This repository provides code in supplement to "Separating internal and externally-forced contributions to global temperature variability using a Bayesian stochastic energy balance framework" (M. Schillinger et al. 2022) published in Chaos: An Interdisciplinary Journal of Nonlinear Science. A preprint of the manuscript is available at https://aip.scitation.org/doi/full/10.1063/5.0106123 .
Authors: M. Schillinger, B. Ellerhoff, R. Scheichl, K. Rehfeld
Responsible for this repository: Beatrice Ellerhoff (@bellerhoff), Maybritt Schillinger (@m-schilinger)
| directories | description |
|---|---|
./output/ |
contains pre-processed data from stochastic multibox EBM fit to data |
./output-spectra/ |
contains pre-processed spectra of stochastic multibox EBM fit |
./plots/ |
empty directory to store created figures |
| scripts | description |
|---|---|
init.R |
load metadata, required libraries and plotting settings |
T2_params.R |
Summary of estimated parameters of 2-box fit, creates Table 2 |
F2_hadcrut_demo.R |
Application of workflow to observations, creates Figure 2 |
F3-F4_spectra_variance.R |
Spectral analysis and emulation of timescale-depenedent variance, creates Figure 3 and 4 |
FS5-FS6-TS3_hadcrut_supp.R |
Supplementary validity check of choice of parameters and sampling of internal variance, creates Figues 5 and 6, as well as Table 3 (in Appendix) |
F7_plot_fits.R |
Forced response from 2-box fit for all considered runs, Figure 7 in Appendix |
FS8_large_ensemble.R |
Comparison of fitted forced response to forced variability from large ensemble |
| additional files | description |
|---|---|
.gitignore |
Information for GIT version control to not add several file extensions to version control (e.g. *.png, *.pdf) |
license.md/ license.html |
Licensing information |
README.md |
General README |
Running the code in this repository requires the following R packages:
ClimBayesfrom https://github.com/paleovar/ClimBayes (latest release v.0.1.1)dplyrggplot2tibblePaleoSpecRColorBrewerzooggpubrpurrrstringrlatex2exptidyr
The directory ./output/ contains the pre-processed data, which we obtained from fitting the stochastic two-box energy balance model (EBM) to global mean surface temperature (GMST) data using the ClimBayes package in R. The target data (Table 1 of submitted manuscript) is available from the data holdings of the Climate Research Programme’s Working, from Schmidt et al., Eby et al. (https://climate.uvic.ca/EMICAR5/participants.html), and Morice et al.. The ClimBayes package provides detailed information on how to prepare and fit the target data (see vignettes). To reproduce our runs, use the ebm_fit_config.yml in the respective ./output/ directory and the ebm_fit() function from ClimBayes. We used n_boxes=2 and detrending=2.
ebm_fit(temp_data, forc_data, start_year, end_year, n_boxes, config, detrending, config_file)
Please see the data availability and acknowledgment statement of the submitted manuscript (https://arxiv.org/abs/2206.14573) and the ClimBayes package.
We acknowledge the R Core team and all package developers of packages used in this study. We thank them for their time and dedication to provide R and the packages to the public. Please see citation() for details on the R Core Team and citation("pkgname") for details on the developers of individual packages.
Please report bugs to the authors (beatrice-marie.ellerhoff(at)uni-tuebingen.de, maybritt.schillinger(at)stat.math.ethz.ch).
Beatrice Ellerhoff and Maybritt Schillinger, June 2022