This repository contains code used for a paper published in Frontiers in Marine Science, titled "Decadal variability of oxygen uptake, export, and storage in the Labrador Sea from observations and CMIP6 models", available at https://www.frontiersin.org/articles/10.3389/fmars.2023.1202299/. The paper was originally submitted on April 7, 2023, and accepted for publication on October 9.
The notebooks included in the repository show examples of the calculation of a "model score", which assesses the ability of models to reproduces observed oxygen data in the Labrador Sea. There are currently two main notebooks showing examples of the score calculation:
- cmip6_score, which uses OpenDAP URLs to directly access files from the server for the MIROC E2SL and NOAA GFDL CM4 models and can thus be run without any additional downloads
- cmip6_score_local_all, which includes data for more models (eight as of Nov 4, 2023), but requires the CMIP6 data to be downloaded separately to run the notebook. There is also a reduced example for the local code, cmip6_score_local, which uses data only from three of the models, which have output fields on a regular grid.
- Added new notebook for the score calculation with local files structure, cmip6_score_local_all; this notebook calculates the score for eight of the nine models used in the paper.
- Updated other notebooks to optimize code
Optimized code to run faster, and fixed some small errors
Uploaded new CSV files of Labrador Sea oxygen content and updated score calculation in both notebooks to reflect changes made during paper revisions. Added link to publication in readme and notebooks
Added second notebook for calculation of scores accessing CMIP6 files directly from the OpenDAP servers without downloading
Added code for the calculation of the model score. Current version includes examples for CMIP6 gridded model data from MIROC E2SL, NCAR CESM2 and NOAA GFDL. Note that the CMIP data are currently not uploaded on GitHub, so to run the code the user would have to separately download the data.