developed at NABE performed better than the default parameter values when applied at BATS. Solutions developed at BATS performed worse than the default values at NABE. The models lacked sufficient ecological complexity to function well at BATS. Errors in the model were masked by errors in the calibrated parameters and the models did not perform well with regard to independent data. The models were well suited to reproducing the NABE data, and the calibrated models performed relatively well at BATS. The models were sensitive to the underlying physical forcing. Although the ecosystem models were originally calibrated within a poor representation of the physical environment at BATS, results from experiments using an improved physical model support the conclusion that the ecosystem models lacked the required complexity at that site.
Ward, Ben Andrew
18f31342-f12c-4ff7-9ab7-64bbaba4d917
Ward, Ben Andrew
18f31342-f12c-4ff7-9ab7-64bbaba4d917

Ward, Ben Andrew (2009) Marine ecosystem model analysis using data assimilation. University of Southampton, School of Ocean and Earth Science, Doctoral Thesis, 204pp.

Record type: Thesis (Doctoral)

Abstract

Numerical modelling of the marine ecosystem requires the aggregation of diverse chemical and biological species into broad categories. To avoid large bias errors it is preferable to resolve as many explicit state variables and processes as possible. The cost of this increased complexity is greater uncertainty in model parameters and output. When comparing models, the importance of quantifying both bias error and the variability of unconstrained solutions was revealed as two marine ecosystem models were calibrated to data. Results demonstrated that all prior parameter information must include realistic error estimates if model uncertainty is to be quantified. Five simple ecosystem models were calibrated to observations from two North Atlantic sites; the Bermuda Atlantic Time-series Study (BATS) and the North Atlantic Bloom Experiment (NABE). Model-data mists were reduced by between 45 and 50%. The addition of model complexity (a parameterised microbial loop, a variable chlorophyll a to nitrogen ratio and dissolved organic nitrogen) led to larger improvements in model performance at BATS relative to NABE. Calibrated parameter values developed at NABE performed better than the default parameter values when applied at BATS. Solutions developed at BATS performed worse than the default values at NABE. The models lacked sufficient ecological complexity to function well at BATS. Errors in the model were masked by errors in the calibrated parameters and the models did not perform well with regard to independent data. The models were well suited to reproducing the NABE data, and the calibrated models performed relatively well at BATS. The models were sensitive to the underlying physical forcing. Although the ecosystem models were originally calibrated within a poor representation of the physical environment at BATS, results from experiments using an improved physical model support the conclusion that the ecosystem models lacked the required complexity at that site.

Text
Ward_2009_PhD.pdf - Other
Download (33MB)

More information

Published date: October 2009
Organisations: University of Southampton

Identifiers

Local EPrints ID: 145089
URI: http://eprints.soton.ac.uk/id/eprint/145089
PURE UUID: a20507fe-7bed-4f31-b968-0c64842f0dec

Catalogue record

Date deposited: 15 Apr 2010 15:37
Last modified: 14 Mar 2024 00:49

Export record

Contributors

Author: Ben Andrew Ward

Download statistics

Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.

View more statistics

Atom RSS 1.0 RSS 2.0

Contact ePrints Soton: eprints@soton.ac.uk

ePrints Soton supports OAI 2.0 with a base URL of http://eprints.soton.ac.uk/cgi/oai2

This repository has been built using EPrints software, developed at the University of Southampton, but available to everyone to use.

We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we will assume that you are happy to receive cookies on the University of Southampton website.

×