Use of genetic algorithms for ocean model parameter optimisation: a case study using PISCES-v2_RC for North Atlantic particulate organic carbon

Falls, Marcus; Bernardello, Raffaele; Castrillo, Miguel; Acosta, Mario; Llort, Joan; Galí, Martí

doi:https://doi.org/10.5194/gmd-15-5713-2022

Articles | Volume 15, issue 14

https://doi.org/10.5194/gmd-15-5713-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

Special issue:

Nucleus for European Modelling of the Ocean - NEMO

https://doi.org/10.5194/gmd-15-5713-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 15, issue 14

Development and technical paper

|

22 Jul 2022

Development and technical paper |

| 22 Jul 2022

Use of genetic algorithms for ocean model parameter optimisation: a case study using PISCES-v2_RC for North Atlantic particulate organic carbon

Marcus Falls, Raffaele Bernardello, Miguel Castrillo, Mario Acosta, Joan Llort, and Martí Galí

Data sets

ARGO Data International Argo Program (Global Ocean Observing System) https://argo.ucsd.edu

Datasets for the comparison between POC estimated from BGC-Argo floats and PISCES model simulations (1.0.0) M. Galí, R. Benardello, M. Falls, H. Claustre, and O. Aumont https://doi.org/10.5281/zenodo.5139602

Model code and software

NEMO and PISCES code Nucleus for European Modelling of the Ocean (NEMO) https://www.nemo-ocean.eu/

PISCES-v2 Martí Galí https://earth.bsc.es/gitlab/mgalitap/p1d_share/-/tree/gapoc

Genetic Algorithm Workflow Marcus Falls https://earth.bsc.es/gitlab/cp/genetic_algorithm_pisces1d

Genetic Algorithm Pisces 1D Workflow and config files Marcus Falls https://doi.org/10.5281/zenodo.5205760

PISCES-v2 1D configuration used to study POC dynamics as observed by BGC-Argo floats (1.0.0) Martí Galí https://doi.org/10.5281/zenodo.5243343

Short summary

This paper describes and tests a method which uses a genetic algorithm (GA), a type of optimisation algorithm, on an ocean biogeochemical model. The aim is to produce a set of numerical parameters that best reflect the observed data of particulate organic carbon in a specific region of the ocean. We show that the GA can provide optimised model parameters in a robust and efficient manner and can also help detect model limitations, ultimately leading to a reduction in the model uncertainties.