The University of Southampton
University of Southampton Institutional Repository

Split-domain calibration of an ecosystem model using satellite ocean colour data

Split-domain calibration of an ecosystem model using satellite ocean colour data
Split-domain calibration of an ecosystem model using satellite ocean colour data
The application of satellite ocean colour data to the calibration of plankton ecosystem models for large geographic domains, over which their ideal parameters cannot be assumed to be invariant, is investigated. A method is presented for seeking the number and geographic scope of parameter sets which allows the best fit to validation data to be achieved. These are independent data not used in the parameter estimation process. The goodness-of-fit of the optimally calibrated model to the validation data is an objective measure of merit for the model, together with its external forcing data. Importantly, this is a statistic which can be used for comparative evaluation of different models. The method makes use of observations from multiple locations, referred to as stations, distributed across the geographic domain. It relies on a technique for finding groups of stations which can be aggregated for parameter estimation purposes with minimal increase in the resulting misfit between model and observations.
The results of testing this split-domain calibration method for a simple zero dimensional model, using observations from 30 stations in the North Atlantic, are presented. The stations are divided into separate calibration and validation sets. One year of ocean colour data from each station were used in conjunction with a climatological estimate of the station’s annual nitrate maximum. The results demonstrate the practical utility of the method and imply that an optimal fit of the model to the validation data would be given by two parameter sets. The corresponding division of the North Atlantic domain into two provinces allows a misfit-based cost to be achieved which is 25% lower than that for the single parameter set obtained using all of the calibration stations. In general, parameters are poorly constrained, contributing to a high degree of uncertainty in model output for unobserved variables. This suggests that limited progress towards a definitive model calibration can be made without including other types of observations.
ecosystem modelling, parameter estimation, data assimilation, biogeochemical provinces
0924-7963
141-179
Hemmings, J.C.P.
ebf33f54-d2b2-4ab3-9ac8-fd9dc9ae6a7f
Srokosz, M.A.
1e0442ce-679f-43f2-8fe4-9a0f0174d483
Challenor, P.
a7e71e56-8391-442c-b140-6e4b90c33547
Fasham, M.J.R.
7fb86485-8cfc-4199-bde4-2276abefdf2e
Hemmings, J.C.P.
ebf33f54-d2b2-4ab3-9ac8-fd9dc9ae6a7f
Srokosz, M.A.
1e0442ce-679f-43f2-8fe4-9a0f0174d483
Challenor, P.
a7e71e56-8391-442c-b140-6e4b90c33547
Fasham, M.J.R.
7fb86485-8cfc-4199-bde4-2276abefdf2e

Hemmings, J.C.P., Srokosz, M.A., Challenor, P. and Fasham, M.J.R. (2004) Split-domain calibration of an ecosystem model using satellite ocean colour data. Journal of Marine Systems, 50 (3-4), 141-179. (doi:10.1016/j.jmarsys.2004.02.003).

Record type: Article

Abstract

The application of satellite ocean colour data to the calibration of plankton ecosystem models for large geographic domains, over which their ideal parameters cannot be assumed to be invariant, is investigated. A method is presented for seeking the number and geographic scope of parameter sets which allows the best fit to validation data to be achieved. These are independent data not used in the parameter estimation process. The goodness-of-fit of the optimally calibrated model to the validation data is an objective measure of merit for the model, together with its external forcing data. Importantly, this is a statistic which can be used for comparative evaluation of different models. The method makes use of observations from multiple locations, referred to as stations, distributed across the geographic domain. It relies on a technique for finding groups of stations which can be aggregated for parameter estimation purposes with minimal increase in the resulting misfit between model and observations.
The results of testing this split-domain calibration method for a simple zero dimensional model, using observations from 30 stations in the North Atlantic, are presented. The stations are divided into separate calibration and validation sets. One year of ocean colour data from each station were used in conjunction with a climatological estimate of the station’s annual nitrate maximum. The results demonstrate the practical utility of the method and imply that an optimal fit of the model to the validation data would be given by two parameter sets. The corresponding division of the North Atlantic domain into two provinces allows a misfit-based cost to be achieved which is 25% lower than that for the single parameter set obtained using all of the calibration stations. In general, parameters are poorly constrained, contributing to a high degree of uncertainty in model output for unobserved variables. This suggests that limited progress towards a definitive model calibration can be made without including other types of observations.

Text
Hemmings_revised.pdf - Other
Download (1MB)

More information

Published date: 2004
Keywords: ecosystem modelling, parameter estimation, data assimilation, biogeochemical provinces

Identifiers

Local EPrints ID: 217
URI: http://eprints.soton.ac.uk/id/eprint/217
ISSN: 0924-7963
PURE UUID: b80cad5a-5e13-4561-8878-6daa62f75ddb

Catalogue record

Date deposited: 05 Nov 2004
Last modified: 15 Mar 2024 04:37

Export record

Altmetrics

Contributors

Author: J.C.P. Hemmings
Author: M.A. Srokosz
Author: P. Challenor
Author: M.J.R. Fasham

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.

×