There is a newer version of the record available.

Published September 17, 2021 | Version 1
Other Open

Code and data archive to accompany "A derivative-free optimisation method for global ocean biogeochemical models", Oliver et. al. 2021

  • 1. Department of Earth Sciences, University of Oxford
  • 2. Mathematical Institute, University of Oxford
  • 3. GEOMAR Helmholtz-Zentrum für Ozeanforschung Kiel
  • 4. School of GeoSciences, University of Edinburgh

Description

This archive is to accompany the article:

A derivative-free optimisation method for global ocean biogeochemical models,
Sophy Oliver, Coralia Cartis, Iris Kriest, Simon Tett, and Samar Khatiwala.

The optimisation framework used in this study to couple any climate model to any optimiser was originally sourced from
Tett et al. (2013). The current version of the software is available at https://github.com/SimonTett/ModelOptimisation. 
This was edited to couple MOPS to CMA-ES and DFO-LS, which is available at https://github.com/SophyOliver/OptClim
(see the user guide OptClim_UserGuide.pdf). 
The original source code of MOPS coupled to the TMM, as well as the separate CMA-ES optimisation framework were from
the Supplement of Kriest et al. 2017 (https://doi.org/10.5194/gmd-10-127-2017), though the script cmaes.cpp was edited to make it compatible with OptClim (available in OptClim/cmaes). 
The most recent TMM source code is available at https://github.com/samarkhatiwala/tmm, 
and most recent DFO-LS source code is available at https://github.com/numericalalgorithmsgroup/dfols.

In this archive:

OptClim/OptClimSO and OptClim/OxfordMOPS contain python Optclim scripts necessary for the Optclim package.

OptClim/Configurations/OxfordMOPS_Configs contain:
- ReadOnlyFiles (Files and Code specifically used to run the global ocean biogeochemical model MOPS model with
  the Transport Matrix Method, which have been edited to differ from the versions downloaded from the sources above.)
- RunCode (runscripts to run the MOPS model with the TMM)
- TWIN_Configs (JSON files required for each optimisation experiment carried out).

OptClim/OxfordMOPS_EXP contains data for each iteration of all optimisation experiments carried out.

OptClim/OPTCLIMSO_PlottingScripts contains MATLAB plotting scripts used to create results figures of these experiments.

OptClim/OptCLim_UserGuide.pdf is a user guide on how to use OptClim.

Files

OptClim.zip

Files (142.3 MB)

Name Size Download all
md5:8f624f5f9adab19661703b7cabc6fafa
142.3 MB Preview Download